Report No. 22913-GE Georgia Public Expenditure Review November 25, 2002 Poverty Reduction and Economic Management Unit Europe and Central Asia Unit Document of the World Bank ammmmmmmmmm CURRENCY AND EQUIVALENT UNITS (Exchange Rate Effective as of November 26, 2002) Currency Unit = Lari US$1.00 = GEL 2.23490000 WEIGHTS AND MEASURES Metric System FISCAL YEAR January 1 to December 31 ACRONYMS AND ABBREVIATIONS ALOS Average Length of Hospital IEF Institutional Expenditure Stay Survey ASCUDA Automated System for IFRC International Federation of Red Customs Data Cross BBP Basic Benefits Package IGO Inspector General Office CEE Central & Eastern Europe IMF International Monetary Fund CIS Commonwealth of IMR. Infant Mortality Rate Independent States LBSBP Law on Budget System and CIT Corporate Income Tax Budget Powers DSA Debt Sustainability Analysis LGSG Law on Local Governance and ECA Europe and Central Asia Self-Governance EF Equalization Fund LGU Local Government Units EPI Expanded Program of LTI Large Taxpayer Inspectorate Immunization MDF Municipal Development Fund EU European Union MDGs Millennium Development FAD Fiscal Affairs Department Goals FIAS Foreign Investor Advisory MMR Maternal Mortality Rate Service MoAg. Ministry of Agriculture FSU Former Soviet Union MoD Ministry of Defense FY Fiscal Year MoE Ministry of Economy GDP Gross Domestic Product MoF Ministry of Finance GE Georgia MoR Ministry of Refugees GEL Georgian Lari Mol Ministry of Interior GFS Government Financial MoLHSA Ministry of Labor, Health and Statistics Social Affairs GAU Georgia Agriculture University MoJ Ministry of Justice GTU Georgia Technical University MoT Ministry of Transport GUSSIF Georgian United State Social MoTR Ministry of Tax Revenue Insurance Fund MTEF Medium-Term Expenditure GYLA Georgian Young Lawyers' Framework Association NBG National Bank of Georgia HEES Household Education NGOs Non-governmental Expenditure Survey Organizations HF Health Fund NPV Net Present Value HIPC Heavily Indebted Poor O&M Operation and Maintenance Countries OECD Organization for Economic IBRD International Bank for Cooperation and Development Reconstruction & O/w Of which Development PER Public Expenditure Review IDA International Development PETS Public Expenditure Tracking Association Survey IDPs Internally Displaced Persons PHD Public Health Department ift SMIC State Medical Insurance PHRD Policy and Human Resources Company Development SOE State-Owned Enterprise PIT Personal Income Tax TB Treasury Bill PlUs ProjectImplementation Units UN United Nations PPP Purchasing Power Parity US United States PRSP Poverty Reduction Strategy VAT Value Added Tax Paper WB World Bank SAC Structural Adjustment Credit WDR World Development Report SDS State Department of Statistics WFP World Food Program SGU Self-Government Units WTO World Trade Organization SIF Social Insurance Fund SMEs Small & Medium-Sized Enterprises Vice President: Johannes F. Linn Country Director: Donna Dowsett-Coirolo Sector Director: Cheryl W. Gray Sector Manager: Samuel Otoo Task Team Leader: Rocio Castro iv V CONTENTS CURRENCY AND EQUIVALENT UNITS EXECUTIVE SUM M ARY ............................................................................................................................... i CHAPTER 1. Overview of Fiscal Developments ........................................................................................... I Background ............................. ..................... 1 Past Budgetary Trends............... . ................ ....... .. .. ...1 Fiscal Adjustment ................. . ............ 2 Overall Level ofPublic Spending. .. .. . ...4 Composition ofPublic Spending.. . ..... . . ... ..5 External Debt Burden........... .. .........6 Medium-Term Fiscal Outlook ........................................7 External Debt Sustainability.. . ... . ................ . 7 Conclusion. .... ....... 9 CHAPTER 2. Revenue Policy and Administration ................................................................................ 10 Trends and Structure.................................10 International Comparisons .... .... .. . 10 Revenue Structure and Outcomes ................ .. 11 Performance Evaluation ..... .............................. ... .........12 The Tax Environment......... ................... . . ........ .. 12 Tax Revenue Potential . . .. ......... ................. .. .13 Tax Policy ........................................... .......16 Tax Administration ............................ .............. 21 Tax Department. ........ .... ..... 23 Customs Administration ....... .. 27 Key Reform Areas ......... .............. .......... ..........31 Short-term Reform Priorities ........ .. 31 A Longer-term Agenda. .. . . . 32 CHAPTER 3. Budget M anagement Issues ................................................................................................... 34 Introduction ........................................... ........ 34 Credibility of the Budget ..........................................34 Budget Preparation ................................... .. .. .. .. ...37 Description of the Process............... . . . . .... 38 Formulation of Budget Ceilings ... ........ . ................................ .39 Protected Items. .... ....... ...... ............ ...40 Investment Budget and Targeted Programs ..... . . ............ 43 Off-budget Accounts.. ..... ........ ... . ........ . . .. 44 Actions Taken to Date.. ....... ..............44 Future Actions Needed in Budget Preparation ..... . .45 Strengthening the Process for Setting Ceilings.. .. . . .......45 Reforming Use ofProtected Items . .......... .. .............46 Moving Toward Programmatic Budgeting. .... .. ..46 Forward Year Program Planning..... .............. .47 Budget Execution............................. ..................48 Description of the Process........ . ....... .................... .48 Sources of Volatility.. . . .......... ......49 Commitment Control . .. . ..................51 Actions Taken to Date........................... ........... 52 Further Actions Needed on Budget Execution 53 Predictability ofReleases 53 Vi Transferring Accountability to Ministries .. ... . ..... ... .................. ... . ... 53 Accounting for Commitments .......... .. . . ... . .. .. .... . .....53 Approaches to Improving Expenditure Management .. ......... .........54 Improvements within a Short-term Horizon . . . . 54 Building on a More Robust Resource Envelope. A Longer-term Horizon... .. ......... 55 CHAPTER 4. Inter-Governmental Fiscal Relations...................................................................................56 Introduction...................................... ...........56 Administrative Structure, Budgeting and Accountability..... ................56 The Legal and Political Framework............... ............ ......56 Budgeting .. .. . .. ....58 Local Governance and Self-Governance ............. .........59 Main Issues and Option for Reform . .. . . ... ........ 60 Expenditure Responsibilities ........................ ...............61 Assignment ofResponsibilities..... .. ...................61 Local Expenditure Patterns . . . ...64 Functional Composition ofLocal Expenditures............ .........65 Economic Composition ofLocal Expenditures. ............... ........67 Local Spending Disparities Across Regions and Rayons.. 67 Main Issues and Options . . ..... ..69 Assignment of Revenues.............................. ........71 The Statutory Assignment ofRevenues . ............................ 71 State Taxes........... . ....... .72 Local "Own" Taxes and Fees...................... . ... .. ......73 Patterns ofLocal Government Revenues . . .. .................... . . .. ..... 74 Aggregate Trends and Shares .... ................ 74 Composition ofLocal Revenues. 75 Disparities in Local Revenue Collections......... .77 Main Issues and Options..........................78 Fiscal Imbalances and Transfers..................................... 79 Fiscal Imbalances. ..................... ...... .... ..... . . .... 79 Transfer Mechanisms . .............. .. ...82 Main Issues and Optionsfor Reform.............. . ...... . . ...... 83 CHAPTER 5. Expenditures in the Health Sector...................................... 85 Health Policy in Georgia.......................................... 85 Overview of the Health Sector......................................5.....7 Health Outcomes................................ . ...... 85 Coverage and Utilization of.Health Services ........ ........ ..... 86 Health Expenditures ............................................. 87 Public Expenditures in Health s...... . .. ...................... ....... 87 Sources.. T n s............... ...... .... ... .. . . . .. .... .... .. . . 87 Health Care Purchasing Agents . ...... . .. .. . .......89 Resource Allocation e......... ...o.. .... . .. .. ...... .. ...90 Private Expenditures on Health Care ........ ...... ... . ....... ... .90 Efficiency of Public Expenditures ............. . ....................79 Equity in Health Expenditures ............................... ........... 92 Public Spending on Health ..... ....... .................... . . .. .92 Out-of-Pocket Spending on Health ..... ............. ....... ............. 93 Reform Issues and Options .................................................................. 94 Improving Efficiency in the Utilization ofPublic Funds...................94 Enhancing Equity in the Utilization of.Health Services.................95 Exploring Innovative Optionsfor Increasing Public Funds ............ ..95 CHAPTER 6. Expenditures in the Education Sector.................... ...............97 Introduction ..................................................................97 Vii Overview of the Sector ......................... ..................97 Enrollment Rates.... .............. ......................97 Education Expenditures.................................... .....99 Public Expenditures on Education .......... ......................99 Actual versus Planned Expenditures on Education.............. ....101 Resource Allocation by Economic Classification................. 102 Resource Allocation by Level........... ......... ...............104 Resource Allocation by Region................................105 Private Expenditures in Education....... .............. . . ........106 Efficiency and Effectiveness of Public Expenditures.......................109 Equity Effects of Public Spending on Education ..........................111 Conclusion and Policy Recommendations...................... .....112 Raising Allocative Efficiency in the Education Sector... .......... ... 112 Improve Equity in Education. ........ ....................... .114 Improving the Quality and Relevance of Education .......................... .. 114 CHAPTER 7. Social Protection ................................................................................. 116 Public Pensions................................................116 Current Situation. Arrears Problem ..................... ..... ........ . . . .....116 The Pension System .......... ...................... ........... ............118 Administration..... ....... .................. ............... 119 General Tax Revenue Performance.......... .......... ......... .. 120 Pensions and Poverty...... ............................... . . 123 Recommendations........................................... 124 Pay Pensions Regularly and Raise its Level in Real Terms.................124 Improve the Administration ofthe Pension Fund......... ........ ...... 125 Social Protection of Internally Displaced Persons (IDPs) ..................................126 IDPs in Georgia - Numbers, Placement and Socio-Economic Situation . ... 126 Government Assistance to IDPs and its Impact on their Wefare............. 127 Issues....................................... .... .... 129 Recommendations............................................130 Potential Savings ............. ......................................131 Poverty Benefits ..................................................................................... ........ 132 Current Situation........ ............. ......................132 Issues ......Sys . . ................... ............. ...133 Recommendations. ........................ .. .. .... ...... 134 Map IBRD 28703 viii ix Tables Table I Tax Structure for CIS Countries, 2000......................... ..vi Table 1.1 Selected Macro-economic Indicators......................................2 Table 1 2 Consolidated General Government Budget...................................3 Table 1.3 Revenues and Expenditures, International Comparison................. ..........4 Table 1.4 Public Expenditures by Level of Government................................5 Table 1.5 External Debt and Debt Service after Debt Rescheduling, 2000-2007 ......... . .... .8 Table 1.6 Medium-Term Fiscal Framework, 2000-2007.......................... ....9 Table 2.1 General Government Revenue................... ................11 Table 2.2 Tax Structure for CIS Countries, 2000....................................12 Table 2.3 Actual Revenues and Potential Revenue Estimates from Petroleum Products, 2001...... 14 Table 2.4 Projected and Actual Revenues from Tobacco Imports and Production, 2001................... Table 2 5 Actual Collection vs. Potential Revenue Estimates, 2001........................15 Table 2.6 Projected vs. Actual Excise Tax on Tobacco Products in 2001........... ................ 19 Table 2.7 Total Revenue from Diesel and Gasoline..................................20 Table 2 8 Advances and Arrears by LTI, by Sector......................... ........23 Table 2.9 Total Stock of Arrears by Sector, 2001....................................24 Table 2.10 Total Tax Collected by LTI, 2001......evenue.Estimates.f2.................... ..........26 Table 2.11 Officially Registered Major Transit Goods as Compared to Import .................8.....1 Table 2.12 Mirror Statistics; Trade with 10 Major Partner Countries, 2001 ........... .........9....1 Table 3.1 Comparison of 2000/2001 Budget versus Actual Expenditures ............ ........5. .. 1 Table 3.2 Execution Rate of the 2001 Central Budget by Economic Classification..............6....2 Table 3.3 Comparison of the Impact of Budget Cuts in 2001 on Selected Non-Salary Expenses. 37 Table 3.4 IMF Estimate of Expenditure Arrears for End-20002001 ......................37 Table 3.5 Budget Process for 2002 ....Coll......................... ...............39 Table 3.6 Protected Items in 2001 Budgets 0 .................. .......... ............42 Table 3.7 Relative Changes in Central Government Expenditures (1999-2001)...............42 Table 4.1 Territorial Self-Government Units.................................... ...59 Table 4 2 Distribution of Public Expenditures between Levels of Government, 1995-2001.............. Table 4.3 Composition of Sub-national Expenditures Functional Classification, 1997-2001.............. 66 Table 4.4 Disparities in Expenditure per Capita, by Rayon...............................68 Table 4.5 Government Revenues, 1997-2001........................................ .6 Table 4.6 Composition of Revenues by Territorial Units, 2001...................... .... 76 Table 4 7 Gross Domestic Product, by Region Budget.by.Economic.Classi ....................................... 81 Table 4.8 Poverty Incidence by Region Ara2..............0..... ............ .....................81 Table 5.1 Health Outcomes in Georgia and Other CIS Countries (2000) ............ ........ 6 .... 3 Table 3 2 Public Expenditures on Health (1997-2001) ..... ........... ......... ............8...4 Table 5.3 SMIC Revenues: Planned and Actual (1998-99) .............................89 Table 5.4 Incidence of Illnesses (not requiring Hospitalization) ............................. 93 Table 5.5 Total Annual Expenditures on Out-Patient Care by the Non-Chronically I0........ .........93 Table 5.6 Total Annual Out-of-Pocket Expenditures on Inpatient Care ............ ....... .....94 Table 6.1 Percent of Age Group Children Enrollment in Education, 2000 .............................. .........97 Table 6.2 Basic Education Indicators of Georgia............................................................. .......98 Table 6.3 Georgia: Overall Public Spending on Education, 1996-200 1 ............. ........100 Table 6 4 Total Budget Transfers to Local Governments......... ............................1 Table 6.5 Planned versus Actual Expenditure on Education, 1997-2001......................................102 Table 6.6 Percent of Public Expenditures on Education by Level and Budget Category, 2000 ..... ....103 Table 6.7 Distribution of Overall Education Expenditures by Level of Education, 1999..................104 Table 6.8 Local and Central Education Expenditures by Level of Education, 1998-2000.... .....105 Table 6 9 Average Local Expenditures in General Education by Region in Ill ....................106 Table 6. 10 Household Expenditures by Level of Education and Types of Spending.. ........... 108 Table 6.11 Student-Teacher Ratios in Georgia and Selected Countries.......................... 109 Table 7.1 Pension Fund, Revenues and Expenditures, 1996-2000........................7.....10 Table 7.2 Pensions Relative to Average Wage and Poverty Line, 2000 ..................... .........119 Table 7.3 Potential Contributions and Collection Rate, 2000 ............................0...10 X Table 7.4 Wage Employees' Average Number of Years of Contribution...........................121 Table 7.5 Absolute Amounts of Social Insurance Contributions .................... .....122 Table 7.6 Financing Needs of the GUSSIF and Sources of their Financing: An Estimate ...................123 Table 7.7 The Poverty Benefit Recipients in Georgia: January-November 2001 ......... .........133 Figures Figure 1 General Government Revenues..........................................v Figure 2 Poverty impact of Pensions ................................... ...... . xiii Figure 1.1 Public Expenditure by Functional Classification, 2001 ................. .................5 Figure 1.2 Public Expenditure by Economic Classification, 2000.................... ..........................6 Figure 2.1 General Government Revenue as % GDP...................................10 Figure 2.2 General Government Revenues and Per Capita GDP in CIS.............. ...........11 Figure 3.1 Budget Execution Rates for 2001 ............................................. 41 Figure 3.2 Actual Monthly Revenue Collection for 2000-2001 ................................50 Figure 3.3 Cash Releases to Ministries in 2001 ......................................51 Figure 4.1 Comparative Shares of Local Governments in Public Expenditures ............ ..........56 Figure 4.2 Disparity in Per Capita Local Govt. Expenditure Coefficient of Variations, 1997-2001.........67 Figure 4.3 Per Capita Local Government Expenditure Coefficient of Variation .......................67 Figure 4.4 Local Government Expenditure, by region (1997-2001) ............................68 Figure 4.5 Tax and Fees Revenues as Share of GDP .............................74 Figure 4.6 Share in Tax Revenue, 1997-2001 .......................................74 Figure 4.7 Disparity in Per Capita Revenue Collection .................................77 Figure 4.8 Per Capita Tax Collection, 2001 ........................................77 Figure 5.1 Allocation of Municipal Funds.... .........................................92 Figure 7.1 Simulating the Poverty Impact of the Timely Payment of Pensions ......... ...........124 Figure 7.2 IDPs Monthly Expenditures in Lari with and without Government Cash Assistance............128 Box and Diagram Box 4.1 Expenditure Assignments ..................... .................63 Diagram 1.1 Georgia: Administrative Structure ....................... ................58 Annexes Annex 1.1 Gross Domestic Product, Growth and Structure Annex 2.1 General Government Revenue (million lari) Annex 2.2 General Government Revenue (percent of GDP) Annex 2.3 Turnover Limits and Actual Collection of VAT, LTI 2001 Annex 2.4 Number of Registered Tax Payers by Types of Taxes Annex 2.5 Number of Active Tax Payers by Types of Taxes Annex 2.6 Registration and Identification of Tax Payers Annex 2.7 Stock of Tax Arrears of the State Tax Department and State Funds, 1999-2000 Annex 2.8 Stock of Tax Arrears, 1999-2001 Annex 3.1 Georgia - Central Budget - List of Credits Disbursements in 2001 Annex 3.2 Monthly Revenue Collections Annex 4.1 Expenditures of Georgia Territorial Units, 2000 Annex 4.2 Disparity of per Capita Local Government Expenditures, 1997-2001 Annex 4.3 Per Capita Local Expenditures and GDP, by Region Annex 4.4 Government Revenues 1997-2001 Annex 4.5 Disparities in Rayon Revenues, per source, 2001 Annex 5.1 Health Service Utilization in Europe and Central Asia Annex 5.2 Health Expenditures in the ECA Region Annex 5.3 Distribution of BBP across Public Purchasers Annex 5A Georgia Pilot Public Expenditure Tracking Survey Table 1: Quarterly Appropriations and Actual Disbursements Xi Table 2: SMIC Programs in 1999-2000, Budget Execution and Arrears Table 3- State Preventive Health Program Budget Revision and Execution, 2001 Figure 1: Kutaisi Children's Hospital - Monthly Cash Flow and Arrears Annex 6 1 Sources of Revenue and Expenditures in Public Education Institutions by level of Education Annex 6.2 Government Expenditures on Educational Institutions by Budget Classification Annex 7.1 Budget of the Georgian United State Social Insurance Fund for 2001 and 2001 Annex 8.1 Georgia Public Expenditure Review Workshop Annex 8.2 Georgia PER Workshop, List of Participants xii xiii Acknowledgements This Public Expenditure Review was prepared by Rocio Castro (Senior Country Economist for Georgia and Task Team Leader, ECSPE) based on the findings of two World Bank missions carried out in December 2001 and February/March 2002. Shantayanan Devarajan (peer reviewer, Chief Economist, HDNVP) and Allister Moon (team member, Lead Economist, ECSPE) provided useful advise at various stages of report preparation. Allister Moon and Bernard Myers (Consultant) worked on expenditure management issues; Joao Oliveira (Senior Economist, ECSPE) on inter-governmental relations; Luis Alvaro Sanchez (Consultant) and Michael Engelschalk (Senior Public Sector Specialist, PRMPS) on revenue issues. Contributions for the social sector chapters were provided by Halil Dundar (Education Economist, ECSHD) on education; Mukesh Chawla (Senior Human Development Economist, ECSHD) and Sarbani Chakraborty (Health Specialist, ECSHD) on health; and Aleksandra Posarac (Senior Economist, ECSHD) on social protection. Afsaneh Sedghi (Research Analyst, ECSPE) contributed with data analysis and regional comparisons. Dolly Teju (Program Assistant, ECSPE) assisted in the production of the report, including editorial support. The team received excellent assistance from the Resident Mission. Sergo Vashakmadze (Economist) worked closely with team members during the missions, Elene Imnadze (Public Sector Management Specialist) commented on the education chapter, and Tamar Gotsadze (Project Officer) contributed to the health expenditure tracking exercise. Special thanks to Guranda Elashvili (Team Assistant) who assisted in coordinating appointments and organizing the workshop. The team is also grateful to David Gentry (US Treasury Adviser at the Ministry of Finance) who provided useful comments to the expenditure management chapter. The findings and recommendations of this report were discussed with the Georgian Government in two stages. First, through bilateral discussions held in July 2002 and subsequently in a follow-up workshop which took place on September 21-22, 2002. The workshop involved relevant government agencies and foreign advisers working in the fiscal area. xiv Georgia: Public Expenditure Review November 25, 2002 Executive Summary 1. Following the devastating economic collapse of the post-independence years, Georgia has been able to restore and sustain economic growth and price stability since 1996. However, at around US$600, Georgia's per-capita income is still less than 40 percent of what it was at independence. While the incidence of poverty is relatively low (23 percent of the population in 2000), it is estimated that 60 percent of the population are at risk of falling into poverty. In addition, health sector outcomes have deteriorated sharply over the past decade, while there are signs that the quality of education is rapidly declining. Moreover, the poor have suffered the most from inadequate access to social services and from crippling basic infrastructure. 2. Substantial fiscal adjustment has taken place in the context of the stabilization program launched in the mid 1990s. Adjustment initially relied on massive expenditure cuts and later on increased domestic revenue mobilization. Still, Georgia's overall public expenditure level (18.5 percent of GDP in 2001) is among the lowest in the CIS region. If this level were to be reduced any further, essential programs could be put at risk. Public expenditures on maintenance and rehabilitation of basic infrastructure are negligible and those on the social sectors are below the regional average. On the other hand, the resource envelope is not likely to increase much in the medium-term. While there is some scope for raising domestic revenue, such scope is limited by Georgia's initial conditions and currently low income level. In addition, a considerable share of available resources will continue to be devoted to service the external debt, even if Georgia were to obtain debt relief under concessional terms. 3. Like other low-income countries, Georgia's ability to reduce poverty and meet the Millennium Development Goals (MDGs) by 2015 will depend crucially on its ability to sustain per-capita GDP growth and to use effectively and efficiently its scarce public resources. Indeed, increased public expenditures will not result in improved social outcomes unless current productivity levels are improved. This Public Expenditure Review (PER) shows that weaknesses in Georgia's public expenditure management systems have led to serious inefficiencies and inequities in the use of public resources and have prevented the necessary restructuring in public expenditures. The first three chapters explore sources of these systemic problems in the context of macroeconomic management, revenue policy and administration, and budget management systems. Subsequent chapters explore the scope for strengthening budget management systems in the areas most critical for poverty reduction and attainment of MDGs: local governments and social sector spending. These chapters both corroborate and supplement with further detail the core constraints on public expenditure restructuring seen in the earlier chapters, but also map out the complementary components of a reform strategy which recognizes the need for parallel action by central agencies, line ministries and local governments. Given that identified weaknesses cannot all be addressed at once, the PER offers an approach to sequence expenditure management reforms in a phased manner, building on progress in improving revenue predictability. Constraints to Effective Expenditure Management Unpredictability of Funds 4. Unpredictability of the resource envelope has been a major constraint to proper budgeting in Georgia. The problem starts at the stage of budget formulation with unrealistic, often politically-driven, revenue forecasts for the budget year, which inevitably require expenditure cuts within year. For example, in 2000 and 2001, budget execution was 63 and 82 percent, respectively, of the original budget. To the extent that some expenditure categories, such as wages and pensions, cannot be cut within an individual annual cycle, arrears on these categories (so-called protected items) have persisted over time. Although the Ministry of Finance has been able to set more conservative, revenue forecast over the last 1 Georgia: Public Expenditure Review November 25, 2002 couple of years, and therefore reduce the occurrence of arrears, revenues continue to fall short of projections. Moreover, there is also a great deal of uncertainty in cash releases to spending agencies within year. 5. Uncertainty about resource availability can be explained, in part, by volatile revenue flows, both of domestic revenue and foreign donor financing. Since domestic revenues are the primary source of government finances in Georgia, related shortfalls/fluctuations can have a significant impact. The uncertainty over domestic revenue does not appear to be associated to an intrinsically volatile tax base or poor technical capacity to forecast revenue, but rather to political and governance factors. Indeed, political considerations seem to drive the establishment of ambitious annual revenue targets as well as the constantly changing tax policy framework. In addition, inadequate, often corrupt, administrative practices in the Georgian context have caused instability of revenue flows within year. 6. However, revenue volatility is not the only factor explaining unpredictability in program financing. Monthly expenditure releases have in general been more volatile than monthly revenue collections, implying that poor cash-flow management plays an important role in explaining unpredictability of funds to spending units within year. Analysis of budget deviations shows that volatility in spending allocations and ineffective implementation of hard budget constraints also contribute to poor predictability at the spending unit level. 7. On balance, the unpredictability of resources has distorted budget outcomes and translated into a very short-term budget horizon, even shorter than a year when it comes to budget execution. Moreover, it has undermined the credibility of the budget and precluded the development of strategic budgeting. Lack of Strategic Planning 8. In addition to the problem of poor resource predictability, the budget formulation process itself does not provide an adequate framework for strategic allocation of public funds. There is little or no discussion of overall government priorities based on explicit links between sector objectives, programs, and policies. The absence of such a strategic framework for resource allocation has created a vacuum which, by default, has been filled by the criteria of protected items. Protected items consist of a list of specific expenditure categories which because of its mandatory nature cannot, in principle, be sequestered within a given budget cycle and, consequently, have first right on available funds. Such protected items include mostly economic expenditure categories (such as wages, debt service, and pensions), but also a few programs (such as poverty benefits and state health care programs). Together they account for about 85-90 percent of the central budget. 9. While the criteria of protected items has served as a useful tool to impose some order on budget execution in a cash-starved environment -and particularly to reduce the scope for arrears' accumulation, it has also indirectly prevented the development of more strategic budgeting. As currently defined 'protected items' cannot easily be linked to specific outputs/outcomes as would sector programs. In addition, to the extent that they account for the bulk of the budget, very little is actually left for discretionary allocation, making the budget rather inflexible, especially in the short term. Given the short term horizon in which spending units are forced to operate, they have really no incentives to generate savings, e.g., through rationalization of staffing, and re-allocate expenditures within their own budgets. The challenge is to develop a phase of strategy formulation within budget preparation which links the whole of public expenditure to sector objectives and programs. This will remove the need for separate definition of protected items and allow this instrument to be gradually phased out, as the coverage and realism of the budget strategy is improved over time. ii Georgia: Public Expenditure Review November 25,2002 Weaknesses in budget execution 10. The uncertainty over revenue flows has been exacerbated by weak cash-flow management. Cash releases do not follow quarterly allocations which have tended to be overstated. As a result, the MoF ends up holding a tight rein on cash releases to control expenditures, even on a daily basis. The implications of the protected items criteria on budget execution are clear. When a large proportion of expenditure are protected, the rest is likely to be more volatile. For example, central government transfers to the local level are not protected and have been subject to very low budget execution. Also, despite recent progress in reducing the size of reported arrears on 'protected items', the system is not able to capture commitments on non-protected expenditure categories. This implies that additional domestic arrears could be building up on account of unpaid contracts or utility services. Approaches to Improving Expenditure Management 11. Improving the predictability of resources available to line ministries is central to any improvement in expenditure management. Predictability of budget resources is at present constrained by considerable uncertainty on the revenue side of the budget. An appropriate strategy for improving expenditure management depends critically on the degree of predictability in aggregate resources, including both domestic and external financing. At present, forecasting of revenue even only one year ahead is subject to a high degree of uncertainty. This effectively limits the horizon for any firm planning of expenditure. Proposals for reform need to take account of these highly constrained initial conditions. Consequently, the suggestions below distinguish between measures which might be appropriate in the immediate context and others which will become feasible only as the predictability of overall resources is improved. Improvements within a Short-Term Horizon 12. Prima facie, there appears to be some scope for increasing predictability of resources for budget organizations, even given the uncertainties on budget revenue. Monthly expenditure releases have in general been more volatile than monthly revenue collection, suggesting scope for increased predictability of releases. The following measures seem relevant, including ongoing reforms and some additional suggestions: * Improving cash-flow management. Setting more conservative quarterly allocations during the first half of the year using within year borrowing to smooth expenditure, as already planned. Minimizing costs of such borrowing will require careful management. * Strengthening treasury functions (ongoing). Bringing revenue accounts into the treasury, introducing comprehensive commitment control. (Note that elsewhere in the region, such reforms have taken several years to complete). * Improving the match between protected items and priority programs. Earmarking of 'protected items' should be regarded at best as a temporary measure, to be phased out as the predictability of budget financing is improved. In the mean time improvements in the system could also be achieved by enhancing the match between protected items and priority programs. * Building incentives for budget rationalization. In the absence of explicit protection for complementary inputs, line ministries have little incentive to cut programs and reduce staffing. Ministries that make a serious effort at restructuring programs consistent with budget ceilings need assurance that such programs will be financed adequately, either through inclusion of the program fully as protected items or through specific commitment during the budget approval process. iii Georgia: Public EiWenditure Review November 25,2002 Developing a phase of budget strategy formulation within the budget cycle. A sustainable restructuring of the budget typically requires a longer term horizon for planning. The budget cycle should include a budget strategy phase, in which ministerial ceilings are decided by Government in the context of the aggregate resource envelope and competing sectoral objectives and options. Initially, this might be based on a strategy framework paper prepared by the MoF, primarily focused on the aggregate resource envelope and the immediate annual horizon. Building on a More Robust Resource Envelope: A Longer-Term Horizon 13. The measures discussed above should contribute to increasing the planning horizon for budget managers, but major progress will depend on improvements in revenue management and forecasting. As such progress is made, more ambitious reforms in expenditure management can be planned. * Extending the budget horizon. Over subsequent budget cycles, the horizon of the budget framework could be extended to the medium term, e.g., three years, with increasing depth of sectoral analysis and greater engagement of other ministries in the formulation process. * Developing bottom up sector programs. Currently, sectors with capacity for realistic expenditure planning are frustrated by the lack of institutions linking sector policy/plans to budgets, and other sectors have no incentive to develop such capacity. The medium term framework exercise would replace many of the functions intended in the present indicative plan and could use more effectively existing central and sectoral planning capacity. * Link with the Poverty Reduction Strategy Paper (PRSP). As the primary link between policy, sector planning and budgets, the framework exercise will need to be closely coordinated with development of the PRSP. Given Georgia's current budget system, this will take several years to implement. * External financing. Uncertainty in the timing of external financing, i.e., from policy- based adjustment lending, is itself a source of poor predictability in aggregate financing. A more explicit budget strategy from Government combined with a more timely implementation of agreed reforms, would ensure greater specificity on external financing. The Revenue Question 14. The Georgian authorities perceive uncertainty in revenue as an important constraint on expenditure planning. It is therefore important to understand how far revenue performance has contributed to an environment of weak expenditure planning, and also to see how far this environment can be expected to improve over time, both in terms of a realistic outlook on the future level of revenue and issues affecting volatility. The Level of Revenue Collection 15. Georgia has a reputation of being a poor revenue performer because of both its chronic inability to meet revenue targets and low revenue to GDP ratio relative to other countries within the CIS region. Closer examination of revenue trends, however, shows that Georgia's revenue collection, 15.9 percent of GDP in 2001, has improved considerably since 1995 and, today, the country's performance is close to iv Georgia: Public Expenditure Review November 25, 2002 most low-income countries within the CIS region.' In fact, there has been a convergence of revenue performance: while Georgia's performance has been improving, that of other low-income CIS countries has been declining or increasing only modestly. Most of the observed gains in Georgia reflect increases in VAT collections, suggesting that the tax is buoyant and highly elastic to real GDP growth. While the level of per-capita income (and growth) may be one important determinant of revenue performance, tax yields may vary due to a number of factors such as initial conditions, access to natural resource tax revenue (like oil in Azerbaijan), tax system and tax rates, as well as the administrative capacity. Although the scope of this PER did not permit a detailed analysis of these factors, some conclusions can be drawn from available information on the structure of taxation across CIS countries. General Government Revenues (in percent of GDP) % GDP 45- 40 . 30 - 25- 10 5 0 1995 1996 1997 1998 1999 2000 16. By and large, CIS countries have similar tax systems. The distribution of revenue by type of taxes shows that Georgia is pretty much aligned with most CIS countries regarding the collection of the Value Added Tax (VAT)2 and income taxes, partly because tax rates are fairly uniform across countries. There are, however, large differences in collections of excise and social security taxes, with Moldova and Uzbekistan scoring the highest ratios and Tajikistan the lowest. 17. In the case of Georgia, revenue from excise taxes, currently amounting to 1.5 percent of GDP, has declined considerably since 1999, reflecting reductions in tax rates and, perhaps most importantly, weakening governance. Although porous borders in Georgia may explain low collection of taxes on cigarettes, for example, there is little justification for low collection of petroleum taxes, since these products are transported in bulk through the railways and therefore are not suitable for easy smuggling unless there is collusion with tax inspectors. 18. Georgia's reputation as a poor revenue performer, on account of its revenue to GDP ratio alone, may not be completely warranted in light of its recent track record. Still, it would appear that there is scope for raising revenue collection by 2-4 percentage points of GDP, particularly from excise and VAT taxes.. In the short term, revenue from (excise) taxes on petroleum products and cigarettes can and should be increased not only to raise the resource envelope, but most importantly to demonstrate Georgia's commitment to improving governance and the overall tax environment. It is estimated that additional tax revenue from these excisable products could amount to 2 percent of GDP (or about US$65 million). This ' Among the low-income CIS countries, Georgia's revenue to GDP ratio is close to that of Azerbaijan, Armenia, Tajikistan, Kyrgyz Republic. Moldova and Uzbekistan have higher revenue ratios, in both cases reflecting large contributions from excise and social security tax collections. 2 The standard VAT rate across CIS countries is 20 percent. v Georgia: Public Expenditure Review November-25, 2002 represents three times the annual disbursements from IDA-financed adjustment operations. In addition, Georgia can bring VAT collections (5.5 percent of GDP in 2001) in line with the average for transition economies (6.4 percent of GDP) through the elimination of exemptions and better administration, but this may take some time to achieve (see below). Table 1: Tax Structure for CIS Countries, 2000 Total Total VAT Excise Trade Profit Income SS Tax NonTax Armenia 6.5 2.5 0.8 2.0 1.4 2.3 17.7 1.2 Azerbaijan 4.1 0.5 2.1 2.7 2.0 2.3 14.6 6.3 Georgia 4.9 1.5 0.9 1.3 1.8 2.4 14.3 0.8 Kirguiz 4.8 2.4 0.4 1.1 1.1 3.5 15.8 2.6 Moldova 8.2 4.1 1.4 1.7 1.1 6.1 26.6 3.7 Tayikistdn 2.5 0.5 1.5 - 1.8 1.6 12.9 0.6 UzbekistAn 7.6 7.8 0.7 3.2 4.2 - 26.7 1.8 Average 6.2 2.5 1.3 2.1 2.6 3.9 21.6 2.5 CIS Unstable Tax Policy Framework 19. Georgia's Tax Code, adopted in June 1997, initially provided a comprehensive framework for taxation, which was considered one of the best in the CIS region. But, the proliferation of innumerable amendments4, mostly providing exemptions and tax reductions on VAT and excises, has seriously undermined the consistency of the code over time. The history of tax policy changes in Georgia since the adoption of the tax code demonstrates an unhealthy focus on short-term measures and a complete disregard for the integrity of the system. New taxation schemes, mostly initiated by Parliament and driven by influential interest groups, are too often introduced at any stage of the budget cycle without proper assessment of their fiscal impact or indication of their duration. On the other hand, efforts to meet targets fiscal agreed under the IMF programs have tended to rely on tax policy measures to compensate for revenue shortfalls. In addition to complicating tax compliance and administration, such a short term approach has made the tax policy framework unstable both for the tax administration and for tax payers. Administrative-related Distortions 20. An exclusive focus on revenue targets as a measure of administrative performance has had unintended adverse consequences. As it is commonly known, tax payments in Georgia are often 'negotiated' between tax payers and tax inspectors. In this context, tax inspectors have resorted to the practice of requesting advanced payments against future tax liabilities, in order to meet their monthly and quarterly revenue targets.! The stock of advance payments as of end 2001 amounted GELl02.million (or 1.6 percent of GDP). While the bulk of these advance payments relates to legitimate advances (e.g., profit taxes and outstanding VAT claims), it also includes such negotiated advanced payments. Tax 3 The on-going adjustment operation, the Third Adjustment Credit (SAC 3), amounts to approximately US$60 million. 4 Between July 1997 and April 2001, 113 amendments to the code were passed, of which 58 percent related to VAT and excises. 5 Targets are set by type of tax and level of government. To the extent that the reporting of these payments is manipulated, the economic significance of tax collection by type of taxes is put into question. vi Georgia: Public Expenditure Review November 25, 2002 payers go along with such a scheme in exchange for lower tax assessments and the ability to draw on these funds at will. On their part, tax officials are able to prevent sanctions, which may include their dismissal, and to supplement their incomes. This practice not only tends to undermine revenue collection but also makes revenue flows unpredictable even within a month's horizon. 21. In parallel, it is found that significant tax arrears have accumulated over the years, mostly owed by state-owned companies, and very little effort is made to recover them. At the end of 2001, the stock of tax arrears (including social security contributions) amounted to GEL394 million (or 6 percent of GDP). 22. The situation described above is one where a small, captive tax base is squeezed to the limit while the rest is outside the tax net. Under the circumstances, it is not surprising to find that the number of registered tax payers in Georgia is low by international standards. Toward Better Revenue Mobilization 23. The above discussion seems to suggest that Georgia has reached a point where it is probably more important to improve the tax environment and the way that taxes are collected, than to raise the envelope per-se. This by no means implies that efforts to raise revenue should not continue, but rather that such efforts must be grounded in a deeper understanding of long-term implications for future growth and revenue mobilization. In this vein, reliance on tax policy measures to meet short-term revenue targets are not a desirable option. The level of taxation in Georgia, for those few who pay taxes, is quite high and there is increasing internal political pressure to lower taxation, often in a rather anarchical fashion. 24. The priority today is to stabilize the system and ensure the sustainability of the revenue gains already achieved by addressing distortions in tax policy and tax administration. In the short term, specific measures could be taken, some of which are already ongoing, to stabilize revenues, improve the tax environment, and facilitate tax administration: * Limit tax policy changes, by making changes only with each annual budget and with appropriate estimates of revenue impact. * Strengthen VAT performance by (a) raising the VAT threshold from GEL 24,000 to 100,000 (or about US$50,000), while limiting voluntary registration of businesses below the threshold; (b) reducing exemptions,7 and (c) begin implementing VAT refunds for exporters. * Increase excise tax collections from petroleum products and cigarettes by enforcing payments from large petroleum importers and domestic cigarette producers. * Introduce a broader range of criteria to measure administrative performance, by including e.g., number of audits, recovery of tax arrears, number of registered tax payers. * Eliminate the practice of advance payments, by transferring all revenue accounts from the National Bank of Georgia (NBG) to the treasury (on-going), thereby eliminating the scope for tax off-sets when transit accounts were kept at the NBG. Also introduce 'a first come first serve' system to refund advance payments and thus limit tax payers' ability to draw on their advance payments at will. 25. In the medium to longer term, efforts must continue to develop the institutional capacity of the tax administration, especially of customs administration where much needs to be done. As progress is made in broadening the tax base and establishing a business-friendly tax environment, it would be 6 It is estimated that state-owned companies account for about 40 percent of outstanding tax arrears. 7 These measures were included in a tax package prepared by the MoF in 2001, which was not formally presented to Parliament, as well as in a current proposal. vii Georgia: Public Expenditure Review November 25,2002 possible to consider lowering the tax burden, e.g., through reductions on social security taxes and the VAT rate. Key elements of a medium-term agenda include: * Simplhfy the tax regime for small businesses, by implementing the proposed simplified and presumptive taxes once the administration for VAT has been strengthened; * Implement a phased program of customs reform, within a realistic time horizon, with immediate focus on improving the coordination between the tax department and customs, and establishing basic procedures such as a customs manual. * Increase tax payers' registration, with emphasis on quality tax payers that may have escaped the tax net. * Strengthen the basis for self-assessment, by adopting a functional organization which separates the collection, audit, and enforcement functions within the tax department. Inter-Governmental Fiscal Relations 26. An analysis of inter-governmental fiscal relations offers further insights as to how weaknesses in Georgia's budget management systems operate at the local level. This is a relevant dimension as local governments account for around 30 percent of total public spending and are particularly important for the delivery of basic social services (e.g., primary and secondary education, primary health care, water, electricity). Local governments absorb 80 percent and 30 percent of the consolidated budget for education and health, respectively. 27. According to the legislation, local governments enjoy a great deal of autonomy in preparing and executing their own budgets. In practice, however, local governments have remained highly subordinated to the center through a vertical administrative structure, high dependence on state shared taxes and transfers, and lack of authority to regulate local fees and taxes. Being dependent on the center, local governments suffer from the same kind of systemic problems described earlier, such as unpredictability of revenue and lack of strategic planning, perhaps compounded by the fact that they have little or no control over their own affairs. 28. Under the current system of revenue assignment and transfers, resources are unpredictable, incentives for increased tax collection effort are inadequate, and the distribution of resources across local government units is highly inequitable. Main issues include: (i) arbitrary allocation (through 'negotiations') of state shared-taxes and central government transfers, which account for the bulk of local self-government financing; (ii) low execution of central government transfers which are not protected expenditures, (iii) lack of authority to set fees, local taxes, and tariffs according to local conditions; and (iv) exclusive reliance on the derivation principle for the main shared taxes, the Personal Income Tax (PIT) and Corporate Income Tax (CIT)), which exacerbates inequities in the distribution of financial resources. 29. On the expenditure side, a number of functions have been delegated to the local level but without ensuring proper funding from the center. Under the criteria of 'protected items', the center guarantees the funding of specific expenditure categories such as teachers salaries, leaving local governments to finance goods and services and other complementary expenditures from their own resources. At the same time, the center regulates expenditures by enforcing national norms. Such rigidities create wrong incentives and perpetuate inefficiencies in the allocation of public resources. Also, decisions on financing and service delivery of exclusive responsibilities are "arranged" between the rayons and the self-government units. 30. The consequences of the system described above are duplication or under-provision of basic public services (e.g., water and electricity), under-funded "delegated functions" and accumulation of viii Georgia: Public Expenditure Review November 25, 2002 arrears, and inefficient use of resources. Furthermore, there are large disparities in per-capita spending by regions and by rayons, with poor local government units spending much less than well-off ones. Options for Reform 31. It is recognized that fiscal decentralization is a long term process and that Georgia is only at the early stages in its implementation. Nevertheless, it is important that the Government defines a clear strategy as to how it plans to establish the necessary building blocks for achieving fiscal decentralization over the long haul. A key challenge is to strike the right balance between the administrative capacity of local authorities and the degree of fiscal autonomy they are granted. In the meantime, the Government needs to take measures to improve the predictability of funding, improve equity in the distribution of resources, as well as providing incentives for increased tax effort at the local level. Establish revenue predictability by adopting a transparent 'formula-based' mechanism for making transfers to the local level (see below). * Reduce inequities across local government units by distributing central government transfers on the basis of a transparent 'equalization' formula. * Improve incentives to boost the tax effort by re-assigning the land and property taxes as local taxes; assigning a flat PIT tax rate to the local level and reassigning the progressive PIT and the CIT to the equalization fund; transferring ownership of public utilities to local governments; and allowing local governments to set fees, taxes and local tariffs within an established range. * Clarify expenditure responsibilities/competences by distributing functions/competences between rayons and self-government units according to clear rules specified in the organic law; and by reassigning to the central government the financing competency of programs such as primary education (rather than specific items such a teachers' salaries). * Increase technical and allocative efficiency of expenditures, as discussed elsewhere, by gradually phasing out the criteria of 'protected items' on delegated responsibilities. Instead, the central government should provide 'conditional grants' for the delivery of well-specified programs. Public Expenditures in the Social Sectors 32. Given the importance of the social sectors for reducing poverty and for attaining the MDGs, this PER provides a detailed analysis of the efficiency and equity of public expenditures on health, education and social protection. 33. In principle, scarce public resources should be used to finance the provision of goods and services with high externalities. In health, typical examples of such goods include the prevention (vaccinations) and treatment of communicable diseases (tuberculosis, HIV/AIDS). In education, externalities may result from basic education and basic research in higher education, but there is really no strong justification for public financing of vocational or higher education. The state can play an important role in financing and promoting access to social services by the poor who typically cannot afford them. Health 34. While comparing favorably with other CIS countries, health outcomes in Georgia have worsened significantly since independence. The infant mortality rate has increased by 16 percent to 24 deaths per thousand births and the maternal mortality rate from 32 to 51 per 100,000 live births. Also, the number of new cases of tuberculosis has tripled and malaria is reemerging. ix Georgia: Public Expenditure Review November 25, 2002 35. At about US$23 per-capita, Georgia's total expenditure on health is one of the lowest in the CIS region.8 Of this amount, out-of pocket expenditures are about US$15.? While recovering substantially since the post-independence collapse, from US$0.8 per-capita in 1994 to about US$ 8 per-capita in 2001 (or 1.3 percent of GDP), this level of spending is still well below the CIS average (3.5 percent of GDP). 36. The Government's health sector policy aims at allocating limited public funds for a Basic Benefits Package (BBP), promoting private sources of financing for medical services not included in the BBP, and downsizing and restructuring the health system. The goal is to provide the BBP to all Georgians and to increase the share of the population enrolled in the social health insurance system so as to cover all Georgians by 2010. It also aims at increasing public financing of health services to 4 percent of GDP by 2005. Public Spending on Health 37. The unpredictability of funds, as described earlier, has been a major problem for public health sector financing which comprises funding from the central government, payroll taxes,10 and municipalities. For example, the average execution of the central budget for the health sector was between 45-62 percent during 1997-2000. In 2001, budget execution improved substantially and reached approximately 87 percent. In addition to the problem of budget execution, the flow of funds into the health sector is extremely erratic. For example, in 1999, 30 percent of the annual budget for health care was released in May, while during the other months, funding levels varied between 1-14 percent.. 38. Efficiency. Georgia's scarce public expenditures on health care are not well targeted to meet the efficiency criteria of public financing. For example, the BBP includes a number of "private goods" such as treatment of heart disease, cancer, kidney transplants and hemodialysis. Since public resources available to finance the BBP are limited, the majority of items on the BBP, including goods with social externalities such as vaccinations, are under-funded and informal payments for items covered through the BBP are common. This has obvious negative equity implications. There are also serious technical inefficiencies resulting from an oversupply of health facilities and personnel which is a drain on the Government's funds. Despite on-going efforts to downsize the number of facilities, especially of hospitals, the bed-to-population ratio is still much too high (4: 1,000 population) even when compared with more developed countries. Similarly there are 3.9 doctors per 1,000 population while the average for the Europe and Central Asia (ECA) region is 2.5. 12 In addition, the cost-effectiveness of the interventions supported through the BBP is limited. The clinical protocols used for the BBP are outdated and the mechanisms adopted to contract with providers for the BBP do not create any incentives for efficiency or quality at the facility level. 39. Equity. Georgia's household data show that the poor have a greater incidence of illness - chronic or acute - compared to the rich. At the same time, fewer of the poor who suffer from an illness seek treatment as compared to the rich. Shortage of money is consistently mentioned as the most common reason for the lack of medical treatment. In order to improve access to healthcare by the poor and to provide financial safeguards for those among the poor who do seek health care, it is imperative that the available resources be targeted toward the poor. This could be achieved by orienting resources to the Above only to Azerbaijan, Kyrgyz Republic and Tajikistan. 9 There are no reliable estimates of out-of-pocket expenditures on health care. Estimates range from US$15 to US$28, or 66 to 87 percent of total health spending. 10 3 plus I percent of the payroll tax go to the state-run Social Medical Insurance Company (SMIC). 1 World Bank. 2001. Georgia Health Care Financing Note. The World Bank, Washington DC. 12 Other indicators such as hospital occupancy rate and staffing ratios are also evidence of technical inefficiencies. x Georgia: Public Expenditure Review November 25, 2002 poorer regions - like Imereti - and thus ensure higher per capita allocations of municipal funding for health than available at present. 40. Given the problems with allocative and technical inefficiency, there is no guarantee that further large infusion of public funds into the health system will produce benefits,in terms of improved health outcomes, especially among the poor. Better targeting of public expenditures on health, the development of a primary health care system, and rationalization of health facilities and personnel in line with health care needs in the country are required. Education 41. Enrollment rates are high for a country with Georgia's income level. Enrollments rates for compulsory education have remained high (90 percent), albeit declining, but those for non-compulsory education (e.g., pre-school and tertiary education) have declined sharply (to 54 percent). Such enrollment trends partly reflect the fact that the financing of education in Georgia has become increasingly privatized during the transition, especially for non-compulsory education.13 Available information indicates that overall expenditure levels on education were 4.9 of GDP in 2000, which is close to OECD levels (at 5.1 percent of GDP). Of this amount, over one half were out-of-pocket expenditures by households and were mostly spent on higher education. 42. While Georgia's enrollment rates are relatively high, there are signs that quality at all levels is deteriorating due to a shortage of teaching materials, poor maintenance, and inadequately trained and unmotivated teachers (because of very low salaries).14 43. Although increasing since the mid 1990s, public expenditures on education (around 2.2 percent of GDP) are also below the average for the CIS region (4.6 percent of GDP). However, the fact that the overall level of financing for education, including private resources, is close to OECD levels, suggests that the fundamental problem in the education sector in Georgia is not low financing but rather the inefficient use of resources and inability to address equity issues. Public Spending on Education 44. Like in the health sector, poor budget execution has been a serious problem for the financing of education. During 1997-99, the execution of the education budget averaged 82-90 percent for local budget and 68 percent for the central budget- although in 2000, execution was I 10 percent. By economic classification, items such as salaries are better executed, while spending on non-salary items are often significantly reduced. For example, in 2001 the Ministry of Education received only 58 percent of the budget for office expenditures. Also, even though teacher salaries are protected items, arrears have been reported in many districts with adverse consequences on teachers' motivation and the quality of education. As discussed elsewhere, unrealistic revenue estimation and budget planning are the primary reasons for the non-fulfillment of planned expenditures. The Ministry of Education, local governments and education institutions cannot make realistic plans. 45. Efficiency. The analysis indicates that public resources in the education sector are not efficiently used. Less than 50 percent of the total education budget is spent in general education and about 30 13 In the case of pre-schooling, the decline in enrolment rates reflects a substantial drop in school facilities. 14 Key performance indicators, such learning outcomes, are not currently available. xi Georgia: Public Expenditure Review November 25, 2002 percent in tertiary education.1s Not only are many of the existing tertiary public institutions obsolete and unsuitable to serve the demands of a market economy, but even if they were suitable, their financing with public funds cannot be really justified given that related externalities are likely to be low.16 In addition, lack of strategic restructuring at the input level has translated into significant technical inefficiencies resulting from low student/teacher ratios (10:1 in general education) and excessive number of education facilities, especially in vocational and tertiary education. Significant saving could be generated by increasing the student/teacher ratio17 and consolidating school facilities. Such savings could be used to improve teachers salaries and the supply of teaching materials and supplies, which have been kept rather depressed. 46. Equity. In a situation where the financing of education has been greatly privatized, the poor have suffered the most both in terms of access and quality of education. Enrollment rates are lower among the poor and poor children are more likely to have prolonged absences from school. Further, poor families face considerable problems of access to education due to recent changes in funding patterns especially in non-compulsory education (e.g., the rising costs of education, informal payments and payments for private tutoring). The same is true of children in many rural areas where school quality has deteriorated dramatically by shrinking public resources. A challenge facing the authorities is assuring the access of the poor to education. Policy options targeted to the poor may include provision of: (i) free of charge basic education materials (e.g., textbooks); (ii) education and training subsidies for poor students (e.g., allocating scholarships for higher education). A student loan policy may not be an option given that both labor and financial markets are not yet well functioning. Social Protection 47. Social protection programs in Georgia comprise (i) pensions; (ii) cash transfers to internally displaced persons (IDPs); and (iii) a limited poverty benefit program. These programs provide regular monthly transfers to about 1.2 million Georgians or over 20 percent of the population. Their total cost accounts for 4.2 percent of GDP or around 20 percent of total public spending. Pensions 48. Georgia's pension system covers about 860,000 pensioners and accounted for 2.7 percent of GDP in 2000. A major problem with the pension system is that benefits are not paid regularly and large arrears have accumulated over the years. As of August 2000, pension arrears stood at about US$50million 8 (or 1.6 percent of GDP ). While pensions have been paid more regularly since then, the situation remains very fragile. The financial difficulties of the pension system do not reflect an excessive number of pensioners (or high level of the benefit), but rather poor revenue collection of the social security tax and persistent shortfalls in budget transfers from the center. 49. Poverty Impact. Despite its financial problems, the current pension system has been effective in reducing poverty. The recent World Bank Poverty Update points out that other things being equal, if pensions had not been paid at all, poverty rates would have been 9 to 21 percent higher over the 1998- 15 In 1999, Georgia devoted 48.5 percent of the total education budget to general education, 20.6 percent to higher education, 11.1 percent to boarding schools and orphanages, 7.7 percent to kindergartens, 6.7 percent to out-of- school institutions, 3.8 percent to administration, and 1.5 percent to vocational and technical education. 16 There are about 84 vocational schools, 85 public specialized secondary schools, and 24 public higher education institutions. ' For example, it is estimated that an increase in the student/teacher ratio in general education to OEC levels (from 10 to 16) could generate cost-savings equivalent to 1.3 percent of GDP, or over half of the public education budget. a Excluding arrears on pensions for the military and other law enforcement agencies estimated at GEL17.7 million. xii Georgia: Public Expenditure Review November 25, 2002 2000 period. Such a result is quite remarkable, especially given the low level of pensions. Depending on the quarter (actually, regularity of pensions payment) during the observed period, each percent of pension expenditures helps eliminate 3.1 to 7.2 percent of poverty among the total population. 30% -*- Poverty Rate if Flat Pension Paid on Time -U- Poverty Rate Actual S20% 0. 0 C. 0 1% 5% 0% Internally Displaced People (IDPs) 50. Cash transfers to over 270,000 IDPs is the second largest social protection program operated in the country, accounting for about 1 percent of GDP. The relative generosity of the program and weak administration may have created incentives for abuse and leakage of funds, resulting in waste of scarce public resources. Furthermore, while the assistance is found to improve welfare of the IDP households, its untargeted nature raises equity concerns both in the case of local versus IDP population, as well as within the IDP population itself. Poverty Benefits 51. The poverty benefit program, which represents 0.4 percent of GDP, involves cash assistance vulnerable groups such as single elderly households, elderly couples and orphans whose both parents have died. Except for orphans, other eligible categories qualify only if they can prove that there is no-one to support them. Currently, the benefit is 22 lari per month for a single recipient households and 35 lari for other households. Relative to Georgian standards, these levels are generous. In November 2001, there were 49,574 households receiving the benefit, a 10 percent increase since the beginning of the year. 52. Targeting. Although found to be well targeted, the program is characterized by both errors of exclusion and inclusion. Other categories of the population which are equally vulnerable to poverty (such as families with many children, with no bread-winner, and with disabled members) are excluded from the program because of budget constraints. Problems of inclusion are mostly related to poor administration and fraud: some of the recipients are found to have other relatives to support them. 53. In sum, social protection programs in Georgia have been found to be effective in alleviating poverty. However, the efficiency and effectiveness of these programs could be significantly enhanced. In the short term, this could be achieved mainly by (i) ensuring regular payments of benefits and raising labor pensions in real terms; and (ii) improving administration and monitoring of programs. In the xii Georgia: Public Expenditure Review November 25, 2002 medium term, some of the programs would need to be restructured. For example, the pensions system needs to establish stronger links between benefits and contributions. Also, there is a need to introduce targeting in the IDP program which eventually should be integrated into the limited poverty benefit program, which should be expanded to cover other identified vulnerable groups. Conclusion 54. When the last PER was completed, in 1996, Georgia's government finances had virtually broken down as a result of the economic collapse that followed post-independence. At the time, it was appropriate to emphasize the need to raise the domestic resource envelope and to increase funding for priority expenditures, especially in the social areas. 55. Over the past years, Georgia has made important strides in raising revenue and in protecting budgetary allocations for the social sectors. Nevertheless, the tax policy and administration frameworks have created serious distortions that may threaten future growth and revenue collection. At the same time, scarce public resources for the social sectors have not necessarily been allocated efficiently or equitably. 56. The overall message of this PER is that it is not the level but the efficient management of public expenditures that matters the most at this stage in Georgia's transition. Notably, the establishment of a predictable fiscal framework is an immediate priority and a necessary pre-condition for improved expenditure management. Both central agencies and local governments work with extremely short planning horizons, and the lack of predictability and other weaknesses in budget management mean that agencies have neither incentives nor the necessary information base for the fundamental restructuring which is required to bring public expenditures in line with sector objectives, including MDGs. A second important message is that while efforts should continue to raise domestic revenues, greater focus needs to be placed on improving the way revenues are mobilized with a view to ensuring predictability and sustainability of collections over the medium term. Thirdly, in a situation when the financing of health and education services have been significantly privatized, the focus needs to turn toward ensuring better targeting. This also applies to existing social protection programs such as the poverty benefits and support for IDPs. 57. These messages are important for the Government as it develops its poverty reduction strategy. Indeed, presently, there is no link between the PRSP and the budget process. It is hoped that this PER will provide an analytical framework for the Government and other stakeholders, including external donors, to begin forging those links. xiv Georgia: Public Expenditure Review November 25, 2002 Chapter 1. Overview of Fiscal Developments Background 1.1 The last Bank Public Expenditure Review (PER), completed in June 1996, was carried out when Georgia was just emerging from one of the worst economic recessions/hyperinflations suffered by any Commonwealth of Independent States (CIS) country at post-independence. Between 1991 and 1994, output had fallen by nearly 80 percent and government finances had broken down. The dramatic drop in incomes and the ensuing collapse in government social services led to a rapid emergence of poverty in Georgia after independence. 1.2 The cessation of internal civil conflicts and the start of structural reforms in the mid-1990s, helped to restore economic growth and bring inflation under control. However, after a significant economic rebound in 1996-97, real Gross Domestic Product (GDP) growth slowed down from 10 percent to 3 percent annually and today, Georgia's real GDP is still only 40 percent of what it was at independence. Weaker (and uneven) economic growth since 1998 has been accompanied by an increase in poverty from 14 percent in 1997 to over 20 percent in 1998-2000.1 At the same time, social outcomes, especially in the health sector, have deteriorated sharply over the past decade: infant and maternal mortality have increased substantially, the number of new cases of communicable diseases has climbed up. In education, where outcome indicators are not available, there are signs that quality is rapidly declining. Moreover, the poor have suffered disproportionately from inadequate access to social services and from crippling basic infrastructure. 1.3 Georgia's growth and poverty reduction efforts depend crucially on a sound management of scarce public resources. While substantial fiscal adjustment has taken place under Georgia's stabilization program, primarily relying on expenditure cuts, fiscal sustainability is yet to be established. The Government has been unable to meet its revenue and expenditure targets and large arrears (notably on wages and pensions) have accumulated over the years. Public expenditures in maintenance and rehabilitation of basic infrastructure are negligible and those in the social sectors are among the lowest in the region. On the other hand, an increasing share of fiscal revenue is being devoted to service Georgia's large external debt. Also, a still oversized state apparatus (in terms of number of personnel and facilities), very low paid and inadequately trained staff, lack of clear priorities, and corruption mean that scarce public resources are not used efficiently or effectively. Past Budgetary Trends 1.4 Stabilization. Like most other CIS countries, stabilization objectives have played an important role in defining Georgia's fiscal policy during the transition. The stabilization program, launched in the mid-1990s, involved a substantial fiscal adjustment, aimed at reducing the monetary financing of the budget and stemming a massive accumulation of domestic arrears. The fiscal deficit (on a commitment basis) was drastically curtailed from an average of 35 percent of GDP during 1991-94 to around 7 percent of GDP in 1996-99. In 2000 and 2001, the fiscal deficit was further decreased to 4.1 and 1.9 percent of GDP, respectively. 1.5 The stabilization program was successful in rapidly eliminating hyperinflation and the rate of inflation was brought down to single digits by 1997. A sharp depreciation of the Lari, Georgia's national 1 According to the more generous "official" basket, around 50 percent of the population are below the poverty line. Also, it is estimated that about 60 percent of the population are at risk of falling into poverty. 1 Georgia: Public Expenditure Review November 25, 2002 currency2, in the 1998 Russian crisis, led to a temporary acceleration of the inflation rate during 1999. But, in subsequent years, the inflation rate has been kept at an annual average of below 5 percent while the exchange rate has remained relatively stable. Given the low degree of monetization of the Georgian economy, money supply has increased faster than nominal GDP during most of the period without causing inflationary pressures. 1.6 Georgia's external position has been weak throughout the transition and worsened in the aftermath of the 1998 Russian crisis. Although improving since then, gross foreign reserves amount to less than two months of imports of goods and services. Table 1.1: Selected Macro-economic Indicators 1995 1996 1997 1998 1999 2000 2001 Annual Real GDP Growth (%) 2.6 10.5 10.6 2.9 3.0 1.9 4.5 GDP Level (1990=100) 29.6 32.7 36.1 37.2 38.3 39.0 40.8 Average Annual Inflation, CPI (%) 162.7 39.3 7.0 3.6 19.1 4.0 4.7 Money Supply Growth (%) 135.2 42.0 46.0 -1.0 20.7 39.0 18.5 FDI (million USD) 6.3 54.4 236.3 221.0 61.7 152.6 96.1 Budget Deficit, as % of GDP -8.1 -7.3 -6.8 -6.1 -6.7 -4.1 -1.9 Tax Revenues, as % GDP 7.0 10.6 12.7 12.8 13.8 14.3 14.8 Current Account, as % of GDP -11.3 -8.9 -10.5 -10.7 -8.5 -5.3 -5.6 Gross Foreign Reserves (Import Cover)" 2.3 2.2 1.5 1.0 1.2 1.0 1.4 Exchange Rate, GELIUS$ (Average) 1.280 1.250 1.297 1.39 2.02 1.98 2.07 Broad Money, as % of GDP 7.4 6.7 8.0 7.3 7.8 10.4 11.3 1/ In terms of imports of goods and services Source: World Bank and International Monetary Fund (IMF). Fiscal Adjustment 1.7 The reduction in Georgia's large fiscal deficit was initially achieved through a dramatic adjustment in general government expenditures in 1995, when expenditures were cut by nearly 20 percentage points of GDP. Such a reduction reflected mostly the virtual elimination of a number of costly subsidies3 as well as a substantial downsizing of the public sector. 1.8 During the 1996-98 period, the bulk of the adjustment came through increased revenue mobilization. Notably, the economic rebound of 1996-97, when real GDP growth reached over 10 percent annually, was accompanied by buoyant growth in revenue. General government revenue collections more than doubled in 1996-97 and continued to grow faster than nominal GDP until 1998, suggesting that elasticity of revenue of GDP is higher than one. In percent of GDP, general government revenues increased considerably from 7.7 percent in 1995 to 15 percent 1998.4 With the slowdown in economic growth during 1999-2000, the revenue of GDP ratio stagnated at around 15 percent in those years. In 2001, the ratio reached 15.9 percent, to some extent reflecting a pick up in economic growth (4.5 percent). 1.9 The improvement of fiscal revenues since 1996 allowed some increases in expenditure commitments until 1998 (to about 23 percent of GDP). Nonetheless, as revenue increases began to tail- 2 The lari was introduced in October 1995 to replace the rubble-based coupon. For example, subsidies for bread and gas. 4 The nominal GDP series were revised down from 1996 onwards to reflect lower estimates for the informal economy. The 1995 nominal GDP was adjusted accordingly to make the ratios comparable. 2 Georgia: Public Expenditure Review November 25, 2002 off, further fiscal adjustment relied primarily on expenditure cuts. By 2001, expenditure commitments had been brought down to 18.5 percent of GDP, slightly below the level reached in 1995. 1.10 Role of external donor financing. In the immediate post-independence years, external donor assistance was an important source of budgetary financing, but its importance has declined considerably since the mid-1990s. Total external grants and credits for the budget have decreased from an estimated 7.3 percent of GDP in 1995 to around 1-3 percent of GDP currently, depending on whether or not untied budgetary financing is available. Notably, funding from grants has gone, down from nearly 3 percent of GDP in 1995 to less than 1 percent in 2001.' As a share of overall government expenditures, external grants and credits account for only about 10-15 percent of the total. Table 1.2: Consolidated General Government Budget 1995 1996 1997 1998 1999 2000 2001 (in % of GDP) General Government Total Revenue 10.6 13.8 14.4 15.6 15.4 15.3 16.6 Revenue 7.7 11.9 14.1 14.8 14.8 15.0 15.9 Grants 2.9 1.8 0.3 0.9 0.9 0.3 0.7 Total Expenditures 18.8 21.1 21.2 21.8 22.1 19.4 18.5 Current Expenditures 16.0 19.2 19.2 19.8 20.0 18.3 16.5 Non-interest Expend. 13.8 17.7 17.4 17.5 17.4 15.3 14.9 Interest Expenditures 2.2 1.5 1.8 2.3 2.8 3.0 1.8 Capital Expenditures" 2.7 1.9 2.0 2.0 2.1 1.0 1.8 Overall Balance (Commit.) -8.1 -7.3 -6.8 -6.2 -6.7 -4.1 -1.9 Primary Balance (Commit.) -5.9 -5.8 -4.9 -3.9 -4.1 -1.1 -0.1 Overall Balance (Cash) -7.0 -7.1 -6.2 -4.9 -5.0 -2.6 -1.7 Financing 7.0 7.1 6.2 4.9 5.0 2.6 1.7 Privatization 0.1 0.5 0.6 1.5 0.9 0.3 0.1 Domestic Financing 2.4 4.1 3.5 1.9 2.3 2.2 -0.4 External Financing 4.4 2.5 2.1 1.5 1.7 0.0 2.0 Disbursement 4.4 2.5 2.4 2.0 2.9 1.1 2.5 Amortization 0.0 0.0 -0.3 -0.5 -1.2 -1.1 -4.2 Debt Rescheduling 3.6 Source: World Bank and IMF Staff Report. 1/ Includes Net Lending. 1.11 Arrears and contingent liabilities. Despite Georgia's efforts to narrow fiscal imbalances, expenditure commitments were not adjusted to in line with persistent shortfalls in revenue targets. As a result, domestic expenditure arrears (notably on wages and pensions), though smaller than previously, continued to build up throughout the period, particularly during 1998 and 1999. Between 1996 and mid- 2000, domestic arrears reached about 5 percent of GDP. In the context of a substantial revision of the 2000 budget in mid-year6, revenue targets were reduced to more realistic levels, budget execution was broadly on track, and no further arrears were accumulated in the second half of that year. In 2001, some outstanding arrears (e.g., on poverty benefits) were paid back, but on a net basis there was still a small accumulation of arrears amounting to 0.1 percent of GDP. 5 This may reflect difficulties in accounting for grant-financed projects which do not pass through the budget. 6 After the newly-elected government took office. 3 Georgia: Public Expenditure Review November 25, 2002 1.12 It should be noted that officially recorded domestic arrears refer only to the so-called 'protected items' which include expenditure categories (such as wages, pensions, debt service) which have a mandatory nature. However, there is currently no system in place to account for expenditure commitments on non-protected items and it is reported that there exist additional outstanding obligations on contracts and utility services. In addition, while there are currently no government guarantees or explicit subsidies to public enterprises, there may be some contingent liabilities resulting from public enterprises debt. Overall Level ofPublic Spending 1.13 While conceptually it is difficult to ascertain whether a particular level of spending is adequate or not, international comparisons with countries within the same region and countries with similar per-capita income levels can offer a useful point of reference. 1.14 At 18.5 percent of GDP in 2001, Georgia's pubic expenditure level is in line with the average for all low-income countries (17 percent of GDP in 1998). It is also close to most low income countries in the CIS region, but considerably below the average for the ECA region (35 percent of GDP in 2000) and for the CIS countries (28.2 percent of GDP in 2000), which include countries with higher income per- capita such as Russia and Belarus. Table 1.3: Revenues and Expenditures, International Comparison (in percent of GDP) Revenue" Expenditure 1995 2000 1995 2000 Georgia 7.7 15.1 18.8 19.4 CIS Region 25.2 24.9 31.5 28.2 ECA Region 32.5 32.5 37.2 35.3 Low-Income All 13.9" 17.0" 1/ Refers to 1998 and to Central Government only. 2/ Excluding grants. Source: Staff Estimates. 1.15 In the absence of significant external donor flows to finance the budget, Georgia's ability to fund expenditures has relied mainly on its capacity to collect domestic revenue. As mentioned earlier, Georgia's revenue to GDP ratio has increased considerably since 1995 and today, the country's performance is close to most low-income countries within the CIS region. In fact, there has been a convergence of revenue performance: while Georgia's performance has been improving, that of other CIS countries has been declining or improving only modestly. Most of the observed gains in Georgia reflect increases in VAT collections, suggesting that the tax is buoyant and highly elastic to real GDP growth. While the level of income (and rate of growth) may be one important determinant of revenue performance, tax yields may vary due to a number of factors such as initial conditions, access to natural resource tax revenue (like oil in Azerbaijan), tax system and tax rates, as well as the administrative capacity. ' Among the low-income CIS countries, Georgia's revenue to GDP ratio is close to that of Azerbaijan, Armenia, Tajikistan, and Kyrgyz Republic. Moldova and Uzbekistan have higher revenue ratios in both cases reflecting large contributions from excise taxes and social security. 4 Georgia: Public Expenditure Review November 25, 2002 Composition ofPublic Spending 1.16 By level of government. Georgia's general government expenditures are divided under three broad categories. The central government, the state funds (or extra-budgetary funds), and the local government. The state budget includes expenditures of the central government and the state funds. In turn, the state funds include the pension fund and the road fund, which are financed from specific taxes (payroll and excise taxes) and central transfers.8 According to Table 1.4, in 2001, the state budget absorbed 67 percent of total general government spending, 46 percent was for central government expenditures and 20 percent for special funds, while local government expenditures accounted for 33 percent. The share of local governments has increased significantly since 1996 when it accounted for 17 percent of the total Government expenditures. 1.17 There are also off-budget expenditures financed with so-called "special resources", mostly related to service fees paid to public institutions (e.g., in tertiary education). These resources do not pass through the budget, but were reported for the first time in the 2002 budget, for information purposes. These special resources account for around 2 percent of GDP.9 Table 1.4: Public Expenditures by Level of Government 2001 percent in Total State Budget 783 67.0 Central budget" 543 46.5 State funds2' 240 20.5 Local budget2' 390 33.0 General Government 1173 100.0 1/ Excludes transfers to local government and to state funds and includes State Medical Insurance Company (SMIC). 2/ Includes transfers from central government. Source: Ministry of Finance. 1.18 Functional and economic classification. The way data are reported at various government/institutional levels makes it extremely difficult to assess the composition of spending by economic and/or functional classification, let alone analyze trends in the structure of spending. For example, the state funds do not report spending by economic categories and external donor financing is not broken down by sectors. Also, expenditures from local governments are not classified functionally according to the Government Financial Statistics (GFS) methodology, but according to government organizations. Despite these limitations, some estimates have been made to illustrate the composition of the consolidated budget for 2000 and 2001, but they should be taken as rough approximations. Figure 1.1: Public Expenditure by Functional Classification, 2001 1.19 An overall picture of the structure of public spending by function was constructed by %Shares 15 distributing available data for central and local government (as well as the state funds) under a common sectoral classification. Figure 1.1 shows the composition of expenditures of the state (central plus state funds) and local budgets in 2001. According to these data, the largest > share of public spending is absorbed by social ' The employment fund was abolished in 2002. 9 If included in the budget, these would be added to non-tax revenue bringing the total revenue to 17.9 percent of GDP in 2001. 5 Georgia: Public Expenditure Review November 25, 2002 insurance and welfare (24 percent), followed by education (13 percent), general services (8 percent), public order (9 percent), and health (7 percent). The structure of spending does not change much if external donor financing and special resources are included. 1.20 The estimated breakdown of expenditures by economic classificationo in 2000 shows that the lion share of Georgia's total government Figure 1.2: Public Expenditure by Economic expenditures (about 91 percent) is accounted Clasifiction, 2000 for by recurrent costs. Among recurrent % Shares 25 costs, the largest proportion (24 percent) went 20 to finance current transfers (e.g., pensions, 15 poverty benefits, assistance to internally 10 displaced people), followed by goods and services (18 percent), debt service (17 percent), personnel costs (14 percent), and subsidies (7 percent). It is estimated that about 8.8 percent of the general budget was Fiur 1.2 PulcLniueyEooi allocated to finance capital expenditures. _ Almost the entire capital budget was financed by external donors, mainly IDA credits. Georgia's capital budget accounts for about one percent of GDP, which is roughly what the country spends in servicing its external debt. 1.21 It is important to note that Georgia's heavy external debt burden has meant that an important proportion, about one fifth of overall expenditures have been directed to service the external debt. As a result, non-interest current expenditures have decreased considerably from 17.2 percent of GDP in 1999 to 14.9 of GDP in 2001. External Debt Burden 1.22 Georgia can be classified as a debt distressed country on account of the fiscal burden of its external debt. Its total government and government-guaranteed external debt at end-2000 was US$1.6 billion" (or 53 percent of GDP), compared with US$1 billion in 1994. Much of the increase reflects new concessional multilateral debt which accounts for about 50 percent (less than 5 percent in 1994). Debt owed to bilateral creditors accounted for the rest, with debt to CIS countries, mainly Turkmenistan and Russia, accounting for about 30 percent of the total. In net present value (NPV) terms, the total debt stock at end-2000 was about US$1.3 billion, equivalent to 128 percent of exports of goods and services and 356 percent of central government revenues. The high NPV of the debt relative to the nominal level reflects the impact of non-concessional debt acquired in the immediate post-independence years. 1.23 Despite undergoing several rescheduling agreements with its bilateral creditors since 1995, Georgia's actual debt service payments (i.e., after debt rescheduling) accounted for a significant share of central government revenues in the 1999-2000 period (around 37 percent). Moreover, a large stock of external arrears (mostly owed to Turkmenistan) has accumulated over the years which at end-2000 represented a further 35 percent of central government revenue. Even after application of the last Paris to To obtain a breakdown of expenditures by economic classification, disaggregated data for central and local governments was consolidated and the pension fund was added to current transfers. Excluding about US$250 million in debts owed by energy utilities but not guaranteed by the state. 6 Georgia: Public Expenditure Review November 25, 2002 Club rescheduling agreement, signed in March 2001, Georgia's debt service payments would still absorb around 30 percent of central budget revenue during the 2001-2002 period.12 Medium-Term Fiscal Outlook 1.24 Under the Fund's medium term macroeconomic framework,13 it is anticipated that Georgia will be able to sustain moderate real GDP growth rates, averaging 4 percent per annum, provided that the Government continues to make steady progress in implementing structural reforms and that the external environment remains reasonably favorable. These growth rates are lower than those achieved in the period after the post-independence economic collapse, but are somewhat higher than the rates recorded in the years that followed the Russian crisis. 1.25 Assuming the continuation of prudent and fiscal monetary policiesl4, the average rate of inflation is projected to be stable at around 5 percent per year and the real exchange rate is projected to remain constant. Gross reserves are projected to increase to 1.7 months of imports in 2003 and remain around that level through 2007. 1.26 The fiscal deficit (on a commitment basis) is projected to be kept at around 1 percent of GDP and the cash deficit at around 2 percent of GDP, assuming that the stock of domestic payment arrears are fully cleared by 2006. The tax-to-GDP ratio for the general government is assumed to increase by half a percentage point of GDP each year, resulting in an increase of central government revenue from 10.5 percent of GDP in 2000 to 13 percent of GDP by 2005." External imbalances will narrow somewhat in 2002 but will widen thereafter and financing gaps are anticipated during 2003-2007, reflecting the expiration of the consolidation period under the last Paris Club agreement.'6 Official grants and loans through 2005 are assumed to average US$55 and US$80 million, respectively. External Debt Sustainability 1.27 Under a base case scenario, which is consistent with the medium-term macroeconomic framework described above, the fiscal burden of the external debt will be unsustainable unless further debt restructuring is provided under concessional terms. The last rescheduling agreement with the Paris Club provided debt relief during the consolidated period of 2001-2002,'1 bringing debt service payments to 25-30 percent of central government revenue. If Georgia were unable to obtain further debt rescheduling, debt service obligations would rise substantially to over 40 percent of the central government revenue during 2003-2004. Although the debt NPV -to-exports ratio would remain below the 12 The rescheduling agreement involves 2001-2002 maturities to be repaid in 20 years, with a grace period of 3 years and an interest rate of 4 percent. 13 See recent IMF Staff Reports. 14 Monetary targets envisaged under the current macroeconomic framework assume a gradual decrease in the velocity of circulation. 1s These revenue projections do not include the fiscal impact of royalties from protected oil and gas pipelines. 16 The balance of payments figures from 2000 have been substantially revised upward to reflect new estimates by NBG which include: higher exports, lower imports, and reduced net receipts and transfers. 17 While the terms of this agreement, so-called 'enhanced' Houston terms, are considered non-concessional, they imply a 27 percent grant element. Debt service obligations due in 2001-2002 would be reduced by US$240 million, with roughly 80 percent of the reduction coming from the non-Paris Club members, in particular Turkmenistan. It is assumed that arrears to Turkmenistan of US$180 million are paid over 10 years. 7 Georgia: Public Expenditure Review November 25, 2002 150 percent solvency threshold throughout the 2001-2010 period, the debt NPV-to-central government revenue ratio would remain well above 250 percent until 2004, indicating severe debt distress.'8 Table 1. 5: External Debt and Debt Service after Debt Rescheduling, 2000-2007 2000 2001 2002 2003 2004 2005 2006 2007 (US$ million) Total External Debt' 1,610 1,685 1,801 1,981 2,065 2,006 2,095 2,147 Net present value 1,110 1,048 1,095 1,188 1,262 1,059 1,131 1,158 % of exports of goods and services 131.3 110.7 102.9 108.4 107.6 84.1 83.3 78.4 % of Central Gov. Revenue 356.1 305.1 297.7 289.9 277.1 208.9. 201.0 185.4 Total Debt Service" 116 90 124 151 179 178 179 187 % of exports of goods and services 10.5 8.5 10.7 1.5 11.6 9.6 8.1 10.5 % of Central Gov. Revenue 36.4 25.4 31.8 32.8 32.0 25.5 21.3 19.4 1/ Based on 2001 Paris Club agreement and additional flow rescheduling of 2004-05 maturities and Naples stock-of- debt operation in 2005 Payments of principal arrears to Turkmenistan are assumed to be paid over 10 years. Source: IMF StaffReports. 1.28 Alternative Debt Strategies. The most recent debt sustainability analysis (DSA) carried out in collaboration with the Fund indicates that concessional debt relief will be required to keep debt service obligations at manageable levels beyond 2002, when the current Paris Club rescheduling agreement is due to expire. Under the base case scenario, Georgia's external debt could become sustainable if concessional debt relief under Naples terms is provided after 2002 (i.e., a flow rescheduling of 2003-2004 maturities followed by a stock-of-debt operation in 2005). From a fiscal point of view, Georgia's external debt would become sustainable by 2005, reflecting the stock of debt operation under Naples, when the net present value of the debt would fall to below 250 percent of central government revenue and the debt service obligations to around 20 percent of projected central government revenue. 1.29 It should be noted, however, that in the event of an adverse scenario, where less favorable external environment would lead to lower real GDP and export growth (between I and 1.5 percentage points lower than the base case scenario), the debt NPV-to-revenue ratio would remain above the HIPC threshold of 250 percent through 2010, even after Naples treatment. 1.30 The medium-term macroeconomic framework discussed above, shows that even under relatively optimistic assumptions, i.e., moderate revenue increases and further debt rescheduling on concessional (Naples) terms, there will be little room for expanding non-interest public spending over the medium term. Non-interest current expenditures will remain at around 15 percent of GDP through 2004 and increase only gradually to 18 percent of GDP by 2007. Capital expenditures will remain depressed at around 1-2 percent of GDP throughout the forecasting period, reflecting continued but modest external donor financing. 18 The debt projections presented here differ from the ones presented in the Board Paper on the External Debt and Fiscal Sustainability of IDA-only CIS countries (IDA/SecM2001-0102 dated February 13, 2001). First, the full effect of the Paris Club rescheduling granted by the creditors on March 6, 2001, is incorporated in the present projections. Second, export numbers for 2000 were revised upwards (by about 50 percent) to previously unregistered export proceeds. Both factors have significantly improved the debt burden indicators. 8 Georgia: Public Expenditure Review November 25, 2002 Table 1.6: Medium Term Fiscal Framework, 2000-2007 2000 2001 2002 2003 2004 2005 2006 2007 (in % of GDP) General Government Total Revenue & 15.3 16.6 17.3 17.7 18.4 18.7 19.3 20.0 Revenue 15.0 15.9 16.2 16.7 17.3 17.9 18.6 19.2 Grants 0.3 0.7 1.1 1.1 1.1 0.7 0.8 0.8 Total Expenditures 19.4 18.5 18.4 18.9 19.3 20.3 20.2 20.9 Current Expenditures 18.3 16.8 16.8 17.2 17.7 18.4 18.7 19.5 Non-interest Expend. 15.3 14.9 14.4 15.0 15.6 16.6 17.2 18.1 Interest Expenditures 3.0 1.8 2.4 2.2 2.0 1.8 1.5 1.4 Capital expenditures" 1.0 1.8 1.6 1.7 1.6 1.9 1.5 1.4 Overall Balance (commit) -4.1 -1.9 -1.1 -1.2 -0.9 -1.6 -0.9 -0.9 Overall Balance (cash) -2.6 -1.7 -1.8 -2.0 -1.7 -2.4 -1.7 -0.9 Financing 2.6 1.7 1.8 2.0 1.7 2.4 1.7 0.9 Privatization 0.3 0.1 0.3 0.4 0.4 0.4 0.3 0.3 Domestic Financing 2.2 -0.4 0.8 0.5 0.4 0.4 0.3 0.3 External Financing 0.0 2.0 0.8 1.1 0.9 1.7 1.4 1.1 Disbursement 1.1 2.5 2.0 2.4 2.2 2.4 2.0 1.8 Amortization -1.1 -4.2 -5.1 -2.7 -2.5 -1.9 -1.6 -1.5 Debt Reschedulinge .. 36 3.8 1.4 13 1.2 10 09 I/ Includes net lending. 2/ Based on 2001 Paris Club agreement and additional flow rescheduling of 2004-05 maturities and Naples stock-of- debt operation in 2005. Payments of principal arrears to Turkmenistan are assumed to be paid over 10 years. Source: IMF Staff Report. Conclusion 1.31 Georgia's fiscal stance under the current macroeconomic framework seems sufficiently conservative to ensure price and exchange rate stability over the medium term. Annual projected cash fiscal deficits of around 2 percent ofGDP, are consistent with domestic financing of around 0.4 percent of GDP. Monetary targets assume gradual decreases in the velocity of circulation, so money supply and credit to the private sector are projected to continue growing above nominal GDP. The fact that monetization in Georgia is quite low, however, would suggest that there may be room for further monetary expansion than what is envisaged under the program without compromising stabilization objectives. This would imply that there may be some margin to accommodate additional domestic financing of the fiscal deficit or increased credit for the private sector. In the absence of a model to simulate alternative scenarios, the calibration of how much margin there may be and where it should be allocated can only be done in the context of annual reviews of the macroeconomic program. In any event, a conservative fiscal stance seems appropriate at this juncture given the uncertainties about the true level of arrears and possibly contingent liabilities that may result from the public enterprise sector. These are areas which will have to be more carefully examined in future programs. 1.32 Clearly, Georgia should explore further opportunities to establish external debt sustainability as soon as possible and preferably earlier than 2005 in order to free much needed public resources for developmental use. 1.33 While projected increases in revenue are rather modest, there is still a question of whether even these increases could be achievable in coming years, given Georgia's structural and political conditions (such as the level of income, size of the informal economy, security, etc.) and the country's limited institutional capacity. These issues are further analyzed in the next chapter on revenue mobilization. 9 Georgia: Public Expenditure Review November 25, 2002 Chapter 2. Revenue Policy and Administration 2.1 Georgia has improved its revenue performance, but it has yet to develop a tax system that facilitates compliance and guarantees a level playing field for all taxpayers. Georgia's ratio of revenue to GDP over the last years has been growing more rapidly than most CIS countries and, today, is close to other low income countries in the region. Increased revenue collection in Georgia has, however, been accompanied by considerable distortions generated by inadequate practices in both tax policy and tax administration which, if continued, are likely to undermine economic growth and revenue mobilization. Since the authorities seem committed to continue increasing revenues gradually over the medium term, this would seem to be an appropriate time to focus attention on improving the quality of the tax environment. Priorities would have to be: (a) broadening the tax base by drastically reducing special exemptions and registering new taxpayers with revenue potential, and (b) improving voluntary compliance based on self-assessment. If this is done, the government would be able to respond effectively to pressures to reduce the overall tax burden, generate a level-playing field for all taxpayers, and foster economic growth. Revenue institutions have already received considerable advice from international donors in policy and administration. Most of this advice has only been partially implemented. The absorptive capacity for institutional reform in the country is small and there is considerable resistance from revenue institutions and society to change. Indeed, implementing appropriate revenue reforms in an environment that readily accommodates corruption and political pressures is not easy. Strong political commitment will be needed to (a) ensure the sustainability of an improved revenue performance and that marginal gains continue over the medium term until a satisfactory fiscal balance is obtained, and (b) create a tax system which facilitates private sector development. Trends and Structure International Comparisons 2.2 Revenue performance in Georgia collapsed after independence as a result of the break-up of economic ties with FSU countries and internal war. After reaching a low point in 1994, revenue performance increased rapidly until 1998 and more gradually since.'9 During the last two years revenue as a Graph 2.1. Georgia: General Government percentage of GDP has grown by about 0.4 percentage Revenue as % GDP points of GDP per year, which is still a strong performance.20 In 2001, tax revenues as a percentage of 17 GDP reached 14.8 percent of GDP, including all of tax 16 revenues at the national and local levels and the social security contributions. On balance, Georgia has had a good 14 track record in increasing revenue collection within the CIS 13 region. As can be seen in Graph 2.2, Georgia's belongs in a197 9M 9 2W 01 group with other low per-capita income CIS countries.21 2.3 Still, Georgia inherited expectations on public expenditures which cannot readily be financed with available domestic resources. While much can be gained by streamlining expenditure, improving 1 The elimination of budgetary/tax offsets leading to more effective cash collection arguably could have led to an increase on officially reported revenues. 20 The government claims that, in 2001, revenue improvements were obtained in spite of erosions to the tax base coming from special treatments given to agriculture, electricity and high mountainous regions On the other hand, if account is taken of increases in the stock of the debts to taxpayers (from refunds and advance payments) revenue growth is more modest 21 The exceptions are Moldova and Uzbekistan 10 Georgia: Public Expenditure Review November 25, 2002 effectiveness and reducing waste and corruption, the authorities realize the need to continue improving Graph 2.2 General Government Revenues and revenue performance. The government has set a Per Capita GDP in CIS target of 0.5 percent per annum. This target appears reasonable and yet ambitious, as rarely do countries Rev 0 show such sustained gains from administrative % improvements, once easy early gains have been GDP 30 obtained. The authorities are also aware that besides 20 generating sustainable public revenues over the medium term, the design and implementation of revenue policy should be such that they create an 0 adequate environment for investment and economic 500 1000 1500 2000 adequte. Per Capita GDP (US$) activity. Revenue Structure and Outcomes 2.4 Indirect taxes. Indirect taxes (VAT, excises, import duties) are the most important source of tax revenue, generating around 7.9 percent of GDP or about 54 percent of total tax revenues in 2001. VAT collection has improved rapidly from 4.3 percent of GDP in 1999 to 5.5 percent in 2001, showing that the tax is buoyant, despite increasing exemptions. Although VAT performance (yield/standard rate) is in line with that of other CIS countries, gains in VAT performance would be feasible by improving administration and expanding the base. Excise performance on the other hand has deteriorated from 2.3 percent of GDP in 1999 to 1.6 in 2001, below reasonable international standards. Customs duties produced 0.8 percent of GDP in 2001. Overall, improvements in indirect taxation are feasible and necessary to help deal with weaknesses elsewhere in the tax system. Table 2.1: General Government Revenue (percent of GDP) 1999 2000 2001 Total Revenue (ex. grants) 14.8 15.0 15.9 Total Tax Revenue (include special state funds) 13.9 14.3 14 6 Taxes on Income 1.8 1.8 2.1 Taxes on Profits 1.0 1.3 1.0 VAT 4.3 48 5.5 From Tax Dept 2.5 2.8 3.1 From Customs 1.8 2.1 2.4 Excises 2.1 1.5 1.4 From Tax Dept 0.3 0.2 0.3 From Customs 1.8 1.3 1.0 Custom Duties 0.6 0.9 0.8 Local Tax 1.5 1.4 1.4 Non-tax Revenue 0.8 0.7 1.3 Special State Funds Revenue 2.7 2.4 2.4 Road Fund 0.6 0.4 0.5 O/w. gasoline excise 0.2 0.1 0.2 Employment Fund 0.1 0.1 0.1 Social Security Fund 2.0 1.9 1.8 Source: NBG. 11 Georgia: Public Expenditure Review November 25, 2002 2.5 Income Taxes and SS contributions. Corporate income tax (CIT) revenue has stood around I percent of GDP since 1995, reflecting Georgia's weak industrial base following the loss of market links with the former Soviet Union. Today, many enterprises remain paralyzed or operate at a low level of activity. This experience parallels that of other transition economies, where a crisis of State-owned Enterprises (SOEs) profitability eroded the tax base for CIT. PIT and social security contributions have been holding steady and in 2001 contributed 4.0 percent of GDP. Performance appears low, given the high rate of social security contributions (30 percent), but the high rate provides incentives for underreporting and informal sector activities to escape the tax payment. 2.6 Local and other Taxes. Local tax revenues have held steady around 1.4 percent of GDP. They include a variety of small taxes and surcharges that produce little in revenue but increase the cost of compliance and administration. (Issues of local taxation are addressed in detail in the inter-government finance Chapter 4). Table 2.2: Tax Structure for CIS Countries, 2000 VAT Excise Trade Profit Income SS Total Total Tax NonTax Armenia 6.5 2.5 0.8 2.0 1.4 2.3 17.7 1.2 Azerbaijan 4.1 0.5 2.1 2.7 2.0 2.3 14.6 6.3 Georgia 4.9 1.5 0.9 1.3 1.8 2.4 14.3 0.8 Kirguiz 4.8 2.4 0.4 1.1 1.1 3.5 15.8 2.6 Moldova 82 4.1 1.4 1.7 1.1 6.1 26.6 3.7 Tayikistdn 2.5 0.5 1.5 - 1.8 1.6 12.9 0.6 Uzbekistin 7.6 7.8 0.7 3.2 4.2 - 26.7 1.8 Average CIS 6.2 2.5 1.3 2.1 2.6 3.9 21.6 2.5 Performance Evaluation The Tax Environment 2.7 Georgia's difficult environment for revenue collection has to be taken into account in assessing performance and considering possibilities of reform. 2.8 A porous border and transit trade. Georgia does not fully control its external borders. Abkhasia, bordering with Russia, operates independently of the central government and does not contribute revenues to the central budget. Adjara, with an important port on the black sea, has the status of a quasi- independent region, with a special arrangement for transferring revenue. South Ossetia does not have a well-defined external border with Russia and is conduit for open smuggling. The border with Azerbaijan is quite open and can be crossed at will, without any significant control, specially on the Georgian side. Additionally, Georgia is a transit route from east, west, north and south. It is an important route to and from Azerbaijan and countries to the east and practically the only effective exit route by land for Armenia. For Russia is an alternative route to the black sea. These factors not only make it difficult to control the flow of goods and their destination, but create special national and local interests that stand in the way of reform. 2.9 Poor payments discipline. Like other transition economies, payments discipline is low in Georgia. There are significant budgetary arrears, as reported elsewhere in this report. These arrears in turn weaken financial discipline in the economy. In sectors like electricity and gas, the effective cash collection ratios are very low. Moreover, the non- implementation of bankruptcy procedures has led to the continued operation of enterprises that do not honor their financial obligations. All of these factors 12 Georgia: Public Expenditure Review November 25, 2002 bode ill for revenue performance, especially when the tax system has already moved from cash to accrual accounting. The consequence is substantial increments in the level of arrears with limited possibilities for collection. 2.10 A drastic economic transition. Of all of the CIS countries, Georgia's economy dropped the most during transition, by 70 to 80 percent. The bulk of the industrial infrastructure was disconnected from the markets in Russia for inputs, outputs and technology and with the loss of the Russian markets, agriculture reversed to subsistence. The growing sectors of the economy have been trade, services and agriculture. Internal conflict and the displacement of a significant percentage of the population and have led to a reduction of the taxable basis of the economy and the ability to produce revenues. To build a new and healthy tax base the country needs an environment that favors the creation of new economic activity, including through foreign investment, by the creation of a fair-playing field for all. 2.11 Corruption and political interests. Reportedly, powerful political forces preclude improving collection in areas where enforcement would appear straightforward, such a fuel imports. Issues of corruption seem to permeate revenue collection from top to bottom. The culprits are not only revenue officials, but society at large that condones corruption and uses it to its advantage. Reportedly revenue institutions are prone to capture by special interests, as evidenced by special treatments and the uneven application of the law.22 Broads systemic changes will take time, meanwhile attention to pay reform, human resource management and training and developing the architecture of internal and external control are necessary activities. 2.12 An improved macroeconomic environment. On the positive side, the macroeconomic environment has improved considerably. Inflation has been down to one digit level for several years now and economic growth has resumed, at moderate rates. Low inflation and economic growth are favorable factors for taxation. The economy seems to be coming out of the Russian crisis and reaching greater integration into the world and regional economy. The main export market already is Turkey. The revenue performance shows that the revenue elasticity with respect to GDP is greater than one, suggesting that the tax design is buoyant. However, continued erosion of the tax design through exemptions risk destroying this positive characteristic. Tax Revenue Potential 2.13 Estimates of revenue potential permit a rough evaluation of the overall revenue performance- compliance, enforcement, etc. In Georgia, early estimates of revenue potential focusing mostly on excises and VAT23, suggested exorbitant revenue short-falls. One, for instance, estimated that in 1998 only 19 percent of cigarette excises were collected. More recent estimates by the MoF, based on levels of domestic consumption adjusted for price changes, yield lower but still considerable revenue short-fall estimates. For petroleum products (gasoline and diesel), MoF estimates that combined VAT and excise payments are only 56 percent of the potential, or a shortfall of around GEL83 million in 2001. Given that all fuel is imported in bulk by few traders through the railway system, such high level of evasion suggest 22 Several studies have document the problems of state capture and corruption in Georgia. Some studies have ranked the country amongst the worst in the Former Soviet Union (FSU). , See Hellman, J., G. Jones and D. Kaufmann, 2000, 'Seize the State, Seize the Day: State Capture, Corruption, and Influence in Transition," World Bank Policy Research Paper 2444 (http-//www.worldbank.org/wbi/eoveernance/). Georgia ranked the highest in state capture, judiciary capture, and administrative corruption severity-in the latter by a high margin. The IMF has also documented how state capture has operate in the area of taxation leading to exemptions, special treatments, and a general perception that the law is manipulated (with the cooperation of parliament) and not implemented fairly. (See IMF- Georgia Recent Economic Development and Selected Issues, November 2001). 23 Some of these estimates can be found in the Barents Web page. 13 Georgia: Public Expenditure Review November 25, 2002 serious collusion to evade taxes. Likewise for cigarettes, collections would according to the MoF only 57 percent of the potential, with a shortfall of GEL39 million. Smuggling of cigarettes is certainly easy and difficult to control in Georgia; but, what stands out in Georgia is that there is also a substantial shortfall in the collection of excise for domestic cigarette production. (Tables 2.3 and 2.4). Table 2.3: Actual Revenues and Potential Revenue Estimates from Petroleum Products, 2001 (GEL millions) Registered Shortfall from Consumption Imports or Potential Revenues Actual Revenues Projection Estimate" Production" Value" (GEL millions) (GEL millions) (GEL millions) (thousand (thousand (GEL tons) tons) millions) Excise VAT Total Excise VAT Total Excise VAT Total Fuel 105.0 82.0 187.0 56 0 48 4 104.4 -49.0 -33.6 -82.6 Gasoline 400 260 210 80.0 58 0 138.0 43 7 38.7 82.4 -36.3 -19.3 -55.6 Diesel 250 77 95 250 24.0 49.0 123 9.7 22.0 -12 7 -14.3 -27.0 1/ Value is estimated based on GEL525 per ton of gasoline and GEL384 per ton of diesel. 2/ There is a time lag between imports registration and actual collection of taxes 3/ Based on Car Park data estimates. Source: MoTR. Table 2.4: Projected and Actual Revenues from Tobacco Imports and Production, 2001 Registered Actual Imports or Revenue Revenues Production Tax Rate Projection in 2001 Shortfall from Consumption (million (GEL per (GEL (GEL Projection (million packs)" Packs) pack) million) million) (GEL million) Tobacco 316.0 69.0 30.4 -38.6 Imported 94.9 61.1 36.0 15.1 -20.9 Non-filtered 10.0 0.0 0.2 2.0 Filtered 84.9 61.1 0.4 34.0 Domestic 221.1 90.3 33.0 15.3 -17.7 Non-filtered 86.3 24.9 0.1 6.0 Filtered 134.8 65.4 0.2 27.0 1/ Based on HH survey data. Source: MoTR. 2.14 Some calculations put VAT collections at 54 percent of potential revenue, based on estimates of national consumption--a revenue short-fall in the amount of GEL305 million. This compares with VAT collection of only 43 percent in 1998.24 Estimates by the MoF put 2001 VAT collection at 68 percent of potential, still a significant short-fall of GEL167 million. Although there is substantial discrepancy 2 Based on updated Barents calculations. 14 Georgia: Public Expenditure Review November 25, 2002 between these two estimates, even the lowest estimate implies a substantial revenue loss. The MoF calculations also show that customs is weaker than the tax office at collecting VAT and excises. 2.15 The MoF calculates that only around 40 percent of total potential of the combined PIT and social security contribution is collected. This is interesting since PIT and social security contributions tend to have high compliance rates because they are collected through withholding. Again, the problem seems substantial underreporting of wages or large informal economy. Additional factors include poor coverage of private entrepreneurs and small legal entities. Table 2.5: Actual Collection vs. Potential Revenue Estimates, 2001 Actual vs Actual vs. Baseline Potential Actual Baseline Potential GEL % share in GEL % share in GEL % share in millions GDP millions GDP millions GDP % % Total PIT 464.7 7.10 348.5 5.30 135.8 2.1 29.2 39 Salaries and wages 334.9 5.1 251.2 3.8 Self-employed 88.1 1.3 66 1 1.0 Dividends 41.7 06 31.3 0.5 Total CIT 148.9 2.3 111.7 1.7 66.5 1 44.6 59.5 Total VAT 697.3 10.7 522.9 8.0 355 9 5 4 51 680 Domestic Production 369.9 5.7 277 4 4 2 200.7 3 0 54 3 72 4 Imports 327.4 5.0 245.5 3 7 155.1 2.3 47.4 63.2 Total Excise 227.7 3.5 170.8 2.6 894 1.3 39.3 52.4 Customs Duties 1169 1.8 87.7 1.3 55 0.8 47 627 Total Social Taxes 50443 771 378.2 58 1606 24 318 424 Social Security Fund 481.3 7.4 361 5 5 155 5 2.3 32.3 43.1 Employment Fund 23 04 17.3 0.3 5.1 0.1 22.2 29.6 Other 240.2 3.7 180.2 2.7 86 1.3 35 8 47 8 Total Taxes 2400 36.7 1800 27.4 949 14.3 39.5 52.7 Tax Department 1683 25.7 1262.2 19.2 638.5 9.6 37.9 50.6 Customs Department 672 10.3 504 7 7 289.5 4.4 43.1 57.5 Road Fund 45 0.7 33.8 0.5 21 0.3 46.7 62.3 Source: Ministry of Tax Revenue (MoTR). 2.16 Even if the above estimates are ballpark figures, they suggest two points as to where to focus in improving collection. Focus on customs should help improve the collection of VAT and excises, as the MoF estimates show custom is a weaker performer than the tax department. The second is the need to improve collection of social security contributions and PIT. This requires strengthening collections 15 Georgia: Public Expenditure Review November 25, 2002 among private entrepreneurs and small legal entities and tight enforcement among large enterprises, through special programs that target collusion to report lower wages. Also, assignment of the PIT to local governments, if properly designed, can help improve PIT (and SS contributions) collection efforts. Experience of other countries shows these improvements are feasible. 2.17 The estimates of potential collection calculated by the MoF have to be seen as indicative of the potential gaps and pointing to directions on how to improve the collection efforts. These are not targets than can easily be achieved even with substantial improvements in compliance and enforcement. They indicate, however, that there is scope for improvement and that meeting the target of continued marginal increases in collections is a feasible objective. 2.18 What is the true tax burden? It is worthwhile pointing out that the discrepancy between the revenue potential and the actual collections does not all remain with the taxpayers. Indeed, experience of other countries show, and Georgia is not likely to be an exemption, that part of that shortfall is actually transferred from taxpayers to agents that "facilitate" taxation inside or outside the revenue agencies. Indeed, survey information on side-payment in Georgia would suggest that taxpayer payments are likely to exceed substantially treasury receipts. Thus addressing issues of corruption, capture harassment has to be an integral part of reform. 2.19 Exemptions and special treatments. Tax base erosion is an indirect way of assessing impact and equity.25 Reportedly, pressure on exemptions are on the increase. This behavior arises often when tax rates are high. In VAT exemptions have grown recently. In 2001, 23 percent of the total turnover in the VAT tax returns is reported as non-taxable21 and the majority of the VAT exempt products and services are also exempt from customs duties. In fact, close to 40 percent of all imports do not pay taxes or duties, mainly imports by humanitarian organizations. Special treatments granted by type of enterprise/organization, rather than product, often lead to distortions and are difficult to control. The MoF estimates that the overall level of recorded tax and contribution privileges in 2001 amounted to 30 percent of total revenues. 2.20 Negative protection. The way the tax regime is enforced can have effects not considered in its design. For instance, smuggling and underreporting affect negatively domestic production, especially of low value added products. This "negative protection" in turn prevents the development of a solid and growing tax base. Weakness in collecting PIT and social security contributions and high rates induce the growth of the informal economy. These factors shift the burden of taxation to the formal sector of the economy. VAT and excises on the other hand, if they are well implemented, by focusing on consumption cover both formal and informal sector. Tax Policy 2.21 Georgia was one of the first CIS countries to codify its tax legislation in a comprehensive Code. However, since its adoption in 1997, there have been numerous amendments which have considerably reduced the consistency of the Code. Some of the mayor changes in recent amendments include: i) changes in the tax rates for tobacco products and the tax rates of the motor vehicles ownership tax; ii) repealing provisions in the Code allowing the tax administration to seize and sell delinquent taxpayers' property; iii) introduction of exemptions from property taxation for enterprises and physical persons in 25 Ideally an analysis of impact and equity would have to look at the net result of the tax system on taxpayers income and wealth. Information of this is not available in Georgia. 26 This does not take account of the evasion induced effects of the exemptions or returns that may be misclassified as zero rated. 16 Georgia: Public Expenditure Review November 25, 2002 mountainous regions. The IMF carried out a review of tax policy in 2000 and a number of the recommendations from this review were actually included in a tax reform package prepared by the Ministry of Finance in September 2001. However, this package was not submitted to Parliament and a revised version is currently under consideration. In the meantime, as part of the IMF program, Parliament passed some amendments to the tax code which included the removal of tax exemptions for NGOs and provisions to allow the seizure of property by tax inspectors in case of default by a taxpayer. Key issues remaining on the tax policy side are: 2.22 Unstable tax policy framework. The history of tax policy changes in Georgia since the adoption of the tax code demonstrates a lack of long-term policy planning and a focus on short-term policy measures, disregarding the general consistency of the Code. Such approach has led to constant changes to Article 273 (on transitional provisions). Even fundamental policy changes are not introduced as permanent features of the tax system, but as temporary ones. A typical example is the cigarette taxation, which has been modified six times since the Code entered into force. New taxation schemes are often introduced late in the year, for short periods of time and without clarification as to their duration. The current taxation scheme for tobacco products was introduced for the year 2001 only on December 2000 and was extended for another year through another amendment to the Code on December 2001. An even worse case is the excise tax on the importation of pyrolisis liquid products which was set at a rate of 400 GEL per ton on December 2000 and reduced to GEL50 per ton less then four month later. There are numerous similar examples of short-term tax policy measures and frequent changes of tax legislation.27 Such an approach neither allows the business community to calculate its tax burden over a longer period of time, nor does it permit the revenue authorities to design appropriate taxation strategies and develop a long-term planning of resource mobilization. The strong influence of lobbies in Parliament28 and the obvious tendency of parliamentarians to further narrow the tax base by granting sector and specific exemptions and rate reductions also contributes significantly to the low quality of tax policy making in Georgia. 2.23 VAT Threshold. Currently, VAT registration is mandatory for businesses with an annual taxable turnover of 24,000 GEL or more (and voluntary for a businesses below this threshold).29 As a result of this low threshold, the tax administration has to deal with a large number of small businesses who contribute little to total VAT revenues. An increase of the threshold from GEL 24,000 to 100,000 would reduce the number of mandatory taxpayers from around 13,000 to 3,200 while reducing collections by GEL 8 million or 2.3 percent of VAT collections.30 A reduction in the number of taxpayers could substantially facilitate the administration of the tax and help combat VAT evasion by permitting a more comprehensive cross-checking of VAT invoices.3 However, this result can only be achieved if the scope for voluntary registration is reduced. The Ministry of Finance therefore should consider to limit voluntary registration, e.g., by excluding businesses with a turnover below GEL 50,000. 2.24 VAT Distortions. There is increasing frustration with the performance of VAT and the distortions its creates because the tax net is narrow and businesses are often unable to deduct VAT payments on their inputs. First, despite the low threshold, the number of 17, 000 businesses registered is quite low by international standards. Second, a true VAT, which is supposed to avoid tax cascading and economic distortions, requires a prompt and full refund of the part of the tax on inputs which exceeds the tax due on outputs. This is especially important for exporters. In Georgia the amount of unpaid VAT 27 An interesting account of changes in the value added tax and excise tax legislation is given in IMF, Georgia: Recent Economic Developments and Selected Issues, November 2001, tables 11-2 and 11-3. 28 See the analysis in IMF, Georgia: RecentEconomic Developments and Selected Issues, Nov. 2001, Chapter II. 29 In hindsight, the VAT threshold should have been much higher when the VAT was adopted. 30 Earlier estimates by the MoF put the revenue loss at 23 percent of VA collections. 3 For example, it would help screen shell companies created for the very purpose to evade tax payments. 17 Georgia: Public Expenditure Review November 25, 2002 refunds is large (about 29 million GEL at the end of 2001). Tax inspectors should eliminate the practice of treating VAT as advanced payments against future tax liabilities in order to meet their monthly revenue targets (see section on tax administration below). Third, while many countries have introduced limited exemptions or reduced rates in their VAT laws to reduce regressive elements of the tax, the scope of tax privileges in the Georgian VAT continues to increase, and the country has embarked on the dangerous path to use tax privileges as a way to compensate for administrative or legal deficiencies.32 2.25 Frustration with the distortionary effects of the VAT have caused some policy makers to consider whether to replace the VAT with a sales tax. The objective would be to reduce compliance risks by applying the tax to one stage of the business cycle only. There are serious concerns regarding this idea. VAT despite its relatively low efficiency has become the main revenue source, contributing 45 percent to total gross tax revenues in 2001. Experience in other countries shows that sales taxes have a far lower revenue potential than the VAT, because it does not capture the total value added in the production and distribution phases and their rates normally are not higher than 5 percent--because of administrative difficulties. In addition, compliance risks and compliance management challenges would not be reduced because collection would have to rely on the retail sector which is more difficult to administer. Rather than replacing the VAT with a sales tax, the focus should be to improve VAT administration and actually implement the key principles of the tax, such as an effective refund system for exporters33. As performance of the tax improves, consideration could be given later to lowering the VAT rate. 2.26 Proposed simplified tax. The MoF plans to introduce a simplified tax (of 7 percent on gross income) for those businesses which are below the VAT threshold and have some basic accounting in place. At the same time, MoF plans to modify the presumptive tax for individual enterprises (which raises relatively little revenueS34) by broadening the base. A fixed tax, like the current presumptive tax, would be based on the nature of the business activity, the size of the business and the business location; and will include more types of small businesses than the presumptive tax. Although the Foreign Investor Advisory Service (FIAS) report35) considers a fixed tax to be very complicated, it need not be so. The fixed tax, if well designed, can be transparent and easy to administer tax. It offers no scope for negotiation to taxpayers, does not require detailed bookkeeping, and could reduce the opportunity for corruption and the compliance costs for taxpayers. Parliament rejected the proposed simplified tax because it considers the rate (7 percent) too high and the coverage too narrow. According to some parliamentarians, the scope of the tax should extend to some larger businesses, which clearly reflects the interest of certain business sectors to simplify and reduce taxation. Presumptive taxation based on gross figures should be limited to small businesses with insufficient bookkeeping, while larger businesses should keep books and records and be taxed on a net basis. There is no good reason to extend the scope of the simplified tax to larger tax payers. While the simplification of the tax regime for small businesses is a welcome initiative, its implementation will be costly administratively. The MoF should probably consider introducing these changes once other taxation issues such as VAT and excises have been properly addressed. 2.27 Excise Taxation. Due to its open borders and weak administrative capacity Georgia faces major problems collecting excise taxes. Reduction in excise tax rates has been the preferred method to improve compliance, but with no positive results so far. Georgian excise taxes are actually very low by international standards, and the focus should more be on efficiently enforcing the excise tax regime. 32 See the discussion of exemptions for utilities and the oil and gas sector in IMF, Georgia: Recent Economic Developments and Selected Issues, section II F. 3 Concerns regarding the replacement of VAT by a sales tax also have been raised in the IMF tax policy review 2000. The mission fully agrees with these concerns. 34 In 2000 presumptive tax collection was only GEL5 million or 0.7 percent of total tax revenues. 3 Georgia: Study of Administrative Barriers to Investment, FIAS, December 2001. 18 Georgia: Public Expenditure Review November 25, 2002 Compared to the CIS country average, excise tax revenues in Georgia are low; in 2000 excises in Georgia contributed 1.5 percent to GDP, while the CIS average was 2.1 percent. Looking at neighboring countries, excise revenue performance is much higher in Armenia with 2.5 percent of GDP and somewhat higher in Russia with 1.9 percent of GDP; it is much lower, however, in Azerbaijan with only 0.5 percent of GDP (which is together with Tajikistan the lowest figure in the CIS region). The fact that Georgia has managed to accumulate a surprisingly high level of tax arrears in an area where arrears normally should not build up - according to IMF data the amount of tax arrears on excises was equivalent to 2.7 percent of GDP by beginning of 2000 - shows, however, that excise revenue increases will also depend on the ability and support of the tax administration to collect revenues from major businesses in the oil and cigarette industry. 2.28 Tobacco products. The taxation regime for tobacco products has been modified again starting January 2001, and the overall tax burden has been reduced considerably. In addition, the harmonized taxation of imported and locally produced tobacco has been abolished and a preferential treatment for local products has been introduced. The tax burden on domestically produced tobacco products has been reduced by 50 percent, for imported non-filtered cigarettes total tax burden has been reduced by one third. As predicted by the IMF tax policy review this has not contributed significantly to reducing smuggling and the actual tax collection is less than 50 percent of projected collection: Table 2.6: Projected vs. Actual Excise Tax on Tobacco Products in 2001 (GEL million) Projection Actual Collection Imported 36.0 15.1 Domestic 33.0 15.3 2.29 Even with reduced rates smuggled cigarettes still are sold on the domestic market at far lower prices than officially imported/produced cigarettes. For filtered cigarettes - the preferred tobacco product consumed in Georgia - the lowest retail price of a smuggled pack is around 35 tetri, while a locally produced and taxed pack is not available below 70 tetri. This clearly shows that the rate reduction did not contribute significantly to making officially produced and distributed cigarettes more competitive to smuggled products. A far better enforcement of the tobacco tax regime through intensified controls of the domestic distribution channels would be necessary to reduce possibilities for selling smuggled goods. It is also surprising that the estimated evasion in the case of domestically produced cigarettes is extremely high, although the number of domestic producers is small with three major and seven smaller factories. 2.30 Petroleum products. Considering the sharp price increases mid-2000 for gasoline and diesel the actual tax burden on petroleum products has decreased due to the shift from an ad valorem to a specific excise tax system in April 2000. In addition, the excise on heavy oil distillates has been reduced from 200 to 100 GEL per ton as of March 16, 2001. Revenues from gasoline and diesel taxation have considerably increased in 2001 compared to 2000 figures, which is mainly due to higher VAT collection: 2.31 Despite this 28 percent increase in tax collection from petroleum products, total tax collection in this sector is low and revenue loss through smuggling and evasion is considerable. The Ministry of Revenues has estimated that, based on the number of registered cars and their estimated consumption, potential tax revenues for 2001 were GEL138 million for gasoline and GEL49 million for diesel. Actual collection in 2001 thus amounted to only 52 percent of the estimated revenue potential for gasoline taxation and 39 percent for diesel taxation. 36 See IMF, Recent Economic trends and Selected Issues, Nov. 2001, p. 31. 19 Georgia: Public Expenditure Review November 25,2002 Table 2.7: Total Revenue from Diesel and Gasoline (GEL million) 2000 2001 Gasoline 64.3 71.3 VAT 24.2 33.4 Excise 40.1 37.9 Diesel 5.8 18.9 VAT 1.3 8.3 Excise 4.4 10.6 Total 70.1 90.2 Source: NGB, Treasury Revenue Accounts. 2.32 As already highlighted by the IMF tax policy review 2000, excise taxes on cars and alcohol are extremely low by international standards. As car excise taxes normally are among the easier to collect taxes, there should be less reason to reduce the tax burden as a way to facilitate collection and reduce evasion. The actual excise tax burden on alcohol has been considerably reduced when moving from the fixed tax system to standard excise tax legislation in January 1999. The IMF has recommended an increase in the excise tax rates, which has not been considered so far. 2.33 Income and social tax. The high tax burden of the personal income tax (PIT) and the social security tax provides a strong incentive to evade the payment of these taxes. Although the personal income tax has reasonably progressive rates (from 12 percent to a maximum of 20 percent), the marginal cost of taxes for both employees and employers creates strong incentives not to formalize the labor contract: employees prefer current to future consumption, while employers seek to reduce costs and increase competitiveness. Overall, the taxation rate of the PIT and the social security tax over the net wage is 68 percent. This implies that for each additional GEL paid to worker in net wage, there is 0.68 GEL to be paid in taxes if the contract is formalized. Financing of the pension system continues to suffer from low compliance in the area of social taxes. (For more details see Social Protection Chapter). 2.34 Corporate and income tax exemptions. The Tax Code currently includes a number of exemptions from corporate and personal income tax, which narrow the tax base, increase the discretion of tax inspectors and the potential for corruption. The IMF has recommended to review and abolish many of these exemptions.37 The Ministry of Finance has prepared an amendment to the Code eliminating most of the current exemptions from personal and corporate income tax. This includes in particular the exemptions from CIT for enterprises in mountainous regions, the exemption of profit generated by energy renewable sources, consumer appliances and energy saving equipment. However, this proposal to amend the Code will still have to be finally presented to the Parliament, after it was withdrawn in September 2001. 2.35 Administrative provisions for tax enforcement An essential feature of a good tax code is a clear definition of tax administration procedures and rights and obligations of taxpayers and tax officials. A reasonable balance needs to be defined between the interests of the taxpayer to simplify taxation procedures and reduce administrative burden and the interest of the tax administration to effectively enforce taxation. In Georgia, the possibility to enforce tax collection has been unduly restricted by reducing the powers given to the tax administration in chapter 42 of the tax code to seize and sell delinquent taxpayer property. As a consequence the only remaining enforcement measure, which does not require a court ruling, is the freezing of a taxpayer's bank account. Considering the absence of specialized tax courts and the weakness of the court system in Georgia, this does not provide the tax 3 See IMF tax policy review 2000, p. 17 ff and 21 ff. 20 Georgia: Public Expenditure Review November 25, 2002 administration with sufficient means to improve its compliance management. Enforcement powers of the tax administration should be harmonized with current practice in Organization for Economic Co- operation and Development (OECD) countries. 2.36 Abolishing nuisance taxes. Earlier Bank and IMF reports have recommended the elimination of nuisance taxes because they typically have extremely low revenue yield and are a burden for small businesses. The tax package prepared Ministry of Finance included the elimination of some of these nuisance taxes, (e.g., the tax on economic activities, the resort tax, the hotel tax, the advertisement tax, and the tax on the use of local symbols), but no progress has been made partly because these taxes are assigned to local governments. However, due to their very limited revenue potential, they contribute less than 10 percent of total local revenues. Considering the administrative and compliance costs of these taxes the actual revenue gains might even be negative, these taxes should be eliminated. Tax Administration 2.37 Until April 2002, the main agency responsible for the collection of tax revenues in Georgia was the Ministry of Tax Revenue, which covered domestic taxes, customs and social security contributions. Donors have provided Georgia with significant support to reform its tax administration. Major initiatives have included: the creation of a large taxpayer unit (LTI) at the tax department; the creation of the Ministry of Revenues; the restructuring of the tax department; special programs to improve excise tax collection; hiring of PSI services; and various strategies have been developed to improve customs with donor support. It is difficult to assess the impact of all of these efforts on the country's tax administration capacity. However, there is a feeling among the donor community that results have fallen short of expectations. The government partially concurs with this assessment, but considers that expectations may have been excessively ambitious in the Georgian context. Experience is showing that tax reform in FSU countries is a long-term endeavor and that short-term ambitious goals often remain unmet. It would then seem appropriate to take stock of the lessons learnt from these initiatives to continue efforts to improve taxation practices so that domestic and foreign business can operate in a friendlier environment while contributing the resources needed for the budget. 2.38 The Ministry of Tax Revenue. The Ministry of Tax Revenue (MoTR) was created in 1999 with the objective of facilitating reforms in customs and tax administration and relieve the MoF of the daily oversight over revenue collection. Since its inception, the MoTR was to be a temporary organization and there was never a clear outline of a long-term arrangement, as could have been for instance the creation of a unified revenue authority to be later returned under the oversight of the MoF. Thus, the MoTR ended up implementing a number of isolated initiatives driven by the authorities in charge and/or by external donors. Notably, the MoTR sought to centralize critical functions from specialized departments. For instance, it created a ministerial police unit (the extraordinary legion) responsible both for tax and customs investigation, it sought direct control of the regional customs offices, it created special groups for excises, and it developed an office of internal control. This resulted in an increase the number of people working at the ministry (by 800), without much to show for it. While efforts to advance a reform of the tax administration offices produced partial results, reform at customs has been stagnant. More importantly, no progress has been made in integrating the work of tax and customs administration. 2.39 As of May 2002, the MoTR has been merged with the MoF. This raises questions as to the long term vision and transitional arrangements to avoid undesirable disruptions. In particular, in the Georgian context (a small, open economy with porous borders and substantial transit trade), a strong coordination between taxes and customs is key to proper collection of important taxes like VAT and excises. 21 Georgia: Public Expenditure Review November 25, 2002 2.40 Revenue targets as tool to measure performance. Meeting revenue targets constitutes the most important standard to measure administrative performance and failure to do so can lead to the dismissal of tax officials, including ministers. Revenue targets are set by tax and by level of government (central, local, special funds) and distributed across customs and tax offices. The targets are set annually and broken down quarterly and monthly, often without due consideration of seasonal patterns. 2.41 It appears that the exclusive focus on revenue targets has had unintended adverse consequences. Notably, it has reduced the importance assigned to improving administrative practices, delayed full implementation of self-assessment practices, and slowed down adoption of the functional organization. Moreover, it has led to undesirable practices, such as the call on taxpayer to advance payments so as to meet monthly revenue targets.38 Advanced payments are payments requested from taxpayers against future tax obligations. Revenue targets and advances or negotiated payments introduce two significant distortions that undermine the revenue system40: (a) efforts to meet the revenue targets lead to false reporting on individual tax to such an extent that revenue information by tax is not credible; and (b) advance payments introduce a high level of uncertainty in the revenue flows received by treasury since tax payers are able to reverse advance payments. 2.42 Pressure from foreign investors has led government to forbid the practice at the LTI. However, advances continue to be actively used in other tax offices. The advances have so far been managed through accounts of the tax offices at the National Bank of Georgia (NBG). Treasury reforms underway will centralize for all revenues in treasury. When this reform is implemented, the incentives for advance payments would be seriously curtailed, as the local tax offices could no longer guarantee return to the advance payments to taxpayers. 2.43 A new structure of incentives and performance targets has to be developed to support the adoption of modern administrative practices and improvements in problem areas, such as auditing, enforcement, taxpayer service, and dispute resolution. Such performance criteria, used by many other tax administrations, could include a reduction of tax arrears; the time needed to carry out a tax audit; the time needed to answer questions from taxpayers; and an increase in the number of registered taxpayers. 2.44 Internal control. Revenue administration operates in a highly corrupt environment in which tax favors are bought from tax inspectors. To control such practices the Government has introduced the office of the Inspector General at MoTR. The office has been in operation for a year and has just produced its first annual report. Reportedly, however, this office has limited power to access information and audit performance at the local tax and customs offices. A priority is therefore to revise the legal empowerment of the Inspector General Office so as to provide tool concomitant with its intended mandate. It is also important that the Inspector General Office develops its own vision of adequate performance and thus identify the main risks areas in the revenue institutions as background to its auditing work and as an input to reform and institutional strengthening efforts in these agencies. Also, it is necessary to coordinate the work with Chamber of Control that also oversees the revenue administration. 38 Advances clearly burden the compliant taxpayer. Cooperation with the authorities could perhaps be seen as a way of buying protection from enforcement practices. 39 These practice may lead enterprises to borrow from banks to advance payments. Also, reportedly tax offices have used their own administrative resources to meet the targets. Schemes have been developed where local authorities link budgetary payment priorities to payment of tax arrears. 40 Other consequences include leading enterprises to borrow from banks to advance payments. Also, reportedly tax offices have used their own administrative resources to meet the targets. Schemes have been developed where local authorities link budgetary payment priorities to payment of tax arrears. 22 Georgia: Public Expenditure Review November 25, 2002 Table 2.8: Advances and Arrears by LTI, by Sector (GEL million) as end.1999 as of end 2000 as of end 2001 Advances Arrears Advances Arrears Advances Arrears Oil Businesses 1.2 -2.4 1.4 -0.6 1.4 -1.3 Bread Businesses 0.3 -0.6 0.2 -0.1 0.2 0.0 Industry 0.6 -4.4 0.2 -15.4 0.3 -1 1 Energy Sector 1.7 -61.6 6.9 -81.3 3.1 -134.5 Gas Industry 0.0 -0.4 0.0 -4.0 0.0 -12.3 Light Industry 0.9 -1.2 2.5 -2.7 0.9 -2.8 Gasoline Importers 0.9 0.0 0.5 -0.3 0.9 0.0 Trade 0.6 0.0 0.7 0.0 0.9 0.0 Hotels 0.2 -0.1 0.4 -0.1 0.6 -0.1 Transport Industry 3.4 -0.9 4.1 -1.4 2.6 -0.7 Communications 3.4 -2.4 3.5 -6.0 2.8 -5.1 Banks 1.8 -0.2 3.3 0.0 2.6 0.0 Insurance Companies 0.0 0.0 0.2 0.0 0.2 0.0 Other 1.3 0.0 1.7 -0.6 1.4 -0.2 Total 16.3 -74.2 25.4 -112.6 17.8 -166.5 Source: MoTR. Tax Department 2.45 Georgia has adopted a system of self-assessment which has yet to be fully implemented. Donors have provided support to introduce a number of reforms in the tax department, but much remains to be achieved. Most importantly, there is need to significantly improve the relationship between the tax offices and the taxpayers, to create a level playing field which favors investment and economic activity, to effectively implement self-assessment, and to develop a dedicated taxpayer service function. 2.46 Past reform efforts. The reform of the domestic tax administration in the last two years has centered around (a) reducing the number of staff; (b) testing staff; (c) improving remuneration; (d) streamlining the number of offices and administrative levels; and (e) setting up an information system. Several advances were made. The number of tax offices was reduced, the number of staff was cut from 2700 to 1700, testing of employees was undertaken41, a special fund of 3 percent of revenues was created to improve compensation, and information equipment and software were installed. A functional organization and a true self-assessment have not been implemented yet. Practices such as arrears enforcement and auditing have not improved as expected. Direct contacts between taxpayers and tax officials are still frequent, in particular linked to the practice of cameral auditing of tax returns, which has not been abolished and there is still widespread suspicion of corruption and unequal application of tax laws. 2.47 The taxpayer universe. The number of registered and active taxpayers in Georgia is very low.42 There is also a large discrepancy between registered and active taxpayers. These facts point to the 4 The testing of staff was challenged in court. 42 The total number of registered taxpayers is 288,389; 60,000 of which are legal persons. Of the total, the registry considers 183,619 as active-this gives a figure of around 3.1 per inhabitant. (see Barents ) But the number of 23 Georgia: Public Expenditure Review November 25, 2002 possibility to increase the taxable base by (a) expanding the number of registered taxpayers and (b) examining the list of non-filers among registered taxpayers. Countries have used both techniques successfully. Often, registered taxpayers that appear as inactive continue to operate either under a different address or under a different name. Also, it is not unusual that operating businesses have not registered as taxpayers---this tends to be the case when there is smuggling involved. 2.48 The tax department has enough personnel to clean the existing registry and undertake a selective registration drive. Such efforts in the past reportedly were blocked at the local level, but it is time to resume these efforts again, bearing in mind that a registration drive is not simply about incorporating more taxpayers. It is essential to identify quality taxpayers which either underreport the true size of their business to benefit from registration thresholds or do not register at all. Registration efforts should be a permanent activity when registration thresholds are high. The tax agency can invite taxpayers to register and normalize their relationship with the tax system. Also, it can utilize more forceful methods as complements. Table 2.9: Total Stock of Arrears by Sector, 2001 (GEL million) Sector Principal Penalties & Fines Total Energy 148.4 188.8 337.2 O/w LTI 148.1 Industry 110.4 140.5 250.9 Transport 90.2 114.8 205.0 Construction 56.7 72.1 128.8 Agriculture 61.5 78.5 140.0 Budgetary Organizations 25.7 32.6 58.3 Total 492.9 627.3 1268.3 O/w LTI 168.4 2.49 Tax arrears. Prompt actions to recover tax arrears is the best way to establish credibility of the tax administration. Increasing tax arrears is a sign that the tax system is not functioning adequately: either the legal framework or administration. The total stock of tax arrears (including contributions to state funds) at the end of 2001 stood at GEL394 million (GEL1,024.5 million including penalties), equivalent to one year's total revenue collection. It is unclear how much of these arrears can be recovered since dome of the arrears correspond to enterprises that are no longer in operation. Such a staggering stock of arrears plus fines must have serious effects on revenue administration and the operation of the economy. Note also that the amount of penalties exceeds by around 50 percent the stock of arrears. In 2001, the stock of arrears grew by around GEL92 million-roughly 10 percent of national revenue collection and 30 percent growth over the initial stock of arrears. Penalties and fines grew by an even larger amount- GEL70 million, or 47 percent growth over the initial stock of penalties and fines. This suggests that penalty and fine regime needs realignment. As to economic sectors, the main stock of arrears is with energy (30 percent), most of it LTI.43 Reasons for the rapid increase in arrears are: (a) the implementation of accrual methods in an economy with poor payments discipline; and (b) poor enforcement practices, including an unclear role of courts in the collection process. effective taxpayers is much lower. The number of profit taxpayers stand around 20,000, of which between 13,000 and 17,000 also pay VAT and 367 excises. The number of active individual taxpayers (other than through withholding and land tax) stands roughly at around 20,000. This gives a total of around 40,000 registered taxpayers, around 1 percent of the total population, at the level of Nepal and half of India. Given that the MoR employs around 3000 persons, there is a ratio of around 13 taxpayers per employee, a very low ratio. If only tax department employees are included, the ratio is 30, still low. 43 See Table 2.9 for details on other sectors. 24 Georgia: Public Expenditure Review November 25, 2002 2.50 Auditing taxpayers. Further reforms in the audit area will be needed to improve compliance management and better detect cases of tax evasion. The tax administration continues to rely to a large extent on cameral audits, even at the LTI, at the time of filing a tax return, instead of on a selective and well targeted audit program. The main type of audit used is the comprehensive audit of businesses, which is not based on a clear risk assessment. By contrast, modern tax administrations use a variety of audit types to react to specific risks, control requirements, and to increase the total number of audits. In addition to comprehensive audits other tax administrations typically conduct registration checks for new enterprises, VAT refund audits, issue-oriented audits, and single tax audits. Audits are not only meant to detect underreporting of income, but also to identify non-registered taxpayers. 2.51 A specific problem for the audit activities of the tax administration is the provision of Article 217 para. 3 of the Tax Code, which stipulates that a routine (targeted) tax audit of a business can be carried out only once a year. While in principle this provision makes perfect sense, it is interpreted by the LTI as prohibiting the LTI to audit a taxpayer when the same taxpayer has been already audited by the Chamber of Control, even though the Chamber does not focus its audit on tax issues. A clarification of the meaning of the provision and a better exchange of information between the tax administration and the Chamber of Control would be useful. 2.52 The LTI. The creation of an office specialized in large taxpayers was an early initiative to improve tax administration. In 2001 the office administered on average 334 taxpayers and accounted for 38 percent of the revenues collected by the tax department or 27 percent of total revenue collection. The LTI coverage is low for a country with weak revenue administration. Moreover, the LTI is far from being the center of excellence initially envisioned. Although there are rules as to who should be a large taxpayer; reportedly, these rules are not strictly implemented. Local offices continue to be reluctant to let their largest taxpayers go. Recent efforts to increase the coverage of the LTI have led to the removal of 40 taxpayers which did not qualify as large and the addition of 100 that do. The 40 enterprises returned had L62 million in arrears and no assets. Still, the roaster of large taxpayers needs further cleaning. The goal now is for the LTI to cover at least 50 percent of the revenues collected by the tax department. 2.53 At the end of 2001, the LTI office had GEL168 millions in arrears or 43 percent of the total national stock. GEL134.5 million of arrears (or 80 percent of the LTI total) are held by the power sector. Effective cash payment for electricity are very low; thus, as taxes due are calculated on an accrual basis, arrears necessarily emerge. Recently parliament approved that taxes due in the power sector be calculated on cash and not accrual. This is an unfortunate measure as mixing cash and accrual methods provide ample opportunity for evasion and make tax administration even more difficult. This measure artificially reduces arrears and lowers compliance from taxpayers and enforcement efforts from the agency. The priority should be tot improve financial discipline, rather than to accommodate to the lack of it. 2.54 The role of the court system. Courts have a substantial role in collecting taxes, as administrative enforcement rights have been curtailed, but they are ineffective. The number of tax cases in the court system is on the increase-more than 3300 cases are pending. Every month the judiciary manages to resolve 60 cases, while 100 new suits are being filed. Moreover the board responsible for collecting tax delinquencies once the government has won its case in court is ineffective. Only about 15 percent of the amounts won in court since 1998 have been collected. The solution is to switch to administrative tax enforcement, without court involvement as is the practice in Western countries. Clearly, an adequate legal infrastructure would have to be set-up, with due process and use of liens rather than encumbering productive assets. 25 Georgia: Public Expenditure Review November 25, 2002 Table 2.10: Total Tax Collected by LTI, 2001 (GEL million) Number Total Total of Businesses VAT Excise CIT Property PIT Other SSF EF Taxes Paid Arrears Taxpayers 19 Oil Business 4.0 0.0 2.4 0.9 3.6 2.8 3.8 0.2 18.3 1.4 -1.3 Bread 7 Business 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.2 0.0 Food Products 33 Business 4.6 13.9 0.2 0.9 1.4 1.3 1.2 0.0 23.8 1.0 -1.9 20 Industry 6.6 0.0 0.4 0.5 4.5 1.0 3.0 0.0 16.1 0.3 -1.1 20 Energy Sector 24.9 0.0 1.6 1.1 8.6 2.9 6.1 0.2 46.4 3.1 -134.5 Light 41 Industry 4.8 0.0 1.1 0.8 3.4 0.7 1.1 0.0 12.1 0.9 -2.8 22 Trade 09 0.0 2.8 0.1 0.7 1.2 0.5 0.0 6.3 0.9 0.0 5 Hotels 0.4 0.0 0.0 1.0 0.7 1.7 0.8 0.0 4.7 0.6 -0.1 53 Transport 10.4 0.0 10.9 2.8 7.4 2.3 11.8 0.4 48.0 2.6 -0.7 35 Banks 2.2 0.0 3.7 0.3 5.2 1.3 5.0 0.2 18.6 2.6 0.0 Insurance 3 Companies 0.0 0.0 0.1 0.0 0.5 0.2 0.2 0.0 1.2 0.2 0.0 24 Other 2.4 0.0 0.9 0.9 3.0 0.4 1.7 0.1 9.6 1.4 -0.2 17 WB PlUs 0.8 0.0 0.0 0.0 1.0 1.1 0.6 0.0 3.7 0.1 -8.5 340.0 Total 84.3 13.9 26.6 11.1 44.5 25.0 41.5 1.5 254.3 18.9 -168.4 2.55 What are the appeals options for the taxpayers? The appeals process is regulated by Article 256 of the Tax Code. The Code provides for a three-level administrative disputes resolution system, before a taxpayer can appeal an administrative decision in Court. The law sets a timeframe of 20 days for each administrative layer to notify the taxpayer of its decision. In practice the disputes resolution mechanism at the administrative level seems much slower. Efficiency of the administrative appeals system could be increased by reducing the number of review layers from three to two. In addition, special training of tax administration staff responsible for the settlement of disputes will be necessary to improve the quality and neutrality of the administrative appeals process. 2.56 The penalty structure. As discussed above, a substantial part of overall tax arrears is composed of penalties. This is partly due to the high level of penalties in the Tax Code for the under-declaration of a tax liability, which is 100 percent of the under-declared amount for under-declaration of more than GEL10,000 (Article 254 Code). The penalty system results in high penalty charges even in cases where taxpayers do not under-declare intentionally, but misinterpret tax laws and regulations. In other countries the amount of the penalty does not primarily depend on the amount of tax unpaid, but distinguishes between intentional and non-intentional non-payment of taxes. The World Tax Code prepared by Harvard University, e.g., proposes a penalty of up to 25 percent of the tax reportable for simple failure to pay the taxes due, which is increased to up to 45 percent of the underpayment in case the underpayment results from negligence or intentional disregard of rules and regulations or substantial understatement of income, and up to 75 percent of the underpayment in case of tax fraud.44 The Russian Tax Code provides for a penalty between 20 percent of the unpaid amount of tax in case of unintentional non-payment and 40 4 See Hussey/Lubick, Basic World Tax Code and Commentary, Sec. 563. 26 Georgia: Public Expenditure Review November 25, 2002 percent in case of intentional non-payment (Article 122 Tax Code RF). Georgia therefore should consider substantially reducing the penalty for non-payment of tax due except for cases of tax fraud. Customs Administration 2.57 The performance of Customs in Georgia is affected both by a difficult external environment and a internal reluctance to reform. Both factors underlie a certain pessimism about the possibility of developing professional customs services in Georgia. During last two years reform strategies were designed but hardly implemented. Moreover, the MoTR, in an effort to improve revenue performance, transferred key functions, such as anti-smuggling, to headquarters and established dual subordination (to headquarters and to the Head of Customs) for the regional offices. Today, the government engages once again, with donor cooperation, an effort to reform customs. Proposed reforms bear the imprint of past recommendations, showing that the areas that need work have been identified. Once again, what is at stake is the willingness and ability to undertake and complete. Customs reform continues to be necessary to improve revenue collection as well as to protect domestic producers from the unfair competition coming from underreporting and outright smuggling. 2.58 PSI. The introduction of PSI services in 1999 was part of the initial effort to improve customs performance. The original design of the contract was considered exemplary as it included support for modernization and for anti smuggling. Application of PSI has not been smooth. To begin, the original PSI design included a minimum threshold of U$5,000, which allowed importers to evade by breaking down imports in smaller lots. While the minimum threshold eventually was reduced to U$2,500, at the same time an erosion of the PSI responsibilities took place by exempting selected goods, such as cars, from PSI. Beginning of 2002 the PSI contract covered just around 35 percent of all imports. Moreover, there has been limited reconciliation between PSI assessments and customs effective evaluation. PSI reference prices are not mandatory. Customs officials continue to argue that PSI is ineffective and that they are equipped to carry out the task by themselves and the contract with ITS, which expired in April 2002, has not been extended. However, despite shortcomings, PSI has developed expertise and a data base of reference prices. Transfer of this expertise has not been properly carried out. Hence, it may be premature to cancel PSI service. Rather, an effort should be made to rein in the customs authority by developing proper governance mechanisms that assure cooperation with the MoTR and MoF. A clear MoF interest in the PSI scheme and the proper reconciliation of data would be a key condition for a new PSI scheme, however. 2.59 Entry points. Georgia does not clear goods at border entry points. Goods are taken to warehouses where they are cleared after payment. The entry points do not have adequate facilities, often operating in makeshift arrangements and without proper communication with the warehouses and headquarters. The exception is the border with Turkey where there was already a secure frontier during soviet times. Because of poor installations at customs and the lack of central control, not all of the traffic coming into the country is reported.45 Moreover, due to a lack of rotation of personnel, close relations with local interests impairs proper customs work. The development of proper border points and the cooperation with border guards is thus a priority for the improvement of customs performance. These improvements require budgetary support and help from donors. Such designs will have to accommodate the facts that some borders are highly porous or even contested. But in any case, it is reported that border posts of neighboring countries are better equipped and manned, even if subject to similar uncertainties as in Georgia. Eventually, border posts would have to play a greater role in controlling incoming trade and thus relieving pressure from clearing points. 45 Today, 86 percents of imports entry Georgia through four custom points: Poti, Adjara, the Tblisi airport, and Gardabani. According to customs authorities, inflows through the smaller entry points are severely underreported. 27 Georgia: Public Expenditure Review November 25, 2002 Table 2.11: Officially Registered Major Transit Goods as Compared to Import (Quantities in tons), 2001 Transit Import Transit/Import Crude oil 10310211.3 2,307.7 4,467.7 Oil Products 3994989.3 376,979.6 10.6 Wheat 299402.2 82,384.4 3.6 Sugar 214406.9 110,845.9 1.9 Aluminum 152411.2 0.0 Aluminum scrap 146676.9 0.0 Soya 120415.7 22.2 5,433.9 Ferrous Scrap metal 70246.2 704.3 99.7 Rice 51486.2 530.5 97.0 Meat and meat products 51131.3 14,799.1 3.5 Flour 46236.4 80,715.3 0.6 Copper scrap 31627.5 0.0 Ethyl Alcohol 28416.7 4,767.3 6.0 Oil lubricants 26888.4 2,676.4 10.0 Cotton 25080.9 0.0 Soya oil 18647.2 17.5 1,068.1 Beer 14621.0 181.7 80.5 Minerals 13403.4 106.1 126.3 Chemicals 13350.0 2,302.2 5.8 Maize (corn). 13203.4 354.8 37.2 Tomatoes 12056.2 938.4 12.8 Sunflower oil 11949.5 4,699.1 2.5 Pipes 11775.7 902.1 13.1 Fertilizers 11554.0 9.9 1,168.6 Humanitarian Aid 10618.5 89,460.7 0.1 Tobacco 10588.2 1,294.8 8.2 Barley 8000.1 11.2 716.9 Coffee 6936.8 1,323.7 5.2 Aluminum 6780.5 76.9 88.1 Source: Georgia Customs Department. 2.60 Controlling transit trade. The geographic and political situation in the area determines trade flows through Georgia. It is basically the only trade route to the West for Azerbaijan and Armenia. It is the conduit east by land for Turkish products. It is a supply route for Russia and western products make their route to central Asia through Georgia. Even if relations between Armenia and Azerbaijan were normalized, the role of Georgia as trade route will continue to be important. There is also an interment of goods for processing and re-exporting. But the country is hardly prepared to deal with transit trade. Certainly there is legislation establishing procedures to handle transit trade, but observations by external experts repeatedly conclude that customs agency is not prepared professionally to develop and apply them. Indeed, some experts conclude that transit trade is the single biggest risk in the country. Table 2.11 shows ratios of transit trade to domestic imports for some basic categories of goods. It can be seen that these ratios are very high for staples, which are difficult to monitor and thus easy to enter into 28 Georgia: Public Expenditure Review November 25, 2002 the country. Also, an examination of entry and clearing points show how goods travel within the country before they are cleared. The country does not have the adequate infrastructure to track such movements. 2.61 Clearing of Goods. The customs warehouses are also poorly equipped and staffed.46 Georgia lacks an information system that properly records imports and exports. Thus it lacks an effective means of control and management. Also, despite a basic legal framework, there has never been a proper development of customs procedures and much less a willingness to implement whichever exist. Lacking proper established procedures and with poor implementation is then practically impossible to establish a paper trail that will serve to record and control trade flows. Moreover, there are no appropriate means of communication between entry points and warehouses and between both and district offices and headquarters. Table 2.12: Mirror Statistics; Trade with 10 Major Partner Countries, 2001 Exports I Imports SDS Date Partner Data SDS Data Partner Data Russia 73.5 138.8 91.3 164.1 Turkey 68.7 112.0 105.0 146.2 USA 9.5 27.6 27.8 118.8 Germany 7.9 8.3 69.1 75.9 UK 22.9 24.6 25.1 27.6 Azerbaijan 10.6 11.1 73.1 91.7 Switzerland 15.6 16.2 14.0 15.5 Ukraine 11.7 12.3 49.5 53.8 Italy 8.6 8.9 25.5 28.8 Turkmenistan 28.9 30.3 17.6 19.7 Total 10 Partners 257.8 390.2 498.2 742.3 Other Countries 62.2 62.8 185.9 185.9 Total Exports 320.0 453.0 684.1 928.2 Source Data from State Department of Statistics (SDS) and United Nations (UN) Trade Partner database 2.62 An examination of recorded trade flows in Georgia and in partner countries shows that overall Georgia statistics underreport both exports and imports, except in the case of trade with Russia (Table 2.12). A variety of incentives would drive this behavior. Underreporting of exports represents illegal exports (wood) or transfers of funds abroad. Underreporting of imports means either the failure of customs officials to outright not recording the imports or under-valuation. In any case, it is likely that in Georgia smuggling comes directly through the customs and does not have to make its way through tortuous routes. 2.63 The tariff structure and exempt goods. The current tariff structure is simple and has three levels: 0, 5 and 12 percent. However, because of exemptions and special privileges only 30 percent of all imports are fully taxed. The rest were either partially taxed (25 percent) or not taxed at all (39 percent.) Reasons for non-taxation include: privileges granted to foreign investment under earlier regimes, grants, and humanitarian aid. Other exemptions cover 16 percent of imports. A large number of small exemptions covering such a significant percentage of imports make revenue administration difficult. 2.64 Human Resources and Management practices. The rotation of high level management at Customs is extraordinary. Over the last two years seven individuals have held the position of chairman of customs. Given the propensity for each new chairman to outline its own strategy and disregard previous 46 Over 80 percent of goods today are cleared at Tbilisi, the railways, Adjara and the airport. 29 Georgia: Public Expenditure Review November 25,2002 guidelines, it is not surprising that so little has been achieved during this period, despite the constant emphasis on reform. In the meantime, customs has continued to operate as in the past. The agency currently employs 1400 people, 400 at headquarters. The staff hardly receives any specialized training and simple things like wearing a uniform become insurmountable tasks. The lack of proper information management and communications makes very difficult for headquarters to coordinate with field offices. The low level of remuneration, in an environment with limited control, procedures and oversight, provides ample opportunities for malfeasance. 2.65 A New Customs Reform. The government is in the process of preparing yet a new customs reform strategy, under the leadership its recently appointed Head, which is ambitious and to be implemented in short period of time. The strategy is based on the following pillars (a) improving human resources, including through reductions in the number of customs officials to 1000,47 testing for qualifications, possibly hiring new agents, and developing comprehensive training pro rams; (b) upgrading legislation, with support from the EU, to bring it in line with European standards, § (c) upgrading the information systems, through the full implementation of Automated System for Customs Data (ASYCUDA) (d) transferring collection to the banking sector,49 (e) upgrading checkpoints and customs offices, (f) international integration. A complementary task would have to be the development of detailed, operational internal procedures. The upgrading of customs points will require procuring the needed resources either from the budget or from donors. Georgia will have to take full advantage of opportunities provided by its World Trade Organization (WTO) membership and also establish representation in the countries of its main trade partners. 2.66 The customs department has also outlined a promising plan to control imports of fuel products. Smuggling of fuel products has been controversial in Georgia given the limited number of importers and the difficulties of smuggling bulk quantities other than through the railway. The Customs office plan considers cooperation with Azerbaijan to monitor imports from the point of destination and then, within the country, continuous monitoring, through constant communication with headquarters, as the fuel is delivered. The plan also considers the development of a detailed set of procedures for customs officials. 2.67 Effective anti-smuggling is needed to develop a professional customs service in Georgia. As the plan to dismantle the MoTR goes ahead, developing anti-smuggling capacity at customs would need to go beyond transferring personnel of the extraordinary legion, now at headquarters, to customs. It would require the creation of a small elite force, protected from special interests and with clear accountability norms. The force would have to cooperate with the border guard and with the national police. In a small open economy like Georgia, it is extremely important that the intelligence and enforcement services of customs and taxes operate in coordination.50 Unfortunately efforts to do so under the umbrella of the MoTR did not yield the desired results. But the task continues to be needed nonetheless. At the present time it is unclear how in the plans of government or the donors such cooperation would come about. 47 Some would argue that 800 would be enough. 48 The current legal framework for customs is weak. Also, it does not provide customs officials with adequate powers to carry out its tasks. Reportedly, customs authorities do not have the authority to inspect transit airplanes. Also, procedures do not instruct officials to inspect containers classified as empty. 49 The removal of direct cash collections by customs will favor transparency and integrity but will require making banking services available at entry and customs points. so When goods can easily enter the country because of porous borders or weak customs, it is necessary to undertake enforcement in the domestic markets. To be effective, close cooperation between customs and taxes is necessary. 30 Georgia: Public Expenditure Review November 25, 2002 Key Reform Areas 2.68 The public perceFtion of the quality and fairness of tax and customs 'administration in Georgia is generally very negative.5 Substantial and visible improvements on the ground will be needed to begin dispelling this perception. This also requires a political commitment to abolish practices which protect and support special interests of taxpayer groups by introducing special exemptions in the tax legislation, thus eroding the tax base, or/and executing pressure on the revenue authorities to grant favorable treatment to specific taxpayers. It will also be necessary to reduce the incentives for revenue officials to participate in corrupt practices and to develop the necessary control mechanisms to detect and punish such behavior. 2.69 Efforts to reduce capture and corruption are to be complemented by long-term strategies to improve the tax policy design and build revenue administration capacity. Tax policy reforms should focus on overall policy design issues instead of exclusively discussing the level of tax rates. Eventual tax rate reductions will only be feasible if accompanied by broadening of the tax base and administrative improvements. Key to improving administration is the effective implementation of self-assessment and the fair and equitable treatment of all taxpayers. Short-term Reform Priorities 2.70 While substantial capacity building in tax and customs requires long-term strategies, there are a number of essential short-term reform initiatives which should be launched immediately to improve revenue performance and reduce tax-related distortions. 2.71 Tax policy. The main challenge is to stabilize the tax policy framework, and avoid ad-hoc short- term policy measures. In general, the revenue impact of tax exemptions should be properly analyzed, and no further exemptions/tax reductions should be introduced without such analysis is explicitly presented in Parliament. Any tax policy changes should be taken in the context of the annual budget. It also recommended that the 2001 tax package prepared by MoF be submitted to Parliament, including key elements such as raising the VAT threshold to GEL 100,000 (with a limit on voluntary registration) and elimination of the nuisance taxes. 2.72 Tax administration. A number of actions could be take to support long-term reform efforts: * Discontinue the practice of soliciting advanced payments to meet revenue targets and design a new performance measurement system with appropriate indicators, supplemented by special incentives to improve revenue administration practices; * Centralize revenue accounts in the Treasury and make payments on "a first come first served basis"; * Begin implementation of special program to control imports through the railway system, especially of petroleum products; * Increase coverage of LTI and focus on improving LTI performance. 51 This also applies to foreign investors, as numerous critical articles on taxation in Georgia published by the AmCham newspaper demonstrate. 31 Georgia: Public Expenditure Review November 25, 2002 A Longer-term Agenda 2.73 A more comprehensive reform program for the medium and long-term reform of the Georgian revenue system will then need to consider the issues below. 2.74 Broadening the base and lowering tax rates. While some taxes may be relatively high and may promote non-compliance - especially the general VAT rate of 20 percent and the combined tax burden on labor - taxes from excisable products are not fully exploited. A longer term tax policy reform objective for a poor economy like Georgia should be to reduce the tax burden on consumption and labor. However, this can only be achieved by (a) broadening the tax base of VAT and profit/income taxes; (b) increasing collection by improving the efficiency and effectiveness of tax and customs administration. 2.75 Past experience with tax policy reform in Georgia has shown that mere tax rate reductions without corresponding improvements in enforcement and compliance management will not contribute to increasing tax compliance. Rate reductions therefore do seem not feasible as long as revenue losses from the rate reduction cannot be compensated by a broader tax base and a better enforcement. Tax policy reform in Georgia therefore will need to mirror experience with tax reform in OECD countries in the last two decades, where rate reductions (mainly in the area of direct taxation) were achieved through base broadening and improving tax administration. 2.76 VAT Reform. The VAT should not be replaced by a sales tax. Rather, the VAT as the mainstay of the revenue system in Georgia should be strengthened. The VAT design appears buoyant, albeit if its base has been eroded by exemptions, privileges, and fraudulent practices involving both tax inspectors and tax payers. Increasing the threshold and reducing the number of taxpayers will help improve its administration. Potential decreases in revenues can be compensated by reducing exemptions. The implementation of a true VAT necessarily has to ensure refunds for exporters and zero rated goods. On the administrative side, it is important to advance existing initiatives to improve cross-checking, monitor registration, and regulate invoices. 2.77 Simplfication. The introduction of a simplified tax and the elimination of nuisance taxes, as proposed by Government, should facilitate administration and reduce the administrative burden on small businesses. In Georgia, nuisance taxes are local taxes generating little revenue. The best would be to eliminate these taxes and find alternative (more solid) own revenue sources for local governments, such as the land and property tax, which are not currently collected centrally (see Chapter IV on Inter- governmental Fiscal Relations). 2.78 The elimination of the Ministry of Revenue. The transfer of responsibilities of the MoTR to the MoF requires a clear and precise blueprint of medium term design to avoid the ad hoc institutional build up when the MoTR was created. It may still remain the best option to think in terms of a Revenue Authority in the long-term. 2.79 Addressing corruption. The creation of an Inspector General Office (IGO) has been a step in the right direction. The work of the IGO should be provided with the appropriate legal and technical instruments to carry out its function. Technically, it is important to develop accurate assessments of where the opportunities for corruption arise, through an analysis of the business process and the use of indirect statistical methods. Legally, the IGO must have the powers to access relevant information from tax-offices and taxpayers. It should also be clear to the agencies and to the public how the recommendations of the IGO would be implemented. The role of the Chamber of Control in evaluating tax performance could help as the IGO builds up strength. The government needs to consider if the current profile of corruption requires development of legal instruments, other than those dealing with corrupt practices in the public sector, to address corruption in the revenue agencies. 32 Georgia: Public Expenditure Review November 25, 2002 2.80 Making effective a functional organization. The centerpiece of a modem approach to tax administration is self-assessment. To properly implement self-assessment requires changing the culture, both in government and society, of how taxes are calculated and collected. The direct contact between officials and taxpayers should be reduced, with emphasis shifted to taxpayer services, quick attention to arrears enforcement and selective but effective auditing. Internal control and anti-corruption services should help keep taxpayers and officials honest. Appeals mechanisms should serve to protect taxpayers rights. The extensive advise provided by donors has already acquainted the authorities with the principles of self-assessment. However, the reform agenda continues to be broad and will take time to implement. The following issues would seem to require special attention: * Registration. The current registry needs to be reviewed with emphasis on quality taxpayers that may be hiding as small or not even registered. * Arrears enforcement. The current stock of arrears plus fines and penalties is large but a large portion of it might not be collectable. It is necessary to make a realistic assessment of what can be collected from the stock and develop timely methods to prevent new arrears from aging, setting clear priorities. * Auditing. There should be a sustained effort to build the quality of auditing. Special attention initially could be placed on critical aspects of the VAT such as VAT refunds, cross-checking of credits and fake invoices. Important to good auditing is the development of risks profiles to guide selection and improve effectiveness. Greater information management capacities available now have to be used to develop such profiles. * LTI. The LTI in not a center of excellence. Efforts to update the roaster of large taxpayers and to reach a coverage of at least 50 percent of the revenues collected by the tax agency are worthwhile, but they have to be sustained. The LTI has to take a more proactive attitude to performance and reform and it is good place to begin developing new incentive mechanisms away from simple revenue targeting. 2.81 Customs Reform. The critical areas of customs reform have been already identified and articulated in previous and current plans. Now as in the past, the envisaged time frame for implementation is not realistic and could lead to frustration and fatigue. It is advisable that the customs authorities focus on obtaining some early wins in a few of the many areas that require attention. The current proposal to focus on a comprehensive program to control fuel imports through the railways is a good example. This and other efforts to produce early wins, have to be complemented with sustained efforts to advance modernization of clearing points, the development of communications and information systems, human resources and training, and the development of clear procedures. that provide a track record to monitor performance. The recent cancellation of PSI services is worrisome and will test the ability of the customs agency to assess import prices. Contracting of PSI may have to be reconsidered. The idea of reducing personnel and testing functionaries is good, but as the experience in taxes show, it needs to be complemented with development of procedures, training, and an appropriate incentive and accountability structure. Otherwise the reform may come to naught. 33 Georgia: Public Expenditure Review November 25, 2002 Chapter 3. Budget Management Issues Introduction 3.1 A key test of the effectiveness of any budgetary system is its success in translating Government policy into budgets, expenditure and service delivery. The Georgian budgetary system shows significant weaknesses according to this test, in spite of some recent improvements, as will be documented in greater detail in specific sectors in later chapters. The purpose of this chapter is, first, to explore some of the features of this weakness, beginning from an assessment of how closely actual expenditure follows from allocations approved in the budget. Subsequent sections examine the sources of the problem in particular practices of budget formulation and execution. The agenda of potential reforms required in each of these areas is large and demanding: the final section of the chapter suggests how the complexity of this agenda may be broken into manageable steps, recognizing the sequential nature of the required reforms. Credibility of the Budget 3.2 At its core a budget should be able to translate policies of the government into public spending to carry out specific functions. One important indicator of the credibility of the budget is the variance between the initial budget amounts and the actual spending at year-end. The greater the variation, the less credible is the budget and the less effort line ministries are likely to put into the budget preparation process. 3.3 There may be various reasons for the discrepancies. First, the budget preparation may be flawed such that the genuine priorities of the government never get fully reflected in the budget to begin with. Second, the resources available to carry out new policies may have been exaggerated, so that some priorities are eventually squeezed out by others. And third, budget execution procedures may be such that priorities are easily changed during the course of the year in ways that were not initially intended. In Georgia all three of these factors are present to a degree. 3.4 There are indeed significant deviations between the initial voted budget and the actual expenditures in Georgia. Because the budget is revised throughout the year, the deviations between expenditures and "budget" gradually narrow during the year. Of course, it is the initial budget that is the basis of planning for line ministries, and it is this that would have had the most extensive debate and review from Parliament. Because of the frequently changing political and economic landscape in Georgia, it is hard to make generalizations about budgetary variation across the years, i.e., patterns in one year may not carry through into the next. With that caveat, it is useful to see the magnitude of the variations, and the types of spending that are favored most or suffer most during budget execution. 3.5 Table 3.1 below compares budget and actual expenditure by functional classification for the years 2000 and 2001. For 2001, actual central government expenditures for the year were about 18 percent below the original budget amount. This still represents some improvement over the prior year, when the budget was revised downward by 25 percent in July alone and actual spending fell another 16 percent below that. The 2001 data shows that the reductions were broadly comparable across the different functions of government. There were a few exceptions: two sectors-Industry/Construction and Environment/Other Economic Services dropped significantly (64 percent and 63 percent), while two others, Transport and Energy, increased dramatically (possibly due to unplanned asset sales). Among the remaining functions, Housing/Community Services and Agriculture were 80 percent and 77 percent of budget, respectively. But spending for Social Insurance/Welfare and Defense were much closer to budget (100 percent and 93 percent). 34 Georgia: Public Expenditure Review November 25, 2002 Table 3.1: Comparison of 2000/2001 Budget versus Actual Expenditures 2000 2001 Expenditure Expenditure As% As% Original Actual Original Original Actual Original Function Budget Expenditure Budget Budget Expenditure Budget General Government Services" 213,7029 101,579.3 48 195,155 8 174,3462 89 Defence 49,1333 28,8985 59 39,8478 37,110 1 93 Public Order & Security 95,558.2 63,851.1 67 87,5257 80,3487 92 Education 42,0540 26,888.0 64 35,1067 31,6666 90 Health 34,0507 20,4525 60 38,064.9 33,5196 88 Social Insurance & Social Welfare 136,3400 109,313 5 80 110,347.1 110,7787 100 Housing & Community Services 4,851.0 4,203.3 87 4,791.4 3,8222 80 Culture, Religion & Sports 25,131.6 21,1990 84 22,387 1 19,9819 89 Fuel and Energy 27,321.1 25,306 2 93 3,513.7 8,970 4 255 Agriculture, Forestry, Fisheries 26,6269 12,312 8 46 19,6414 15,105 6 77 & Hunting Industry & Construction 1,016.3 331.1 33 9067 583 9 64 Transport & Communications 4,014.1 9,666 0 241 914.7 2,2419 245 Environment & Other Economic 3,7179 1,701 1 46 3,667.0 2,2953 63 Services Not Classified Elsewherev 410,682.0 255,671 0 62 292,0050 177,9486 61 Total Central Budget 1,074,2000 681,373 5 63 853,875 0 698,7197 82 1/ Donor-financed projects are classified under General Government Services regardless of the sector that benefits 2/ Principal payments on foreign and domestic debt, as well as interest payments are included in Not Classified Elsewhere for 2000 3.6 Biases in spending toward particular economic classifications may be one of the factors (besides shifting policy priorities) that contribute to the changes in the functional allocation of expenditure. As Table 3.2 shows, some aggregate line items tend to be executed at a much higher rate than others. For example, expenditures on wages/salaries and for food were both executed at or above 100 percent of budget. In contrast, the line items for office expenditures were executed at only 40 percent on average, and for inventory/uniforms at only 52 percent. Likewise, resources devoted to utilities were 19 percent below budget. Subsidies and current transfers were 8 percent over budget. 3.7 Aggregate data by economic classification may also hide biases that may occur in spending across ministries. Table 3.3 examines the budget execution rate across several ministries for two illustrative line items - office expenditures and utilities. The table shows that while both items experienced cuts, these cuts fell harder on some ministries than others. For example, with office expenditures Ministry of Environment spent only 7 percent of budget, but Foreign Affairs spent 90 percent. Despite the wide variability though, there is no discernable pattern in which ministries suffered most. Ministries that one might suspect to be more politically powerful experienced reductions in spending much the same as others. 35 Georgia: Public Expenditure Review November 25, 2002 Table 3: Execution Rate of the 2001 Central Budget by Economic Classification kxeection of OrIgImal 04get Rwised Budet Aotuat Expenditvr Original Sudget TOTAL EXPENDITURE 853,875,000 751,346,600 698,719,725 82% CURRENT EXPENDITURE 731,046,500 645,956,455 578,146,242 79% ages and salaries 79,796,600 87,456,205 81,636,184 102% Salary 66,651,700 72,447,453 68,453,878 103% Bonus 754,700 976,600 944,320 125% Salary supplement 10,224,600 11,830,256 10,208,764 100% Emergency assistance 2,165,600 2,201,896 2,029,222 94% Employers contribution 19,204,400 19,735,410 17,912,165 93% State Social Security Fund 17,181,731 17,637,069 16,032,266 93% State Medical Insurance Company 1,533,233 1,594,151 1,431,723 93% Employment Fund 489,436 504,191 448,176 92% Business trips 4,944,700 6,463,194 4,736,520 96% Other goods and services 207,221,800 107,705,606 90,568,780 44% o w. Office expenditures 12,905,900 8,821,460 5,112,210 40% o w Utilities 14,021,100 12,824,127 11,366,965 81% o w Transport and equipment maintenance 13,417,400 11,723,406 10,543,289 79% o w Food 19,151,700 19,318,825 19,110,466 100% o.w Medicines 762,700 550,200 182,607 24% o w. Inventory and uniforms 4,280,200 2,774,087 2,215,706 52% o.w. Other expenditures 142,682,800 51,693,501 42,037,537 29% Interest payments 156,000,000 135,900,000 114,987,623 74% Subsidies and current transfers 218,606,500 248,946,733 235,175,469 108% o w Subsidies 53,012,500 52,801,800 48,763,898 92% o w Current transfers 165,594,000 196,144,933 186,411,571 113% Targeted programs 45,272,500 39,749,306 33,129,500 73% CAPITAL EXPENDITURE 2,519,200 3,517,345 2,854,459 113% Purchase of machinery and equipment 869,800 1,234,967 903,005 104% Purchase, construction and reconstruction of buildings 750,000 357,600 158,121 21% Major repairs of buildings, machinery and equipment 899,400 1,924,778 1,793,333 199% PROJECTS FINANCED BY DONORS 120,309,300 101,872,800 117,719,025 98% 3.8 Accumulation of arrears has also been one of the symptoms in Georgia of an ineffective budget system. Although reporting by the Georgian Treasury may show actual expenditures at certain level, it fails to capture the commitments that were entered into but not yet paid. The data on the true magnitude of arrears is rather imprecise. However, estimates by the IMF put the total stock of arrears by end-2000 at GEL320 million, of which the two largest chunks were for wages/salaries (GEL 1l8 million) and the Social Security Fund (GEL83 million). (See Table 3.4.) Although some within MoF have questioned the accuracy of these figures, no reliable verification method exists. Other estimates from MoF put the total stock of Central Government arrears at just under 200 million, plus roughly 40 million for Local Government. Though this sum is somewhat smaller, it is still indicative of a substantial problem. 36 Georgia: Public Expenditure Review November 25, 2002 Table 3.3: Comparison of the Impact of Budget Cuts in 2001 On Selected Non-Salary Expenses Figures Represent Actual Expenditures as a percent of Original Budget Spending as % of Original Budget Ministry Office Expenditures Utilities Parliament 34 64 Chamber of Control 36 42 State Chancellery 38 n.a. Education 58 89 Environment 7 62 Economy 17 56 Defense 23 40 Justice 58 110 Culture 30 84 Refugees 26 52 Foreign Affairs 90 83 Security 69 104 Agriculture 55 72 Labor & Health 20 45 Transport 22 51 Urbanization 19 61 Finance 28 52 Internal Affairs 26 94 Table 3.4: IMF Estimate of Expenditure Arrears for End-2000 (GEL millions) Current Expenditures 319.9 Wages and Salaries 118.0 Goods and Services 55.6 Subsidies and Transfers 52.8 Interest Payments 0.2 Extra-Budgetary Expenditures 83 4 o/w Social Security Fund 83 4 Local Government (non-wage) 10.0 3.9 For 2001, MoF estimates that payments on past arrears totaled GEL27.2 million instead of the 81.6 that was planned. Unfortunately, new arrears were introduced simultaneously in the amount of GELI 14.4 million - resulting in a net increase in the arrears stock of 87.2 million. It is estimated that 64 million of this is for Central Government, while the balance is primarily for the social security fund. Because of the way MoF defines arrears, there is a risk that some may actually be higher still. For example, if actual utility costs incurred by a ministry exceeded the amount in the budget, this would not have been included as an arrear - only the amount up to the budget is the arrear. Budget Preparation 3.10 The budget allocation process in Georgia is set against a history of unrealistic revenue forecasts and, until recently, continued accumulation of large payment arrears. As of end 2000, recorded 37 Georgia: Public Expenditure Review November 25, 2002 government arrears amounted to around 5 percent of GDP (or $150 million). As a way to cope with a volatile resource envelope and to manage cash in the short term, both budget preparation and execution have been strongly guided by the concept of protected items. Protected items are mostly recurrent expenditures (e.g., wages, pensions, debt service, etc), which account for 85-90 percent of the recurrent budget, thus leaving very little for discretionary allocation. The investment budget, which is basically financed with donor funds, is not well integrated into the budget process and runs in parallel with the medium term development plan prepared by the Ministry of the Economy. In this environment, the budget preparation process does not provide an adequate framework for a strategic allocation of public funds. Description of the Process 3.11 Budget preparation consists of two stages - first is the submission of the main aggregate parameters to Parliament, and second is the communication of the budget law along with the detailed line items. (See Table 3.5 for a full description of the current steps in budget preparation.) 3.12 Budget preparation begins in March of the budget year with the preparation of aggregate budget parameters, although there are often delays. This document, to be submitted to Parliament by June 1 for approval, outlines overall macroeconomic conditions, along with the total resource envelope for the proposed budget and the expected distribution of resources across major functional categories. Produced principally by the MoF, the parameters for 2002 presented a thorough background on the macroeconomic situation, including sources of economic growth, tax policy changes, revenue forecast, expenditure levels, and sources of financing the deficit. To produce the document MoF, Ministry of Economy, and Ministry of Revenue Collections, and other ministries of the government must reach agreement on an overall resource envelope. Each of the three principal economic ministries contributes their estimates of the appropriate resource target to use. During the past year, MoF estimates were lower than those of the Ministry of Economy but higher than those of the Ministry of Tax Revenue. 3.13 After the overall resource envelope is agreed to by Parliament, ceilings for the individual ministries are drawn up by MoF. Although the functional allocations should, in principle, form the basis for setting individual ceilings, in practice there is much less of a linkage. Indeed, the functional allocations in the parameters document do not appear to be used much beyond MoF. Both the Government and Parliament have shown little interest in debating them, and instead prefer to concentrate on actual ministry ceilings. 3.14 Parliament plays a very active role in reviewing the budget. The Parliamentary Budget Committee is supported by professional staff who assist it in analyzing the budget. Each of the Committee's last three annual reports have contained chapters addressing the macroeconomic foundations of the State budget, revenues, expenditures, the deficit, and transfers. The Committee's Report on the 2002 State Budget, for example, raised a number of questions. For example, it questioned the realism of the budget given the revenue forecasts it was based on (which assumed substantial increases in monthly collections over the prior year.) It also cast doubt on the efficiency that would come from placing relatively large amounts of funds into the otherwise unregulated President's Fund. As a demonstration of the independence of Parliament, the Committee recommended that the budget only be adopted after substantial revision. 38 Georgia: Public Expenditure Review November 25, 2002 Table 3.5: Budget Process for 2002 Current Steps in Georgian Budget 2002 Plan 2002 WB Proposed Modifications Proposal Preparation Actual MoF reviews savings resulting January from policy changes agreed to during prior budget cycle. Line ministries submit proposals for March March (no longer needed) na targeted programs to Min of Economy MoF begins working on budget parameters, March March MoF seeks input/comment February including proposed envelope, functional from line ministries before allocations, and size of targeted programs. finalizing parameters proposal Govt Committee reaches agreement on May End-May Decision of Govt Committee April resource envelope, functional allocations, includes explicit discussion and and targeted program amount agreement on ceilings for budget year, forecast of spending for the forward year, and major policy changes to achieve savings MoF submits Budget Parameters to Early June July (no change) Early May Parliament for approval Parliament approves Parameters (approval is End-June August (no change) End-May informal) MoF transmits general budget instructions Early July July (no change) mid-June to line ministries MoF converts functional allocations into Beg -July Beg - MoF transmits ceilings to line mid-June budget ceilings and communicates them to August ministries that include targeted line ministries programs. Also communicates indicative ceilings for the forward year and proposals for savings to be achieved. Line ministries prioritize targeted programs End-July Targeted programs are mid-July to be funded and transmit to MoF/MoE submitted with overall budget request, but reviewed concurrently by MoE. Line ministries submit budget requests to Early Late Budget requests now include mid-July MoF in conformity with ceilings August August targeted programs. Line ministries also propose any major policy changes for the forward year. MoF reviews requests and holds budget 2"d-half Late (no change) mid-July discussions with line ministries August August MoF submits proposed budget to President Beg - Beg - (no change) mid- August September September President submits initial Budget proposal to End End (no change) Early Parliament September September September Parliament approves final Budget December February (no change) December Formulation of Budget Ceilings 3.15 A consistent problem has been that despite increasingly conservative revenue estimates by MoF, the projections have still tended to exceed actual collections during the year. Consequently, the budgets submitted to Parliament are potentially flawed because they have tended to over-estimate the amount of resources that can be collected. 39 Georgia: Public Expenditure Review November 25, 2002 3.16 Moreover, line ministries are given little opportunity to contest the ceilings they are given by the MoF and there is no opportunity for the government as whole to decide where the offsetting savings/reductions should come from. MoF's meetings with individual ministries are not conducive for making trade-offs across ministries and sectors. Failure to create an appropriate context in which line ministries can provide input into the allocation of ceilings, carries the risk that the budget allocations themselves will fail to reflect accurately the true and underlying cost structure of the ministry. The potential disconnect between the budget ceiling and the actual cost of activities could potentially lead to the creation of arrears or deferred maintenance that will come due at some point in the future. 3.17 In the past, the MoF had de-facto control over intra-sectoral allocations as well. As ceilings tended to be ignored by line ministries in their budget submissions, the MoF assumed responsibility for establishing line ministries' detailed allocations so that they conformed with the overall resource envelope. This practice effectively removed responsibility from line ministries for developing activities and plans consistent with their funding. The preparation of the 2002 budget was the first time in recent years that MoF communicated hard ceilings to line ministries and requested them to develop their own budgets within that amount. Unfortunately, for the 2002 budget, even though ministries were given clear ceilings within which to plan, in most cases the ceilings were communicated too late in the budget cycle to permit a proper review of activities and reprioritization that may have been warranted. Moreover, the ceilings were merely set as some percentage of the prior years budget. Though MoF did make distinctions between priority and non-priority ministries, the allocation-setting still reflected incremental budgeting rather than a substantive policy review. Protected Items 3.18 In light of the consistent shortfalls in revenue, the MoF has resorted to the concept of protected items, partly, to avoid further build up of arrears. In the text of the annual budget law for 2001 it defined the protected items as: "salaries, pensions, social allowances, scholarships and other similar disbursements, food, maintenance of military servants during compulsory service period, medicine acquisition, state debts servicing, obligations stipulated in international and bilateral agreements ratified by the Parliament of Georgia, agreement of the executive branch of the Government of Georgia and the European union, implementation of measures financed by the president's fund and the state reserve fund, healthcare programs." 3.19 As indicated in Table 3.6 below, the sum of these items made up about 89 percent of the total current expenditures in the 2001 revised central budget. The rationale behind the selection of protected categories is fairly clear, as they represent activities to which the government has some explicit or implicit obligation to fulfill. 3.20 Knowing that actual resources during the year tend to always be less than those budgeted, the MoF has created rules that help it to enforce discipline and to ensure that these activities are given priority during budget execution. In practice, that goal appears to be fulfilled because in recent years, MoF has made the cash releases for protected items a priority. In fact, Parliament by voting for the protected items in the budget law, is giving its affirmation to the MoF's cash arbitrage during the year that leads it to favor some spending over others. (See Figure 3.1.) But while creating this segmentation of the budget is understandable as a short term measure to deal with crisis, it is undesirable as budget management practice over the long term. The use of protected items should be reformed because of its impact on strategic budget allocation at the beginning of the year and on fiscal control during the year. 40 Georgia: Public Expenditure Review November 25, 2002 Figure 3.1: Budget Execution Rates for 2001 120% - S100%- 60% - cL m420% - x- E 0 0% 0% - Protected Items Public Debt Non-Protected (excl. debt service) Service Domestic 3.21 Strategic budget allocation suffers because a substantial portion of the budget is already fenced off and excluded from genuine priority setting and budget rationalization. Not surprisingly, MoF officials have made the point since most of the budget is already locked up in protected categories, they have little discretion during budget preparation. Furthermore, a substantial portion of the balance is linked to donor- funded investment projects or to other explicit obligations such as payment of utilities and maintenance of embassies. 3.22 By automatically establishing categories as protected, the incentive and urgency for policy reforms that reduce their cost is subtly eroded. The real question is not whether these are or are not obligations of the government; but whether they need to be at the magnitude that they are currently. For example, given the current cost of maintaining refugees, should that amount be higher or lower, and what would be the policy actions needed in order to develop a more cost-effective program over the long-term. 3.23 The fact that protected items already constitute the overwhelming bulk of current expenditures is an indication the system is unsustainable and needs reform. The solution is not to arbitrarily reduce the categories that are included as protected, especially since some things that are currently unprotected are nevertheless mandatory expenditures of the government, e.g., utilities for public offices. Yet, a greater proportion of the budget needs to be on the table for discussion so that the budget better reflects the long- term development priorities of the country. Rather than waiting for dramatic improvements in revenue administration to alleviate the problem, an alternative approach is to introduce rigorous policy and program reviews into the budget process so that government can begin to reduce the size of the commitments it has. Indeed, this may be the only way to control the growth rate of protected items. Table 3.7 is instructive because it shows that even though total (central government) expenditures have remained flat since 1999, protected categories such as salaries as well as subsidies/transfers have been growing. 41 Georgia: Public Expenditure Review November 25, 2002 Table 3.6: Protected Items in 2001 Budgets Revised Budget Percent (GEL millions) Share Wages and Salaries 87.5 Total Transfer for the State Social Insurance Fund 54.8 Stipends, Social and Other Similar Allowances 166.7 Food Expenditures 19.3 Maintenance of Military Servants during Compulsory Service 5.4 Purchase of Medicines 0.6 Public Debt Service 192.3 International Inter-Governmental Obligations 2.0 State Health-Care Programs 34.6 President's and Reserve Funds 16.4 Total Protected Items 579.5 Total Current Expenditure 653.1 Protected as Percentage 89% Revenues & Grants 655.0 Protected as Percentage 88% Other Commitments Utilities 12.8 Embassies 16.1 Transfers to Local Governments 42.8 Total Protected plus Other 651.2 As percentage of Current Expenditure 100% Table 3.7: Relative Changes in Central Government Expenditure (1999-2001) GEL millions) 1999 2001 % Growth Expenditure Expenditure 1999-2001 Current Expenditures 625.4 578.1 -8 Wages and Salaries 72.4 81.6 13 Employer's Contribution 9.4 17.9 90 Business Trips 10.0 4.7 -53 Other Goods and Services 187.5 123.7 -34 Interest Payments 143.0 115.0 -20 Subsidies and Other Current Transfers 203.1 235.2 16 Capital Expenditures 8.2 2.9 -65 Donor-Financed Projects 58.7 117.7 101 (excluding principal payments) Total 692.3 698.7 1 3.24 In future, government needs to develop a policymaking framework that enables it to effectively address the overall scope and nature of its activities. This in turn would necessitate adopting a multi-year budget horizon, given that many of the structural and institutional changes needed to bring commitments in line with resources cannot be implemented successfully within a single budget year. 3.25 The current practice of protected items has an adverse effect on strategic allocation and priority- setting by line ministries as well. On the basis of past experience, line ministries know that the only 42 Georgia: Public Expenditure Review November 25, 2002 assured appropriations are those for protected items. They know that unprotected categories are more likely to be subject to shortfalls in cash. When preparing their budget, it may be wise sometimes for a line ministry to reallocate funds from wages and salaries to capital expenditure, or to cease certain activities altogether so that others can be expanded. Yet, under the current system, they may find themselves penalized because of either a sequestration of funds or because of the non-release of cash. In other words, shifting resources from protected to unprotected categories opens a potential risk that their overall resources will decline during the next year. Thus, line ministries have little incentive to engage in the kind of priority-setting and rationalization of budgets that is necessary in any country from time to time for public services to improve. 3.26 The practice of protected items can affect not only the strategic allocation decisions of line ministries, but it may also distort the composition of inputs that line ministries ultimately utilize. Given that cash release decisions follow the protected categories, line ministries may find themselves with heavy inputs of human resources but little access to supplies and other operating inputs. All line ministries recognize that maintaining staff without adequate materials with which to work results in inefficient and ineffective operations. And yet the current system has perpetuated the tendency to shift the relative shares of spending from operating supplies to wages and salaries. Investment Budget and Targeted Programs 3.27 Budget preparation is also marked by distinct and parallel processes for the medium term development plan, which comprises the so-called " targeted programs", and for donor-financed investment projects. 3.28 "Targeted programs ", which account for about 5 percent of the central government's current expenditures for 2002, are a subset of the medium term development plan prepared by the Ministry of Economy every year.52 Beginning in March, line ministries submit documentation to the Ministry of Economy with proposals for targeted programs to be conducted the following year. Some of these are new projects, while others are ongoing. Funding for targeted programs may be either domestic or externally provided. Line ministries submit their proposals to MoE, with no fiscal constraints. This list of projects, approved by Presidential decree, are eventually presented in a five year Indicative Plan produced by the MoE called "Principles of Social and Economic Development of the Country." The plan essentially represents a medium term development plan for the nation - except for the fact that only a fraction of the programs can actually be funded. It is not until June that MoF communicates an approximate envelope for all targeted programs. Although line ministries may have invested considerable time in developing their proposals, it is only at this stage that they are asked to reduce their funding requests. 3.29 If there is a distinction between the kinds of activities funded through the budget for targeted programs and those that are more appropriately requested as part of the main budget process, it is vaguely defined. Indeed, there may be no formal guidance as to what activities are to be funded through targeted programs. One common thread may be donor participation or implementation through external organizations (i.e., non civil servants). One purpose that seems to be served through these programs is that it may be providing line ministries with a more distinct opportunity to "make their case" for additional funding. To the extent that this creates a competition for funding and results in better proposals being forwarded, it is a helpful element of the budget process. However, more harmonization between 52 MoF estimates and definitions of targeted programs differs from that of Ministry of Economy. Among the most important differences are that Ministry of Economy includes social security/social welfare programs and all health activities among the targeted programs. This and other differences give Ministry of Economy a total budget for targeted programs of 388 million compared to 34.4 million (5 percent of current expenditure) for MoF. 43 Georgia: Public Expenditure Review November 25,2002 MoF and MoE is needed so that a consistent set of economic criteria are set for funding decisions - regardless of whether it is through the main budget or the targeted programs budget. 3.30 Investment spending is also a separate component of the budget and is not well-integrated into the overall budget structure. The overwhelming bulk of investment spending is financed by donors, and in 2001 it totaled GEL 118 million or 14 percent of the central budget, including counterpart funding from Georgia. (In contrast, domestically financed capital expenditure was at only about GEL 3 million.) Even though investment spending cuts across a variety of functional areas, the MoF accounting system categories all of the expenditure as general government services. Consequently, it would be difficult for Parliament and civil society to see a complete picture of spending priorities across functions and the relative contribution of investment spending. Moreover, based on preliminary observations of the budget preparation process at the MoF, it does not appear that there are any strategic discussions within government about the appropriate level of investment spending or the sectors that should most benefit from it. Instead, the process is heavily influenced by the priorities of donors - including the World Bank -with the government often in a reactive rather than initiating role. 3.31 Targeted programs and investment spending are two areas of the budget that most clearly imply a multi-year implementation. Yet, the Georgian budget preparation is absent any explicit consideration of forward-year budget planning - setting aside agreements with the IMF on a fiscal framework. The Indicative Plan, as well as the Poverty Reduction Strategy Paper (PRSP), reflect multi-year planning, but both are in practice detached from the budget process. Unpredictability in revenues have been a rationale for avoiding multi-year budget planning. However, enhancing the impact and utility of the Indicative Plans and the PRSP will only be possible if they can be linked in some way to the budgetary process. Off-budget accounts 3.32 As noted earlier, expenditures by line ministries are financed not only through direct appropriations by Parliament, but also by the resources that line ministries are able to raise on their own through fees-for-service. Until the 2002 budget, expenditures from these sources were essentially off- budget. However, beginning in 2002 Parliament formally approved these expenditures as part of the State Budget. These special resources are estimated to total GEL 111.2 million, but the true amount would be difficulty to know with certainty. Line ministries are responsible for reporting this data to MoF, but there are no measures or legal basis by which to assure that the amounts reported are accurate and complete. A draft organic law includes provisions that would strengthen MoF's authority in monitoring the magnitude and use of such funds. Actions Taken To Date 3.33 Budget preparation processes have been rapidly evolving over the past two years at MoF. Although the Ministry has long prepared a document for Parliament containing macroeconomic and budgetary parameters, the amount of analytical detail going into the document has been steadily increasing. Today, MoF provides an excellent summary table comparing the current and proposed budgets by various categories of expenditure. The table provides the reader with a very clear picture of the different components of revenue and expenditure and how they have changed. 3.34 Moreover, as noted earlier, MoF has begun to transfer greater responsibility for budget preparation to line ministries. In contrast to the practice in past years, line ministries received clear ceilings by which they could begin to prepare their budgets. While not all line ministries were willing to submit proposals within the envelope provided, MoF remained firm and introduced discipline into the process of budget preparation. In addition, line ministries for the first time were able to assume full responsibility for the allocations within the envelope given by MoF. Over time, the quality of budget 44 Georgia: Public Expenditure Review November 25, 2002 preparation by line ministries is certain to improve as a result. Though the budget sent to Parliament is still heavily detailed by line item, the MoF has begun the process of dialoguing with Parliament about the structure of the budget submission. 3.35 With respect to the comprehensiveness of the budget, MoF took important first steps to integrate into the budget the special resources that line ministries generate. Eventually, MoF will want to determine how best to integrate own-source funds into appropriation decisions and still preserve incentives for spending units to report and collect them. The computerized database on budgets and expenditures being developed within the Budget Office also seems to have as an aim to develop a comprehensive look at all public expenditure - including donor spending, state funds, and local budgets. 3.36 In the past MoF has also taken steps to address the scope of the public sector and to encourage a reduction in the size of public sector employment. Using budget levels in part as a tool, MoF succeeded in encouraging some degree of staff reductions at central government. Although efforts have also been made to reduce employment at local government levels, MoF's leverage with which to encourage it is more limited. Future Actions Needed in Budget Preparation 3.37 The government should begin taking steps to address four areas of budget preparation in order to improve strategic management over the medium term: (1) Strengthening the process for setting budget ceilings; (2) Reforming the scope and usage of protected items; (3) Moving toward programmatic budgeting; and (4) Introducing forward year planning into the budget cycle. 3.38 Moreover, while each of these actions may only be fully implemented over the medium term, it is important that the government begin laying the foundation for them immediately. Strengthening the process for setting ceilings 3.39 The strategic nature of budget allocation would be helped by broadening the participation in setting the ceilings beyond the MoF. Budget allocations should represent the strategic priorities of the Government as a whole. While there may be institutional circumstances that may hinder this from occurring in Georgia in the short term, there are transitional steps that Georgia can begin building into budget preparation. Specific actions that the government can take to do this include: (a) Deepening the level of analysis that goes into developing the proposed budget ceilings and making the process more transparent within government. With the help of a resident technical assistance advisor, the MoF has already begun building a database of information on past budgets and expenditures in order to do comparative analysis for planning purposes. This type of analysis should be encouraged, and the basis for setting ceilings should be explainable to line ministries. (b) Assuring that ceilings and/or functional allocations have been expressly endorsed by the President and Parliament, and that line ministries have had an opportunity to participate before they are final. In this way, it should be clear that they represent the government's strategic priorities and not merely the intent of MoF. (c) Providing an opportunity for bottoms-up input into the overall allocation setting. Line ministries may have information regarding commitments that may be useful for 45 Georgia: Public Expenditure Review November 25,2002 MoF/President to take into consideration before setting ceilings. In addition, it could be beneficial to allow some flexibility to increase ceilings marginally as incentive for those ministries that have done a particularly strong job of preparing their requests or in demonstrating strong performance the prior year. Reforming Use of Protected Items 3.40 Over the medium term, the Government should phase out the use of protected items as it develops a realistic and comprehensive budget framework as the basis for annual budget preparation. During the transition, while protected items continue to be used, the following actions are recommended: (a) Revise the definition of protected items to more accurately reflect the full extent of mandatory expenditures and/or implied commitments of the government during the year. In doing so, the government should endeavor to get a full picture of the existing base of commitments. (b) Incorporate into the budget preparation some form of policy review to agree on options for reducing the size and rate of increase of protected items as a share of the budget. As part of the policy review, they should identify specific types of expenditure for which programmatic, or policy changes can be implemented over time to reduce or limit the level of the expenditure in the future. At the beginning of each year's budget cycle, protected expenditures should be viewed as part of the menu of options for expenditure reduction. (c) As an input to budget preparation by line ministries, MoF should give assurance that cash releases during the year will be linked to programmatic objectives and activities and not merely to economic categories of expenditure. MoF should examine ways in which line ministries can be guaranteed a certain level of resources for priority activities, along with the flexibility to allocate the funding to different economic categories consistent with the budget provisions. Moving toward programmatic budgeting 3.41 Over the longer term, Georgia should move toward presenting and monitoring the State budget in a programmatic format, where resources are clearly linked to accomplishment of specific objectives. As part of that process, MoF should begin taking actions to foster better integration and harmony between the main budget and the treatment of targeted programs, special resources, and investment projects. (a) Targeted programs: In the short-term MoE should work with MoF to ensure that an overall envelope is established for targeted programs so that line ministry preparation of targeted programs is immediately guided by a realistic resource envelope. MoE/MoF can decide on whether it might also be appropriate to provide some guidance as to the likely distribution of that envelope across major functions/sectors, so that this can be factored into LM planning. Line ministries could still be permitted to submit program proposals that exceed that amount, but with the full understanding of what the expected parameters are. Over the medium term, MoF/MoE should examine how to harmonize treatment of programmatic activities. For example, activities currently funded through an organizational heading in the main recurrent budget could be presented and monitored in the form of programs. Perhaps, starting with an initial pilot ministry, MoF could begin making a programmatic structure and accompanying justification a mainstream part of the budget process. 46 Georgia: Pubic Expenditure Review November 25,2002 (b) Investment projects: To facilitate transparency and strategic allocation decisions, the government should plan ways to integrate investment spending into the same system of functional and economic classification as the rest of the budget. In doing so, the government budget system should be able to report on the total budget for a particular sector regardless of whether it is financed from donors or domestic resources. Secondly, the government should plan over the medium term how to integrate decisions about new investment spending into a framework of strategic allocation-setting for the budget as a whole. In doing so, the government may first need to articulate its own medium term budgetary priorities, and then engaging donors in dialogue over how they can contribute to those priorities. While the PRSP and/or the Indicative Plan could potentially serve as a foundation for planning investment spending, each of them would first need to be better integrated into the annual budget process. (c) Special resources: The government should continue in its efforts to integrate own-source revenues fully into the budget process. The publication of this information in the budget document has been an important first step. However, MoF should work with the Chamber of Control to ensure the accuracy of reporting, and to ensure that the new organic law contains sufficient provisions to enforce the use of Treasury accounts for all funds received by public institutions. MoF should also be prepared to exercise sanctions for resources that are not accurately accounted for. Drawing upon appropriate technical assistance, MoF should follow-through on developing formulas and procedures for motivating line ministries to maximize opportunities for collecting revenue and including such information in appropriation decisions. Forward Year Program Planning 3.42 Given the lack of success that Georgia has had in recent years in accurately estimating the size of the annual budget, introducing a medium term budgetary framework may seem premature. Yet, in reality, there are some aspects of forward year planning and budget forecasting that Georgia can and should begin implementing as a tool to create savings within the budget and to begin enhancing the accountability of line ministries. Two important facts underpin reasons why some form of forward year planning is needed in Georgia. First, some changes that are needed to bring the level of aggregate commitments in line with the available resources may not be implementable within a single budget year. Some programmatic or policy changes may require the line ministry concerned to have a sufficient lead time to implement it in a phased approach or to plan for alternative actions. This might include changes in policies governing refugee allowances, funding for Universities, or other activities. Even if only one year beyond the budget year, MoF should have some framework through which programmatic decisions are made by government today (year x-1) that are intended to have financial impact not in the immediate budget year (x), but in the year thereafter (x+l). Such a framework should encourage follow-up by MoF on the decisions (to ensure they are being implemented), and to assure that the budget allocations reflect those agreements that were reached. A second reason to have a multi-year planning horizon is to provide a mechanism by which line ministries who do achieve savings in one year, can be assured that they will be able to retain them for other priority activities. While ministries may be able to identify potential budgetary savings that could be achieved over a one to two year period there may be little incentive to embark upon them if there is a risk that it will merely result in a lower appropriation when done. 3.43 MoF should build this forward-year planning into the budget process, with clarification of the relevant roles of MoF, the line ministries, and the President or Government Committee. In particular, the program decisions would require input beyond that of MoF to be meaningful. In addition, budget analysts within MoF would need to increase their programmatic understanding of the ministries whose 47 Georgia: Public Expenditure Review November 25, 2002 budgets they review. They would also be monitoring program execution, as an input into the following year's budget preparation. Budget Execution 3.44 Like budget preparation, budget execution is characterized by procedures that enable MoF to manage the continued uncertainty and shortfalls in revenue. Although these procedures have in some ways helped to contain expenditures within the available resources, they also have an adverse effect on the quality of public service management. Greater predictability in revenue collection would most certainly facilitate budget management for MoF, but there are also actions that the government can take they would help to improve budget management even in times of uncertain revenue. In particular, there is concern that some of the current strategies, while giving a semblance of fiscal control, may actually be promoting a gradual buildup of arrears. Going forward, it will be important for MoF to ensure that cash releases are better matched to the underlying commitments that the government has entered into. MoF has already taken some initial steps to do so. In addition, fiscal control could be enhanced by carefully managing short-term borrowing tools and making cash releases to line ministries more predictable. Description of the Process 3.45 Once the budget is approved MoF prepares a quarterly allocation plan to guide the execution of the budget. These quarterly allocations represent the amount of budget authority each line ministry and spending unit will have. The MoF develops the plan for the full year, based on its understanding of the seasonal needs of particular ministries, and its own forecasts of the revenue flows during the year. Many ministries prefer that their allocations be weighted toward the early quarters so that they have greater assurance of receiving funding before any sequestration can go into effect. The MoF takes all of these factors into account. Overall, the methodology and objectives of the MoF's quarterly allocation plan seem sound and it is an appropriate first step in budget execution, though it has been ineffective, as explained below. 3.46 In principle, the monthly cash releases53 to ministries or authorizations to spend are to be based on the quarterly allocation plan. Based on these cash releases, spending units are able to issue payment orders against purchases of goods and services. The regional Treasuries then issue the checks or cash upon verification that the budgetary organization indeed has cash available for that particular line item of the budget. The central Treasury maintains a comprehensive database on the amount of cash release for each budgetary organization and line item of the budget. In practice, unused balances by budgetary organizations are rarely an issue, and so total payments made by the regional Treasuries (on behalf of individual budgetary organizations) are equal to the cash releases they received. 3.47 While releases were intended to mirror the quarterly plan, in the past couple of years the actual practice has been different. Because of revenue shortfalls, MoF has often been late or unable to fulfill the cash release agreements that were implied in the quarterly plan. For example, it was only in December 2001 that Treasury was able to release cash for goods and services provided for in the 2nd quarters allocation plan. Generally, the tardiness and/or shortfalls in releases have been limited to goods and services and to investment spending. Since May 2000 MoF has made a commitment to ensure that releases for some categories of spending were made in full. Most of these categories are defined in the annual budget law as "protected" (as noted earlier). However, there are also some categories that lack the legal status of "protected" but are nevertheless honored on a consistent basis due to the implied 53 Though referred to here as "cash releases," in fact, there are more like cash limits or spending authorizations that are assigned to each ministry. The cash itself remains within the Treasury system until a payment to a supplier needs to be made. 48 Georgia: Public Expenditure Review November 25, 2002 commitment that exists, such as allowances for staff of foreign missions. In addition, MoF maintains an internal prioritization of which expenditures among the protected items should be paid first. 3.48 In addition to its cash release, the MoF also uses sequestration to ensure budgetary control. MoF on its own may reduce budgetary authority by up to 10 percent for unprotected items. When it does so, its reductions are to be across-the-board and are not to vary by ministry. If the gap is greater than 10 percent then MoF will submit sequestration proposals to Parliament for their approval. At this stage any changes in the initial budget can be made that are deemed necessary; some ministries may be cut more than others. For 2001 the Parliament reduced the budget by GEL109 million or about 12 percent. In some years, such as 2000, there have been multiple revisions of the budget that were approved by Parliament. 3.49 Given the lag that occurs with sequestration, it is the cash release system that becomes the frontline mechanism for controlling the level of expenditure. A committee within MoF, including the Minister himself, participate in the decisions about the amount and timing of cash releases. Their intent is to meet commitments on "protected" items and to ensure that there is a sufficient buffer of cash available from period to period. However, the analysis in Table 3.4 showed that spending for goods and services is not necessarily consistent from ministry to ministry, thus implying that some judgments are being made within MoF about relative priorities. The Chamber of Control has also alluded to this in one of its reports. Sources of Volatility 3.50 Expenditure management is affected by volatility in both the pattern of domestic revenue collection and in the receipt of foreign financing. In 2000, the shortfall in foreign financing had the largest impact on the budget. Instead of receiving GELI 19.5 million, the government received only GEL44.5 million. Privatization proceeds were also GEL21 million less than projected (GEL13 million instead of GEL34 million). While tax collections were also below plan 'by GELl5 million, it was a relatively small shortfall in percentage terms (3.3 percent). In contrast, in 2001 foreign financing exceeded plan by 10 percent of GELl5.4 million, but tax collections were still short of budget - this time by 9 percent or GEL46.4 million. 49 Georgia: Public Expenditure Review November 25, 2002 3.51 Domestic revenue collection introduces uncertainty in budget management not only because of the fact that it consistently falls short of annual projections, but also because of the lack of reliable monthly patterns. Although Figure3.2 only shows two years data, it is illustrative of the kinds of fluctuation from year to year that occur. Some months with high revenue collections do not experience them the next year. As discussed elsewhere, much of this volatility stems from governance problems in Figure 2: Actual Monthly Revenue Collection for 2000-2001 80,000 70,000 60,000 50 ,000 Z/Y ar20 0 40,000 Year 2001 30,000 20,000 10,000 0 1 2 3 4 5 6 7 8 9 10 11 12 the revenue collection institutions. Since domestic revenues - rather than foreign financing - are the primary means for financing recurrent government operations, these fluctuations can have significant impact. 3.52 Yet even with the fluctuations in monthly revenue and financing, our analysis indicates that cash releases for (or spending by) line ministries is considerably more volatile. Figure 3.3 shows the percent deviation in total resources (revenues and external financing) each month relative to the average for the year. It also shows the percent deviation in spending for three sample ministries compared to the average monthly spending for the year. Of the three, the Ministry of Agriculture's pattern of spending fluctuated the most from month to month. Average monthly spending for the year was GEL7 million, but actual spending ranged from 0 to more than double that amount in any given month. Under such circumstances it would appear to be extremely difficult for line ministries to plan purchases and activities. Ministry of Agriculture officials have stressed that because weather/seasons dictate when they must commence certain activities, it is vital for them to have greater control over the timing of cash releases. In addition, procurement regulations are forcing them to demonstrate that resources will be available before the tender process begins - a task which has been difficult to do under the current funding practices. Moreover, if contracts are tendered and resources fail to materialize it does create difficulties for both the ministry and the supplier. Postponing the tender to the following year may also fail to solve the problem, but instead merely shift the liability into a different period. 50 Georgia: Public Expenditure Review November 25, 2002 3.53 While line ministries commend MoF for paying wages and salaries on time - a practice not followed prior to May 2000 - they are quick to point out that maintaining staff without materials and supplies to use is grossly inefficient and ineffective. Moreover, because of the subjective nature of the decisions, the responsibility for whether policies or activities are carried out risks being shifted from the line ministries - where it should rest - to the MoF since it is they who determine the cash releases. There is some evidence that this may be occurring, as some line ministries do attribute responsibility for funding shortfalls to the MoF. Figure 3: Cash Releases to Ministries in 2001 were much more volative than resource inflows (Monthly deviation from the average) 200% A 150% Awltuftre w Exqendlture 100% Health 4 Expenditure 4 Jr Education 1 Co50% mitentCo 0 Revenrues,Grants. E -7 0% 3.4 Fsa oto hudb rmr objcti uno budge exe utin ntesrac,teeitn fJan01 F eb Lepi_ -50% - I Revenues plus Financing Education Expend. Health Expend. - AgricultueEpn Commitment Control 3.54 Fiscal control should be a primary objective of budget execution. On the surface, the existing system in Georgia of tightly controlled cash releases based on the prioritization of the MoF appears to be a credible system for controlling expenditure. Yet, there are aspects of budget execution that may be subtly undermining fiscal discipline and commitment control. Data on arrears does not necessarily reveal what may be taking place. Rather, the data tends to show that the government has done an excellent job of stopping the arrears buildup that had been occurring during the mid and late 1990s. Still, the government needs to pay special attention to assure that arrears are not accumulating outside of the existing monitoring system. 3.55 Protected items are a dominant feature of budget preparation, but the practice may also indirectly contribute to a weakening of fiscal control. One problem created by the distinction is that some line items, while not legally protected are nevertheless predictable and ongoing costs that must be paid. So for example in the case of utilities for government offices, the risk is that their not being protected could potentially lead to an accumulation of arrears. In the case of telephone services, one government official suggested it would not be surprising if at some point the company were to sue for payment. Indeed, there 51 Georgia: Public Expenditure Review November 25, 2002 may be other instances in which the government receives services from suppliers on implicit credits, but where the obligation is not appropriately captured in the government accounting system. 3.56 Another evidence that the current system lacks adequate identification of government commitments may be the Road Fund. Officials there indicated that the Road Fund had in recent years entered into multi-year contractual commitments with private sector firms for them to provide regular maintenance and repair to Georgia's road network. These contracts were entered into with expectation of a certain funding level. Since then, the actual cash releases to the Road Fund have been insufficient to cover the cost of the work being performed. Setting aside any issues of whether the contract should have been entered into in the form that it was, the reality is that it exists and must be dealt with. Yet, this apparent accumulation of arrears is largely outside of the accounting and monitoring by the MoF. 3.57 Uncertainty over the timing of cash releases may also potentially contribute to future arrears. If ministries anticipate funding arriving at certain periods and enter into purchase contracts on that basis, they may eventually incur an arrear if the funding does not arrive. In practice, line ministries suggested that they guarded against this by not committing to purchase until they are sure funding is available. Yet, they are aware that suppliers are now able to sue the government for payment, a practice which appears to be institutionalized by the fact that a separate line item of GEL4 million was created in the 2001 budget specifically for settlement of such lawsuits. It is not clear, however, if there is any legal or regulatory reason preventing line ministries from committing amounts up to the quarterly allocation established by MoF. Actions Taken to Date 3.58 The government of Georgia has taken some actions to address the problem of revenue volatility. In implementing its cash releases MoF sometimes is able to up a small cash margin to from month to month. In addition, in 2001 MoF began issuing T-bills as intra-month method of financing to begin smoothing expenditure management. The government should however carefully manage the volume of 54 T-bills, given the very high domestic interest rates. Eventually, MoF hopes to lengthen the maturities so that they are useful for financing across months. Also, over the medium-term MoF believes that the institutional reforms in revenue administration scheduled to be implemented in 2002 will help considerably in bringing greater predictability. During this period Treasury should begin assuming responsibilities for accounting of revenue - currently performed by the NBG. By taking over this role MoF expects to have greater access to information on the timing and flow of taxes, and to reduce sources of instability that have been created under the old system. (A fuller discussion of revenue administration issues is found in Chapter 2.) 3.59 To tighten control over commitments and ensure that funds are available to honor ongoing commitments of the government, MoF is also adopting a new commitment registration system. Some parts of the system have already been piloted but the bulk of the work remains to be implemented in 2002. Beginning in 2001, Treasury begun allowing spending units to "register" commitments that exceeded GEL5000. Once registered, Treasury would essentially guarantee that funds would be available to pay it once the goods or services were received. Because this criteria excludes a wide variety of expenditures, under the new system MoF will implement registration for all commitments, including ongoing commitments such as utilities and personnel costs. The IMF has provided a resident advisor to assist MoF in designing and implementing the system. The ministry is still seeking staff, hardware, and software to finalize the new system across the regional Treasuries. 54 The credibility of the government is very low and lenders charge a large risk premium, annual T-bill rates have been over 50 percent reaching 80 percent recently. 52 Georgia: Public Expenditure Review November 25, 2002 Further Actions Needed on Budget Execution 3.60 Building on a number of positive steps already taken by the MoF, further progress would be needed to: (a) increase the predictability of cash releases to line ministries during the year, (b) transfer greater accountability to line ministries for budget management, (c) implement commitment accounting fully. Predictability of Releases 3.61 Efforts need to continue to bring revenue accounting under the authority of the Treasury. This should improve monthly revenue forecasting over the medium term and reduce some of the volatility observed to date. Likewise, if and when the aggregate tax forecasts becomes more reliable, it may be appropriate for MoF to use T-bills or temporary central bank financing to smooth some of the annual seasonality in collections. 3.62 In addition, in the short-term, MoF should refine its approach to developing quarterly allocations so as to ensure that they more directly reflect the expected resources that will be made available to ministries. Given that recent history has repeatedly shown that budget estimates for the year have needed to be adjusted downward, adopting a more conservative approach to quarterly allocations for the first half of the year - although difficult for ministries to absorb - should enhance predictability. As revenues - and external financing - actually materialize, quarterly allocations can be expanded. The important thing is that, once the quarterly allocations are established by MoF, line ministries need to be able to plan procurements based on them, and with assurance that once commitments come due, MoF will honor their payment. This need not be very different from what MoF is already prepared to do in implementing its new commitment control system. Finally, while MoF could continue to use T-bills to smooth the intra- monthly cycle of revenue collection, it is recommended that MoF reach appropriate agreements with NGB that would enhance access to revolving lines of credit, as a lower cost alternative to the T-bill market. Transferring Accountability to Ministries 3.63 As a complementary measure to improving predictability of cash releases, MoF may consider ways for giving line ministries greater responsibility for prioritizing their aggregate (non-salary) cash release amount. Over the medium term, it will be critical for MoF to extricate itself from the task of deciding how much will be released for each economic classification. Instead, it must begin to shift the accountability to line ministries for them to determine how best to manage and prioritize the overall cash release they receive. Treasury can play an active role in monitoring and enforcement to assure responsible use by line ministries of this new freedom. For example, if a ministry were to pay for trips prior rather than their utility bills, it might be appropriate to withdraw the flexibility from them. But ultimately, the goal must be to increase the accountability of budget organizations in the management of their own resources. Accounting for Commitments 3.64 MoF should follow-through as quickly as possible on its plan to introduce comprehensive commitment control. Such controls should ensure that not only procurement contracts are registered, but also ongoing commitments such as personnel costs, utilities, and debt service. Although the MoF is optimistic that this can be implemented in 2002, a substantial amount of training and preparation must occur for this to be realistic for end-2002. MoF can begin developing training for accounting staff on the 53 Georgia: Public Expenditure Review November 25, 2002 new procedures. The verification process that will trigger the creation of a liability also needs to be reviewed to ensure that the information needed for ex-post audits is available as well. Approaches to Improving Expenditure Management 3.65 Improving the predictability of resources available to line ministries is central to any improvement in expenditure management. Predictability of budget resources is at present constrained by considerable uncertainty on the revenue side of the budget. An appropriate strategy for improving expenditure management depends critically on the degree of predictability in aggregate resources, including both domestic revenue, and external and domestic financing. At present, forecasting of revenue even only one year ahead is subject to a high degree of uncertainty. This effectively limits the horizon for any firm planning of expenditure. Proposals for reform need to take account of these highly constrained initial conditions. Consequently, the suggestions below distinguish between measures which might be appropriate in the immediate context and others which will become feasible only as the predictability of overall resources is improved. Improvements within a Short-Term Horizon 3.66 Prima facie, there appears to be some scope for increasing predictability of resources for budget organizations, even given the uncertainties on budget revenue. Monthly expenditure releases have in general been more volatile than monthly revenue collection, suggesting scope for increased predictability of releases. The following measures seem relevant, including ongoing reforms and some additional suggestions: * Improving cash-flow management. (i) Setting more conservative quarterly allocations during the first half of the year (ii) using within year borrowing to smooth expenditure, as already planned. Minimizing costs of such borrowing will require careful management. * Strengthening treasury functions (ongoing). Bringing revenue accounts into the treasury, introducing comprehensive commitment control. (Note that elsewhere in the region, such reforms have taken several years to complete). * Improving the match between protected items and priority programs. Earmarking of 'protected items' should be regarded at best as a temporary measure, to be phased out as the predictability of budget financing is improved. In the mean time improvements in the system could also be achieved by enhancing the match between protected items and priority programs. * Building incentives for budget rationalization. In the absence of explicit protection for complementary inputs, line ministries have little incentive to cut programs and reduce staffing. Ministries that make a serious effort at restructuring programs consistent with budget ceilings need assurance that such programs will be financed adequately, either through inclusion of the program fully as protected items or through specific commitment during the budget approval process. * Developing a phase of budget strategy formulation within the budget cycle. A sustainable restructuring of the budget typically requires a longer term horizon for planning. The budget cycle should include a budget strategy phase, in which ministerial ceilings are decided by Government in the context of the aggregate resource envelope and competing sectoral objectives and options. Initially, this might be based on a strategy framework paper prepared by the MoF, primarily focused on the aggregate resource envelope and the immediate annual horizon. 54 Georgia: Public Expenditure Review November 25, 2002 Building on a More Robust Resource Envelope: A Longer-Term Horizon 3.67 The measures discussed above should contribute to increasing the planning horizon for budget managers, but major progress will depend on improvements in revenue management and forecasting. As such progress is made, more ambitious reforms in expenditure management can be planned. * Extending the budget horizon. Over subsequent budget cycles, the horizon of the budget framework could be extended to the medium term, e.g., three years, with increasing depth of sectoral analysis and greater engagement of other ministries in the formulation process. * Developing bottom up sector programs. Currently, sectors with capacity for realistic expenditure planning are frustrated by the lack of institutions linking sector policy/plans to budgets, and other sectors have no incentive to develop such capacity. The medium term framework exercise would replace many of the functions intended in the present indicative plan and could use more effectively existing central and sectoral planning capacity. * Link with the Poverty Reduction Strategy Paper (PRSP). As the primary link between policy, sector planning and budgets, the framework exercise will need to be closely coordinated with development of the PRSP. Given Georgia's current budget system, this will take several years to implement. * External financing. Uncertainty in the timing of external financing, i.e., from policy- based adjustment lending, is itself a source of poor predictability in aggregate financing. A more explicit budget strategy from Government combined with a more timely implementation of agreed reforms, would ensure greater specificity on external financing. 55 Georgia: Public Expendkure Review November 25, 2002 Chapter 4. Inter-Governmental Fiscal Relations Introduction 4.1 Although systematically increasing, the importance of local governments in Georgia is still Figure 4.1 Georgia: Comparative Shares of Local Gov. ropely ecome in t an in Public Expenditures (European Unitary Countries) Evrteles,othes of loa 60- governments in total public expenditures in Georgia has 5050 increased substantially, from 16.6 percent in 1996 to 27.6 percent in 40 2001, essentially reflecting the x40- delegation of expenditure 51_ responsibilities from the central to local level. This share in Georgia is not much different from that found in 20 some more advanced transition Eastern European countries, such as the Czech Republic and Poland, but it is much smaller than in Northern 0 European countries- which have a long tradition on decentralized fiscal policies-but it is higher than the -Southern European countries (see Figure 4.1). Administrative Structure, Budgeting and Accountability 4.2 While the legal basis by and large reflects the European Charter of Local Self-Governance, the practice is in many respects a continuation of the old Soviet system of rigid vertical, nested public administration, budgeting and control. Weak policy implementation has led to a detachment of the legal framework from the actual practice.s6 Such detachment has generated confusion of authority, operational inefficiency, and lack of accountability. The Legal and Political Framework 4.3 The current legislation of the State of Georgia establishes that the country operates under a decentralized fiscal administration.57 The legal framework does not consider the break away region of 5 A higher share of local expenditures does not necessarily mean that there is more autonomy at the local level. 56 Lack of political will to clearly define attributions and making authorities accountable is also a factor. Article 2 of the Constitution (1995) establishes that matters of local importance are conducted by the citizens through their respective local self-government, whose powers are regulated by the "Organic Law on Local Governance and Self-Governance" approved by Parliament and sanctioned by the President in October 1997, with amendments in 1999, 2000 and 2001. The legal framework under which the Local Governments and the Self- Governments operate include the "Law on Budget Authorities and Budgetary Powers" (approved in May 1996 and now under revision); the '.'Tax Code" approved in July 1997); the Laws on the "System Fees" and "Local Fees" (approved in February and May of 1998, respectively); the "Law on the Capital City, Tbilisi" (approved in February 1998); the Law on the Organization of Economic Activity of the Local Government and Self-Governments" 56 Georgia: Public Expenditure Review November 25, 2002 South Ossetia (about 50,000 inhabitants), but recognizes the two Autonomous Republics of Abkhazia (517,000 inhabitants) and Adjara (394,000 inhabitants). Abkhazia is claiming its complete independence from Georgia (and does not have a seat in the Parliament) and Adjara is struggling to keep its full fiscal autonomy. Some observers consider Georgia as having a four government levels, with autonomous republics forming part of the third level, just below the State (or central) Government (Diagram 1.1). 4.4 The other ten administrative regions (the City of Tbilisi being considered one of them), in which the rest of the country's territory is divided, complete the rest of the third level government. These regions, however, have neither legal status, nor any budget of their own, nor have authority to perform any service delivery function. Their "governors",ss who are appointed directly by the President,' are very influential in intergovernmental fiscal relations and have been used by the central government as a convenient political instrument to control the territory. 4.5 The second level Government, the 56 rayons (or districts), although having a established legal status, a directly elected Council (the Sakrebulo), and a budget of their own, have little fiscal autonomy and accountability. Their Chief Executive Officer (the Gamgebeli) and the Head of the Finance Department report directly (and are allegiant) to the central government-the former is directly appointed by the President and the latter appointed by the Gamgebeli with the approval of the Ministry of Finance.60 Therefore, at least in terms of administrative and fiscal autonomy, decentralization down to the rayon level does not yet exist in Georgia.6' Regional as well as rayon level of administration are effectively only coordinators of deconcentrated functions of the central government in the territory. While regional offices have no legal executive powers, the rayons are in charge of both carrying out State delegated functions and coordinating the actions of their respective self-government units. 4.6 The first level government, the 1031 self-government territorial units (comprising the 1024 local governments such as towns, villages and community of villages, and the seven larger Cities with special status,62) is governed by a directly elected Local Council (the local Sakrebulo, a local legislative body) and an indirectly elected Head of the Local Council (who exercise the local executive powers). This is the only level of government that could be considered as having the legal status to exercise local self- governance (through their exclusive, delegated and voluntary authorities) in line with the European Charter of Local Self-Government and subsidiarity principle. As all its representatives are elected by local citizens, they could be horizontally accountable to the local population. In practice, however, their (approved in November 1997); and the "Presidential Decree on Transfer of the State Property to the Local Government Bodies" (issued in May 1999). 58 Like the regional "governor", these Mayors are appointed by the President and though not exercising any direct executive function, they are considered the highest public official of the City administration. 5 Apparently the original idea was for the regional governor to act as a "Prefect", like in many European countries (e.g., France), whose role essentially is to oversee the legality of the acts undertaken by the local elected bodies. In Georgia, however, the regional governor is very powerful politically and controls all important contacts of rayons' representatives with the central government. One "governor" reported that nobody can be hired or fired at the local level without his consent. 60 Until 1998, the Head of the Finance Department of the rayons (as well as the whole staff of the rayons' Finance Department) also used to be paid by the State budget. 6' However, the new "Amendment Law on Local Government and Self-Government" approved in August 2001 (not yet enforced) establishes two important innovations: (a) the election of the rayon Council members would now be based on a majoritary rather than proportional system (as it has been so far); and (b) that the rayon's Gamgebeli would be nominated by the rayon's Council although still approved (and dismissed) by the President. While still insufficient, these innovations would be significant advance in Georgia. 62 Including Tblisi, Poti, Rustavi, Kutaisi, Sukhumi, Batumi, and Tskhivali. The last three are capital cities of Abkazia, Adjara, and South Ossetia, respectively, and are not under the control of the Georgian State. 57 Georgia: Public Expenditure Review November25, 2002 power is very much limited by the rayons' authorities, since it is only the latter who directly Diagram 1.1. Georgia: Administrative Structure - dt f/' (mtone aton) laih Repalh;) (an phbb) 1 SDthanab .17 tis 2nhA - 394 tmb o ilo 5 raw ray=m S6 r-Y communicates-through the regional "governor"-with the central authorities, and who directly receives the transfers from the center to be redistributed to the local self-governments on a negotiated basis (see below). Budgeting 4.7 The Parliament of Georgia only approves the State Budget, and the new Law on Local Governance and Self-Governance (LGSG) establishes that the local budget shall be independent from any other budget, including the State Budget. But, this independence is rather limited, because the completion of most local budgets crucially depends on the approval of transfers from the center which can be subject to several months delay. In practice, the formulation of budgets in Georgia is still a top-down process based on negotiated gap-fill transfer mechanism. On the expenditure side, expenditure norms are established by the center. On the revenue side, the size of revenues for local self-governments is the result of political negotiations rather than technical considerations. This system is highly perverse, since it is politically driven, optimistic budget revenues prevent major fiscal adjustments and generate arrears accumulation and lack of credibility.63 4.8 The new budget systems law ("Law on Budget System and Budget Powers", LBSBP), currently being drafted, may have some inconsistencies with the LGSG that should be removed. For instance, while the LGSG establishes that "unused resources during the fiscal year shall be used in the following year", the current LBSBP (Article 18) still requires that "in case of increase in the revenue and/or decrease in the expenditures, the appropriate funds shall be transferred to the State budget". Claw-back of 63 Regulation on local budgeting in Georgia is still incomplete. Article 48, Transitional Provisions, of the LGSG establishes that by November 1, 2001 a draft law on Local Budget should be submitted to Parliament. As this requirement has not been fulfilled yet, uncertainties still remain on the budgetary process. Urgency should be given to regulating local budget execution, monitoring (including the reporting system), and auditing. The LGSG established the same deadline for the submission of the "Equalization Mechanism Based on Formula" and of the "Rules for the Ownership, Disposal, and Use of Local Property", which are also overdue. 58 Georgia: Public Expenditure Review November 25, 2002 local budget revenue should be avoided by the State, since this represent a perverse incentive for the local tax effort. Also, while the LGSG (Article 34) establishes that "local self-governance bodies may be authorized to borrow for social needs", a version of the draft Law on Budget System and Budget Powers (LBSBP) states that "local authorities are entitled to borrow only from the Government." An outright ban on access to credit by local governments may not be warranted, as their low creditworthiness should in itself keep commercial borrowing in check. Instead, the law should ban the concession of guarantees by the central government (or the central bank) and by the local governments to any entity. In the long run, access to borrowing by local governments may represent an important, and efficient, mechanism for financing development projects. Local Governance and Self-Governance 4.9 The new "Organic Law on Local Governance and Self-Governance" (LGSG) establishes a distinction between Local Government Units (LGU), which correspond to the rayons' administration, and the Self-Government Units (SGU), which correspond to local governments representing the Cities, towns, association of villages, and villages. From the legal point of view (Article 1, LGSG), the role of the LGU authorities is "to resolve the local issues of State importance delegated by the State organs", while the role of the SGU authorities is "to resolve, in conformity with the law, matters of local importance independently and at their own responsibility". 4.10 Although their respective Local Councils are formed by elected representatives, these two levels of subnational authorities are quite distinct both from the political and the fiscal perspectives. Politically, while the LGUs have their Chief Executive Officer (the Gamgebeli) and the Head of the Finance allegiant to the central government, the SGUs have their Local Executive Officers (the Head of the Local Council and its executive office) chosen by and among the elected members of the respective local Councils. Fiscally, they are distinct because, while most of local taxes and fees are legally assigned to the SGUs' Councils, all their shared-tax and transfers cannot be taken for granted, since they first flow through the LGUs' Councils and have to be submitted to negotiations before re-distribution. 4.11 The current administrative structure is aimed at ensuring that LGUs are in line with the central government, under a top-down fiscal policy approach. Thus, LGUs operate as deconcentrated branches of the central government down to the Self-Government Units (SGUs) level. The SGUs decisions have, to a large extent, been curbed by the LGU Council that controls most of their revenues (shared taxes and transfers) and interferes on what services are to be delivered at the rayon level, including city/town/village. Table 4.1: Territorial Self-Government Units" Population Number of Percent Cumulative Number of Percent Cumulative Size (in Units Distributed Percent Inhabitants Distributed Percent thousands) Distributed Distributed < 1 212 219 21 9 129709 3.0 3.0 1-2 274 28.4 .3 407052 9.4 12.4 2-5 354 36.6 8,7.0 1045022 24 2 36.6 5-10 89 9.2 96.2 550117 127 493 10-50 32 3.3 99.5 552400 8 621 50-100 1 0.1 99.6 64700 - 1.5 636 100-1000 3 0.3 999 478100 I 1 747 1000< 1 0.1 1000 1095000 25 3 100.0 Total 966 100.0 4322100 100.0 1/ Except the conflict zones Source: State Department of Statistics. 59 Georgia: Public Expenditure Review November 25,2002 4.12 Most of the SGUs are fragmented into a large number of small communities, leading to inefficiencies in the delivery of public services, mainly because they lack technical and administrative capacities and minimum economies of scale (Table 4.1). The distribution of SGUs by size is quite skewed. The number of SGUs is too large for the size of the population, with less than 5,000 inhabitants per SGU on average. Of the 966 SGUs (including Abkhazia), 87 percent have a population smaller than 5,000 and about half a population smaller than 2,000. Main Issues and Options for Reform 4.13 There are a number of problems with the current government structure for local governments. First, the government structure is highly hierarchical, involves too many layers of authority, and responsibilities are not well defined. The most powerful decision-making authorities at the local level are still those whose appointment and dismissal depend directly on the central government (the regions' "Governors", the rayons' "Gamgebelis" and "Head of the Financial Department"). In the other end, there is lack of horizontal accountability to the local population, since the main Chief Executive Officers are not directly elected for the key decision-making positions that they hold. 4.14 The self-government units are excessively fragmented, which prevent the efficiency gains of a decentralized fiscal structure. There is also lack of coordination and cooperation among the self- government units, and little cooperation between the central government, the rayon level and the local self-governments. Especially in the case of the small units, cooperation in the provision of public services would be essential. 4.15 The assignment of responsibilities and resource sources is asymmetric. Decisions for adjustments in the final delivery of services is basically negotiated between the rayon authorities and the local self-governments. There is, therefore, no basic regulated criteria for the allocation of responsibilities and resource sources between the rayons and their respective local self-governments, which has diffused the responsibilities on delivery of services and damaged accountability. 4.16 In order to establish the basis for a sustainable decentralized fiscal structure of governance, the Government may consider reviewing the design of the decentralization strategy, to empower the territorial self-governing units with a meaningful degree of autonomy and authority to perform their functions and make them accountable to the local population. The following areas of reform of the administrative structure may be considered: i. promoting the voluntary amalgamation of small self-government units into economically and financially viable self-governing units, through strong financial and legal incentives.64 The central government might consider using its policy on the transfer of State assets to local governments, its privatization policy, and the Municipal Development Fund (MDF) to promote amalgamation. Special concessions could be granted to self-governing units which can prove to be financially and economically sustainable in the long term after amalgamation. The central government could accelerate its privatization program in order to facilitate outsourcing of services to the private sector. Private sector delivery (and financing), by covering more than one small self-governing unit, in turn, may help overcome the inefficiencies associated with the lack of operational scale of self-governing units that will remain small anyway. Additionally, the central government might promote the formation of cooperatives or joint stock companies, by the association of contiguous self-governing units, for the delivery of particular public services with the purpose of increasing operational scale and efficiency. 64 Some isolated areas of the mountainous territories may still require special treatment. 60 Georgia: Public Expenditure Review November 25, 2002 ii. establishing a multilevel government coordinating agency, led by the Ministry of Finance, to implement fiscal decentralization. This institutional reform is essential for the sustainability of fiscal decentralization. The mandate of this agency should be to: (a) serve as a forum for initial conflict resolution between the center and the self-government units, and also among the latter; (b) oversee the implementation of the decentralization agenda (as well as provide feedback to the government on the scope and pace of the decentralization process); and (c) disseminate information and build capacity of both central and local officials. iii. regulating assignment of responsibilities and revenue sources asymmetrically according to local conditions. For instance, local self-governing units with higher administrative and technical capacities should be encouraged to perform public service delivery responsibilities for those contiguous units with lesser capacities. This could be done either on a contractual basis, association, or amalgamation as suggested above. The respective revenue sources should also be transferred following the same pattern of assignment of service delivery responsibilities. Expenditure Responsibilities 4.17 In Georgia, subnational governments are assigned with relevant competences in the most public service areas except for monetary policy, foreign affairs and trade, and statistical services, which fall under the jurisdiction of central government agencies (mainly Ministries). The new Law on Local Self- Governance and Governance (adopted in August 2001)65 shifted some authorities from the local self- governments back to the rayon level of government (including health and education services). Assignment ofResponsibilities 4.18 In general, Georgia conforms to conventional intergovernmental fiscal assignment of expenditure responsibilities and, to a certain extent, to the subsidiarity principle. The new law specifies that while the local self-government units are assigned with exclusive, delegated and "voluntary" responsibilities, the local governments (the rayons) are assigned only with delegated responsibilities.67 The seven Cities with special status, which are not subordinated to rayon administration, are assigned with both responsibilities for self-governing units' and rayons'. 4.19 The exclusive responsibilities assigned to local self-governments (as well as to cities with special status) comprise services related to local police; local infrastructure (e.g., maintenance and expansion of local roads, streets, bridges, parks, street lighting); local forest and landscape; cemeteries; local transportation; primary health care (ambulatory and ambulance services); local information services; 65 As compared to the 1997 organic Law. 66 This principle, applied by the European Union (EU), assigns public functions to the lowest level of government, as long as it can perform them more efficiently than higher levels of government. This tends to increase efficiency, since preferences and costs vary from place to place. Except for national defense, foreign policy, and few others, local decision-makers are in a better position to decide what services should be provided. Also, local taxpayers should pay for the services they consume (equivalence principle). However, frequently local needs and tax capacity differ, and consumption of certain public services may be considered of national interest (e.g., preventive vaccination campaigns, eradication of illiteracy, elimination of poverty), requiring a separation between delivery and financing-while delivery may be better performed locally, financing may be provided by the central or higher-level authority. 67 Delegated responsibilities (on functions like education and health, or specific competences, such as regulation of some specific aspects of policy, financing, or delivery of services) are those transferred by the central government to the local administration. 61 Georgia: Public Expenditure Review November 25, 2002 utilities (e.g., water and sewerage, solid waste, energy); irrigation; city spatial planning; local tourism; and housing and community services. 4.20 Local self-government units are supposed to finance the current and capital expenditures related to their exclusive responsibilities out of there own revenues sources,68 and autonomously decide on priorities according to their local needs and preferences. In practice, however, they are not autonomous, because their own revenues are low and most of the small self-governing units are highly dependent on State transfer to complement their local basic needs. As state transfers (also low) flow down by stages- first to the rayon -regional and rayon authorities end up exerting considerable influence on the local self- government allocation decisions. 4.21 Under the current administrative structure, the State, by retaining the power of influence over the local self-governing units (either indirectly through appointed top officials or directly by exercising the power of final approval of tariffs for utility services), actually obliterates the intended local government autonomy and ends up determining resource allocations even in the sphere of local exclusive responsibilities. Subsidies to utilities and transportation usually absorbs most of the meager local self- governing unit's resources, leaving other basic needs under-provided and the local self-governing units totally exposed to and dependent on transfers to be negotiated with the rayons' authorities. Frequently, the scarcity of resources vis-A-vis commitments has led to accumulation of arrears by self-governing units. 4.22 There are a few State delegated functions that the law attributes to local self-governing units directly (e.g., registration of civil acts-by keeping the books on and issuing certificates for birth, marriage, and death events-administrative activities related to military conscriptions, veterinary and agricultural related services, including sanitation and local environmental protection), but even those functions may not be enforced because of weak implementation capacity. Regarding the responsibility for State delegated functions, the understanding is that they can be asymmetrically arranged between the rayon and its respective self-governing units, so that rayons' authorities are ultimately responsible for the delivery of public services. 4.23 The rayon administration is also responsible for other state delegated functions: approval and control over the execution of local budgets (including that of local self-governing units'); decisions on local taxes and fees; pre-university education (e.g., nursery, kindergarten, primary, secondary, boarding and technical schools); health care (e.g., hospitals, outpatient, clinics and primary care); land use matters; rayon police and fire brigade; natural disaster and calamity related matters; programs for local social- economic development; cultural activities (museums, exhibitions, music schools, libraries, cinema and theaters); sports and recreational activities; the use of natural resources; and environment and ecological related matters in the jurisdiction. 61 See Section on "Revenue assignment". 62 Georgia: Public Expenditure Review November 25, 2002 Box 4.1: Expenditure Assign ments Rayons Local Self-Government (Delegated Responsibility) Units Responsibility State Exclusive Resp. Delegated Resp. Monetary Policy Nat Bank of Georgia Statistical Serv State Dept. of Statistics Defence MoD - Coordination of some - Minitary admin Activ military admin Activities, (conscriptions) including conscriptions Justice/Internal Affairs and MoJ -Rayon police services - Local police services - Regist Of civil acts Security Mol -Fire brigade -Natural disaster and calamity related matters Economic Policies MoF -Approval&control over MoE execution of local budget MoR -Intr.& abolish local tax&fee Agriculture MoA -Veterinary and agric related -Veterinary& agricultural services related services -Land use related matters Regional Development MoT - Programs for the social- - Local infrastr (local roads, economic develop streets, st lighting, bridges) Traffic and Public Transport MoT - Local Transport - Local transport subsidies Education -All university &res. institutes -All (O&M+K) pre-university -Teachers' salary education, including kindergarten, nursery, initial,basic,secondary, boarding, and technical colleges Health care - Doctors' salary -Hospitals, outpatient clinics -Ambulatory care -Primary care -Ambulance service Culture, Information Leisure - Min of Culture -Museum, exhib, music -Local Information Serv Recreation and Parks school, library, cinema & -Local Archives theater,cult & sport recrt -Maint bldg cult events Public Utilities -Regulate by State -Water treatm &supply -Delivered by off-bdgt -Irrigation systems companies (in general SOE) -Energy (electr. & gas) -Sewerage, solid waste Urban Development and - Regulated by the State -City & special planning Tourism -Local tourism Housing & Comm Serv - Regul By the State - Housing & comm serv Social Security and Welfare Nat. Sec Min. (Social Sec -Personal serv for the elderly System) - Social Fund & handicapped -Nurseries, homeless, families & child care -Housing for the poor Social Protection MoLHSA -Social Assist to the poor, incl unemployment relief Environment & Public Min of Environment -Environ &Ecol Rel matter -Local forest & landscape -Local envir protection Sanitation -Natural resource uses -Cemeteries -Sanitation 63 Georgia: Public Expenditure Review November 25, 2002 4.24 In most cases, delegated functions include financing and delivery of public services, involving recurrent and capital expenditures. The Law prescribes that the local governments are to be compensated with adequate financial means for the delegated functions. However, there are some pending issues on the financing of state delegated functions. Although resources are in principle not earmarked, the budgets in Georgia operates under a "protected items" regime, which means that certain budget line items cannot be sequestered even in face of severe revenue short-fall. Under this regime, local governments are fully compensated by the central budget for individual expenditure items such as teachers' salaries. Such expenditure rigidities tend to create wrong incentives and lead to the perpetuation of inefficiencies in the allocation of public resources. For efficiency reasons, the center should not finance the totality of the costs of one specific expenditure category through grants, because it impedes the optimum combination of production factors at the local level. If some expenditures really deserve to be protected, these should be only a few selected meritorious programs which can generate high social rates of return. Budget resources for these programs should be disbursed conditional to verifiable agreed outcomes. 4.25 The reported under-funding of delegated functions, with arrears being accumulated, seems to be a problem related to the process of budget formulation and execution, and poor definition and attribution of authority. Local budget revenues (both own and shared) have been persistently overstated at the programming phase, owing to political vested interests mainly on protected items which prevent adjustment of budgetary commitments to resource availability; and pressures from the center to raise revenues in order to close budget gaps. During the execution phase, as the tax administration is totally centralized and most important expenditure items are protected, there is little economic incentive for adjustments of local budgets. 4.26 Local self-governing units also have residual "voluntary" powers on any matter of local interest. These are minor authorities that can be exercised at the community level, on a subsidiarity principle basis, but not many are. Local Expenditure Patterns 4.27 While increasing since the mid-1990s, local government expenditures (rayons and local self- governing units) as a share of GDP are still low for European standards-5 percent average during 1996- 2001 in Georgia, against about 10 percent of many Central European countries (Table 4.2).69 As a share of general government expenditure, local budgets have substantially increased to 27.6 percent in 2001, from 16.6 percent in 1996.'0 On the latter, the current Georgia local governments' budgets do not differ much from the Eastern European countries. 69 In Georgia "local government budgets" includes both the "Local Governments" (i.e., only delegated functions of rayons) and the "Local Self-Governments" (i.e., exclusive and delegated functions of self-government units). 70 General government expenditures include the central budget, local budgets, and extra-budgetary funds. 64 Georgia: Public Expenditure Review November 25, 2002 Table 4.2: Distribution of Public Expenditures between Levels of Government, 1995-2001 1996 1997 1998 1999 2000 2001 (As percentage of total) Sum Expenditures 100.0 100.0 100.0 100.0 100.0 100.0 1. State Budget 70.9 68.4 65.5 '59.0 56.9 54.7 2. Local Government Budget 16.6 21.6 25.2 24.9 27.1 27.6 3. Extra-budgetary Funds 12.4 10.0 9.3 16.1 16.1 17.7 (As percentage of GDP) Sum Expenditures 20.1 21.4 19.9 22.7 20.1 22.9 1. State Budget 14.2 14.6 13.0 13.4 11.4 12.5 2. Local Government Budgets 3.3 4.6 5.0 5.6 5.4 6.3 3. Extra-budgetary Funds 2.5 2.1 1.8 3.6 3.2 4.0 Source: Ministry of Finance. Functional Composition ofLocal Expenditures 4.28 Except for social protection"-which has steadily increased-the functional composition of local government expenditures has remained fairly stable since 1997 (Table 4.3). The most important expenditures are national economy and education, representing almost one-third of the local budget each.72 Culture and health together absorb another 12 percent of the budget. In total, social and cultural activities (including education, culture, sports, health, and social protection)73 absorb almost half of the budget, which, together with national economy, consumes more than three-quarters of the whole local government budgets. On the remaining, expenditures with local authorities (i.e., the maintenance of the legislative and executive bureaucracies) and law enforcement have consumed between 12-15 percent of local budget resources. 4.29 Subsidies (mainly to utilities and transport services) represent an important component of the local self-governing unit budgets and constitute a major issue related to their exclusive responsibilities for two reasons. First, the provision of utilities and transportation services is highly subsidized. Either fees/tariffs for some services (e.g., garbage collection) are not charged at all, or they are not adjusted to cover the operational costs of the provision facilities (e.g., water supply). Moreover, the collection ratios are extremely low. For example, the collection ratio of water bills in Gori and Mtskheta in the past few years are running at around 20 percent.74 Second, the current arrangement for the asset ownership and Operation and Maintenance (O&M) of these utility companies-where the property rights are still retained by the State, but the management (operation, maintenance, and even expansion) is up to the local 71 Includes social assistance for the poor, unemployed, elderly, handicapped, homeless, nursery services, and families and child care. 72 "National Economy" expenditures includes: housing, public works, agriculture, roads, environment, and subsidies to utilities like water, gas, transportation and communications. 73 These are "protected expenditures" over which the local governments have very limited decision power. 74 Because of lack of cost recovery, practically all utility companies in Georgia are under considerable financial stress. Reported reasons include: (a) the population has been accustomed since the Soviet times to low cost or free provision of public utilities; (b) the unemployed and/or poor (e.g., pensioners whose pensions are not adequately adjusted by the State, or are in arrear) cannot afford utility services; (c) proposed tariff changes have to be approved by both the rayon Councils and central government agencies; (d) politicians are afraid to approve tariff increases and unwilling to change the present situation; and (e) even if decisions were taken towards a more realistic tariff structure, the enforcement system is not effective enough to help with the implementation of sanctions and increase collection. 65 Georgia: Public Expenditure Review November 25,2002 governing units-is ineffective, unsustainable, and doomed to fail. The property and management responsibilities should be placed at the same government level, otherwise it is unlikely that the service facilities will be adequately maintained-let alone expanded, or the needed innovations introduced. The current arrangement tends to lead to a faster depreciation and quick depletion of the capital stock. Table 4.3: Composition of Sub-national Expenditures Functional Classification, 1997-2001 Functional Categories 1997 1998 1999 2000 2001 Average National Economy 29.5 33.9 29.5 27.9 30.5 30.2 Education 33.4 29.7 28.5 31.8 27.3 30.2 Culture 7.3 5.4 5.3 6.8 5.3 6.0 Sports 1.0 0.7 1.1 1.0 1.8 1.1 Health 6.8 6.2 5.0 5.4 5.8 5.8 Social Protection 0.8 1.8 4.3 5 8 8.9 4.3 Archives & Science 0.1 0.1 0.1 0.1 0.1 0.1 Defence 0.2 0.2 0.2 0.2 0.2 0.2 Law Enforcement 3.0 3.6 7.6 3.5 3.9 4.3 Local Authorities (leg. & Ex.) 7.4 7.1 8.6 9.5 8.2 8.2 Expend. Out of spec. Resources 3.1 3.6 2.8 3.3 1.8 2.9 Other Expenditures 6.7 7.2 6.8 4.7 6.1 6.3 Repayment of Budgetary Loans 0.4 0.5 0.0 0.1 0.2 0.2 Total Expenditures 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ministry of Finance 4.30 A notable feature of the local expenditures by function in Georgia is its significant variation across rayons. Annex 4.2 shows that while there was an increase in the average per-capita expenditure during the 1997-2001 period (from GEL33 to GEL56 per capita), there were large disparities both among the aggregate functional categories and across rayons in each functional category. For instance, the period average per capita maximum expenditure was thirty times the average minimum (a maximum GEL437, against a minimum of GELl5, with a GEL45 average). By far the "national economy" is the function showing the largest expenditure disparity among rayons in every year (a maximum of GEL365, against a minimum of zero). 4.31 The above translates into a very skewed distribution of functional expenditures among rayons, simultaneously reflecting differences in revenue capacities, local preferences, and possibly in political clout; with stronger local authorities obtaining better deals in the negotiation process of transfers from the central government. The data show that the differences in expenditure per capita are less pronounced for the more basic expenditures items, like education and health, with a period average coefficient of variation of 0.4 and 0.9, respectively, and more pronounced for items like sports and law enforcement (Figure 4.2). The smaller variation of the former may reflect the more direct influence of the central government through the regulation of national norms and the transfers of grants for education (salary of teachers) and health. On the other hand, the larger variations on the "residual functions" result from the fact that only the better-off local governments are able to spend in social activity like sports and to sustain certain type of social assistance. Nevertheless, although showing relatively low coefficients of variations, the distribution of education and health expenditures across rayons still ranges between GEL9-41 per capita, for education, and between GELO- 10 per capita, for health." 7 This does not mean that the local population receives only this minimum of public services on education and health, because the central government also provides the related services (e.g., tertiary education). 66 Georgia: Public Expenditure Review November 25, 2002 Figure 4.2: Disparity In Per Capita Local Govt Expenditure Fig. 4.3: Per Capita Local Govt. Expenditure Coefficient of Variation, 1997-2001 Coef.of Variation 45 600- 40 500 c 30 400- 20 15- Economic Composition of Local Expenditures 4.32 Figure 4.3 shows that per capita spending in "subsidies", "capital expenditures", "utilities", and "stipends and social assistance" are highly variable across rayons, while per capita expenditures on "wages" and "contributions" are rather uniform. This reflects the fact that wages are set nationally, and the wage bill is a "protected item" in the budget, while most of the other expenditures are not and the local government may adjust them according to local conditions. The composition by economic functions of local expenditures also varies substantially across the country. In most of the regions wages and contribution represent about two-thirds of the total expenditures, while capital expenditure (maintenance of buildings, and investments) is only about 5 percent. Tbilisi and Poti cities again are two major exemptions and 27 percent of their respective budgets is allocated on capital expenditures (see Annex 4.1). 4.33 While the regional or rayon aggregate data by economic and functional classifications provide some information on expenditure priorities and on the input composition for the delivery of local services, they can do little in explaining the relative efficiency of individual local governments. One important issue, for example, is how the size variation of local self-governing units may affect the composition of expenditures, because smaller communities are less able to afford and to spend on some "residual" functions (e.g., culture, sports, arts; even on social assistance) than others functions (e.g., wages, subsidies to utilities) that they may consider higher priorities or that are imposed on them as "protected items." 4.34 As an indication of economies of scale, the share of fixed costs such as internal administration and infrastructure could be expected to decrease with the size of the local government unit. The data for Georgia seems to confirm such a relationship between the size of the locality and the percentage of expenditures in internal administration (proxied by "wages and contributions") and in infrastructure (proxied by "maintenance and services of transport and equipment", "subsidies", and "capital expenditures"). Local Spending Disparities Across Regions and Rayons 4.35 The disparities of per capita local budget spending across regions increased considerably during 1997-2001-with a coefficient of variation of 0.95 in 2001 from 0.57 in 1997 (Annex 4.3). Most of the increases in local budget expenditures occurred in Adjara region and Tbilisi City. The average per capita 67 Georgia: Public Expenditure Review November 25, 2002 Ispending in Adjara region and Tbilisi City were GEL199 and GEL104, respectively, while the average for Figure 4.4: Georgia - Local Government the other nine regions under the State Expenditure, by region (1997-2001) administration was only GEL38 200 (Figure 4.4). These extreme 180. variations in regional public spending 160 160-1 can hardly be justified only by the 140 120- -local tax capacity and economic 120 M 100disparities. For instance, there is no 5 80 correlation between the regional GDP 0 60 ____per capita (as a proxy for the regional 40~ revenue capacity) and the regional 20 per capita spending. The expenditure 60 0 * *~ ~ .~GDP ratios were diverse: 16.5 percent for Adjara and 7.4 percent for Tbilisi, while only about 3 per cent for the other regions. Large disparities cannot be explained either by differences in needs. Other factors may be influencing these regional spending divergences as: (a) tax evasion, which may be larger in some regions than in others; and (b) political clout of Adjara republic and Tbilisi City, which can directly influence political decisions favoring their respective budgets (e.g., fiscal privileges, extraordinary transfers etc.). Table 4.4: Disparities in Expenditure Per Capita, by Rayon (Excet Adjara, Tbilisi, Poti, and Kazbvegi Raion Descriptive 1997 1998 1999 2000 2001 Avg. Statistics 97-2001 Maximum 267 233 710 637 340 437 Minimum 12 12 16 16 19 15 Average 33 36 51 50 56 45 Standard Deviation 35.9 36.1 93.1 83.9 53.6 61 Coefficient Variation 1.10 1.01 1.83 1.68 0.96 1.32 Maximum 43 78 80 67 91 72 Minimum 12 12 16 16 19 15 Average 25 27 33 33 44 32 Standard Deviation 8.4 10.3 12.8 13.2 15.4 12 Coefficient Variation 0.33 0.38 0.39 1 0.39 0.35 10.37 Source: Ministry of Finance. 4.36 The coefficient of variation is even larger among rayons (see Table 4.4), since Poti and Kazbegi rayons are outliers-the former a city Port, which until recently was enjoying certain tax revenue privileges, and the latter, a remote mountainous rayon with special needs, highly dependent on State budget transfers. Excluding Adjara republic, the City of Tbilisi, and Poti and Kazbegi rayon, the distribution of the per capita local expenditures among the remaining 55 rayons present less of a disparity, even though with a coefficient of variation of 0.35 in 2001-a minimum value of GEL19 per capita (Zugididi and Zugididi rayon) against a maximum value of GEL9 per capita (Bolnisi rayon), four times larger. All in all, large per capita expenditure disparities both across regions and across rayons are a major feature associated with local public expenditures in Georgia. 68 Georgia: Public Expenditure Review November 25, 2002 Main Issues and Options 4.37 The Law on Self-Governance and Governance prescribes that the local self-governing units are autonomous in the prioritization of their own exclusive expenditures, but in practice these priorities have been highly affected and distorted by higher level governments. This has led to increased subsidies to "economic services", detracting scarce local resources from other priorities. The law also prescribes that delegated responsibilities are to be fully financed by the State budget, but in practice arrears have accumulated at the local level because expenditure commitments (in general determined by the center) exceed resource availability. Moreover, the system of assigned responsibilities itself presents several weaknesses. 4.38 Poor assignment of authority to local governments on exclusive and delegated responsibilities. Decisions on financing and service delivery are "arranged" between the rayon administration and the self- governing units-without any pre-determined criterion. The consequences are duplication or under- provision of public services, proliferation of unfunded mandates/delegated functions, and accumulation of arrears, especially on those budget items which are of national interest-but which have been pushed down by higher level governments. 4.39 Weak administrative and technical capacities especially at local self-governing level. Since the central administration directly and indirectly controls the spending decisions at the local level, local governments are not empowered to take charge of their own legal responsibilities and build their own capacity, including through training. This has weakened the incentives for efficiency and good governance normally expected from a decentralized fiscal structure. 4.40 Deteriorated delivery of public services, particularly utilities. In most cases, the State has kept most of the ownership of physical assets and has pushed to the local level the full responsibility for operational costs, assets replacement and expansion. In addition, tariff policy for most local public services is open to political influence from both the rayon and the central level. Because of unrealistic tariffs and low collections, the utility companies are mostly bankrupt, accumulating unrecoverable arrears and "cannibalizing" their own assets. Also, low tariffs lead to wastage, which aggravates demand-supply imbalances. Such perverse incentives are not congruent with a sustainable provision of public services and, if not changed, the Georgian population may be facing serious under-provision of basic public services in the near future. 4.41 Considerable disparities in per capita spending by local governments, both across regions and across rayons, and considerable variations across functional categories. Apart from different tax capacities and local preferences for types of services, different degrees of tax evasion across regions and rayons, the political clout that some regions and Cities are able to exercise may also have a strong influence in determining the spending disparities. If the current intergovernmental fiscal system continues, these influences may increase in the future and jeopardize nation-building objectives. 4.42 The most important aspect of the strategy is to put in place an appropriate structure of economic incentives for the functioning of a viable decentralized fiscal system that can be perceived as efficient and equitable, based on rules and immune from political influences. On the expenditure side, such a strategy would include: i. specifying and re-assigning expenditure responsibilities to ensure the benefits of economy of scale and positive externalities, and avoid concurrent competencies, duplication and under-provision of services. The following measures may be considered: * redistribute function/competences between rayons and local self-governments on the basis ofrules properly specified in the organic law; 69 Georgia: Public Expenditure Review November 25, 2002 * re-assign to the central budget the financing competency for primary education, the basic health care package but keep the service delivery competency to the subnational governments, rayons or local, in accordance with their capacities and the scope of the operation; * decentralize decision-making according to the principle of subsidiarity, asymmetrically; i.e., decentralization should proceed to the level of local governments according to their capacity to responsibly take decisions on factors of production and service delivery, going down to the level of the service agency (e.g., the Head of School Boards could be elected by the local school boards or by parents associations)7; ii. eliminate the criterion of "protected items" under which the State functions are delegated to the local governments to increase effectiveness and allocative and technical efficiencies. This would imply discontinuing the current policy of protecting the "wage bill" item in the budgets, and offering local governments the flexibility to reallocate resources, and improve the O&M to wage bill ratio, to increase operational efficiency. The central government may still guarantee the finance but delegate the delivery of some critical public goods and services through the local governments. In this respect, the central budget should provide "conditional grants" for very well-specified "programs" of national interest, where clear verifiable outcome indicators could be identified and monitored by the center." Such delegated functions/competences typically involve programs to promote primary education, primary health care (including vaccination, combat against contagious diseases-tuberculosis, malaria, HIV etc), and programs to reduce poverty (e.g., targeted social assistance to the poor). As long as the local governments deliver the quantity and quality of the required services to the citizens, in accordance with the agreed outcome indicators, the targets are met, regardless of the specific ways the local budget allocated the resources. This will require getting away from the old inefficient input approach and moving towards an output/outcome criterion of public resource allocation. iii. restructuring the financing and delivery of public services (including utilities) to render them sustainable in the long run. The local utility companies are all under enormous financial distress, their assets being depleted, and services are deteriorating. Outsourcing of public services is not yet a common practice in Georgia, and the participation of the private sector both on financing and on provision of public services is very small. If prospects for privatization are not good in the short run, it could be feasible to administratively re-structure existing companies, and to reform the tariffs policies and the system of service delivery consistently with promoting the participation of the private sector both on financing and provision of services. The reformed companies and the private sector (including consumers) should then be persuaded to operate under a new paradigm sustained by a consistent structure of incentives, which would include: * transfer ownership of assets (both social assets and those of the utilities) to the local governments aiming at more efficient use of the capital stock. Ownership of service delivery facilities should be placed at the same level of government where the responsibility for their maintenance lies. While general rules on property rights of local governments is in place, the ownership of assets has not been transferred by the State. 76 Currently, local governments can only propose candidates as head of school boards and chief physicians, but the actual appointments are made by the relevant state Ministries. Also, regarding the protected items, local fovernments can neither change the norms (fixed coefficients), nor wage levels. This may require considering incentive mechanisms to the local governments' performance, as well as punishment measures. 70 Georgia: Public Expenditure Review November 25, 2002 * a utility tariffpolicy based on the principle of full cost recovery, without disregarding the social condition of the population. This means a clear and complete separation of the two policy instruments: (a) cost recovery -which would rationalize the allocation and use the scarce resources and render the system sustainable; and (b) social objectives, e.g., explicit current "grants" to targeted groups with the transparent approval of the local Council; * regulation of specific sectors to facilitate and encourage cooperation among local self-government units for a joint provision of services (either public or private companies, or an association among them), in order to explore economies of scale. The regulation should also contemplate an appropriate outsourcing of public service provision to the private sector; * submission by each local public utility company of its annual business plan, accompanied by previous year report, to be approved by the local government Council. The companies' managers should be given full autonomy to perform, but also be made personally accountable before the law and the local Council. iv. addressing the challenge of large disparities on per capita spending across regions and across rayons to preempt instabilities. The current situation of large and increasing disparities on per capita spending may not be socially or politically tolerable in the long term. Besides devising a more diversified, fairer, and stable source of resources for the local and local-self governments and combating tax evasion in all regions, transfers from the central government (both regular and extraordinary) to support local expenditures programs should be based on general rules and criterion open to the participation of every local government who qualifies to compete for them, regardless of their specific regional political clout. Assignment of Revenues 4.43 Georgia has made important strides towards the access of subnational governments to sources of revenues. There is an established system of tax-sharing, a clear structure of local taxes and fees, and some definition on which aspects of the tax policy the Local Councils can exert their authority. Nevertheless, current arrangements do not yet provide local self-governments with any meaningful revenue autonomy, their revenue remains unpredictable, and tax collection effort incentives are not adequate. The Statutory Assignment ofRevenues 4.44 Georgia's legal framework establishes the public revenue assignment system as follows: (a) state taxes, (b) local taxes and fees; and (c) off-budget revenues. The 1997 Tax Code establishes the tax base, regulates all exemptions and fixes ceiling for local taxes. Local fees are regulated by a specific law (Law # 1401, May 1998). Specific laws deal with off-budget revenues (special funds, utility companies, and resources from some education and health institutions). 4.45 The State of Georgia basically retains the decision-making power (of creating taxes, defining basis and rate ranges, establishing exemptions etc.) on the overall fiscal policy and tax administration. The two autonomous republics and the local self- governments (including the cities with special status) have some prerogative to set local tax and fee rates and, in certain cases, even to make some generic exemptions, but they cannot introduce taxes or fees beyond those specified in the law. 4.46 Regions, as de-concentrated State territorial administration agencies, are neither directly entitled to tax-sharing nor have any tax of their own, and are entirely financed by the State budget. Georgian 71 Georgia: Public Expenditure Review November 25, 2002 rayons are mix institutions, since, besides essentially carrying out State delegated functions, they also coordinate self-government territorial units under their jurisdictions. This hierarchical arrangement of the administrative structure has major implications on local self-governing units revenue autonomy, since it subjects any relevant decisions of local self-governments to negotiations and approval by the rayon authorities (see below). State Taxes 4.47 State taxes are entirely regulated by the State through the Tax Code, and are formed by "non- shared" taxes and shared taxes. The State "non-shared" taxes are the most important sources of tax revenues (about 7 percent of GDP), and account for almost two-thirds of total tax collection in Georgia and about 90 percent of tax revenue of the State budget. 4.48 Non-Shared State Taxes. In Georgia, Customs Duties, Excise, and Value Added Tax (VAT) are non-shared taxes, in principle, but they have often been shared on a temporary and exceptional basis. For example, between 1998 and 2000, Poti, for being considered a strategic port of the country, was retaining 5 percent of customs duties. During the same period the city of Tbilisi was retaining 15 per cent of VAT. Since 2001, these privileges do not exist any longer, but the Adjara Autonomous Republic is still authorized to retain-since July 2000 until 2003-50 percent of all customs duties collected in its jurisdiction. During this period Adjara can also retain 40 percent of both VAT and Excise taxes, when these taxes are collected by the State Customs Agency, and 30 percent when collected by the State Tax Inspectorate. 4.49 Shared State Taxes. The two most important shared taxes in Georgia (about 3 percent of GDP) are the Personal Income Tax (PIT)78 and the Corporate Income Tax (CIT). The current sharing rule is that 15 percent of tax proceeds from these taxes goes to the State budget and 85 percent to the budgets of the territorial units where the tax was collected (i.e., derivation basis).79 The only exception to this rule is again Adjara, which keeps 99 percent of both the PIT and CIT.80 Given the skewed distribution of the country's economic activity and fiscal capacities, however, this almost exclusive reliance on a derivation basis system to allocate shared-tax proceeds raises some efficiency and equity concerns. 4.50 The Georgian PIT is a progressive tax (with a tax rate structure of 12, 15, 17, and 20 percent), initially paid at the workplace (withheld from the wage bill) and adjusted at the place of residence (where the annual income tax declaration is filed). The PIT is generally considered one of the best taxes to be shared or assigned to local governments-since it tends to be a stable tax and a buoyant revenue source for the local government. But, the progressive character of the PIT tax, although an appropriate redistributive central government policy" - collected on a derivation basis tends to exacerbate the 78 Small entrepreneurs (annual turnover of GEL24,000 or less) pay a "Presumptive Tax" in lieu of the Income Tax. Presumptive tax is a fixed amount paid monthly that varies between GEL 10-150 depending on type of business and size of local population. 79 This is the new tax sharing arrangement, in place since January 1, 2001 (for Tbilisi, since July 2000). Previously, these rates were 40 and 60 percent, for the State and local government units, respectively. 80 The exception might be Land Tax, where local governments are supposed to have some influence on determining the tax basis. But apparently central tax inspectors are not equally encouraged to collect local taxes or taxes which are 100 percent shared by local governments, as incentives and penalties affecting these tax inspectors are said to favor State taxes. In the past, tax inspectors used to have a dual subordination both to the central government and local governments. 81 Income distribution is a concern of central governments. 72 Georgia: Public Expenditure Review November 25, 2002 concentration of tax revenues in favor of the already well-off localities, where the richest individuals work and/or reside.82 4.51 The CIT is not an adequate tax to directly fund local governments either. Also in Georgia, it has shown to be a volatile source of revenue,s3 can easily be exportable, and its final incidence is likely to be outside the jurisdiction, implying that this tax could hardly provide a direct link between the taxes paid and the benefits received from public services. The most serious problem with the CIT is that usually larger tax payers collect their taxes at the place where they are registered, or where their headquarters are situated-especially at the most developed urban centers-regardless of where their regular business activities actually take place. In this regard, urban centers like Tbilisi and Poti are the places where most of the CIT collection will occur, with these localities receiving fiscal resources beyond the fair share corresponding to the economic activities actually developed in their jurisdictions. 4.52 All other shared-taxes have been less important as revenue sources for the local governments, barely reaching one percent of the GDP and 20 percent of local tax and fees revenue (average 1997-2001 period. Although these taxes (on land, properties, property transfers, inheritance and gifts, motor vehicle transfer, natural resource exploration, and environmental protection) are 100 percent shared by the local governments, they are considered State shared taxes because they are regulated and administered only by the central government. In this regard, all shared-tax revenues in Georgia are more a sort of transfer mechanism operated by the center rather than a tax instrument that local governments could exercise any real decision power on or somehow influence.84 Of these 100 percent shared-taxes, land tax and property tax are relatively more important, and could become very promising source of local revenue in the near future, if the right incentive structure is put in place-including a reform of the tax base and value assessment system, which could provide more autonomy to local governments.85 Land tax and property tax are typically local taxes: they have unmovable bases, tend to be less distorting to business activities, usually have identifiable links between payment of the tax and the benefits from local public services, are less vulnerable to business cycles, and are not easily exportable. Local "Own" Taxes and Fees 4.53 Local self-governments are explicitly assigned by the Tax Code with six local own taxes (on entrepreneurial activities, on gambling, on health resort stations, on hotel accommodations, on advertisement, and on the use of local symbols), and by the Law on Local Fees with four fees (trade permit, outdoor advertisement, permit to restrict the use of public places, and car parking). The Law on Local Fees also authorizes the local governments (rayons) to charge three other fees (on construction permit, on local transportation services, and on the changes of building exterior fagade). 82 Since the bulk of the PIT is collected at the place of work, a derivation based revenue allocation introduces a major allocative distortion, since usually workers' families benefit from services provided in the places where they live. This particular problem could be solved by shifting the shared PIT from being collected in the place of work to be collected at the place of residence. 8 It has fluctuated from 0.8 percent of GDP in 1997 to 1.0 percent in 1998-99 and 1.3 percent in 2000, but down again to 1.0 percent in 2001. 8 Despite the enormous pressures that the central government has exercised on local authorities-through different means, including implicit conditionality and political suasion-as a tool to increase collection of both shared and local taxes. International experience has shown, however, that these type of stimulus are never a replacement for an effective structure of economic incentives. 85 The current valuation assessment system for the land tax is based on fixed presumed values, established on tables issued by the central government, which can vary according to the type of use, quality of land and with the location of the lots. Regarding the property tax (real estate tax), enterprises and individuals are charged different rates (1 percent and 0.1 percent, respectively) on the "historical value" (i.e., book value) of the property Both these procedures has implied outdated value assessment which disregards market dynamics. 73 Georgia: Public Expenditure Review November 25, 2002 4.54 The contributions of these local taxes and fees for the large majority of the local budgets in Georgia are very limited (about 0.5 percent of GDP, and 10 percent of local government revenue) (Table A3.1). Some of these taxes and fees are local specific, like the tax on health resort station or the use of local symbols. Others are simply not applicable to small communities, since only urban centers of certain size can have opportunities to collect some revenue from these taxes--e.g., hotel tax, advertisement tax, or parking fee. Nevertheless, the average-sized and larger cities are apparently not taking full advantage of these local taxes and fees to increase their marginal contribution to the local budgets. One reason may be that it is easier for local authorities to make a case with the central government for larger transfers than to persuade local voters to pay additional contributions for local services In any case, the existence of this legal structure of local taxes and fees (including the possibility for the local self-governments and local governments to choose the tax rates in a bandwidth) is already a virtue of the Georgian fiscal system. The central government should, however, closely monitor the implementation of this system by the local governments, realistically updating the coverage and rates, and provide the right incentives to encourage them to use the system to its full potential. One incentive in this direction is to exclude actual collection of local taxes and fees from any calculation of local tax capacity when considering transfers to local governments. This exclusion would encourage the local governments to maximize tax collection from local taxes and fees, without the fear that the central government could indirectly claw back the outcomes of their local tax collection effort. Patterns ofLocal Government Revenues Aggregate Trends and Shares 4.55 Local government revenues from taxes and fees increased to 5.2 percent of GDP in 2001, from 3 percent in 1997 (Figure 4.5). In terms of the total tax revenue local government has expanded its share, from 28 percent to 41 percent between 1997 and 2001 (Figure 4.6). One factor explaining this expansion of local government during the period is the increase in the statutory share of State taxes, while their revenue from both local tax and fees remained small and stagnant (Annex 4.4). 4.56 Local government revenue in Georgia is excessively dependent on State shared taxes (about 79 percent). During the period 1997-2001, about 35 percent of total tax collection and half of the non-tax Figure 4.5: GeorgIa - Tax and Fees Revenue as Figure 4.6: Georgia - Share in Tax Revenue, share of GOP 1997-2001 80 70 60 0. 1100 0050 s"- 80 40 -60 J Local Gvt 0 Central Gv 2&o -- -- -- 20 10 *0 - - Ceral GA 1997 1998 1999 2000 2001 1997 1998 1999 2000 2001 revenue was assigned to local governments (Table 4.5). The contribution of "own local revenues", including local taxes and local fees, was quite low (less than 10 percent of total local revenues). Other non-tax revenue (another 10 percent) fluctuated during the period, and actually decreased to 0.3 percent 86 During the period, Tbilisi, Poti, and Adjara Autonomous Republic exceptionally also shared on the State non- shared taxes (Customs tax, Excise, and VAT). 74 Georgia: Public Expenditure Review November 25, 2002 of GDP in 2001 from 0.7 percent in 1998 (Annex 4.4). The contribution of these non-tax revenue items tend to be unsustainable, since privatization/sales proceeds are only temporary sources. On balance, the general structure and trend of local revenues has shown that local authorities have not yet been provided with adequate incentives to increase efficiency and accountability, since "own local revenue" are still insignificant and local budget too much dependent on shared taxes. Further increases in local government revenues in Georgia will be very much dependent on factors outside the scope of local governments' actions-like general economic growth, discrete increases in shared-tax ratio, improvement in tax payer compliance, and performance of the State tax administration agencies-and only marginally will depend on the local authorities to enforce local taxes and fees. Composition ofLocal Revenues 4.57 The composition of local government budget inflows varies from region to region and among rayons in the same region, but almost the totality of the rayons are either dependent on tax sharing revenue or on transfers from the central government. (Table 4.6). In 2001, local taxes and fees financed about 7 percent of local budgets on average. This corresponds to less than 0.6 percent of the national GDP." Except for Adjara Autonomous Republic, the rayons of Tskaltubo and Borjomi, and the cities of Tbilisi, Poti and Rustavi-which finance between 9 and 12 percent of their budgets with revenue from these sources-rayons have had only an insignificant contribution from local taxes and fees. In more than half of the rayons this contribution has been less than 2 percent. 4.58 Data for 2001 shows that Adjara was, on an exceptional basis, the only subnational government sharing VAT, Excises and Customs duties, which together financed about 50 percent of its budget. The shared taxes, PIT, CIT and Land Tax have been the most significant sources of revenue in the larger and more economically developed centers like Tbilisi, Khashuri rayon, Poti, Samtredia rayon, and Zestafoni rayon-where they finance more than half of the budgets. Land Tax and Property Tax have been the most important of the "own local taxes." Land Tax has assumed more importance for the local governments especially in the Khaketi region-where it contributes to more than 20 percent of the rayon budgets. Property Tax has some budgetary significance (between 10 and 20 percent) in only eight local governments; apparently even Tbilisi has lagged behind and has not exploited the full potential contribution of this tax. Because of the very nature of the Tax on Natural Resources and Tax on Pollution, their bases are more dispersed geographically and they have only been of some relevance in few special places, like Sagarejo rayon, Bolnisi rayon, and Borjomi rayon (in the case of natural resources), and Tetritskaro rayon (in case of pollution). Or 0.8 percent if the other non-tax revenues are added. 75 Georgia: Public Expenditure Review November 25,2002 Table 4.5: Government Revenues, 1997-2001 Total Revenues" Types of Taxes Average 1998-2001 Average 1997-2001 Central Local Total Central Local Total State non-Shared Taxes 79.4 14.8 54.9 91.1 8.9 100 Customs Tax 96 30 72 842 158 100 Excise 172 18 113 93.8 62 100 VAT 52.7 100 364 91 1 8.9 100 State Shared Taxes 11.1 64.3 31.4 23.5 76.5 100 Income Tax 73 256 144 332 668 100 Profit Tax 38 15.7 8.4 301 699 100 Land Tax 79 30 100 100 Property Tax 83 3 1 100 100 Tax on Transfer of Property 1.0 04 100 100 Pollution Tax 40 1 5 100 100 Natural Resource Tax 18 07 100 100 Local Taxes 5.9 2.2 100 100 Tax on Economic Activity 24 1 0 100 100 Gambling Tax 0.4 02 100 100 Resort Tax 00 00 100 100 Hotel Tax 0.0 00 100 100 Advertisement Tax 0.1 00 100 100 Tax on Use of Local Symbols 0.0 00 100 100 Total Tax Revenues 90 5 85.0 886 65 3 34 7 100 Non-Tax Revenues 9.5 15.0 11.4 51.6 48.4 100 Fees 3.8 14 100 100 Other Non-tax Revenues2 95 11.2 100 589 41 1 100 Total Tax and Fees Revenues 91 89 90 62.6 37.4 100 Total Revenues 1000 1000 1000 62.2 378 100 1/ Exclude extra-budgetary special state funds (pension, employment, and road funds) and central transfers 2/ Proceeds from privatization of State properties (including Sales of agricultural and non-agricultural land), traffic policy fines, and special resources. Source: Ministry of Finance, World Bank, IMF. Table 4.6: Composition of Revenues by Territorial Units, 2001 (perce tage of total inf ows) Of Of which Total which Own tax and Sum Terntonal Units by Tax Shared Local Non-tax non-tax Other Of which of "inflows" Regions Revenues Tax Tax Fees Revenues Revenues Inflows" Transfers II Adjara 89.8 87.2 2 6 7.7 1.0 985 1 5 1 5 1000 III Tbilisi 909 82.1 8 8 2.8 4 9 98.6 1.4 0 0 1000 IV Kakheti 598 58 3 1 5 0 9 4 2 64.8 35 2 238 1000 V Imereti 42.1 419 0.2 1 1 10.6 67.7 323 248 1000 VI Samegrelo and 64.9 624 2 5 0 9 5 0 63.4 36.6 274 1000 Zemo Svaneti VII Shida Kartli 825 82.5 00 18 7 5 644 356 255 1000 VIII Kvemo Kartli 523 48.2 4.2 1.0 6.6 878 122 5 8 1000 IX Guria 429 426 0 3 0 8 5 0 435 565 42.4 1000 X Samtskhe- 883 874 1.0 0.8 5 7 55.1 44.9 345 1000 Javakheti I XI Mtskheta- 478 47.4 0 4 0 4 2 8 499 50 1 184 1000 Mtianeti XII Racha-Lechkhu 46.1 457 0.4 0 4 1 8 317 683 526 1000 Kvemo Svanett Total 560 52.3 3.7 3.5 4 3 862 13 8 9 6 1000 Maximum 986 930 107 4 9 19 5 1000 83 9 73 3 1000 Minimum 15.8 156 0 0 0 0 0 0 16.1 0 0 0 0 1000 Average 480 458 2 2 0 7 5 1 53 9 461 335 1000 St Dev 236 218 2 8 0.8 3.9 25.2 25 2 217 0 0 Coef Var. 0 5 05 1.2 1.1 0 8 05 0.5 0 6 0 0 1/ Includes transfers, borrowing, loans, and other inflows. Source: Ministry of Finance. 76 Georgia: Public Expenditure Review November 25, 2002 Disparities in Local Revenue Collections 4.59 An outstanding characteristic of local government units' in Georgia are the enormous disparities in per capita revenue collections across rayons. The largest disparity is between Adjara Republic, Poti and Tbilisi cities as compared to the rest of the country. While total revenues per capita (tax and non-tax) -- are GEL272 in Adjara, Figure 4.7: Georgia - Disparity in Per Capita Revenue Collection GEL 763 in Poti, and o -GEL124 in Tbilisi, the country's average is 250- -only GEL25-with a maximum of GEL96 in Bolnisi rayon and a 200 mminimum of 6 in Chkhorotsku rayon (Figure 4.7). Actually, apart from Adjara, Poti _J and Tbilisi, the 100- ....... ..... . ...... .- "-disparities in total per capita revenue are less pronounced-the coefficient of variation drops from 1.3 to 0.7 when these three 2 3outliers are removed from the data set C q (Annex 4.5). 4.60 Across territorial units, even excluding the outliers (Adjara, Poti, and Tbilisi), the per capita collection coefficients of variation are still large for all taxes; the smallest being for Income Tax (0.7) and the largest for Gambling Tax (7.5). The maximum per capita income tax collection (Kazbegi rayon) is about one hundred times as large as the minimum (Ninotsminda rayon), while the tax on gambling can only be collected in very few larger cities Figure 4.8: Georgia - Per Capita Tax Collection, 2001 (Coeff. of Variation across (Annex 4.5). In the particular cases of the terrrlnit most important shared taxes (on income and on profit) the coefficients of variation 80- in per capita collection are considerably increased when the outliers are included, 70 a E probably because of great differentials in tax capacity. Across territorial units, the coefficients of variation are in general much larger for the local taxes than for the shared taxes, possibly indicating, besides diverse local tax bases, distinct collection efforts exerted by local 0 Cagencies of the State tax administration (Figure 4.8). The data show that, except for the case of the tax on economic Sactivities, , the big majority of the territorial units collect nothing of these local taxes. 77 Georgia: Public Expenditure Review November 25, 2002 4.61 In the particular case of local fees, the coefficient of variation almost quadruples when the three outliners are included in the calculations, suggesting that the economic basis and opportunities for charging local fees in these outliers are incomparably higher than in the rest of the country. All in all, these large disparities in the distribution of per capita collections may be denoting a very distinct revenue mobilization capacity among territorial units. This would indicate that the full reliance on the derivation principle may not be the most efficient and equitable system of revenue allocation among the territorial units in Georgia, suggesting the need for a more transparent, well-designed equalization transfer system. The latter could work as a compensatory mechanism, reducing fiscal disparities among territorial units, and providing all Georgian citizens similar opportunity to access the same standard of public services regardless of where they live. Main Issues and Options 4.62 The current system of revenue assignment in Georgia does mot provide the predictability and autonomy needed to ensure effective local government. Main issues include: (a) local self-governing units are totally dependent on annual negotiations with the rayon authorities to complement their "budget needs" from State shared taxes and transfers. These negotiations undermine any incentive to increase the tax effort and thus remove from the local self-governments the sense of revenue autonomy and accountability; (b) the exclusive reliance on derivation principle for the main State shared taxes (PIT and CI) promotes allocative inefficiencies and exacerbates inequities; (c) local taxes and fees are outdated and need to be reformed in order to be more productive and provide local governments with the right enforcement incentives; and (d) the current attribution of tax administration responsibilities are still too elusive, shifting back and forth between local and central authorities. Owing to political pressures and interests, sometimes both local and central authorities involvements have gone beyond their legal mandates in certain aspects but fallen short in others.88 Moreover, the issue of the dual subordination of local inspectors employed by the State Tax Department has not been resolved entirely. In order to properly address the issues associated to the revenue assignment system, the government of Georgia may consider the following options: (i) assigning fixed State tax shares directly to the local self-governments. This should be explicitly regulated by the organic law, to improve revenue predictability and discontinue the current negotiations between rayons and local self-governments on State-shared tax arrangements. This should revert the perception that revenue increases resulting from local self- government collection effort could be clawed back by higher level authorities. (i) updating and reforming some of the 100 percent shared taxes, local taxes and fees to consider market-based value assessment and ad-valorem rates as far as possible. All specific local taxes and fees rates should be reviewed and updated. In the particular cases of land and property taxes, the government may consider converting these taxes into local taxes and allowing local authorities to exercise some discretion at the margin (e.g., choosing the tax rates in a bandwidth), promoting updated cadastres, and evaluating the feasibility of individual vs. mass assessment valuation criteria. (Centrally regulated fixed/specific nominal values should be avoided to the extent possible or, when deemed necessary, used as a transitional measure.) The government may also consider gradually allowing local (and local self) governments to administer the collection of their own local taxes and fees, starting with the most administratively and technically advanced self-governing units. This policy should enhance the tax-benefit link, the tax effort incentive, local tax autonomy, and political accountability at the local level. 8 Some local authorities report being responsible for pension contribution and payment arrears, while they are unaware about the potential and problems concerning local taxes and fees. Meanwhile, some central officials stationed in the territorial attribute tax revenue shortfalls to inaction of local authorities. 78 Georgia: Public Expenditure Review November 25, 2002 Especially in the cases of the entrepreneurial tax, land tax and property tax, this approach could establish a business- and wealth-related revenue base for the local self-government units, which could serve as an important complementary source of financing. (iii) reviewing the derivation basis of the State shared-taxes. This is the most critical step towards addressing the issue of the existing fiscal disparities in Georgia. One possible option is to reassign the local share of both PIT and CIT to an Equalization Fund (EF), while preserving the same source of funding for the local governments (i.e., without changing the sharing ratios to avoid any major disruption in the aggregate financial flows between the central and local budgets). The basic reform should involve the following changes: , * splitting the PIT into two components. one national progressive tax and one local proportional tax. The national progressive rate would be similar to the current one, only slightly adjusted downwards to reflect the fiscal neutrality of the reform and keeping the PIT tax burden the same. The proceeds from this component should be reassigned to a special "Equalization Fund" (EF). The EF would only function as a transit account, just serving the purpose of calculating and immediately distributing the collected funds to the subnational governments according to a pre-established equalization mechanism (see below). The local proportional rate would be a local tax. This would be a flat proportional rate (say, a maximum 2 percent local income tax) that the local self- governing authorities could choose to charge. This local flat rate component of the PIT would be collected on a derivation basis to encourage local tax effort and help solve vertical imbalances. (This will provide an enormous positive incentive for increased collaboration between local and central authorities in tax administration as well); * reassigning CIT local share entirely to a "EF." In the face of the difficulties to properly apportion the CIT tax collection in Georgia, the best alternative to avoid the current distortions and unfairness of the derivation basis revenue allocation is also to reassign the aggregate CIT local share to the EF. (iv) reviewing and properly attributing responsibiUties in the tax administration. In Georgia, administration of State and local taxes (shared and non-shared) is an exclusive responsibility of the State Tax Administration, but it is not uncommon to see the involvement, voluntary or induced by central government, of local authorities in these services. This has led to major confusion of responsibilities, while the system remains unaccountable. The local authorities should be made legally and politically accountable only for the collection of taxes and fees on which they can exert significant decision-making, and should be clearly discharged from State responsibilities which have not been specifically delegated to them by law-for example, responsibilities for the collection of State taxes, for the contributions and payments of pensions, and other State functions which local governments have sometimes claimed authority on, or central authorities have pushed down responsibility for failed implementation. Clarification of these responsibilities is critical to improve transparency, efficiency and accountability, which are essential elements for proper self-governance of a fiscally decentralized system. Fiscal Imbalances and Transfers Fiscal Imbalances 4.63 The data show substantial disparities in both per capita expenditures and per capita revenue collections across local government units in Georgia. Also, local governments claim that revenues are not enough to cover needs. These are key intergovernmental fiscal issues associated to the efficiency and equity of the system of delivery of public services. The relevant dimension of the inequality challenge is 79 Georgia: Public Expenditure Review November 25, 2002 the potential mismatch of resource and responsibilities between the State and the local governments (the vertical imbalance) and among the local governments themselves (the horizontal imbalance). 4.64 Vertical fiscal imbalances measured through the (ex-post) budget deficit may be a biased measure towards the central government, since the center can often absorb some responsibilities of the local governments (especially capital expenditures) because it can borrow more freely to finance public expenditures. A more relevant measure of the vertical fiscal imbalance is the size of the structural (ex- ante) deficits. The latter reflects each government level's ability to finance expenditures from their own revenue sources. But this measure can vary substantially according to the definition of "own revenue sources". If own revenue sources include both all shared taxes plus local tax and fees- for 2000 and 2001 the index of vertical imbalance for the local governments was 0.8 and 0.9, where the value 1 for the index would represent a perfect vertical balance. This indicator of vertical imbalance would compare favorably to other decentralized countries. However, if one considers a stricter definition for "own revenue sources"-by taking into account only the sources over which the local governments actually have some degree of autonomy-the index would fall to below 0.1 percent, indicating an enormous vertical imbalance against the local governments in Georgia. 4.65 Horizontal fiscal imbalances result from mismatches between tax capacity and expenditure needs across local governments. However, the observed considerable discrepancies in per-capita revenue collection and expenditures are only ex-post measures, and may be biased by distinct tax efforts and distorted allocation incentives. Relevant measures of horizontal imbalances, useful for corrective policies, have to focus on potential (i.e., ex-ante) tax capacities (or tax bases) and expenditure needs, rather than on actual tax collection and cash outlays. 4.66 Sub-national governments in Georgia differ substantially in terms of their ability to raise revenues. Their economic bases are distinct in terms of natural resource endowment, level of agricultural and industrial development, employment, productivity, which directly affect the local tax capacity. For example, Table 4.7 shows the per capita composition of the GDP by region, though on a very aggregate level. Even at this level of aggregation the table shows that the per capita GDP in Samtskhe-Javakheti region is only about 60 percent of the per capita GDP in the Tbilisi region, and about two-thirds of Guria and Shida Kartli. The composition of output is even more variable among regions, with urban based activities, like industry and services, being more developed in Tbilisi, Shida Kartli and Guria as compared to Samtskhe-Javakheti and Samegrelo. Wage differentials are also large among regions. The maximum regional average wage (Adjara) is more than twice the minimum regional average wage (Imereti and Qvemo Svaneti). These discrepancies in tax bases are probably much larger at a more disaggregated level-among the rayons and at the local self-government level. 4.67 Also public expenditure needs among local governments in Georgia are substantially distinct, because of unequal cost of public service provision (e.g., civil security, transportation, pollution, sanitation, housing, energy, water supply) and the share of local population with special needs (e.g., refugees and internally displaced population, children to be educated, incidence of contagious disease, elderly and poor people). For example, certain localities of Georgia accommodate larger percentage of refugees than the rest of the country. The city of Tbilisi, for being the capital city and a large industrial and metropolitan area, has special needs mainly related to transportation, housing, sanitation. The mountainous areas of the country are colder and scarcely populated, having higher energy consumption needs, higher education, transportation and communication costs. Also, poverty incidence varies significantly according to the region. For instance, the poverty incidence in Samtskhe-Javakheti region is 4 times as high as that in Samegrelo. 80 Georgia: Public Expenditure Review November 25, 2002 Table 4.7: Gross Domestic Product, By Region (G L per capita), 2001 i 82 1 6 1 - cc U 0 6 aU 36 6 G 2 0 1 3 S 75 Imeret &4ahalch5v3 79 192 34 128 1 38 62 57 12 142 656 Totda 25tket-3879 63 7 16 85 171 148 4436 51 68 416 1342 693 Region 1996 1997 1 19 000 Kakheti 513 89 13 132 81 8 16 63 12 . 801 7 Tbilisi 5 31 2. 98 217 236 72 5 12 566 168 121 Shida Katl48i 9 227 9.6 14 4 39 55 241 1413 69 Kvemo Kartli 27 1.9 1 4 101918 52 45 27 10 4.69 Samtskhe-Javakheti 31693 539.5 9 14.7 2 242 980 .3 AjaraA22 6 1 2 3 136227511.3 294.128 1324 Guria 416 93 21 2 142 14 3 135.7 24.6 1423 5 Samegrelo&ZemoSvaneti 345 70 78 31 130 111 26 29 55 165 1040 65 Imereiti & Racha-lech, Qv 366 79 192 34 172 132 38 62 57 270 1402 56 Sv Total 1257 63 1168 153 1171 1148 1 36 151 168 1316 1342 93 Source: State Department of Statistics. Table 4.8: Poverty Incidence by Region (ercent) Region 1996 1997 1998 1999 2000 Kakheti 12.0 13.2 11.6 15.0 17.9 Tbilisi 12.6 9.6 15.5 13.5 11.7 Shida Karthi 16.9 9.6 14.9 12.9 13.5 Kvemo Kartli 14.9 10.9 11.8 12.2 14.6 Samtskhe-Javakheti 16.3 9.5 14.7 24.2 29.3 Adjara 13.2 7.7 11.3 14.1 13.4 Guria 161 14.2 13.7 11.6 12.5 Samegrelo 9.4 4.6 13.2 14.2 7.3 Intereti 123.0 115.3 120.5 20.3 13.8 Source: Poverty Monitoring in Georgia, Re. 2000-State Department of Statistics. 4.68 The incidence of these factors associated to tax capacities and public service expenditures has a differential impact on horizontal fiscal imbalances. The current system of tax assignment solely based on the derivation principle does not take these factors into account and tend, instead, to exacerbate the existing disparities, since tax bases tend to be inversely correlated with regional poverty indicators. 89 Thus, there is a risk that in the process of economic development, the poorest regions are left behind and this may not be in the interest of the nation in the long run. A balance between efficiency and equity B9 For example, poverty incidence indexes (Table 4.8) are negatively correlated both with per capita GDP and average wage (Table 4.7). 81 Georgia: Public Expenditure Review November 25, 2002 concerns can be effectively stricken mainly through compensatory grants or transfers to reduce horizontal fiscal disparities, so that every region could provide similar standards of public services to every citizen. Once a reasonable equalization mechanism is in place the problem of the vertical imbalances should be resolved mainly through an adequate degree of revenue autonomy to the local governments so that they can respond to the specific demands of local population at their own cost. Transfer Mechanisms 4.69 The Georgian government operates a system of transfers to local governments intended to cover the costs of delegated functions to the local governments by complementing their revenues from the State shared and local taxes and fees. The basic approach for transfers in Georgia is to "fill the gap" between the expenditures and local revenues. Although the transfer system is not regulated in detail, these complementary transfers are in principle legally mandated, since the organic Law on Local Governance and Self-Governance requires that delegated responsibilities must be entirely funded, either by State shared taxes or transfers. 4.70 In practice, the amount of transfer is negotiated on an annual basis with the rayons, and the final aggregate amount is approved by Parliament in the context of the annual budget law. Own local taxes and fees are meant to finance mainly the communal/ utilities costs. But, as the local self-governments also participate in the delivery of most public services, they also receive transfers-an undefined share of the State shared tax-also through negotiations with the respective rayons. Since the basic principle is to finance State delegated expenditures, until recently all these (complementary) transfers used to be earmarked, especially to cover "protected expenditures" (e.g., teacher wages). It is reported that starting in 2002 these transfers are not earmarked anymore, although their amounts are determined as before on the basis of specific expenditure norms and the "gap-filling" concept. 4.71 Until mid-2000, there was an accumulation of arrears on these delegated responsibilities at local levels, because the State budget fell short of fulfilling its commitments almost systematically. For example, in 1998, 1999 and 2000, the execution ratio of the budgeted transfers to local governments were only 62 percent, 69 percent, and 79 percent, respectively. During that period, there was a great variation in the execution ratios across rayons, without any apparent relation with local tax capacity or needs, reaffirming the perception of ad hoc allocations of these transfers. Only in 2001, the budgeted transfers is said to have been fully executed. 4.72 In the past the government has tried to get away from the negotiations of transfers and tried to change the system towards a formula driven mechanism, but so far unsuccessfully. In 1999, the IMF proposed some changes in both the tax sharing and the transfer systems, aiming at removing some unjustified regional privileges.0 Some of those proposed changes in the tax sharing system have been implemented since then (e.g., elimination of the 15 percent VAT collected in Tbilisi, and the 5 percent of customs duties collected in Poti), but other proposals related to the increase of the central transfer pool and the introduction of a purely formula driven system for the distribution of the transfers were not implemented. 4.73 Currently, the government is studying alternative formula mechanisms to design a new transfer system and propose it to Parliament. The current government simulations for a new transfer system are based on a proposed formula which takes into account, for each rayon, basically the following: (a) tax capacity (considering all taxes and non-tax revenues) and (b) population (excluding IDPs). The formula includes special adjustment coefficients for each rayon, which considers (i) proportion of population 9 "Georgia: Issues in Intergovernmental Fiscal Relations", FAD/IMF (by W. Mahler, G. Brosio, and C. Donian), June 1999. 82 Georgia: Public Expenditure Review November 25, 2002 residing in high mountains; (ii) proportion of refugees and IDPs; (iii) proportion of veterans and similar; and (iv) a "facilitating factor" to take into account potential increase in revenue.91 Main Issues and Options for Reform 4.74 The most serious concern with the current system of transfer is that it is still based on negotiations and the old Soviet "gap-filling" approach. This generates perverse incentives in that local governments are not encouraged to increase their own revenue, because if they do so their expected transfers will be smaller-this means that local governments are penalized if they are efficient. Although it has been said that the "gap-filling" transfers are based on expenditure norms and revenue collection potentials on rayon individual basis, there is no clear rational for determining the aggregate amount of transfers. 4.75 Most of the recent simulations have only considered distributing a fixed, arbitrary amount of transfer pool, which seems to be historically determined without any specific rational. Moreover, the total amount of transfers is not linked to any specific sources of resource that could be predicted and that could fund the transfers on a sustainable basis. Actually, the total amount of transfer-both the initially budgeted and approved, and the final execution-has fluctuated from year to year, which increases unpredictability of resources for the local governments. 4.76 An approach to properly address these issues is to combine the solution for the current ineffectiveness of the transfer system simultaneously with the solution for the current inefficiencies and unfairness of the derivation basis shared tax system, as follows: (i) reassigning the local government shares of the PIT and the CIT to the EF. (As discussed earlier.) In the future, distinct desired levels of equalization can be achieved by submitting to Parliament adjustments in the sharing ratios of the progressive PIT and the CIT. This option has the advantages of simultaneously: (a) eliminating the inefficiency and unfairness of the current tax sharing system purely based on the derivation principle;92 (b) providing a definite, predictable and sustainable source of funding for the equalization transfer pool; and (c) eliminating the "gap-filling" approach and the annual political negotiations that define the aggregate amount of transfers to the lower level governments. (ii) distributing the EFpurely on the basis of an equalization formula. The formula should exclusively target fiscal equalization, and preserve autonomy;93 as such the resources should be transferred on a non-conditional, non-earmarked basis. The equalization formula should be based on few indicators of key determinant factors for the tax capacity and expenditure needs, as mentioned above. The formula should be as simple as possible (to be accessible and transparent), and the indicators entering it should be out of the control by any beneficiary, and be immune to any bureaucratic or political influence. 4.77 This new system should help establish predictability, autonomy and accountability at all government levels. At the same time, it will open the possibility to define more clearly the strategic 9 The previous IMF simulation distribution formula considered a fixed (lump-sum) element, population, educational needs, mountainous regions' special needs, declining industrial areas' special needs, revenue capacity, and revenue effort. 92 But still keeping the tax effort incentive in place, especially for the local taxes and fees. 9 Which would not eliminate the possibility of the existence of other types of transfers, e.g., conditional capital grants for investment of national interest, conditional transfers for the financing of a national vaccination program to combat contagious disease to be delivered locally, etc. 83 Georgia: Public Expenditure Review November 25, 2002 direction on intergovernmental fiscal relations for the near future in Georgia. Such a strategic orientation, by minimizing negotiations on transfers and eliminating the current distortions of the shared tax system, will allow the central administration to get away from matters of strict local interest and focus on macro management issues. This strategic orientation will also encourage local governments to focus on maximizing collection of their own tax and fees, and on the cost-effectiveness of local service delivery. 84 Georgia: Public Expenditure Review November 25, 2002 Chapter 5. Expenditures in the Health Sector 5.1 Georgia's health sector is characterized by chronic under-funding, excessive facilities and staff, poor quality services, and limited access. Although budgets are low, those modest transfers are further reduced by low transfer compliance by the MoF. More egregious is the high level of inefficiency and low quality in the use of resources, that undermine the system. Finally, equity is seriously compromised by the high out-of-pocket payments that reduce access in generally, but particularly for the most vulnerable. Health Policy in Georgia 5.2 Prior to independence in 1991, the Georgian health system was characterized by strong central government control over all aspects of health care financing and provision. The economic crisis that followed the establishment of an independent state in 1991, led to the collapse of government finances for the social sectors. By 1994, the government was spending less than one dollar per-capita on heath care. 5.3 In 1995, Georgia initiated health reforms aimed at shifting the orientation of the health system towards a market-based system while ensuring a basic level of health care for all. Key goals of the health sector reform program included: promoting equity in access; fostering economic efficiency; and providing freedom of choice for consumers and appropriate autonomy for providers. These policy objectives translated into a set of reforms that completely transformed the health system. New sources of revenue for the health sector were mobilized through the introduction of a social insurance system, and mandating municipalities to contribute to health care. The financing of health services was separated from provision and all health facilities in Georgia were granted autonomy allowing them to contract with a range of purchasers and charge user fees. The role of the Ministry of Health changed from direct financing and provision to policy-making and regulation of the health sector. Primary health care was emphasized over curative care, and the Government made a commitment to restructuring and downsizing the health sector to better meet the needs of the country. 5.4 The basic tenets of the Health Policy defined at the beginning of the health sector reform process in 1994-95 have not changed much over the years. More recently, there has been a renewed emphasis on preventive and primary health care, increasing public expenditures on health care to 4 percent of GDP by 2005, addressing the health care needs of the poor, and increasing the coverage of vulnerable groups under social insurance. Overview of the Health Sector Health Outcomes 5.5 Available data indicates that Georgia's health outcomes measured in terms of life expectancy at birth, infant and under-five mortality rate and maternal mortality rate compare favorably with other CIS countries. (Table 5. 1).94 However, it should be noted that Georgia's health indicators have worsened significantly over the past decade. The infant mortality rate (IMR) has increased from 9 in 1991 to 24 deaths per 1000 live births in 2000. Similarly, the maternal mortality rate (MMR) in Georgia is 51 per 100,000 live births which, while low in comparison to other developing countries, is way up from the MMR of 32.4 reported in 1993. The majority of maternal deaths in Georgia (38 percent ) occur from bleeding, one of the causes of which is anemia which typically needs to be diagnosed and addressed during the prenatal stage. This is yet another indication of poor health system performance. 94 Data on health outcomes is Georgia are not very reliable since disease surveillance is rather weak. 85 Georgia: Public Expenditure Review November 25, 2002 5.6 Similarly, incidence rates of infectious and parasitic diseases in Georgia have increased in the last decade. For example, there has been a huge increase in the number of new tuberculosis cases, from 1,531 in 1991 to 4,515 in 1999. Also, while the number of diphtheria cases has declined from 425 in 1995 to 60 cases, it is still considerably higher than 1992 (when 3 cases were reported). The situation in non- communicable diseases has also worsened. Overall, circulatory system diseases cause about 71 percent of all deaths in Georgia, followed by neoplasms (11 percent). Other social illnesses, particularly cases of drug addiction, are on the rise (from 1,347 in 1984 to 4,873 in 1996). Table 5.1: Health Outcomes in Georgia and Other CIS Countries (2000) Countries GNP per Life Crude Death Maternal Infant Mortality Children capita Expectancy Rate (per Mortality Rate Rate(per 1,000 Mortality Rate (US$95) (years) 1,000) (per 100,000 live births) (under 5, per live births) 1,000 live births) Georgia 620 73 8 51 24 29 Other FSC Countries Armenia 490 74 6 35 25 30 Azerbaijan 460 71 6 43 74 105 Kazakhstan 1,250 65 10 70 60 75 Kyrgyz Republic 300 67 8 65 53 63 Tajikistan 280 69 5 65 54 73 Uzbekistin 720 70 6 21 51 67 ECA 1,990 68 -- -- 20 26 Low-income 420 59 -- -- 77 116 Countries Source: WDI, 2001, UNICEF (for Infant and Under-5 mortality rates) Coverage and Utilization of Health Services 5.7 Under the Soviet system, the entire population was fully covered by the State for all levels of health care. In the immediate post-transition phase (1991-95), the old system of covering the entire population for all services continued on paper, but very few people were actually covered since the amount of public resources flowing into the system was negligible. For the first time since the introduction of the Soviet system, Georgians had to pay substantial amounts of out of pocket funds for health services. 5.8 The health sector reforms of 1995 introduced a basic benefits package covered through the social insurance system and municipal health funds, principally designed to optimize the use of limited resources. The benefits package covers all health services for children less than 3 years of age, rural populations, and for internally displaced persons. For others, the benefits package covers outpatient and emergency care, maternal health services, and specific diseases (tuberculosis and cancer). 5.9 Health services utilization statistics in Georgia are lower than the average for the ECA region and for low and middle income countries. On average, Georgians recorded 3.2 outpatient visits per person per year during 1990-98 , lower than the ECA average of 6 visits and low and middle income countries average of 4 visits per person per year. In the rural and high mountain areas of Georgia, outpatient visits per capita are even lower (less than 1 visit per capita). Similarly, the inpatient admission rate is 4.3 percent, which is also significantly lower than the ECA average of 17 percent and low and middle income countries average of 7 percent. Both outpatient visits and hospital admission rates are much lower compared with the pre-transition period, when health services were provided largely free of cost. 86 Georgia: Public Expenditure Review November 25, 2002 Health Expenditures 5.10 In 2000, Georgia spent approximately GEL220.3 million on health care of which approximately 73 percent is estimated to be private out-of-pocket and 27 percent is public spending. This represents about 4 percent of GDP, equivalent to US$20 per person. In per capita terms, Georgia is one of the lowest spenders in the ECA region, after Azerbaijan ($9), Tajikistan ($13), and Kyrgyz Republic ($15) (Annex 5.1). Public Expenditures on Health 5.11 As mentioned earlier, public expenditures can be justified on efficiency and equity grounds. In the case of the health sector, there are certain health care goods and services which are not likely to be provided at all by the private sector or if provided they are provided at socially sub-optimal levels. Examples include the prevention and treatment of communicable diseases (tuberculosis, HIV/AIDS), health education, and addressing environmental health problems such as pollution. Since the poor cannot always afford health care, the government can also play an important role in financing and promoting equitable access to health care that would improve their productivity and well-being. 5.12 Public expenditures on health in per capita terms have increased over past two years from about $5 in 1997-99 to nearly $8 in 2001, but still represents only 1 percent of GDP compared to 3.5 percent of GDP in the CIS region. An analysis of these expenditures in Georgia shows that almost 65 percent of public spending on health is on preventive and primary health care. While all preventive activities fall in the category for which public financing is justified, it is less clear what percentage of primary health care goods and services are "public" goods and what percentage are "private" goods, and which are a mix. Sources 5.13 The main sources of public sector funds for health care are: (i) general budget financing; (ii) payroll taxes; and (iii) allocations from municipalities. 5.14 General Budget. General budget financing is an important source of health care funds in Georgia. Central government funds are channeled through two budgets General Budget I and General Budget 2. General Budget I is an allocation from the central budget approved under the budget law on an annual basis. There are no earmarked funds for health, and the budget is negotiated between the MoF and the MoLHSA. The final decision on the budgetary allocation lies with the Parliament. There are three recipients of General Budget 1: (i) Ministry of Labor, Health and Social Affairs (MoLHSA); (ii) Department of Public Health (PHD); and (iii) State Medical Insurance Company (SMIC). General Budget 2 is also an allocation from the central budget approved under the budget law, without any earmarked funds for the health sector. The difference is that the recipients of these allocations are (i) Ministry of Defense; (ii) Ministry of Internal Affairs; (iii) Ministry of State Security; (iv) State Border Defense; and (v) Government Security. For both General Budgets I and 2, the main source of funds is government revenues obtained from taxes and other non-tax income, such as revenues from privatization programs. Budgetary allocations can, at times, be tied to some revenue sources as, for example, federal government allocations to the SMIC in 1999 were tied to revenues from the privatization programs. Between 1997 and 2000, the actual amounts allocated to the health sector from general budget financing has averaged between GEL40 million, equivalent to about to approximately 3 percent of general budget expenditures during the period. 87 Georgia: Public Expenditure Review November 25, 2002 Table 5.2: Public Expenditures on Health (1997-2001) 1997 1998 1999 2000 2001 Public Expenditure on Health (GEL millions) 37.7 44.9 42.7 58.6 82.2 o/w Central Budget -- -- -- 20.7 37.7 Local Budgets -- -- -- 17.6 23.5 SMIC (payroll taxes) -- -- -- 20.5 21.0 Public Expenditures on Health (percent of GDP) 0.8 0.8 0.8 1.0 1.3 Public Expenditures on Health (millions, US$) 18.8 22.2 21.1 29.6 39.2 Per capita Public health expenditure ($)" 4.3 5.1 4.8 6.7 8.9 Real Public expenditure per capita (GEL 2000 prices) 10.9 12.6 10.1 13.3 17.8 Real Public expenditures per capita (US$. 2000 prices) 1/ Assumes population of 4 4 million according to 2002 census. Source- World Bank Calculations based on MoLHSA data. 5.15 A major issue in federal funding for the health sector is the gap between planned allocations and actual receipts. Like other ministries, the Ministry of Labor, Health and Social Affairs also submits a budget to the Ministry of Finance each year. However, the amount requested each year far exceed the amounts that are actually received, and the average budget execution levels for the health sector range between 45 - 62 percent. In 2001, budget execution for the health sector has improved substantially and reached approximately 85 percent. In addition to the problem of budget execution, flow of funds into the health sector is extremely erratic. For example, in 1999, 30 percent of the annual budget for health care was released in May, while during the other months, funding levels varied between I - 14 percent.95 Officials at the Ministry of Health, indicate that uncertainty over budget flows is probably more problematic than the low level of funding. 5.16 Payroll Taxes. Georgia introduced social health insurance in 1995, and a semi-autonomous organization - SMIC - was created to facilitate the pooling of resources. Participation in SMIC is mandatory for all formal sector employees from the public and private sectors. Financial support for SMIC comes from employees who make payroll contributions, currently equivalent to 1 percent of salaries, and by employers, who contribute 3 percent of the formal wage bill. Revenues from this "3+1" financing, as it is popularly known, are directly deposited into the accounts of SMIC. 5.17 Planned revenues from payroll taxes tend to be very inadequate to meet all the needs of the benefit package. The formal sector is small, unemployment levels are high, and only about 29 percent of all Georgians are employed in the formal sector. As a result, the dependency ratio is very high and unfavorable for adequate resource mobilization. Tax evasion and poor collection further exacerbate the problem, so much so that actual revenues collected from payroll taxes are significantly less than what is planned at the beginning of the year. Actual payroll tax collections as percentage of planned collections were 55 percent and 68 percent in 1998 and 1999, respectively. * World Bank. 2001. Georgia Health Care Financing Note. The World Bank, Washington DC. 88 Georgia: Public Expenditure Review November 25, 2002 Table 5.3: SMIC Revenues: Planned and Actual (1998-99) In GEL millions 1998 plan 1998 Actual 1999 Plan 1999 Actual Central budget transfer 15.4 11.3 18.5 9.6 Payroll taxes (3+1) 23.0 12.6 24.0 16.3 Other income - - - 0.8 Total 38.4 23.9 42.5 26.7 Source: SMIC, 2001. 5.18 Allocations from Municipalities. Municipalities are the third main source of public sector funds. In 2000-2001, municipalities contributed approximately 30 percent to total public sector allocations for the health sector. There are two elements that determine the volume of local government funding. First, the MoLHSA mandates that local governments spend at least 2.5 GL per capita per year or 10 percent of the municipal budget on health care, whichever is greater. Second, the municipalities may, on a voluntary basis, allocate more resources to health care based on local needs and preferences. Municipalities spend approximately 5.2 percent of total expenditures (average across all municipalities) on health, which is higher than the 2.9 percent to 3.0 percent spent by the state. Some municipalities like Batumi and Tbilisi spend almost 16 percent and 10 percent respectively of total expenditures on health, while others like Kazbegi spend only 0.3 percent. As in the case of general budget financing, budget execution for health care is a problem in municipalities and rarely do municipal health funds receive the funds committed at the beginning of the fiscal year. Health Care Purchasing Agents 5.19 The majority of public sector funds for health care in Georgia (90 percent) are allocated for the financing of a Basic Benefits Package (BBP) managed by the three purchasers (State Medical Insurance Company, Municipal Health Funds and the Public Health Department). Currently there are 28 programs under the BBP and each purchaser is responsible for a different program. The remaining public resources go to the other purchasing agents, like the Ministries of Defense, Internal Affairs, State Security, State Border Defense and Government Security. 5.20 State Medical Insurance Company. The SMIC is responsible for approximately 17 programs included in the BBP (Annex 3) and has [about GEL30 million] per year to fund these programs. The SMIC contracts with public and private health facilities to deliver the BBP. One of the major problems faced by SMIC is chronic under-funding, and erratic flow of funds, to meet the BBP needs. In fact, between 1998-2001, the number of BBP programs managed by SMIC has doubled while its budget in real terms has not grown. Also, despite facing chronic under-funding, SMIC has not cut back on the volume and size of contracts with health facilities. As a result, SMIC has resorted to cut the amounts available for all programs, including the most cost-effective ones such as medical services for orphans, children 0-3, infectious diseases and tuberculosis. Moreover, reimbursements rates during 2000-2001 have been better for less cost-effective programs (90-100 percent), such as psychiatry, cardiosurgery, oncology, than for the most cost-effective ones (60-70 percent). There are also regional differences in SMIC reimbursement and Tbilisi claims have a higher chance of getting reimbursed. 5.21 To combat the problem of recurrent arrears, SMIC recently adopted a program of global budgets whereby there is a cap on the total amounts that SMIC will pay each facility. This has addressed the problem of arrears but has not necessarily contributed to improving the quality of health services provided by facilities. In addition to the problem of under-funding, the clinical protocols used to determine the type of care provided under the SMIC BBP are outdated. Consequently, the unit cost of these services are much higher than these should be. For example, acute respiratory infections among children are mainly treated on an inpatient basis, while international protocols show that respiratory illnesses among children can be treated very cost-effectively in a primary care setting. 89 Georgia: Public Expenditure Review November 25, 2002 5.22 Municipalities. Supported entirely by own funds, that include both state transfers and local collections, the municipalities finance and provide outpatient care, pain killers for patients suffering from oncological diseases, anti-epidemic measures, sanitary surveillance and ambulance services. In addition, the municipalities also provide additional services for vulnerable groups. As in the case of SMIC, municipal health funds contract for services that are less than the amounts that they annually receive from municipalities. Consequently, MHFs have a recurrent problem of arrears with providers. Although the exact magnitude of municipal arrears in the health sector is not known, it is most likely much higher than SMIC arrears. 5.23 Public Health Department. PHD is entirely financed by the state budget, and received about 8 percent of all budgetary expenditure on health in 1999. PHD finances and provides a variety of preventive services, such as prevention of HIV-AIDS and STDs, prevention of micronutrient deficiencies, prevention of cardio-vascular diseases, and of malaria, trauma and drug-addiction. In addition, PHD finances and provides for the health information system, epidemiological surveillance and promotion of health and healthy life-style. Resource Allocation 5.24 Approximately 52 percent of total public expenditures on health are allocated for primary care, 21 percent for secondary, 15 percent for preventive, 4 percent for drugs and 8 percent for other (supporting Ministry of Labor, Health and Social Affairs, other administrative functions). Since the implementation of health sector reforms, public purchasers directly contract with facilities, and expenditures on salaries and wages, equipment, utilities, operations and maintenance have been shifted to the facility level. There are no performance contracts, and facility expenditures are not monitored. Information from two micro- studies of 34 hospitals in Georgia and 6 policlinics in Tbilisi show that these facilities have very high fixed costs (80 percent of total costs), mainly attributable to wages and salaries. The remainder of the costs are attributable to purchasing of drugs and other supplies, and operations and maintenance. 5.25 In terms of economic classification of expenditures, 80 percent of all public spending goes to the payment of salaries and wages of personnel employed in the health sector. About 10 percent of all spending takes place on equipment, followed by about 5 percent on utilities. About 4.5 percent of all spending goes to maintenance, and a very small amount equivalent to 0.5 percent is allocated to expenses related to depreciation. 5.26 Hospital expenditures on drugs and other supplies are inadequate, and it is estimated that hospitals spend GEL3.9 per admission and GELO.5 per patient day on supplies which is far below any reasonable level. The majority of hospitals do not have any funds for capital investment or asset replacement which leads to a gradual consumption of capital assets and results in the breakdown of these hospitals. Hospitals are very often unable to pay their utility bills, which are written off by the government. Finally, official salaries paid to health personnel are extremely low, and far below minimal subsistence level for the general population. Private Expenditures on Health Care 5.27 Private spending consists of out-of-pocket payments in the form of official co-payments, fee-for- service payments and informal payments. Reliable estimates of private expenditures on health are not available for each year. One of the first assessments was made in 1998 based on a 1995 UNICEF household survey, which estimated private expenditure on health to be almost GEL272.7 million. Since 96 Hospital study already quoted. Gotsadze G. and Bennett S. 2001. Financing of Primary Health Care in Georgia, Tbilisi, Georgia. 90 Georgia: Public Expenditure Review November 25, 2002 then, several national household surveys have been conducted, especially in recent years. Results of these surveys vary, and it is difficult to say with certainty how much is spent out-of-pocket on health care. Different sources of data on out-of-pocket spending indicate different figures for out-of-pocket spending, ranging from 66 to 87 percent of total health spending. For example, a recent study by the Department of Statistics, Georgia (SDS) found that out-of-pocket payments were approximately $28 per capita or 73 percent of total health expenditures. In 2000, according to the Tbilisi Household Survey and the Quarterly Survey of Georgian Households, total out-of-pocket expenditures amounted to GEL132 million, or about 66 percent of total spending on health. Most of the out-of-pocket payments go towards the purchase of drugs (53 percent), followed by hospitalization (23 percent), and outpatient services (17 percent). In addition to formal co-payments and fee-for-service payments, informal payments are probably significant and are very much part of the system. Private health insurance in Georgia is not significant, and amounts to less than GELO.5 million per year. Efficiency of Public Expenditures 5.28 Public expenditures on health care do not necessarily meet allocative efficiency criteria. While originally designed to make the most effective use of limited public funds for health care, the BBP has grown and currently covers programs which do not meet the public or merit good criteria. Overall, the public expenditures envelope for health is small, and these small amounts are spread thinly over many programs. Moreover, prioritization between competing programs and needs is ad-hoc and often does not correspond with the criteria of ensuring the highest positive externalities. The net result is that maximum efficacy gains for the society are not realized. 5.29 Georgia inherited a health system characterized by misallocation of resources, underutilization of capacity in some areas and under-supply in others, and shortage of drugs and medical supplies. Since then, several waves of health sector reforms have focused on downsizing, restructuring and privatization of health facilities, especially of hospitals. Despite these attempts, Georgia still has a high number of specialized hospitals, and a bed-to-population ratio of 4.1:1,000 population, significantly higher than ratios in the much better-off developed countries. A bigger problem exists with the reorganization of medical personnel. The number of physicians in Georgia has hardly declined over the years and in 1999, Georgia had 21,520 physicians which translates into 3.9 physicians per 1,000 population, which is higher than the CEE and FSU averages of 2.5 and 3.7 respectively. At the same time, the number of nurses in Georgia is low, with 1.4 nurses per physician compared with 2.4 in the Central and Eastern Europe (CEE) and 2.1 in FSU. 5.30 Efficiency in public spending can also be examined by looking at technical efficiency measures such as the average length of hospital stay, staffing ratios, prescription of drugs, mix of factors of productions, etc. The average length of hospital stay (ALOS) in Georgia is 10.6 days, which compares favorably with other countries in the region, and is slightly below the ECA and low and middle income country averages of 14 and 12 days respectively.97 However, hospital occupancy rates in Georgia are very low (31 percent), indicative of both low levels of utilization as well as excessive number of hospital beds. There is also the severe problem as a result of excessive staffing, as is evident from the ratio of 1.5 physicians per occupied bed. Moreover, the majority of doctors are specialists, which contributes to increasing costs in the production and delivery of health care.98 9 World Development Indicators, 2001. 9 Zoidze A, Gzirishvili D, Gotsadze G. July 1999. Hospital Financing Study for Georgia. Small Applied Research 4, Bethesda, MD: Partnership for Health Reform Project, Abt Associates Inc. 91 Georgia: Public Expenditure Review November 25, 2002 5.31 The health sector reforms designed to restructure the health system through market-based mechanisms, through purchaser-provider spilt and the introduction of contracting, have not really worked. Public purchasers do not have the capacity and lack the political clout to screen health facilities and selectively contract with those that meet specific criteria. This has contributed to a situation where almost all health facilities in Georgia are contracted under public funds. In addition, the small size of public funds compared with out-of-pocket spending has meant that the power of the purchasers has been reduced. Finally, the adoption of accreditation and licensing system has occurred slowly. As a result, there are few facilities that have been closed down by the Government. Equity in Health Expenditures 5.32 A little over one-fifth of the population of Georgia lives below the national poverty line (June, 2000). About half of the poor, i.e., around 10 percent of the population, live in extreme poverty. Ill health is both a dimension of poverty and an important factor in generating poverty. A recent poverty profile of Georgia shows that approximately 60 percent of the population are at risk of falling into poverty, and the costs of seeking health care is one major factor (World Bank, Poverty Assessment, 2001). Public Spending on Health 5.33 In the absence of data on utilization by income quintiles of publicly-provided or financed health services, we examine allocation of municipal budgets, the only public expenditure on health data available by region, and the regional distribution of poverty, to draw inferences on patterns of use. 5.34 The distribution of poverty varies significantly across regions in Georgia. Regardless of which measure of poverty is used, Adjara and Samegrelo have the lowest incidence of poverty, while Imereti - home to almost one-third of Georgia's poor - has the highest (The World Bank, 1999). An equitable distribution of public expenditures would be one in which the poorer regions have a relatively higher per capita allocation of public expenditures compared to the richer regions. However, an examination of the allocation of municipal expenditure and poverty distribution shows that the reverse is true: richer regions, i.e., those with fewer persons below the poverty line, tend to spend more public money per person compared with the poor regions (see figure below). Imereti, with the largest number of poor in the country, spends the least, while Adjara, with the least number of the poor, spends the most. 5.35 Municipal allocation to health varies considerably across the years, however, and to this extent is not a robust indicator of resource use. No other data is available on resource allocation and utilization of health services that can be used to assess the impact on equity of public expenditures on health. Figure 5.1: Allocation of Municipal Funds 70 0 CL 6 0 30 0 50 r o 40 'S :30 0 2 0 0 0 2 4 6 8 10 a. 92 Georgia: Public Expenditure Review November25, 2002 Out-of-Pocket Spending on Health 5.36 Private spending is a major source of funds for the health sector. In 2000, according to the Tbilisi Household Survey and the Quarterly Survey of Georgian Households, total out-of-pocket expenditures amounted to GEL132 million. Out of all types of medical services, inpatient treatment incurred highest costs (average GEL311 per visit) followed by out-patient services (average GEL70 per visit) and emergency care (average GEL44 per episode). 5.37 To assess the burden of out-of-pocket expenditure on the poor, we analyze the data from the Georgia Households Survey (fourth quarter, 2000). Of the total of 39,648 respondents, 19 percent (7,467) reported suffering from illnesses during the recent three months. Of those suffering from an illness, 25 percent required urgent medical care, and 81 percent of these sought medical treatment. The poorer quintiles reported higher incidence of illness and lower percentage of those seeking treatment when ill (Table 5.4). 5.38 Of the several reasons why many of the ill do not seek treatment, shortage of money is the most prominent, affecting over 90 percent of those who do not seek any treatment, predominantly in the poorest two quintiles. Among those who seek treatment, 57 percent report no expenditure on treatment, 33 percent report no expenditure on drugs, while 26 percent report no expenditure overall. For the sample as a whole, the average annual expenditure per individual on illness is GEL47, with the richest spending four times as much as the poor (Table 5.4). Table 5.4: Incidence of Illnesses (not requiring Hospitalization) Quintile Group Percentage Reporting Percentage Requiring Percentage Seeking Care Illnesses Out-Patient Medical Care 1 23.2 21.5 67.2 2 17.4 25.4 79.6 3 18.4 23.7 82.1 4 19.1 23.5 82.1 5 17.7 28.2 87.2 Total 18.8 24.6 80.9 5.39 As far as inpatient care is concerned, of the total of 39,648 respondents to the Georgia Households Survey, only 195 people (0.5 percent) reported being hospitalized. The richest quintile reports twice as many cases of hospitalization than the poorest quintile. Of 195 people who had undergone inpatient treatment, 38 percent (74 respondents) respondents reported no expense on inpatient treatment. Of those who incurred some expenses, about one-fifth reported monthly overall expenditure over GEL500. Table 5.5: Total Annual Expenditures on Out-Patient Care by the Non-Chronically Ill Quintile Total Health Care Expenditures Mean Standard Deviation Minimum Maximum 1 20.9 41.5 0 350 2 32.2 73.0 0 690 3 35.9 79.8 0 700 4 37.2 67.6 0 610 5 86.8 496.3 0 5500 93 Georgia: Public Expenditure Review November 25, 2002 5.40 An analysis of total expenditures related to the hospitalization by income quintiles shows that the richest spend four times on inpatient care as much as the poor, and the average annual expenditure per individual is GEL376 (Table 5.6). The calculations of expenditure on inpatient care by income quintiles shows that just one incidence of hospitalization can potentially completely wipe out the annual income of a household in the poorer quintiles. Table 5.6: Total Annual Out-of-Pocket Expenditures on Inpatient Care Quintile Total Expenditures on Inpatient Care Mean Standard Deviation Minimum Maximum 1 145.9 275.8 0 1213 2 367.9 484.3 0 1850 3 185.3 247.0 0 1090 4 257.9 252.5 0 1000 5 680.0 1413.1 0 5500 5.41 In terms of the percentage of income spent on health care, households in the poorest quintile can expect to spend over 5 percent of their annual income on health care, while those in the three highest quintiles can expect to spend approximately 1 percent of their annual incomes on health care. 5.42 There is little doubt that the poor in Georgia have suffered disproportionately from both the collapse in public spending on social services as well from misallocation of available resources. As out- of-pocket expenditures have become the major source of health financing, access to basic health services for the poor has been adversely affected, and increasingly large numbers of health conditions are going untreated. Shortage of money is consistently mentioned as the most common reason for the lack of medical treatment, providing a strong basis for the inequities in the system. Expectedly, the rich also undergo inpatient care treatment in larger numbers than the poor. Those among the poor who seek medical treatment spend twice or even three times as much as the rich with respect to their income. The data show that very few episodes of illness are sufficient to completely impoverish those with low incomes who seek health care when ill. Reform Issues and Options 5.43 Of the many problems in the health sector in Georgia, perhaps the most significant are related to efficiency in the utilization of existing resources, resource mobilization for the health sector, and equity in access to and utilization of health services. The misallocation of resources is exacerbated by continuous under-funding of the health sector and by the gap between budgeted and actual expenditures. As a result, the poor in Georgia suffer disproportionately and increasingly large numbers of health conditions are going untreated. On the demand-side, the poor experience significant constraints to utilization as out-of- pocket expenditures have become the major source of health financing. On the supply-side, there is variation between regions and urban and rural areas in the quality and volume of services available, and providers do not have any incentives to be efficient or consumer oriented. These are discussed in turn. Improving Efficiency in the Utilization ofPublic Funds 5.44 As discussed earlier, there are two main sources of inefficiency in utilization of resources: (i) technical inefficiency resulting from excessive staffing, excessive number of hospital beds and high levels of excess capacity; (ii) allocative inefficiency resulting from the inclusion of many "provate goods" such as the treatment of heart disease, cancer and haemodialysis. 94 Georgia: Public Expenditure Review November 25, 2002 5.45 Downsizing Facilities and Personnel. The most straightforward and effective way to improve technical efficiency in health spending is to close down hospital beds where possible, reorganize and rationalize existing beds among high-demand services, convert acute care beds to long-term beds where possible, and reduce excess staff in public facilities. 5.46 Replacing the BBP with a Full Overage Preventive and Primary Health Care. The scope of services covered by public funds needs to be re-examined, and highest priority should be given to health services that have high levels of positive externalities and relatively higher public goods characteristics, such as (but not limited to) preventive services and maternal and child health services, and those that possess significant interpersonal utility values, such as (but not limited to) medical services for the vulnerable segments of the population and emergency services for trauma patients. For a variety of reasons (such as asymmetry of information between physicians and patients, widespread understanding of services actually covered in the package, potential for abuse and misuse of the system by service providers and patients alike), planning the provision of services on the basis of a basic benefits package can by itself be problematic and counter-effective. In the Georgian context, where resources are spread so thinly as it is, it is suggested that the guarantee of basic benefits package be withdrawn, and public funds be used to finance the provision of primary care and first-contact health services for all patients on a modest co-payment basis with full waiver for those in the bottom two quintiles. 5.47 To address existing inefficiencies, Georgia will need to develop a strong preventive and primary health care system and introducing the latest international protocols for the treatment of diseases. In the absence of a primary health care system, it is difficult for the government to rationalize the closure of many specialized facilities. The development of an effective and transparent accreditation and licensing system is also necessary to close down hospitals and control the growth of an unregulated private sector. Efforts to develop a regulatory system to be combined with increasing the power and ability of purchasers to contract selectively with providers. In order to do this, the current fragmentation in the health care financing system with multiple public purchasers needs to be addressed, and Georgia should consider the development of single public purchaser system as have other countries with similar or lower levels of public expenditures on health care (Kyrgyz Republic). Enhancing Equity in the Utilization ofHealth Services 5.48 Most health care demanded and consumed by patients in Georgian has to be paid for, and data shows that few episodes of non-chronic illnesses or chronic illness are sufficient to completely impoverish those who already have low incomes. The household data shows that the poor have a greater incidence of illness - chronic or acute - compared to the rich. At the same time, fewer of the poor who suffer from an illness seek treatment as compared to the rich. Shortage of money is consistently mentioned as the most common reason for the lack of medical treatment. In order to improve the utilization of healthcare by the poor and to provide financial safeguards for those among the poor who do seek health care, it is imperative that the available resources be targeted toward the poor. This could be achieved by orienting resources to the poorer regions - like Imereti - and thus ensure higher per capita allocations of municipal funding for health than available at present. Exploring Innovative Options for Increasing Public Funds 5.49 While it is not straightforward to estimate how much public funds Georgia should spend on health care, the present levels seem to be inadequate and insufficient to meet the needs. The 1993 WDR estimates the cost of financing an essential package of services to range between US$13 per capita (for an essential package of health services reflecting a disease burden largely attributable to communicable diseases) and US$22 (for an essential package of health services for middle income countries where the disease burden is attributable to both communicable and non-communicable diseases). On this basis, 95 Georgia: Public Expenditure Review November 25, 2002 Georgia would require US$60 million to US$100 million, which is 55 percent and 2.5 times the present allocations, for services that are currently covered. Increasing coverage would require significantly larger cash inflows. In fact, the Georgian National Health Policy aims to increase the share of the population covered by the state social insurance system by 2005, specifically increasing the number of vulnerable people covered through the program to 500,000, the number of children and adolescents covered from approximately 230,000 to I million, and increasing coverage of the elderly to about 500,000 citizens of Georgia for outpatient, inpatient and emergency care. 5.50 Given the current fiscal situation in Georgia and the low levels of formal sector employment, it is doubtful whether these levels of public resources can reasonably be mobilized. Consequently, new and innovative ways of raising public finances for the health sector have to be explored. There is some experience in Georgia with community-based health financing schemes, and this needs to be examined very carefully. Targeting municipal funds for preventive and maternal and child health services is another option that needs attention and enforcement. Hospital based medical savings plans - into which members subscribe regularly in return for guaranteed treatment for catastrophic care - is yet another option that needs to be developed in the Georgian context. 5.51 Public expenditures on health form a small part of total health expenditures in Georgia, and are being used neither efficiently nor equitably. There is an urgent need to bring about improvements in the efficiency of public funding; at the same time, new and innovative ways of increasing the availability of public funds need to be explored. Enhancing equity in the utilization of health services is another challenge that the policy makers must face. Private out-of-pocket expenditures are the significant source of funding for the health sector, and the challenge for policy makers in Georgia is to channel these expenditures in a way that it frees up public sector resources for preventive care and poverty alleviation measures. 96 Georgia: Public Expenditure Review November 25, 2002 Chapter 6. Expenditures in the Education Sector Introduction 6.1 This chapter examines the role of the state in the funding and provision of education in Georgia. It analyzes the efficiency and equity effects of public spending on education in order to prioritize public spending according to desired education objectives. Specifically, can public spending on education be reduced through rationalization and alternative delivery and financing mechanisms? Can available resources be used more effectively and efficiently to achieve the desired education objectives? And where can additional resources be allocated within the education sector, whose marginal rates of return will be greater? The report also examines the budgetary process and the intergovernmental fiscal relations (e.g., the revenue transfer system) for an equitable, efficient, and accountable policy implementation in the education sector.9 Overview of the Sector 6.2 Georgia's education system is based on the former Soviet pattern, comprising preschools, general education (covering primary education for grades 1-6, lower secondary for grades 7-9, upper secondary education for grades 10-11); vocational schools; specialized secondary; and higher education institutions. Compulsory education covers grades 1-9. In addition, there are a large number of other education facilities, including boarding schools for disabled children and orphanages, out-of-school facilities, including child houses and youth palaces for arts, music, sport and so forth. Enrollment Rates 6.3 At independence, Georgia had relatively good education indicators that were comparable with those in middle-income countries. The adult literacy rate was nearly universal (99.6 percent in 1989). Of the population aged 15 years and older, about 90 percent had completed a minimum of 8-year education in 1988, over 57 have completed some form of secondary education, and 12 percent higher education. Enrollment and completion rates for compulsory basic education (grades 1-9) were more or less universal (over 97 percent) for all children ages 7-14, regardless of gender, poverty level or geographic location. There was also broad equality between the genders in terms of access to education at all levels. Table 6.1: Percent of Age Group Children Enrollment in Education, 2000 Urban Rural Women Men Women Men Preschool (age under 6) Non Poor 23 26 17 16 Poor 6 12 4 18 Basic and compulsory education Non poor 91 92 93 93 (age 7-14) Poor 91 94 86 89 Post-compulsory education (age 15-2 1) Non poor 64 66 53 50 Poor 51 63 41 43 Source: World Bank, 2001, p.40. 9 This summary is based on the background report on the costs and financing of education in Georgia. The background report used all the available data and background studies, but the lack of data or inconsistency represents a serious impediment to analyze accurately the costs and financing of education in Georgia. 97 Georgia: Public Expenditure Review November 25, 2002 6.4 Available data suggest that Georgia still enjoys high enrollment rates for a country with its income level. Enrollment rates in basic education school age population (age 7-14) are almost universal (90 percent in 2000), albeit declining. However, enrollment rates for pre-school and upper secondary education have fallen sharply. The enrollment rate for preschool education dropped from 41.6 percent in 1990 to 24.3 percent in 2000, largely due to the closure of many preschools ( the number of preschool establishments dropped from 2,470 in 1990 to 1,229 in 1999) Table 6.2: Basic Education Indicators of Georgia Percent Level of Education 1990-1991 1995-1996 1999-2000 Change, 1990-2000 Preschool Students 200,000 81,900 74,000 -63 Teachers 22,538 9,368 n.a Schools 2,479 1,322 1,229 -50 Student-Teacher Ratio 8.9 8.7 - General Education Students 878,196 710,900 714,400 -19 Teachers 100,932 84,819 71,681** Schools 3,686 3,219 3,201 -13 Student-Teacher Ratio 8.7 8.4 9.97 Vocational Schools Students 42,200 19,400 16,800 -61 Teachers 2,425 2,129 n.a Schools 159 118 84 -47 Student-Teacher Ratio 17.4 5.6 - Public Specialized Secondary Students 42,863 26,907 29,927 -30 Teachers 5,284 4,818 4,640 Schools 87 77 85 -2 Student-Teacher Ratio 8.1 5.6 6.45 - Private Specialized Secondary Students - 5,144 6,806 - Teachers 372 1,330 - Schools 13 58 - Student-Teacher Ratio - 13.8 5.12 - Public Higher Education Students 103,900 71,707 95,000 -9 Percent evening & 37.5 40.4 28.9 - correspondence Teachers 10,277 9,271 14,484 - Percent full time .80.6 83 64.8 - Schools 19 23 24 - Student-Teacher Ratio 10.1 7.7 6.56 - Private Higher Education Students - 42,000 40,100 - Teachers - .8,809 10,423 - Percent full time - 9.5 29.1 - Schools - 109 162 - Student-Teacher Ratio 4.8 3.85 Notes: * The data are for 1991-1992. ** Data is for the years 1998-1999. 98 Georgia: Public Expenditure Review November 25, 2002 6.5 There are small differences in enrollment rates between the poor and non-poor for compulsory basic education and some differences by location. The gap is however large between the poor and non- poor for preschool and post-secondary education. Enrollment of children of poor families is particularly low for any type of education, especially in rural areas. Girls' enrollment rates are lower than boys among the poor. Apparently, attendance is also low. Among those enrolled, poor students (aged 7-14), especially in rural areas do miss classes more often than the non-poor in urban areas. The main reason for missing school is related to poor quality of education (World Bank, 2001, p. 41), resulting from teacher absenteeism, deterioration of physical facilities, and lack of textbooks and basic equipment and materials. Education Expenditures 6.6 The funding of education in Georgia has become increasingly privatized during the transition. Available information indicates that overall expenditures on education were 4.9 of GDP in 2000, of which 2.7 of GDP was out-of-pocket expenditures by households. While Georgia's overall level of expenditure is comparable with that OECD levels (at 5.1 percent of GDP), the contribution of the public funds (at 2.2 percent of GDP) is well below the average for OECD countries and for the CIS region (4.6 percent of GDP). Public Expenditures on Education 6.7 Public intervention in education is justified in part to correct the market failure that the private sector may not supply the appropriate level of education services. Public spending on education is also justified on the grounds that it will reduce the costs to the poor, so that they can have access to education. Externalities in education may result from basic education and basic research in higher education. Typically, for vocational and higher education, there is no strong justification for public financing, while the state still has the responsibility for providing financing based on equity grounds. Similarly, there is no efficiency and equity justification for boarding schools, except for severely disabled children who need special care and training. This section examines public spending on education in view of these main principles of public finance for education. 6.8 At present, the main sources of funding for the education sector include public sources (i.e., the central and local budgets) and non-budgetary resources. The main sources of non-budgetary revenue include (i) formal student fees in upper secondary education (grades 10-11) and in the tuition track programs of specialized secondary and higher education institutions; and (ii) parental contributions. 6.9 The budget does not report revenue from non-budgetary and there are no reliable data to present a consolidated accounting of the sources of and uses of education funds. According to the Institutional Expenditure Survey (IES)00, carried out by the State Statistical Department in 2000 (see Table Annex 6.1), approximately 74 percent of total funding for public education institutions came from the central and local budgets, 24 percent from "formal" student fees; and the remaining 3 percent from other sources and donations. 6.10 Georgia has been unable to maintain the same level of public financing for education provided during the Soviet times. Public spending on education as a share of GDP fell from approximately 7 percent in 1991 to one percent in 1994. Similarly, the share of the education budget in the consolidated 1oo The Survey was carried out by the State Statistical Department in 2000, covering a period of January - September, 2000. The survey covered all higher education institutions, and randomly selected a group of preschool, primary, secondary, and vocational schools. The survey identified sources of institutional income as follows: (i) public subsidy from the central and local budget; (ii) student fee income; and (iii) other incomes (e.g., income from economic activities, donations). 99 Georgia: Public Expenditure Review November 25, 2002 public budget declined from an estimated 24 percent in 1991 to 7 percent in 1995. Since 1996, public spending has stabilized around 2 percent of GDP and I1-12 percent of the consolidated budget. For 2002, it was planned that the total public budget would increase to 12.8 of the total public budget. Table 6.3: Georgia: Overall Public Spending on Education, 1996-2001 1996 1997 1998 1999 2000 2001 Total Central Budget 558.9 776.8 797.2 904.8 833.9 756.1 Education Expenditure 24.2 38.3 30.5 29.3 26.9 32.0 % of Total Central Budget 4.3 4.9 3.8 3.2 3 2 4.2 Total Local Budget Expenditure 159.4 203.2 252.0 319.8 324.3 n.a. Education Expenditure 47.4 68.2 73.6 91.2 103.2 112.0 % of Total Local Budget Expenditure 29.7 33.6 29.2 28.5 31.8 n.a Total Public Budget Expenditure 718.3 980.0 1,049.2 1,224.6 1,158.2 1,173.3 Total Public Education Expenditure 71.6 106.5 104.1 120.5 130.1 144.0 % of Total Public Budget 10.0 10.9 9.9 9.8 11.2 12.3 Total Public Budget as % of GDP 12.5 14.4 21.9 21.9 18.5 n.a. Total Public Education Expenditure as Percent of GDP 1.3 2.3 2.1 2.1 22 2.2 Nominal GDP (in GEL million) 4,639 5,041 5,666 5,971.0 6,506.0 Source: 1998-2001 data on GDP are provided by MoF, and the data for 1996-1997 are obtained from Orivel (1998). Total public budget excludes debt coverage and amortization. 6.11 Central Budget. The central budget is still the main source of revenues in the financing of education services provided at the national level. As shown in Table 3.3, central spending on education was about 3.2 percent of the total state budget or about 24.4 percent of the total education budget. In addition, in 1999 the state transferred about GEL39 million to local authorities to finance education services provided by local governments. Overall, it is estimated that roughly 57 percent of the consolidated education budget (both central expenditures and central transfers for education) were financed by the central government. 6.12 Local Budget. Local governments (58 regions, 2 autonomous republics and the city of Tbilisi) are required to finance preschools, general schools, some special schools, and out-of-school facilities. Some local funds are also provided for special boarding schools, vocational and technical schools, and higher education institutions located outside Tbilisi. The local education budget accounts for nearly 80 percent of the total education budget. Local governments have two main sources of revenue (a transfer from the central budget and local revenues). Central budget transfers play a prominent role in the financing of education, approximately 40 percent of the local budget for education, especially in poor districts, which do not have the capacity to generate sufficient revenue to finance local public services. Large cities (e.g., Tbilisi) are in a much better position to generate local revenues, whereas some mountainous regions have almost zero revenue from local sources. 6.13 Table 6.4 shows the regional distribution of central transfers for the local education budget. Except the Tbilisi City, all districts depend largely on the central transfer for the financing of primary and secondary education. Despite their decline, central budget transfers are still an important source of local expenditures on education since over 80 percent of the central transfer is spent on teacher salaries. 100 Georgia: Public Expenditure Review November 25, 2002 6.14 Central transfers are important to ensure equity in the financing of education since poor districts often do not have sufficient revenues to finance education. Yet, it appears that local governments often do not receive sufficient resources from the central government to finance education services. More importantly, central transfers are based on "negotiations" between the representatives of a given district and the MoF. Despite the importance of the central government to provide equality of education opportunity, it has less control over the regions and local education expenditures, and does not have the technical capacity to assess the impact of central and local spending on education outputs and outcomes. Table 6.4: Total Budget Transfers to Local Governments by the Central Government and the Share of Education Expenditures in Total Transfer by Regions, 1997-2000 1997 1998 1999 2000 Share of Share of Share of Total Share of Region Total Euain Total EuainTotalToa Transfer Transfer Transfer Education Transfer Education Percent Percent Percent Percent In thousands GEL Abkhazia 2706.0 17 4081.0 16 3664.0 95 3364.1 16 Adjara 6506.0 32 4077.0 61 2500.0 84 n.a. n.a. Tbilisi 12862.4 22 1095.0 - 490.0 - 2000 0 - Kakheti 2150.0 84 2081.0 65 2881.0 79 2366.0 78 Imereti 5759.7 69 5732.0 82 6291.0 86 5225.3 78 Samegrelo & Zemo 3667.0 64 3585.0 78 4904.3 82 4404.0 76 Svaneti Shida Kartli 1074.0 77 671.0 35 1533.0 83 1884.0 63 Kvemo Kartli 1391.0 87 1982.0 30 1005.0 51 1085.0 71 Guria 914.0 88 1471.0 85 2161.0 86 1653.0 80 Samtskhe-Javakheti 2264.0 68 1810.0 85 2742.0 87 2177.0 84 Mtskheta-Mtianeti 2553.0 34 1998.0 39 1607.2 81 1955.2 70 Racha-Lechkhumi 1627.0 65 1137.4 73 1982.0 83 1657.0 60 Total 43474.1 46 29720.4 58 31760.5 83 27770.6 62 Source: MoF. All data are actual expenditures. Actual versus Planned Expenditures on Education 6.15 In Georgia, the execution of the planned budget has been a significant problem for the financing of education. The budget is revised throughout the year. Table 3.5 presents both planned and actual education expenditures for state and local education budgets from 1996 to 2000. The local education budget was executed by 95 percent in 1997, but it dropped to 83 percent in 1998 and then increased to 93 percent in 1999. In 2000, the local actual expenditure on education was about 12 percent above the planned budget, indicating a significant improvement for budget execution over previous years. Central government expenditures also represented similar trends. During 1997-99, the execution of the central budget was about 70 percent the original budget amount, but in 2000 execution was 102 percent of the original budget, but fell to 90 percent in 2001. 6.16 The execution of the planned expenditure varies by region. By economic classification, aggregate line items for salaries tend to be executed at about 100 percent of the budget, while spending on non-salary items are often reduced. For example, the Ministry of Education received only 58 percent of the budget for office expenditures in 2001. Although salaries tend to be one of the protected items in the budget execution, arrears in teacher salaries also often reported in many districts. As discussed in Chapter 2 (Budget Management Issues), unrealistic revenue estimation and budget planning are usually the primary reasons for the non-fulfillment of the planned expenditures. This unpredictability in available 101 Georgia: Public Expenditure Review November 25, 2002 resources for the education sector has a considerable negative effect on the quality of education services, especially resulting in arrears in teacher salaries. The Ministry of Education, local governments and education institutions cannot make realistic plans. Table 6.5: Planned versus Actual Expenditure on Education, 1997-2001 (000, GEL) Central Education Budget Local Education Budget Total Education Budget Year Plan Actual Percent Plan Actual Percent Plan Actual Percent 1997 54,398 38,330 70 71,859 68,249 95 126,257 106,578 84 1998 45,456 30,500 67 88,623 73,585 83 134,079 104,085 78 1999 43,765 29,300 67 97,968 91,235 93 141,732 120,535 85 2000 26,453 26,862 102 92,309 103,217 112 118,762 130,079 110 2001 35,106 31,666 90 n.a. n.a. n.a. n.a. n.a. n.a. Source: Ministry of Finance. Resource Allocation by Economic Classification 6.17 Personnel Expenditures. Georgia spends a higher share, about 80 Vercent, of its available resources on personnel than the average of OECD countries. (see Table 6.6).'o By level of education, personnel expenditures represented 53 percent of preschool expenditures, 96 percent of general education expenditures, 85 percent of post-secondary and higher education expenditures. By comparison, in OECD countries the share of total staff costs is 78 percent of total recurrent expenditures on primary, secondary and non-tertiary education. In tertiary education, the share of all staff costs is only 65 percent of total recurrent expenditures. In Georgia, the share of teacher's salaries within the total education budget has increased from 77 percent in 1998 to 81 percent in 2001. This increase is mainly due to (i) the sharp decline in public spending on education; (ii) overstaffing of both teaching and non-teaching staff; and (iii) protection of teachers' salaries and weak expenditure prioritization. 6.18 Despite the high share of teacher salaries in the total education budget, real teacher salaries have declined sharply during the transition. Monthly wages (at about US$20) are well below subsistence level or 65 percent of the national average. Teachers are also paid on an irregular basis. As a result, teaching motivation is low and absenteeism has soared, as my teachers have taken other informal employment for generating additional income. Also, many teachers have reached retirement age, suggesting that a large number of teachers may leave the profession in the next few years. 10' Expenditure in this section is net expenditure only for educational institutions both on central and local levels and does not include expenditure for administration (i.e., central Ministry of Education and local administration). 102 Georgia: Public Expenditure Review November 25, 2002 Table 6.6: Percent of Public Expenditures on Education by Level and Budget Category, 2000 General Vocational and Budget Categories Preschools Schools Specialized Secondary Higher Education Current Expenditures 90 100 96 99 Expenditures on Goods and Services 90 100 96 95 Wages and salaries 41 75 65 65 Employer's contribution 12 21 19 20 Business travel 0 0 0 0 Other goods and services 38 4 11 10 Office expenses 6 3 3 1 Utilities 10 1 4 8 Food expenses 20 0 2 0 Medical expenses 0 0 0 0 Inventory and uniform 0 0 0 0 Transport expenses 0 0 0 0 Other expenses 1 0 1 1 Program expenses 0 0 1 0 Subsidies 0 0 0 5 Capital Expenditures 10 0 4 1 Total 100 100 100 100 Notes: General schools include primary and secondary general education schools; and vocational and secondary specialized schools include boarding schools, vocational schools, and technikums. 6.19 Teaching Materials, Equipment, and Maintenance. In Georgia, the state no longer finances textbooks and provides negligible funding for teacher training, teaching materials, equipment and supplies. General schools have been relying on the revenue generated from school fees from upper secondary students, community contributions and informal payments for minor school maintenance, supplies, and materials. Parents purchase textbooks, teaching materials and supplies at all levels of education.102 Qualitative reports suggest that many students in primary and secondary education, especially coming from poor households, do not have access to basic textbooks and learning materials, and when available are of poor quality and/or out-of-date. 6.20 Vocational schools also suffer from the lack of basic teaching materials and equipment. Only 22 percent of the teachers in vocational education reported that they had the consumables they needed. About 55 percent said they had no materials at all for practical lessons. Higher education institutions also suffer from totally outdated equipment, laboratories, libraries, and lack of basic materials and supplies. 6.21 Many education facilities have also deteriorated considerably during the transition as a direct result of the lack of funding for maintenance. It is likely that the cost of deferred maintenance would be significantly higher in the coming years because of rapid deterioration of physical infrastructure.03 Finally, public spending on capital expenditures has been negligible. In 2000, capital spending was only one percent of the total preschool budget; 4 percent for general education; and one percent for higher education. 102 The Department of Education of the Tbilisi City provided textbooks free of charge for grade I students for the first time in 2000. It is not clear whether they will continue to implement this policy. 103 The World Bank assisted Georgian Social Investment Fund project financed some capital repairs of education facilities (mostly pre-schools and secondary schools). 103 Georgia: Public Expenditure Review November 25, 2002 6.22 Other Goods and Services. The cost of utilities as share of the total public education budget increased from 1.4 percent in 1997 to 5.2 percent 2000. Similarly, the share of food expenditures increased from 3.4 percent to 5.2 percent during the same period. Food expenditures are mainly for orphanages, childcare houses and special regime boarding schools for the disabled.104 Resource Allocation by Level 6.23 In Georgia, the distribution of education expenditures by type and levels of education suggests considerable inequities and inefficiencies (see Table 6.5). In 1999, Georgia devoted 48.5 percent of the total education budget to general education, 20.6 percent to higher education, 11.1 percent to boarding schools and orphanages, 7.7 percent to kindergartens, 6.7 percent to out-of-school institutions, 3.8 percent to administration, and 1.5 percent to vocational and technical education. These ratios have remained more or less unchanged over the past four years, but Georgia appears to spend a high share of its education budget on special education and boarding schools. By comparison, OECD countries allocate about 70 percent of public education budget to primary and secondary education (Orivel, 2000). Table 6.7: Distribution of Overall Education Expenditures by Level of Education, 1999 Levels of Education Local Central Total Administration 0 11.4 3.8 Kindergartens 11.2 0.9 7.7 General education 72.0 1.6 48.5 Boarding schools & orphanages 6.6 20.1 11.1 Technciums 0 4.6 1.5 Higher Education 3.0 55.5 20.6 Out-of-school institutions 7.2 5.8 6.7 Total 100 100 100 Source: Orivel, 2000 6.24 The central budget is primarily used to finance education institutions at the national level such as boarding schools, special schools, technical schools, higher education institutions. In addition, several selected preschools and secondary schools and a number of out-of-school facilities are financed by the central budget. Overall, approximately 67 percent of the central education budget goes for higher education. A significant share of the central budget is allocated for special and boarding schools, which increased from 9 percent in 1995 to 18 percent in 2000. 6.25 Local governments have the main responsibility for financing preschools and general schools. In 2000, nearly 83 percent of the total local budget was allocated for general schools (i.e., primary, incomplete secondary, and complete secondary schools); 16 percent for preschools and one percent for vocational and higher education institutions. Although the general education budget also includes some funds for vocational schools, the data are not available to document local education expenditures by types of secondary schools. 104 Such an increase is also partially due to the budgetary support from the EU Agriculture Program that provided around I million Euro in 2000 earmarked for boarding schools and orphanages. 104 Georgia: Public Expenditure Review November 25, 2002 Table 6.8: Local and Central Education Expenditures by Level of Education, 1998-2000 1998 1999 2000 1998 1999 2000 Levels of Education Local Budget (000, GEL) Percent Pre-schools 8415.2 10257 11155.2 13 15 16 Elementary education stage 1935.3 2196.2 3137.6 3 3 4 Major education stage 8353.4 10211.2 8439.8 13 15 12 Secondary schools 44068.8 44302.8 47458.4 68 66 67 Higher education 2293.7 263.4 237.5 4 0 0 Others n.a n.a n.a n.a n.a n.a Total 65066.4 67230.6 70428.5 100 100 100 Central Budget (000, GEL) Percent Pre-schools 316.7 113.1 128.4 2 1 1 Elementary education 1102.1 383.4 300.6 6 2 2 Secondary schools and orphanages 801.7 1442.7 1629.3 4 7 8 Vocational schools 1683.1 828.5 575.2 9 4 3 Technicums and colleges 1485.3 2057 1783.4 8 10 9 Higher education 13039.8 13450.8 13119.9 67 67 67 Childcare houses 253.6 609.3 604.7 1 3 3 Special school 50.9 54.9 52 0 0 0 Special regime boarding schools for disabled 749 1179.6 1332.1 4 6 7 Total 19482.2 20119.3 19525.7 100 100 100 Source: MoF. All data in this table are actual. Resource Allocation by Region 6.26 Table 3.7 presents the data on per capita spending on general education by region. On average, local spending per general education student was GELl27 in 1999, but it varied significantly by region. For example, the Tbilisi City spent about GEL235 per student or 85 percent more than the national average of GEL127. On the other hand, the poor regions such as the Shida Kartli and Guria Regions spent only GEL81 and GEL85 per student or 63 percent and 67 percent of the national average. Bezause of such significant variations in the distributions of resources by region, it is likely that education outcomes also vary considerably by region, resulting in inequities in access to quality education in the poor districts, but the data are not yet available to examine enrollment rates and learning outcomes by district. 105 Georgia: Public Expenditure Review November25, 2002 Table 6.9: Average Local Expenditures in General Education by Regions in 1999 Average Local Spending Average Local Spending Regions in General Education in Georgian as percent of National Average, GEL, 1999 1999 Abkhazia 91 72 Adjara 235 185 Tbilisi 164 129 Kakheti Region 103 81 Imereti Region 96 76 Samegrelo and Zemo Svaneti 111 87 Shida Kartli Region 81 63 Kvemo Kartli Region 102 81 Guria Region 85 67 Samtskhe-Javakheti Region 104 82 Mtskheta-Mtianeti Region 154 121 Racha-Lechkhumi Region 195 153 Total 127 100 Note: The local expenditure data are for 1999 and the enrollment data are for 1998-1999 6.27 In summary, there are several major trends in the costs and financing of education in Georgia. First, real public spending on education has declined sharply during the transition. Despite some recent improvements, public spending on education is still one of the lowest in the region. Second, the sector has not yet adjusted effectively to the reduced budgetary support in the form of an overall rationalization or prioritization of education expenditures. The government is coping with the fiscal crisis through across-the-board cuts in education spending and temporary measures (e.g., eroding teachers' wages in real terms, building arrears on energy bills and teacher wages, deferring maintenance, cutting resources for basic education materials and supplies). Third, under-financing (and irregular payment of teachers salaries) continues to have adverse impacts on the effectiveness of teachers, on the provision of basic teaching and learning materials, and on the physical conditions of schools. The quality of education is likely to have deteriorated considerably due to the shortage of basic teaching materials, supplies equipment, poor maintenance and operations, and poor teacher training and effectiveness. It is also likely that learning outcomes have declined considerably due to sharp declines in real resources available for education as indicated by most-input based quality indicators during the transition, but data are not yet available to document the impact of declines in public spending on education. Fourth, out-of-pocket spending has increased sharply in response to the collapse of public spending at the beginning of the transition. Private Expenditures on Education 6.28 Overall private spending on education (including overseas education) was about GEL177 million for a period of nine months or about GEL229 million for a period of 12 months in 2000,105 which was equivalent to 63 percent of total spending on education or 3.8 percent of GDP. Excluding expenditures for overseas education, private spending on education was about GEL121.8 million for a period of 9 months or GEL162.4 million for a period of 12 months, which accounted for roughly over 54 percent of total education spending or 2.7 percent of GDP in 2000.1o6 Private spending on education as a share of GDP exceeds substantially the average of OECD countries when all revenue received from non-budgetary 05 The actual expenditures were GEL 177.0 million for a period of nine months, covering January-September, 2000. '0 This is based on the assumption that the total public spending was approximately GEL130 million and direct private spending was 162.4 for a period of 12 months (excluding GEL55.2 million for overseas education). 106 Georgia: Public Expenditure Review November 25, 2002 sources (including informal payments) are combined. By comparison, in OECD countries the average share of private funds was 10 percent of the total spending on primary, secondary, and post-secondary non-tertiary education. The share of the budgetary support varied considerably by level of education. For example, in higher education, approximately 56 percent of total revenue came from student fees, 41 percent from the budget, and the remaining 3 percent from other sources. Private funding in tertiary education accounted for approximately 23 percent of total spending in OECD countries. (OECD, 2000).0' 6.29 The role of private funding in education has increased because (i) the direct costs of education have been gradually shifted to users since the state no longer finances textbooks, basic teaching materials and supplies; and (ii) the increase in the number of private education providers, especially in higher education. According to official statistics, there were 58 private specialized secondary schools and 162 private higher education institutions in 1999, which accounted for approximately 19 and 30 percent of total enrollment in post-secondary education institutions, respectively. In 2000, it was estimated that private education accounted for approximately 12 percent of all institutional revenues and 10 percent of all institutional expenditures (SSD, 2001). The rapid growth of the private sector has reduced the education sector's burden on the public budget, but has raised questions about the quality of education provided by private institutions in the absence of adequate accreditation and quality assurance systems. 6.30 Despite the importance of revenue from private sources, related data are limited. The recent IES provides preliminary data on both public and private sources of revenue as reported by public institutions. The Household Education Expenditure Survey (HEES)'08 also provides preliminary data on household expenditures on education, including "unofficial fees". Table 6.8 presents household expenditures on education by level of education and types of spending for January-September 2000. 6.31 Formal Fees. According to HES, formal schooling fees accounted fro 58.6 percent of households spending on education. About 72 percent is to finance higher education and studying abroad. Formal fees have been introduced in upper secondary education (grades 10-11) and other non-compulsory education (e.g., preschools, specialized secondary and higher education). The share of revenue from student fees varies significantly by level of education. Revenue from student fees represented about 55 percent of the total revenue of public higher education, 25 percent of specialized secondary education, 12 percent of preschools; and 4 percent of general education. In the 2000-2001 academic year, fee-paying students accounted for about 20 and 26 percent of enrollments in specialized secondary and higher education institutions, respectively. About 30 percent of grade 9 graduates in urban and rural areas are exempted mostly based on merit, but specific criteria for exemption are not clear. All students in mountainous regions are also exempted from paying fees. These funds are mainly used for supplementing salaries for those teachers involved in grades 10-11 and for paying other expenses. 107 The share of private funds in total education budget varies significantly by level of education and by country. For example, the share of private funds in total tertiary education spending ranged from 78 percent in Korea, 49 percent in the United States, 9 percent in Sweden, and to 1 percent in Denmark (OECD, 2000). '0 Carried out by the State Statistical Department in 2000, covering a period of January-September, 2000. 107 Georgia: Public Expenditure Review November 25, 2002 Table 6.10: Household Expenditures by Level of Education and Types of Spending January-September, 2000 Special Type of Expenditures Pre- General Access secondary Higher Extension Studying schools schools courses schools Tutors education courses abroad Total Total (GEL million) Total actual expenses 4.2 47.4 1.2 9.8 11.6 47 0.6 55.2 177 (in percent of total) Schooling fee 85.7 19.0 83.3 52.0 87.9 72.8 100.0 72.5 58.6 Fee for studying in special groups - 7.6 - - - 0.9 - - 2.3 Irregular expenses 14.3 14.6 - 20.4 - 5.5 - 24.6 14.5 Unofficial expenses - 1.5 - 4.1 - 4.5 - - 1.8 Remuneration in kind - 1.1 - 1.0 12.1 0.2 - - 1.2 Schooling articles - 56.5 8.3 21.4 - 16.2 - - 21.7 Total actual expenses 100 100 100 100 100 100 100 100 100 Source: State Department of Statistics (2000) Household Education Expenditure Survey Tbilisi: SDS. 6.32 Informal Payments.'09 Both qualitative reports and household surveys suggest that informal payments account for a significant share of education funding, especially in higher education (see, Lorentzen, 1999; Sharari, 1999; Sealy, 1999; SSD, 2001). In the absence of readily available data, it is difficult to examine the nature, magnitude, and consequences of informal payments in the education sector of Georgia. According to the HEES, it was estimated that "unofficial fees" accounted for approximately GEL3.2 million or 1.8 percent of the estimated household education spending or 3.1 percent of the estimated public institutional spending. 6.33 Other Private Spending on Education. Parents also cover an increasing share of direct expenditures on education (e.g., textbooks, teaching materials, transportation, food). Spending for school materials, supplies and other items accounted and 61 percent of total private spending. 6.34 Private tutoring has become a necessary complement to attend higher education, especially in selective public institutions and programs. The prevalence of private tutoring is partly related to a concern over the poor quality of education in general schools and partly related to teachers' interest to obtain supplementary income. 6.35 It was estimated that approximately 34 percent of non-poor students (aged 7-14) in urban areas receive private lessons, whereas only 20 percent of poor students receive such lessons. In rural areas, about 25 percent of women and 20 percent of men among the non-poor receive private lessons, while only 13 percent of women and 7 percent of men receive private tutoring. These findings suggests that the poor have less ability to pay for private lessons and therefore are the ones who are most affected by the lower quality of primary and secondary, and access to higher education (World Bank, 2001, p. 40). According to the findings of the HEES, it was estimated that household spending for private tutoring was about '09 Informal payments are defined as payments by parents (or students) for educational services that are not explicit or publicly mandated. They are difficult to estimate because they are unofficial, often in the form of gifts, contributions, and bribes. They can be voluntary payments for maintenance, equipment and materials or bribes to be admitted into a specific program or institutions or to obtain higher grades in examinations; and involuntary payments, when requested by teaching staff or administrators. In both cases, they highlight problems with transparency and the probity of official conduct in the education sector. 108 Georgia: Public Expenditure Review November25, 2002 GELl 1.6 million for a period of nine months or 9.5 percent of total private spending in the education sector in 2000. 6.36 Several observations can be made based on the Institutional and Household Education Expenditure Surveys. First, the financing of public education has become dependent heavily on non- budgetary funds during the transition. Second, while the data are not available to examine the impact of this cost shift on the education of the poor, access to higher education is now largely confined to those who can afford private fees or private tuition to meet state entrance examination requirements. It is also likely that the poor have less access "quality schooling" or "supplementary materials" as a result of sharp declines in public spending on education even though the data do not show significant changes in enrollment rates in basic education. Third, considerable differences exist in overall education spending by institution and by region since the ability of individual schools to raise contributions from parents and the local community varies greatly. Efficiency and Effectiveness of Public Expenditures 6.37 Georgia does not use available resources in the education sector as efficiently as possible. The main sources of inefficiency arises from (i) the excessively large number of both teaching and non- teaching staff (ii) the existence of outdated and often narrowly specialized education programs and institutions especially tertiary education. Since the recurrent budget is almost entirely consumed by salaries, food (mainly for preschools and boarding schools) and utility expenses, it is essential to evaluate staffing and infrastructure expenditure patterns of public education aiming at improving efficiency in the use of resources and enhancing the quality of education services Table 6.11: Student-Teacher Ratios in Georgia and Selected Countries Primary Secondary Higher Georgia 10.4 8.0 United Kingdom 22.0 16.7 17.7 Germany 21.6 15.5 12.4 Czech Republic 19.2 15.4 13.5 Ireland 22.6 16.3 16.6 OECD Average 17.0 - - Source: OECD, 2000 6.38 Student/Teacher Ratio. By international standards, student-teacher ratios in Georgia's general education are very low. As shown in Table 3.11, the student teacher ratio is 10:1 in primary and secondary education, compared to a ratio of 17:1 in OECD countries. In Georgia, student-teacher ratios are especially low in general education in rural areas, vocational schools and technicums, boarding schools, and higher education institutions. Recent measures by the government to rationalize the teaching force and improve incentives have succeeded in raising the student/teacher ratio from 8.3 in 1991. These include: reducing the number of part-time teachers, increasing the teaching load, and increasing salaries on a performance basis through a national testing and certification process (Duthieuell, 1998). The present student teacher ratios suggest that considerable cost savings could be achieved by simply reducing the number of staff. For instance, an increase in the student-teacher ratio from 10 to 16 could generate costs savings in the order of 1.3 percent of the GDP. Staff rationalization, however, should take into account, among other things, (i) a urban/rural differentiation; and (ii) employment status (e.g., full/part time employment) (World Bank, 2001, p.56). 6.39 Although the number of teachers in vocational schools under the control of the MoE has been reduced by 50 percent since 1995, student/teacher ratios are still low and make the system expensive to 109 Georgia: Public Expenditure Review November25, 2002 run (Sealy, 1999, p.10). Similarly, the higher education sector is very labor intensive. The public sector employs about 17,000 staff or roughly 2 percent of total employment in the Georgian service sector, resulting in a student/staff ratio of 5:1. Of these, 10,430 are teaching staff, giving a student/teacher ratio of 8:1. However, these overall ratios hide considerable differences in student teacher ratios across institutions and within institutions. For example, at the Georgia Technical University (GTU), the student teacher ratio ranged from 13.2 in informatics to 3.2 in architecture. At the Georgia Agriculture University (GAU), the average student teacher ratio is 3.4 and the department of mechanization has 54 faculty members and no students. This suggests significant efficiency losses in the deployment of resources within the higher education sector. 6.40 In addition to the employment of a large number of teaching staff, some higher education institutions employ a large number of scientific research staff. For example, TSU, GTU, and GAU employed 1,448, 722, of which 118 were scientific research staff or 25 percent, 15 percent or 11 percent of their of total employment, respectively. Given the fact that universities have not received any of the funds to upgrade its basic and advanced research equipment and material basis or even had the funds to maintain subscription to academic journals or attend international conferences, it is questionable to spend a significant share of public spending on university-based scientific staff. A strategy must be developed to integrate science and teaching in higher education to achieve scale and scope economies, reduce the cost of scientific research staff, and improve the effectiveness of basic research in selected key areas. 6.41 Existence of outdated and often narrowly specialized public education programs and institutions. The second main source of inefficiency in the education sector is the building intensity of education production. There are large number of physical facilities, resulting in underutilized space and high costs of utilities (e.g., heating and electricity), especially in rural areas and non-general schools (e.g., vocational, technicums, higher education, boarding schools). As most general schools covers grades 1- 11, the average class sizes are often very small in upper secondary education especially in rural areas. Preliminary data indicate that there will be significant scope for improved efficiency through general school consolidation, especially in rural areas where there have been sharp declines in the number of school age population during the transition. For example, the rationalization of the general schools may generate cost savings around 1.5- 6 percent of the total education budget. It is important to note that school facilities are used quite intensively in urban areas (World Bank, 2001, p.59). A large number of schools have two or three shifts in urban areas. Some schools in rural areas also operate in multi grade teaching, but often without proper teacher training and adequate materials to make it more effective. 6.42 In addition, most school facilities also have poor construction standards with little regard to heating efficiency or maintenance costs. Because of deferred maintenance, the percentage of school buildings reported to be in urgent need of repair or major reconstruction increased from 45 percent in 1992 to 73 percent in 1997 (Perkins, 1998). Again, in the medium to long term, considerable cost savings could be achieved by consolidating education facilities and by the provision of producing education services through more cost-effective delivery mechanisms (i.e., multi-grade teaching in rural areas), even if such reforms will have significant cost implications in the short term. 6.43 Another important factor contributing to inefficiency in the education sector is the existence of a large number of narrowly specialized, small-scale vocational schools and technicums, higher education and research institutions. Many of these institutions also offer programs which are relevant to the needs of the market economy. Despite declines in the number of vocational and specialized secondary schools during the transition, there is still scope to achieve cost savings through consolidation and rationalization of vocational and secondary specialized schools. Similarly, public higher education and research institutions should be rationalized, especially in light of the rapid growth of the private higher education sector. However, a strategy is yet to be developed, and more importantly, the data are not yet available to examine the cost savings resulting from possible rationalization and re-structuring. 110 Georgia: Public Expenditure Review November 25, 2002 6.44 In higher education, many new programs especially in social sciences and professional fields (e.g., economics, business, law, etc.) were established in both state and private institutions to meet the skills demand for the market economy. The education system, however, has not been effective in meeting market demands and continues to focus on narrowly specialized programs designed to meet skills' needs of a planned economy. For example, the results of a recent survey of 60 firms showed the dissatisfaction of the business community with the education system since (i) the education system does not provide sufficient or adequate training in special skills that they are required; and (ii) graduates do not have certain skills and attitudes (e.g., flexibility, team work, ability to learn, independence, etc.) which are required in services (Lorentzen, 1999). 6.45 Learning Outcomes. Data are not readily available to examine the link between changes in the level and composition of public spending and education outcomes. The quality of education services, as measured by crude input indicators (i.e., a well-qualified and highly motivated teaching staff, committed and well-prepared students, and availability of sufficient basic teaching and learning resources, which are all essential inputs for achieving high quality education), has decreased considerably during the transition. First, the quality of teaching staff has deteriorated considerably due to sharp drops in real teacher salaries, deteriorating conditions of the school environment, and poor teacher training (both pre-service and in- service). Such factors have prompted many of the best teachers to leave the system, de-motivated the existing teaching staff and reduced the demand for teacher training. Second, the state no longer provides funding for basic teaching and learning materials and supplies, which are mainly financed by parents. As a consequence, most poor families cannot afford the costs of such basic materials and supplies, and almost all education institutions lack basic learning and teaching materials, equipment, and supplies to perform a minimally acceptable standard during the transition. Shrinking budgetary allocations have also reduced funds for operations and maintenance and repairs, resulting in deterioration in the overall physical conditions of many school buildings, especially in rural areas (see Shahriari, 1999). Third, outdated school curricula and teaching methods, which were inherited from the Soviet system, are still concerns for quality improvement, since they have little relevance to the needs of an emerging market economy and an independent country. 6.46 In summary, the education sector is highly inefficient in the use of available resources. Little structural reform has taken place and commonly employed indicators of system inefficiency remain largely unchanged or have worsened in several cases such as student-teacher ratios. In addition, the sector has not yet been restructured to re-align to meet the needs of the market economy, which in turn results in mismatch between the skills and knowledge needed in the labor market and those produced by the education sector, especially in secondary vocational and higher education. Equity Effects of Public Spending on Education 6.47 Available data indicate the existence of considerable variations in the level of public spending on education. According to the findings from the Household Budget Survey in 1999, it was reported that, on average, poor families spent about GELl5 or just over 2 percent of their cash budget to send a child to school, whereas the non-poor families spent about GEL140 per child or over 5 percent of their total cost expenditures. The distribution of the private costs of education by expenditure item also suggests unequal access to resources. For example, about 10 percent of private expenditures were allocated for private lessons, but almost no poor household reported taking private lessons. Also it is reported that only 7 percent of poor families had a member of 15 years old or more who was a student while 15 percent of non-poor families had a member who was a student in this age group (World Bank, 1999b, pp. 21-22). 6.48 The equity effects of public spending on education are closely associated to disparities in enrollment rates. In Georgia, the data indicate declines in primary and secondary education have been low, but there have been sharp declines in enrollment rates in non-compulsory education.. The declines in Ill Georgia: Public Expenditure Review November 25, 2002 enrollment rates are due to both supply and demand side factors. On the demand side, it was reported that the increased costs of education - both direct expenditures (in the form of school fees, teaching materials, etc) and opportunity cost -has affected poor 's access to education, especially to non-compulsory education. Given the fact that higher education is less affordable by the poor because of the high cost, students coming from poor families are less likely to continue to higher education. Furthermore, it was also reported that the opportunity costs of education in upper grades could be high, especially for poor households in rural areas. As a result, it is likely that fewer poor children can continue their education beyond the compulsory education. The recent poverty assessment also found that there was a link between the level of education and poverty incidence, but it was very weak, except for higher education (World Bank, 1999, p.1 1). It appears that secondary education or vocational education provide little insurance against the risk of poverty. On the supply side, many education facilities (e.g., vocational schools), especially in rural areas, have been closed down or operate with little or equipment and supplies, which is likely to have affected enrollment rates by the poor. 6.49 There are also clear signs that the poor are affected considerably from a rapid deterioration of educational quality, especially in rural areas and disadvantaged regions. The poor do not have access to (i) basic teaching materials and supplies; and (ii) quality teaching staff; and (iii) well-maintained school buildings, laboratories, and facilities. Children of the poorest families may also be at a disadvantage in primary and secondary education because of the inability to afford winter clothing, school meals, textbooks, materials and supplies, and resources to have supplementary education opportunities in the form of private tutoring, or the increased direct costs of higher education in the form of student fees or informal payments in some schools. The increased direct costs of such inputs are likely to determine the probability of completion of secondary school and higher education enrollment. As a result, poor students' access to higher education has deteriorated sharply during the transition (World Bank, 2001). 6.50 Children from poor families often receive an inadequate education, have less access to selective postsecondary education institutions and programs and enter the labor force as unqualified or low qualified workers. The poor do not have access to "quality education" in general education or private tutoring and basic teaching materials and supplies, which largely determine access to higher education, especially "free" higher education. This has serious long-term implications for Georgia's economic development and poverty alleviation efforts. Conclusion and Policy Recommendations 6.51 In Georgia, both the coverage and quality of education have deteriorated partly as a result of the fiscal retrenchment the country underwent over the past decade. At the same time, although the government has introduced a number of reforms to diversify the delivery of education services and the revenues sources of education institutions (e.g., cost recovery in higher education), the sector has not yet been restructured to address the inherited operational inefficiencies in the production of education services. Given the importance of education for both poverty reduction and economic development, policy reforms must be implemented to (i) increase efficiency in the use of the scarce given resources, (ii) reduce inequities in access to education; and (iii) improve the quality of education services.'10 Raising Allocative Efficiency in the Education Sector 6.52 Despite low public spending on education, Georgia spends available limited resources extremely inefficient as suggested by low student -teacher ratios, the existence of a large number of narrowly "o Under the on-going IDA-funded Education Adaptable Program Loan (APL) - Education System Realignment and Strengthening Program (Report No. 20952-GE), approved in March 2001 - the Government aims at addressing some of the efficiency and equity issues identified in this report. 112 Georgia: Public Expenditure Review November 25, 2002 specialized vocational and higher education institutions, and allocation resources within the sector through significant subsidy for postsecondary education. The first priority is to improve efficiency in the education sector. Specific policies and priorities can be grouped into four broad categories: (i) re- allocation of education expenditures from tertiary to primary and secondary; (ii) improve internal efficiency; (iii) shift more costs onto clients; and (iv) provide an enabling environment for the development of the private sector. 6.53 Re-allocation public funds from tertiary to primary and secondary education. Georgia inherited a large number of higher public education institutions, including a large number of narrowly specialized schools, and higher education and research institutions. They are considered to be ineffective in meeting the needs of an emerging market economy. Since Georgia continues to spend a significant share of public resources on such education institutions, it is recommended that re-allocate resources within the education sector to maximize benefits, especially from vocational and technical education, and higher education to primary and secondary education. In 2000, about 48.5 percent of the education budget was allocated for primary and secondary education. This is low by international standards - in OECD countries this ratio is typically around 70 percent. One of the triggers of the Education APL is to increase expenditures for primary and secondary education to 60 percent of the consolidated education budget._Restructuring the systems of vocational and higher education from narrowly specialized vocational and higher education institutions to general academic instruction would not only release funds but also would better prepare students for the demands for a rapidly changing economy by equipping them with better skills in acquiring new job skills. 6.54 Improving Technical Efficiency in the Production of Education Services. There is significant scope for improved technical efficiency through rationalization of staff and education facilities, which are very generous by both OECD and developing country standards. A reduction in the number of teaching staff to OECD levels is likely to result in significant cost savings (about 1.5 percent of GDP). This can be achieved through the introduction of a set of measures such as (i) providing teachers with training in needed fields; (ii) higher teaching load, larger class sizes, where possible; and (iii) multi-grade teaching in rural areas or small school districts. The Government plans to raise the student teacher ratio to 16:1 by 2010, partly by rationalizing teaching staff in general education through a severance payment program supported under Education APL. 6.55 A reduction in the number of school buildings is less likely to produce significant cost savings in primary and secondary education since school buildings in urban areas are generally over-utilized in general schools (many of which already have two or three shifts). Maintaining underutilized rural general schools remains more cost effective than reducing the number of schools and providing all-weather roads and reliable public transport. But other underutilized education facilities (technicums and higher education institutions), for which public financing is not justified, are obvious candidates for downsizing. It is important to note such as to improve the efficiency of service production (reducing staffing ratios and improving the performance of the education system) in the education sector will require a substantial amount of investment before yielding significant cost savings. The Government plans to develop a methodology for more efficient use of school buildings and an incentive plan to encourage local governments to consolidate schools. Both the methodology and the incentive plan will be implemented by the Government during the second phase of the Education APL. 6.56 Reform the Budgeting Process and Resource Allocation Mechanism. The presert resource allocation and mechanism also contributes to an inefficient use of the resources available for education. Specifically, (i) funds are allocated based on notional class sizes rather than on actual number of students; and (ii) the education budget is unpredictable (with low execution rates even for teacher's salaries) and inflexible, and the allocation of central transfers to the local level is non-transparent. The budgetary process is also inadequate as a framework for planning and policymaking. It provides little incentives for 113 Georgia: Public Expenditure Review November 25,2002 efficiency in spending units and fails to require strong accountability in the production of education services. 6.57 Formalizing private financing of education. While formal fess have already been introduced in upper secondary education (grades 10-11), specialized secondary and higher education, there continues to be strong disincentives for public and private institutions to generate revenue from tuition income because of high taxes. The existence large informal payments argues for greater explicit reliance on user charges, especially in higher education. Formal student fees should be increased and designed to target those persons who have the ability to pay. 6.58 Providing an Enabling Environment for the Development of the Private Sector. Despite the significant growth in the number of private higher education institutions -and strong demand for higher education, the share of enrollments in the private sector is still very low, suggesting room for further expansion. Incentives for private higher education institutions could be provided through reduction/ elimination of taxes. Also, there is a need to establish a system of quality assurance (e.g., accreditation) and to analyze what changes in the regulatory and institutional framework should be taken to encourage further expansion of the private sector. Improve Equity in Education 6.59 In a situation where the financing of education has been essentially privatized, the poor have suffered the most both in terms of access and quality of education. Enrollment rates are lower among the poor and poor children are more likely to have prolonged absences from school. Further, poor families face considerable problems of access to education due to recent changes in funding patterns especially in non-compulsory education (e.g., the rising costs of education, informal payments and payments for private tutoring). The same is true of children in many rural areas where school quality has deteriorated dramatically by shrinking public resources. As a result, a challenge facing the authorities is assuring the access of the poor to education. Policy options targeted to the poor may include provision of: (i) free of charge basic education materials (e.g., textbooks); (ii) education and training subsidies particularly for poor students (e.g., allocating some scholarships in higher education on the basis of needs as well as academic merit). Under the Education APL, the Government plans to reduce the existing inequities in access to text books and learning materials through the establishment of a text book rental scheme and of school learning improvement grants. It is also recommended that government defines a rationale for the provision of financial aid to ensure the poor also have access to higher education. (A student loan policy may not be an alternative policy given the fact that both labor and financial markets are not yet well functioning). Improving the Quality and Relevance ofEducation 6.60 Preliminary evidence, mainly based on input based indicators, suggest that the quality of education at all levels has deteriorated during the transition as a result of outdated curricula and teaching methods, lack of textbooks and basic teaching and learning materials, poorly paid and trained teachers, and poorly maintained physical facilities. Better execution of the education, including regular payments of teachers salaries, should help to improve teaching conditions. To improve the quality of education, priority should be given to the following policy options: (i) to improve curricula and teaching methods to reorient the country's education system to meet the new skills needs of a market economy; (ii) introduce a new system of student monitoring and evaluation to reorient the country's monitoring and evaluation process from mainly inputs towards a system which monitors inputs, teaching and learning processes (e.g. teaching and learning methods and their flexible application) and outcomes (what knowledge and skills are acquired by students and how these are relevant to the labor market, the requirements of the higher education system, and societal goals); (iii) Provide basic education materials and supplies, especially for 114 Georgia: Public Expenditure Review November 25, 2002 primary grades and for the poor (iv) provide teacher training. to improve the effectiveness of teachers in introducing new standards and methods; and (v) Improve management capacity, including management information systems, which are vitally important for improved cost efficiency and effectiveness in the use of resources available for education. Under the Education APL the Government plans to address each of the recommended policy actions. 115 Georgia: Public Expenditure Review November 25, 2002 Chapter 7. Social Protection 7.1 Social protection programs are aimed at assisting individuals, households, and communities to better manage systemic and idiosyncratic risks and reducing inequities. The state can adopt policies that allow individuals to reduce and/or mitigate income risks as well as to cope with risks that materialize through consumption-smoothing programs. 7.2 Social protection programs in Georgia comprise (i) pensions, (ii) cash transfers to internally displaced persons (IDPs); and (iii) a limited poverty benefit program. These are essentially risk-coping programs that provide regular monthly transfers to about 1.2 million Georgians or over 25 percent of the population. Their total cost accounts for 4.2 percent of GDP or around 20 percent of total public spending. 7.3 The review presented below shows that social protection programs in Georgia have been found to be effective in alleviating poverty. However, efficiency and effectiveness of these programs could be significantly enhanced. In the short term, this could be achieved mainly by (i) ensuring regular payments of benefits, specially pensions; and by (ii) improving administration and monitoring of social programs. In the medium term, some of the programs would need to be restructured. For example, regarding the pensions system, stronger links need to be established between benefits and contributions. Also, there is a need to introduce targeting in the IDP program which eventually should be integrated into the poverty benefit program. Public Pensions Current Situation: Arrears Problem 7.4 Georgia's pension system is mandatory for all economically active individuals, including self- employed in agriculture. The system comprises old-age, disability and survivors' benefits. An individual is eligible for a labor pension at the age of 60 (women) and 65 (men), after at least five years of contribution history. There is no early retirement provision or favorable counting of years of service, except for two years spent in the Army during regular service. Disability pensions are certified by special commissions under the Ministry of Health, Labor and Social Protection. Family members of a deceased pensioner or a person who qualified for a pension before he/she has died, are eligible for a survivor's pension. Working pensioners cannot receive their pension as long as they are employed. Individuals with less than five years of working history (women at 65, men at 70), invalids since childhood, and other individuals not qualifying for a labor pension are entitled to a social pension. 7.5 Labor and social pensions are flat: currently GEL14 per month and are not adjusted automatically. Certain categories of pensioners- such as war veterans and invalids, political repression victims -receive higher pensions (at present up to GEL45 per month). Labor pensions are financed through a payroll tax. The payment is not individualized, e.g., the tax is paid as a bulk sum on the payroll."' In October 1998, the collection of contributions was transferred from the pension fund to the Ministry of Revenue. Social pensions are supposed to be financed from the central budget. 7.6 Pensions are administered by the Georgian United State Social Insurance Fund (GUSSIF), established in 1991. The fund is an autonomous public administration agency with the central office in .. Eligibility for retirement is still proven by the labor books (or by the two witnesses confirming working history). 116 Georgia: Public Expenditure Review November 25, 2002 Tbilisi and city and local branches covering entire country.12 Most of the pensions are delivered to beneficiaries personally through the postal system. In Tbilisi and other cities some pensions are paid to individual bank accounts. 7.7 The shortage of funds has been a persistent feature of the pension system. Until August 2000, the fund accumulated arrears of GEL97 million"' (1.6 percent of GDP ) with most of arrears incurred in 1998. While pensions have been paid more regularly since then, the situation remains very fragile: as of September 2001, the stock of arrears had increased by a further GEL1O million."4 Table 7.1: Pension Fund, Revenues and Expenditures, 1996-2000 1996 1997 1998 1999 2000" In GEL millions Revenues 111.0 130.8 149.7 163.8 177.5 Employer contributions 60.2 90.6 108.6 111.9 140.0 Employee contribution 2.8 3.4 4.8 4.1 n/a Transfer from the budget 41.7 32.8 30.8 44.7 27.1 Other 6.3 4.0 5.5 3.1 10.8 Expenditures 109.1 130.8 146.9 162.8 177.5 Pensions 94.7 121.0 134.2 148.5 157.4 Other benefits' 9.9 7.6 8.2 10.3 152 Administrative expenditures 4.5 2.2 4.5 3.9 4.9 Arrears accumulated during the year 1.9 9.2 40.3 28.3 17.3 In percent of GDP Revenues 2.9 2.9 3.1 3.0 2.9 Employer contributions 1.6 2.0 2.2 2.0 2.3 Employee contribution 0.1 0.1 0.1 0.1 0.0 Transfer from the budget 1.1 0.7 0.6 0.8 0.4 Other 0.2 0.1 0.1 0.1 0.2 Expenditures 2.9 2.9 3.0 3.0 2.9 Pensions 2.5 2.7 2.8 2.7 2.6 Other benefits 0.3 0.2 0.2 0.1 0.2 Administrative expenditures 0.1 0.0 0.1 0.1 0.1 Arrears accumulated during the year 0.1 0.2 0.8 0.5 0.3 1/ Budget for 2000 includes GEL3.8 million of "other" revenues and 6.6 million rolled over from 1999, on payment of pensions it includes GEL6.3 million of repayment of arrears. 2/ Other benefits include compensation/allowance for maternity and sick leave. Source: The GUSSIF estimates of the budget execution. 7.8 Arrears are unevenly distributed across the regions. The better off regions of Ajaria and Tbilisi are almost current with the pensions, while the poor regions of Guria, Imereti, Kakheti, Samtskhe- Javakheti, and Inner Qartli accumulated 10 to 12 months of benefits due. Arrears in payments are to the local population; there are almost none to internally displaced persons. The uneven regional distribution of arrears was caused by the system of collection of social taxes and payment of benefits inherited from the Soviet times, which, when faced with severe financial crisis, effectively turned a national program of 112 In addition GUSSIF administers sick-leave compensation, and pays military pensions, as well as a poverty benefit. Military pensions and poverty benefit are funded by the state budget. us Excluding arrears on pensions for military and other law enforcement agencies estimated at GEL17.7 million. 114 Because of the poor financial management of the pension system, the data are estimates, at best, and should be read as an indication of trends. 117 Georgia: Public Expenditure Review November 25, 2002 social insurance into a regional (even municipal) program. Until 2001, each region paid pensions from the social taxes it collected and deficit regions had to rely on a surplus of funds collected in other regions. Faced with a chronic deficit, the system collapsed, creating large regional inequities with an adverse impact on poverty in these regions. To address the problem of uneven regional payments, in January 2001 the GUSSIF adopted a new benefit's payment mechanism. Pensions in better-off ("donor") regions are still paid before the pensions in worse-off regions, but if arrears are incurred in other regions, payments for the next period in the better off regions are put on hold until benefits for the previous month in all the regions have been paid. 7.9 In principle, financial performance of the pension system is determined by three groups of factors: (i) the pension system, i.e., the number of beneficiaries and the level of benefits it generates; (ii) efficiency of administration and delivery systems, and (iii) general economic and fiscal performance. The Pension System"5 7.10 The number of pensioners or the general level of pensions do not explain the financial difficulties of the pension system. The number of pensioners has been decreasing (due to increase in retirement age and cleaning of the records) and the pensions are low and have not been indexed for three years. 7.11 The Number of Pensioners. As of May 2001, there were 860,000 recipients of pensions"'6 (about 20 percent of the total population) '-approximately 100,000 less than at the end of 1998. The drop is both the result of the increase in retirement age, as well as cleaning of the GUSSIF rolls of the deceased pensioners, those receiving several pensions at the same time,118 and those not present in the country. About 90 percent (767,000) are labor pensioners (those with at least 5 years of employment). Of this, 6 percent are survivor's, about 20 percent disability pensioners, and 74 percent old-age pensioners. The remaining 10 percent are social pensioners (with less than 5 years of employment, invalids since childhood, orphans and survivor's of recipients of the social pension). 7.12 The total number of pensioners does not seem excessive, as indicated by the dependency ratio (the number of pensioners over the number of people of retirement age) is equal to one)."'9 But, Georgia the share of disability pensioners-about 20 percent of the total, is high compared to other countries in the region where the share is lower than 10 percent. Also, while the relative size of privileged pensioners does not appear to be particularly excessive (about 10 percent of all pensioners), the composition by age could be of some concern: 3 percent of privileged pensions recipients are below 40, and 9.5 percent are between 40 and 60 years of age. '" All data are estimates. Statistics on pensions are unreliable and some of the data that are crucial for expenditure projections (e.g., age specific mortality rates) are completely lacking. Recently, the GUSSIF started updating its data base on pensioners, but this work will take some time to complete. 116 Excluding military and law enforcement retirees whose retirement is regulated separately and pensions funded by the central budget. The estimated number of recipients is about 20,000, with the yearly bill of about GEL20 million (on average over GEL80 per pensioner per month). " Estimated at 4.4 million in the 2002 census. 1" The record is held by a pensioner who was receiving five pensions. "9 When military pensioners are included the ratio increases to 1.03. 118 Georgia: Public Expenditure Review November 25, 2002 Table 7.2: Pensions Relative to Average W ge and Poverty Line, 2000 Type of Pension Benefits in In percent of GEL Average wagel Poverty line7j Basic labor and social pension 14 22 27 Average privileged pension 38 58 74 War participants' survivors 18 28 35 Granted status of war veterans 35 54 68 War veterans younger than 70 40 62 78 War veterans and invalids over 70 45 69 87 Law enforcement agencies pensions 80 116 160 Memorandum Items: Average Net Wage" 65 Alternative Poverty Line 2/ 51.5 1 / Georgian State Department of Statistics Household Survey 2000. 2/ Georgian SDS cost of consumption basket providing daily 2,200 Kcal in lari per month to an equivalent man with working capacity 7.13 Benefits. The basic labor and social pension of GELl4 per month, paid to 90 percent of beneficiaries, is quite low and only covers a little bit more than one quarter of the poverty line. The situation is better for the military and the privileged pensions (Table 7.2). 7.14 While privileged pensions are below the poverty line and the number of privileged pensioners is not excessive, the total bill for this category is substantial. About GEL25 million per year -- 0.4 percent of GDP in 2000-- is needed to cover on average GEL24 per month per privileged pensioner as a top up to their basic pension. This raises equity concerns, given the low level of the basic pension and the fact that the privileged pensions top-up is partially funded (cross-subsidized) from social tax revenues, because of the shortage of the state budget transfers. There is also an issue as to the affordability and sustainability of privileged pensions, given overall limited resources. Administration 7.15 Despite recent positive steps to revamp the system, Georgia is yet to establish a reasonably efficient social insurance administration. The administration of pensions has been characterized by poor record keeping, weak information flows, lack of office equipment,120 outdated accounting practices, lack of internal and external institutional and financial audits, insufficiently skilled staff, poor organization, poor financial management, etc. With an administration in disarray, the GUSSIF was, until recently, unable to produce accurate and updated lists of beneficiaries, let alone to control the financial flows and prevent the leakage of scarce resources. Similarly, the delivery of benefits, which mostly done through direct delivery of cash to pensioners by postmen, is said to be slow and often non-transparent. For instance, the Post Bank which was used to channel the funds into the postal delivery system was reportedly believed to hold the money for some period before making it available for the payment of benefits. Also, anecdotal evidence suggests that tipping postmen for pensions delivery is almost "institutionalized". 121 120 Except for the central office and the Tbilisi regional office, there are no computers in the social insurance offices. Georgia is probably the only country in the region that has not computerized its social insurance administration yet. 121 One half of a lari (50 tetri) in urban areas. 119 Georgia: Public Expenditure Review November25, 2002 7.16 Since May 2000, when the new management of the GUSSIF was appointed, important steps have been taken to streamline the administration and financial management of the system and improve service delivery: the records have been cleaned and the data base computerized (for the time being at the central level); in order to avoid regional inequities in benefits delivery, the payment of pensions was centralized and has been monitored closely; the Post Bank was replaced by a consortium of three commercial banks that won the tender for benefits delivery; new business plan and operational manuals are being developed, the project and a work plan for introduction of individual accounts, corresponding computerization and training of the staff has been completed, etc. General Tax Revenue Performance 7.17 A greatly eroded tax base due to steep economic downfall till 1994 and slow rebounding afterwards, as well weak capacity to collect taxes, have been the key factors behind financial difficulties of the pension system. 7.18 Small Formal Taxable Wage Bill. Economic developments during the 1990s have translated into levels of formal wage employment as well as of average wages. Only 20 percent of the working-age population is wage employed (about 700,000 people) and, accordingly, subject to taxation for social insurance. Self-employed, including farmers, do not contribute at all, although they should because they are included in the mandatory social insurance coverage. The outcome is that there are less contributors to the social insurance system than beneficiaries, resulting in a system support ratio of 0.8. This low ratio has been stable for a couple of years, reflecting the narrow based economic growth in Georgia, as well as a large scope of informal employment (to which high labor taxation is one contributing factor). In addition, formal wages are low: GEL65/83 per month, net and gross respectively. The combination of the two results in a very low wage bill-about 12 percent of GDP according to the household survey conducted by the Georgian State Department of Statistics (SDS). Table 7.3: Potential Contributions and Collection Rate, 2000 Number of Average Persons Gross (000) Wage (GEL) Wage employees 696 83 Public administration 312 54 State owned enterprises 165 78 Private sector 219 128 Potential wage bill (GEL million) 695 Pension Fund payroll tax rate GEL16 for gross wages up to GEL60; 28 percent for 60+ Potential contributions (GEL million) 59.5+43.2+94.2=197 Pension Fund contributions revenues (GEL million) 144 Implicit collection rate (percent) 73 Source: SDS Household survey, and Pension Fund. 7.19 Relatively Low Collection Rate and Widespread Under-Reporting. With the social tax of 28 percent, 122 and assuming a collection rate of 100 percent, the reported wage bill from employment of 12 122 Wages are additionally taxed for unemployment insurance (1 percent) and health insurance (4 percent). In contrast to pensions where the 28 percent tax is treated as a social tax and collected by the tax authorities, the 4 120 Georgia: Public Expenditure Review November 25, 2002 percent of GDP should yield 3.3 percent of GDP in social tax revenues. Actual contributions to the pension fund in 2000 were 73 percent of the potential yield, suggesting that the remaining 27 percent of the employment wage bill reported in the household survey is not reported for fiscal purposes or/and is under-collected by the tax authorities. (see Table 7.3). The low collection rate of social taxes can be explained by: poor enforcement, weak reporting and monitoring, and lack of incentives for workers and employers to comply with the payment of the social tax. 7.20 In October 1998, the collection of social taxes was transferred from the pension fund to the tax authorities. The transfer was not well prepared and contributed to a very poor fiscal performance in 1998. The tax inspectorate went through several reorganizations concerning the collection of the social tax. In principle, employers should pay the social tax due at the point of withdrawing money from its bank account to pay wages- and no money withdrawal should be allowed without payment of the social tax. Yet, wages are paid and arrears in the social tax accumulated, particularly in the case of big state enterprises. The payroll tax is paid in bulk-there are no individualized lists of employees, their wages and social taxes due/paid. 7.21 In an environment where tax payments are "negotiated" between the tax inspectors and tax payers, it seems that the social tax is given a low priority. Moreover, in contrast to other countries, the GUSSIF that is the most interested party in getting its due revenues has been denied any access to the taxpayers records and thus has no role/power to monitor and control the payment/collection of contributions. While the collection itself is purely an organizational issue, reporting, monitoring and control is not. The practice of limiting the GUSSIF to the mere recipient of whatever funds are deposited on its accounts deprives both the GUSSIF and the Government of needed information to better plan and manage social insurance revenues. Table 7.4: Wage Employees' Average Number of Years of Contributions that Correspond to the Expected Flow of Benefits when Reaching Statutory Retirement Age Average Monthly Gross Wage in GEL Years of Contributions Wage Employees 83 11 Public Administration 54 17 State Owned Enterprises 78 12 Private Sector 128 7 Expected average duration of benefit 18 7.22 Weak Incentives for Participation. Incentives for the employers and employees to comply with the social tax obligation are weak. First, the minimum employment (and presumed contribution) history required is low-anyone who can show 5 years of employment recorded in a workbook has the right to receive the basic pension of GEL14 in case of disability or when reaching the statutory age of retirement. Second, benefits are low and unattractive. Third, the imbalance between the amount of social taxes a regular contributor would put into the system and the amount of benefit he/she would receive effectively discourages workers from participating (Table 7.4). This imbalance is caused by strong redistributive character of the pension system effected through universal coverage, flat benefit and low system support ratio. On average, beneficiaries with less than 11 years of contribution history could be considered winners (provided that the estimated number of years of the benefit receipt is correct). percent tax for health insurance is treated as a contribution to (health) insurance and it is collected separately by the health insurance agency (both instruments are levied in exactly the same way). 121 Georgia: Public Expenditure Review November 25, 2002 7.23 Some differentiation in benefits based on contributions paid may improve the compliance, particularly if coupled with a decrease in the social tax rate. Also, as suggested in Table 1.4, there is a case for increasing the minimum number of years of contribution (not employment) history for an old age and a disability benefit from 5 to 10, so as to at least balance the current benefit stream and the minimum required contribution history, that is to eliminate the "winners" from the system. 7.24 High Labor Taxation. High taxation in Georgia discourages compliance from both employers and employees and promotes informal sector activities. Table 1.5 offers an example of the burden pf taxation on waged employment. Out of the gross wage-for instance GEL 100, the worker is obliged to pay personal income tax (20 percent) and its portion of the social insurance contribution and the health fund contribution (1 percent, each). Thus, the worker's net wage is GEL78. The employer's portion of the social tax is 31 percent. For the employer, the total labor cost in this case is GEL 131, of which net wage makes up 59.5 percent, employer's portion of the social tax 23.7 percent and employees part of the taxes 16.8 percent. Overall, the taxation level over the net wage is 68 percent. Hence, for each additional lari paid to worker in net wage, there is GELO.68 to be paid in taxes if the contract is formalized. Hence, high marginal cost of taxes for both employees and employers creates strong incentives not to formalize the labor contract: employees prefer current to future consumption, while employers seek to reduce costs and increase competitiveness. Table 7.5: Absolute Amounts of Social Insurance Contributions on a Gross Wage of 100 lari pe Month Gross Wage 100 Taxes Paid by Employee Pension Fund I Health Fund I Income tax 20 Paid by Employer Pension Fund 27 Health Fund 3 Employment Fund I Net Wage Worker 78 Income Tax State budget 20 Social Tax SIF/HF/EF 33 7.25 Links with the State Budget. Pensions in Georgia are financed from two principal sources: the social tax and the general budget. The social tax should finance social insurance benefits: labor pensions (old-age, disability and survivor's), funeral allowance and sick leave benefits. The central budget is supposed to finance privileged pensions and social pensions23 as well as to cover the GUSSIF deficit. The GUSSIF function is to administer state social insurance benefits, but in practice it also administers social pensions. Furthermore, its benefit delivery network is used for delivery of military pensions and a family poverty benefit, which are funded by the state budget. 123 The program was inherited from former Soviet Union. The term pension is confusing, because the program is non-contributory and the benefit is paid to certain categories of citizens that supposedly were among the most vulnerable during the Soviet times: those who have never been employed and thus have fallen out of the social insurance coverage or to their descendents incapable of supporting themselves, orphans with both parents deceased, and invalids since childhood. As such, the program essentially belongs to the social assistance, not the social insurance domain of the social protection. 122 Georgia: Public Expenditure Review November 25, 2002 Table 7.6: Financing Needs of the GUSSIF and Sources of their Financing: An Estimate (Without Military Pensions and the Poverty Benefi) Number of Sources of financing and beneficiaries needed amount Total needed Actual financing in (thousand) (GEL million) (GEL million) 2000 (GEL million) GUSSIF State GUSSIF State Budget Budget Labor pensions 780 132.0 132.0 Social pensions 90 14.5 14.5 Privileged 88 25.0 25.0 pensionsi Other benefits 1.5 1.5 Administrative 9.0 9.0 costs Total 1 142.5 39.5 182.0 168.6 19.1 1/ Assuming that GUSSIF covers the base benefit of GEL14 per pensioner per month, while the state budget finances additions to the base benefit Source: An estimate based on the Social Insurance Fund (SIF) data. Military pensions not included. They are awarded separately from the SIF system, funded by the budget and only delivered through the Social Insurance Fund Their total cost for 2001 is estimated at GEL20 million. 7.26 Although the division of who/what source should finances what benefits is clear, in reality the financial flows are not separated (moreover, military pensions and poverty benefit are shown as a GUSSIF budget item, which creates confusion), and the social tax has been used to cross subsidize privileges and social pensions, thus compensating for shortfalls in central budget transfers. Table 1.6 summarizes the financing needs of the GUSSIF and the sources of their financing in the year 2000. It shows that the GUSSIF received from the central budget less than one half of the funds it should have received to fund social pensions and special privileges. If the budget had met its obligations towards the GUSSIF, there would have been room to increase the flat labor pension by 20 percent to GELl7. In 2001, the budget transfers envisaged for funding of the social pensions and privileges amounted to only GELI 1.3 million-about one fourth of the required amount. Hence, the social tax had to cover additional GEL30 million needed for financing of the social pensions and privileges of the special categories of retirees. The plan for 2002 envisages even more cross-subsidization, leaving at the same time the labor pension unchanged for the fourth consecutive year. Pensions and Poverty 7.27 A recent World Bank Poverty Update for Georgia, found that families with pensioners (about 51 percent of Georgian familieS124) do not face significantly higher poverty risk than population in general (respective poverty rates are 23 and 20 percent). However, with a poverty rate of 42 percent, single pensioners and pensioners couples without relatives to support them, as well as pensioners living in families with no bread-winning capacity are found to be among the most vulnerable groups in Georgia. This is consistent with the finding that in addition to economic status, a dependency ratio (a number of persons living on each breadwinner's income) is the main determinant of poverty risk in Georgia. 124 This percent age has increased over the years, reflecting one of the common poverty coping strategies in the region: a return to extended family living arrangements. 123 Georgia: Public Expenditure Review November 25, 2002 Figure 7.1: Simulating the Poverty Impact of the Timely Payment of Pensions 30% -*- Poverty Rate if Flat Pension Paid on Time -1- Poverty Rate Actual 20% .5% 0. . 35%, 0% >~ >~ Source: Georgia: Poverty Profile Update, Report No 22350-GE, June 28, 2001 7.28 Despite its financial problems, the current pension system seems to have been effective in reducing poverty. The Poverty Update points out that other things being equal, if pensions had not been paid at all, poverty rates would have been 9 to 21 percent higher over the 1998-2000 period. Such a result is quite remarkable, especially given the low level of pensions. Depending on the quarter (actually, regularity of pensions payment) during the observed period, each percent of pension expenditures helps eliminate 3.1 to 7.2 percent of poverty among the total population. Figure 1.1 presents a simulation of the poverty rate, if pensions had been paid on time, compared to the actual poverty rate. It illustrates the expected result: even with a low level, regular payment of pensions would appreciably lower the poverty rate. This suggests that putting the pension system on a sound footing should be a priority among poverty reduction policies. Recommendations 7.29 In Georgia, the weak ability of the government to collect taxes and the generally poor economic situation constrain the choices for the development of the pension system in the short to medium term. The consequences for pension policy are clear: given the fiscal situation, only a basic pension level can be afforded in the medium term. This leaves the government with at least two broadly defined short to medium run challenges. Pay Pensions Regularly and Raise its Level in Real Terms 7.30 Given limited resources and based on poverty and equity concerns, the Government should give priority to paying labor pensions and social pensions. The top-ups for privileged pensions should be paid out only to the extent the state budget can afford them. Also, pension arrears should be gradually paid out. Provided that the state budget meets its obligations for financing of the social pensions and pension privileges, the GUSSIF should be able to start increasing pensions and paying arrears. In order to stabilize the financial performance of the pension system and improve it, the following is recommended. 7.31 Improve the Social Tax Collection. Alongside efforts to improve revenue collection, specific actions could be taken to increase the collection rate of the social tax itself, as follows: (i) the payment of the tax should be individualized, e.g., the tax reported and paid for each employee individually; (ii) the 124 Georgia: Public Expenditure Review November 25, 2002 existing rule that no wage payment should be allowed without prior payment of the social tax obligations should be strictly enforced; (iii) the social tax payment should be reported and corresponding documentation sent to both the tax authorities and the GUSSIF; (iv) the GUSSIF should be given full authority to monitor and control the payment of the social tax (together with the tax authorities). 1 7.32 In addition, efforts should be made to increase the number of tax payers, particularly farmers and self-employed. However, given the highly corrupt environment in Georgia and the high cost of collecting social insurance contributions from economic groups such as farmers, an evaluation of costs and benefits and a separate action plan for each group of taxpayers should be prepared. Better enforcement would yield better result only if combined with incentives for participation. These may include: (i) lower social tax rates, (ii) differentiation of the social and labor pensions levels with the former lower than the latter and (iii) differentiation of the labor pension based on the contribution history, combined with an increase in the minimum eligibility history (for instance, from 5 to 10 years). Introduction of individual accounts and related distribution of social insurance cards will provide an excellent opportunity for the GUSSIF to register all potential social tax payers. 7.33 Strictly separate financing of the social insurance benefits from those benefits which are the responsibility of the central budget. The current practice of cross-subsidizing from labor pensions to social pensions and privileges is inequitable and has negative poverty impact. Specifically, the system favors paying higher benefits to special categories of pensioners at the expense of maintaining the low level of labor pensions (the later is 2.5 times lower than the former). This policy should be phased out and the state budget should fully fund social pensions and privileges. The separation is consistent with the Government's on-going efforts to introduce individual accounts into the pension system. [Also, the benefits such as military pensions and the poverty family benefit that are only delivered to beneficiaries by the GUSSIF, should not be included in the GUSSIF budget, because their inclusion additionally confuses the budgeting process, as well as the budget execution.] 7.34 Maintain Tight Eligibility Controls. The Government should refrain from actions that would effectively decrease the retirement age? It should also refrain from extending current or introducing new privileges. Moreover, all privileges should be reviewed and individual decisions to grant them [gradually] audited. The rule that any privilege should be affordable and sustainable should be strictly enforced. Another area of concern are disability pensions. In order to decrease currently high number of disability pensioners, the third category disability should be abolished, decisions to award disability regularly audited, and disability certification for social insurance benefits moved back to GUSSIF following the principle that "who pays, decides". Improve the Administration of the Pension Fund 7.35 Further improvements in information and financial management systems and in the benefit delivery mechanism are needed. To that end the following measures is recommended: (i) conduct an institutional and functional review of the social insurance administration and develop an action plan for improvements, including development of a new business plan, operational manuals and procedures, etc.; (ii) introduce management information system comprising a data base on pensioners, those covered by the social insurance (participants) and a data base on employers. Introduce individual accounts. Computerize financial information flows; (iii) improve reporting procedures, introduce international accounting standards, and regular internal and external auditing; (iv) review the current skill mix and adjusting it to the needs of the new administration; and (v) develop a plan to minimize cash handling in the delivery mechanism. 125 For instance, better collection of income tax in comparison to social tax should automatically trigger control of both taxpayers and tax collectors, because both taxes are levied on the same base. 125 Georgia: Public Expenditure Review November 25, 2002 7.36 Successful implementation of the above measures in the short to medium run, would allow Georgia not only to increase pensions and clear arrears, but also introduce a new pension system based on much stronger links between benefits and contributions paid. Social Protection of Internally Displaced Persons (IDPs) 7.37 Social protection of internally displaced persons is the second largest social protection program operated in the country. It provides regular monthly cash income support to IDPs and consumes approximately 1 percent of GDP annually. The program constitutes about one fourth of public spending on social protection and close to 6 percent of total public spending in Georgia. The relative generosity of the program and weak administration may have created incentives for abuse and leakage of funds, resulting in waste of scarce public resources. Furthermore, while the assistance is found to improve welfare of the IDP households, its untargeted nature raises some important equity concerns both in the case of local versus IDP population, as well as within the IDP population itself. IDPs in Georgia: Numbers, Placement and Socio-Economic Situation 7.38 Secessionist movements following the disintegration of the Soviet Union and Georgian independence instigated the eruption of two ethnic conflicts in Abkhazia and South Ossetia that led to the displacement problem the country is continuing to struggle with today.126 7.39 Magnitude and Location. According to the UNHCR, there were 272,100 IDPs living in Georgia at the end of 2000,127 96 percent from Abkhazia. Children account for 27 percent, elderly for 17 percent and working age adults for 56 percent. IDPs are mostly concentrated in the Samegrelo region (42 percent), Tbilisi (33 percent) and the Imereti region (12 percent). In recent years, IDPs have been increasingly moving to urban centers, pushing the number of IDPs in Tbilisi from about 40,000 in 1997 to about 90,000 today.128 7.40 Accommodation. Just over half of IDPs live in private accommodation -- with friends or relatives or in their own rented or purchased accommodation. The reminder reside in collective centers -- mostly hotels and sanatoria, schools or converted factories. According to UNICEF, a disproportionately large share of IDP children live in collective centers.129 Providing adequate shelter for IDPs is a serious issue, particularly for those living in collective centers, for most of the centers are overcrowded and run-down, with dismal sanitary conditions.'30 The 1996 Law on IDPs bars them from purchasing property in Georgia proper by tying this to the loss of the IDP status. While some IDPs have nonetheless purchased property clandestinely, this clause keeps many of them in a provisional and emergency-type lodgings. Moreover, many IDPs staying with friends or relatives have become a heavy burden on their impoverished hosts. 7.41 Economic Status, Poverty and Vulnerability. A special survey conducted by the International Federation of Red Cross (IFRC) in May 2000"' finds that IDP households, and in particular those 126 For more up to date and detailed information on the causes and reality of displacement in Georgia go to http://www.db.idpproiect.or/Sites/1dpProiectDb/idpSurvev.nsf/wCountries/Georgia. 127 A 1996 UNHCR-sponsored survey registered 282,000 IDPs in Georgia. 128 Statement by a Ministry of Refugees official at a conference on IDPs in Tbilisi in July 2001. 129 UNICEF Georgia. Data available at http://www.unicef.org/programme/countryprog/cee-cis/georgia. 130 Some of the international humanitarian organizations have in the past worked on the rehabilitation of collective centers. However, these programs are being phased out in favor of programs fostering self-reliance of IDPs. 131 Internally Displaced Persons- A Socio-Economic Survey for Georgia; A Report based on the Survey, conducted by the IFRC, the Georgian Red Cross Society and the State Department for Statistics of Georgia. 126 Georgia: Public Expenditure Review November 25, 2002 residing in collective centers are more vulnerable than local population with respect to employment and income, land ownership'32 and housing conditions. Single female-headed households with children and households with no one gainfully employed (especially in collective centers) were found to be particularly vulnerable. The IDPs residing in Tbilisi are found to have better employment opportunities than those located in the Samegrelo and Imereti regions. In all the three regions, IDPs in private accommodation fare better than those in collective centers. The Survey found no significant differences in the level of consumption between the local population and IDPs. This appears at odds with the finding that IDPs are at a substantial disadvantage over the local population, and it is explained by the fact that IDPs are much better covered by various public and Non-Govemmental Organizations (NGOs) assistance programs (Figure 1.2). Based on the IFRC survey data, the World Bank "Poverty Update for Georgia"'33 concludes that "poverty among IDPs is higher than among the local population by at most a third". 7.42 Vulnerability Grouping of IDPs. Recent Vulnerability and Capacity Assessment for Georgia (IFRC)' divides the IDPs into four vulnerability groups, reflecting the fact that during eight years of displacement IDP households have adjusted differently to their displacement: (i) IDPs living in private accommodation in and around Tbilisi, which can be viewed as "economically settled, since many of them have adapted themselves to local conditions, found employment and often proper housing"; (ii) IDPs living in private accommodation near the boundary line to Abkhazia, mainly in the Samegrelo and also in the Imereti region, many of whom keep close contacts with Abkhazia and some of them commute to the Gali district to work on their land; (iii) IDPs in collective centers in the Samegrelo region, who are "openly vulnerable"; and (iv) IDPs in collective centers in and around Tbilisi, of which some only maintain nominal residence in collective centers, in order to retain their IDP status, while actually living in new-found accommodation elsewhere. Government Assistance to IDPs and its Impact on their Welfare 7.43 IDP Allowance. In Georgia, a monthly cash allowance is paid to all IDPs. Those in private accommodation are entitled to GEL 14 per person per month, while IDPs in collective centers qualify for GEL 11 per person per month. Until end 2001, the latter received additional support to cover communal services of GEL8.75 per person per month (GEL4.75 for electricity, GEL2 for water and other services and GEL2 for cleaning), which was paid directly to the collective centers. As of January 2002, for IDPs residing in collective centers in Tbilisi the subsidy for electricity is increased to GELl2 per capita per month, raising the total monthly level of assistance to GEL27 per capita. The electricity subsidy is now paid directly to the AES "Telasi", a local electricity distribution company. 7.44 The allowance is financed from the central budget. Each year, the Ministry of Refugees provides the Ministry of Finance with the numbers of both categories of IDPs, on whose basis the budget for the allowances is calculated. The amount allocated for cash IDP allowance in the 2001 budget was GEL44.23 million (0.73 percent of 2000 GDP), while the amount allocated for communal services in collective centers was GELl3 million (0,22 percent of 2000 GDP). 132 Farming is an important source of income in Georgia, both for rural and urban population. According to the law, local authorities are obliged to provide IDPs with a plot of land for temporary use. The IFRC survey finds, however, that among the IDPs residing in rural areas, 71 percent of those placed in collective centers and 33 percent of those residing in private accommodation do not have any access to land. Furthermore, the amount of land available to most IDPs is not sufficient to yield a sizeable income. A UN-commissioned survey on the IDP living conditions conducted by the Georgian Young Lawyers' Association (GYLA) finds that the authorities frequently fail to provide adequate land (provided land is of poor quality or too far away from an IDP settlement) and demand bribes. Furthermore, IDPs often lack training, equipment and seeds to cultivate the land. IDP households are also found to possess less livestock than local households. 33 World Bank (2001), Poverty Update for Georgia, draft. '1" IFRC (2000), Georgia: Vulnerability and Capacity Assessment. 127 Georgia: Public Expenditure Review November 25, 2002 Figure 7.2: IIDPs Monthly Expenditures in Lari with and without Government Cash Assistance 160 U Average Monthly 140 Government 120 Assistance 60 0 Household 40 Expenditure as it would be without 20 Government 0 A' Assistance Local Population IDP Private IoP Collective Source: IFRC (2000), Internally Displaced Persons: A Socio-Economic Survey for Georgia. 7.45 Free Transportation. IDPs in Tbilisi benefit from free travel on metro and public buses, for which they receive a special identification card. From 2001 onwards, the scheme is being funded by the Tbilisi municipal authorities (before it was financed from the central budget). 7.46 Education. IDP children are entitled to education in public schools free of charge. Occasionally, the Ministry of Education makes funds available for IDPs to buy textbooks, but there is no special budget for this. The "Abkhaz Government-in-Exile" whose activities are funded by the central budget operates some schools exclusively for IDP students, especially in the Samegrelo and Imereti regions. These schools are attended by about 3 percent of children living in collective centers and about 1 percent by those residing in private accommodation. Also, IDP pu ?ils rarely attend separate classes in mixed schools (about 4 per cent of children in collective centers)' . Only in Ajara, about one fifth of the IDP pupils go to exclusive IDP schools. 7.47 Access to education is characterized by near equality between IDPs in collective centers, private accommodation and the local population. While more IDPs in collective centers attend kindergarten, IDP children attend school in about the same proportion as the local population. Enrollment rates in higher education do not vary significantly across the three groups. However, there appears to be lower school attendance among the IDP children. Schools catering exclusively for IDPs face more severe financial problems, and problems of finding adequate premises and rehabilitating them. 7.48 Health Services. The state-owned medical insurance company runs a special program with the annual budget of GEL300,000 to cover certain medical services for IDPs. The Ministry of Health also provides funding for rehabilitation and treatment for handicapped IDPs and the 2001 budget allocation for it was GEL128,000. The Abkhaz Government-in-Exile finances six policlinics and one hospital for IDPs, paying staff and other running costs, but no drugs136. The Ministry of Refugees also operates a special fund for "extraordinary cases", for which individuals can apply to cover one-off special expenses, such as special medical expenses or expenses for a funeral. [The amounts JDPs can claim range from GEL50-300, and the 2001 budget allocation for it was GEL300,000.] 135 IFRC (2000), Internally Displaced Persons: A Socio-Economic Survey for Georgia. 136 According to the Ministry of Health, Labor and Social Affairs. 128 Georgia: Public Expenditure Review November 25, 2002 7.49 According to the IFRC study," "IDPs are not significantly different from the local population as regards extend of illness, mortality or method of intervention from illness". However, a slightly higher 138 incidence of stunting is found among the IDP than among the general population Issues 7.50 As mentioned before, public resources in Georgia are scarce, the needs for public moneys numerous, while governance and capacity are poor. Many important sectors or programs to alleviate poverty are inadequately funded or often not funded at all. Competing needs and limited resources require that funds are used efficiently and in an effective manner. In the case of the program for IDPs, there are a number of issues that affect its efficiency and effectiveness, as discussed below. 7.51 Benefit Design and Eligibility Requirements. The program was introduced in the aftermath of the displacement crisis. Its design (universal, flat rate benefit) and only one eligibility requirement (the IDP status) were determined by (i) the more or less similar economic situation in which IDPs found themselves right at the beginning of their displacement; (ii) low capacity to administer more complicated schemes; (iii) expectations that the displacement would be a temporary phenomenon. Contrary to expectations, the displacement turned out to be a long term state (almost a decade so far), with no certain prospects for the return of IDPs to their homes. In the mean time, IDPs have adjusted differently to the displacement, with some of them finding jobs or establishing businesses and settling down. The social and economic differentiation that has taken place has created different vulnerability "geography" among the IDP population and calls for reexamining the benefit design for efficiency, effectiveness and equity reasons. 7.52 Benefit Administration and Benefit Delivery Mechanism. The assistance is administered by the Ministry of Refugees (MoR) which requests funds from the Ministry of Finance based on the number of IDPs. On behalf of the MoR, treasury transfers the funds to the Post Bank (PB) of Georgia. The system is similar to that used for pensions: the allowance is disbursed through the PB branch offices throughout the country and, in order to receive the allowance, registered IDPs need to go to one of the branch offices to draw it. However, the PB cannot guarantee the regular disbursement on specific days, since it receives funds from treasury on irregular basis (the situation has significantly improved since August 2000). Also, there is no system of notification-IDPs typically become aware of the availability of the allowance at the local PB branch though rumors. Funds not drawn accumulate in the individual's account. 7.53 Payment delays and arrears have caused repeated protest by the IDPs, especially in those regions where their concentration is high. These protests notwithstanding, there is no evidence that IDPs have been disadvantaged with regard to the payment of assistance.' According to the IFRC survey, respondents of both local population and IDPs (in private accommodation and collective centers) reported having received 7-8 payments of social transfers in the 12 months prior to the survey.140 There is little difference in social transfers payment arrears between the local population and IDPs. 7.54 Lack of monitoring and control mechanisms in Georgia's highly corrupt environment, creates incentives for leakage and wastage of scarce funds, both regarding the assistance paid to IDPs and the management of collective centers to cover energy and utility costs. Per capita utility costs compensation creates incentives for collective centers management to report as many residents as they can, and does not 137 IFRC, ibid. "3 UNICEF (1999) Multiple Indicator Cluster Survey on the Nutritional Status of Children in Georgia; IFRC (1998), Nutritional Status Survey oflDP Children in Western Georgia. 139 It may even be the other way round. For instance, pensions to IDPs are paid before they are paid to the locals. 140 IFRC (2000), Internally Displaced Persons: A Socio-Economic Survey for Georgia. 129 Georgia: Public Expenditure Review November 25, 2002 stimulate beneficiaries to save electricity and water (when supplied). The new payment arrangement for electricity spent in collective centers in Tbilisi (GEL12 per resident per month payment transferred directly to the electricity distribution company) may further dilute any incentives for cost-savings. 7.55 Equity. Universal, flat, relatively generous cash assistance for IDPs puts at disadvantage local households that are in equally vulnerable situation. Of two equally vulnerable households: one local, one IDP, the IDP one is going to be better-off after receiving the benefit. In addition, the disadvantage of local households may be even greater, because IDPs receive humanitarian assistance, although its amount has substantially decreased over the years. The assistance also maintains inequities among the IDP households by treating everyone equally, irrespective of their economic and social status. 7.56 Horizontal equity concerns (local versus IDP population in the same socio-economic situation) can only be addressed by equalizing access to assistance for both groups of the population. In the short run this will be difficult to achieve due to severe shortage of fiscal funds and a very complex political economy. Vertical equity concerns among the IDP population itself can be addressed by replacing universal flat benefit by targeted and (possibly) differentiated benefit. This will also be difficult to achieve in the short run because: (i) the development of an appropriate targeting mechanism requires further analytical work; (ii) even a simple targeting needs appropriate administrative capacity, which is lacking now and (iii) complex political economy. 7.57 Assistance Dependence and Work Incentives: The assistance may have created a "dependency" syndrome among some of the IDPs, effectively discouraging them from seeking employment more actively and persistently. Recommendations 7.58 There is a significant scope for improving efficiency and effectiveness of public resources allocated to the program. Improvements can be achieved (i) though streamlining the benefit administration and delivery, and (ii) through restructuring of the program and introducing targeting. The former could be undertaken almost immediately. The latter, however, is conditioned upon a political decision and firm political backing and requires time to develop targeting mechanism and build capacity to implement it successfully. 7.59 Streamlining the Benefit Administration. Immediate steps could atke to improve the administration and delivery of the benefit. These include: (i) re-registration of beneficiaries and introduction of mandatory annual re-registration of IDPs; (ii) audit all branches of the PB disbursing the benefit to IDPs and follow with regular quarterly audits; (iii) develop clear, strict rules on who can withdraw the funds from the account with random checks on the spot; (iv) develop monthly reporting procedures of the PB to the MoR, Treasury and the Ministry of Finance; (v) conduct annually beneficiary assessment of the IDPs; (vi) entrust monitoring of benefit delivery to NGOs with results reported quarterly; (vii) review and streamline organization of government agencies involved in administering support to IDPs and develop appropriate operational guidelines; (viii) transfer the administration of the assistance to the local social protection offices (but after improving their capacity to administer the program); (ix) introduce quarterly audits of collective centers; (ix) review and recalculate current utilities compensation amounts and introduce a principle of reimbursing receipts instead of per capita transfer of funds to collective centers and the utility company (for electricity); (x) start developing a utilities payment mechanism in collective centers that would individualize utilities consumption and a responsibility for the bills payments. 7.60 Restructuring the Assistance. The following options to increase self-reliance of IDPs and provide assistance to the most needy ones could be considered: 130 Georgia: Public Expenditure Review November 25, 2002 * Transform the IDP allowance into a poverty benefit for IDPs by introducing targeting mechanism (for instance, based on the IFRC vulnerability grouping). * Introduce self-targeting mechanisms. In addition to introducing mandatory annual re- registration for the receipt of income support, the IDP allowance in full or in part could be tied to participation in public works programs (when and where available). * Abolish special funds for IDPs over and above the monthly cash benefit. This applies for instance for the MoR "fund for extraordinary cases", whose budget is too small to make a difference and whose operation is costly. Instead the resources should be included in the IDP poverty benefit budget. * Focus on employment generation for IDPs. Attempts should be made to increase the employability of IDPs through training programs to update skills or facilitate their job search. In rural areas, where self-employment in agriculture is most common, IDPs can raise household income through improving access to land and to higher quality land in particular. Furthermore, IDPs in rural areas also might benefit from agricultural training to improve the productivity of land use.141 * Offer alternatives to collective centers. Collective centers are not an adequate shelter solution for long term displacement. There is a need to identify ways to help collective center IDPs move into alternative and more "humane" private accommodation if they wish.142 In particular, there is a strong case for abolishing the law preventing IDPs from purchasing property. Enabling IDPs to settle in a more long-term living arrangement helps them to build the very social contacts needed forjob search and income generation. * In the longer term integrate vulnerability criteria for IDP and non-IDP population so as to provide equal access to assistance to equally vulnerable households, irrespective of their residential status. 7.61 Any changes to the current IDP allowance system need to be well communicated with the IDP community and the population as a whole so as to increase its acceptability. The retargeting should be designed such that both the less vulnerable (through better chances in the labor market) and the vulnerable (through a higher and regularly paid benefit) feel they gain from the changes. The fact that the current system works so poorly may be an advantage, since a retargeting of the allowance could be portrayed as an improvement vis-A-vis the current one-size-fits-all allowance. The government should therefore establish a task force, including IDP representatives, to review the current system and propose changes in consultation with stakeholders. For example, there are a number of NGOs focusing on IDP rights advocacy and advisory services who could represent IDP concerns. Potential Savings 7.62 The estimates represented are very tentative. Their purpose is to give a general idea on how much could be saved if the administration and delivery of the cash assistance to IDPs is streamlined, monitored and controlled and if targeting is introduced. 7.63 Improving Administrative Efficiency. Based on the GUSSIF experience with re-registration of beneficiaries, re-registration of IDPs and streamlining of benefit administration and delivery could result in savings of up to 5 percent of total spending on cash assistance (about GEL2.8 million per year). In addition, review and revision of utility subsidies to the collective centers and electricity distribution 141 According to the WFP representative in Georgia, productivity of land use remains low, especially for subsistence farming, and could be raised by training farmers. 142 Not everyone might want to leave collective centers. However, the longer displacement continues the smaller this group can be expected to be. 131 Georgia: Public Expenditure Review November 25, 2002 company, as well as change in the reimbursement mechanism would most probably result in more savings. 7.64 Introducing Targeting and Benefit Differentiation. Different targeting scenarios can be developed, resulting in different savings. Given weak administrative capacity and highly corrupt environment, a targeting mechanism should be simple to administer and relatively "easy" to monitor. Three targeting scenarios resulting in savings of 13-25 percent of the current budget allocation of GEL57 million are presented below. It is assumed that increased administrative costs related to targeting could be covered by savings generated from streamlining of the current benefit administration and delivery. (a) All children up to 18 (assumption: 60 percent are in collective centers, 40 percent in private accommodation) and all elderly over 60 (assumption: 60 percent reside in collective centers, 40 percent in private lodgings) remain eligible. Economically active IDPs: one third of working age individuals residing in collective centers (24,000 IDPs) and one half of working age individuals in private accommodation (43,000 IDPs) are excluded. Benefit levels are unchanged. Savings are about GEL13 million per year or 22 percent of the 2001 budget allocation. Savings would allow inclusion of about 70,000 poor individuals in the poverty benefit program with assistance equal to GEL15 per month per beneficiary. (b) Targeting is done based on economic activity (families with employed/self-employed individuals and those having access to arable land) are excluded. Based on the IFRC regional vulnerability grouping, one half of IDPs in Tbilisi is excluded (45,000 IDPs), one half of those residing in Imereti (16,000 people) and one fifth of those in Samagrelo (30,000 people). Savings are about GELl5 million per year or one fourth of the 2001 budget allocation. (c) One third of IDPs in private accommodation (45,000 individuals or about 12,000 families), based on economic status are excluded. All the others continue to receive the benefit. Savings are GEL7.6 million per year (13 percent). Poverty Benefits Current Situation 7.65 The Benefit and Beneficiaries. There is in Georgia a limited program of targeted poverty assistance to single elderly households, elderly couples and orphans whose both parents have died. Except for the orphans, other eligible categories qualify only if they can prove that there is no-one to support them. Currently, the benefit is GEL22 per month for a single recipient households and GEL35 for other households. Relative to Georgian standards, these levels are generous. In November 2001, there were 49,574 households receiving the benefit, a 10 percent increase since the beginning of the year (see Table 7.1). 7.66 Resources and Financing. The benefit is financed from the state budget for most of the recipients. In 2001, the total central budget outlays were GEL8,6 million, or 0.1 percent of GDP.143 In 14 districts the benefit is financed from the local budgets in the amount of about GEL4 million or 0.6 percent of GDP. As of August 2000, the central budget has been current in financing the benefit. It also cleared 143 For 2001 the central budget financing of the poverty benefit was secured through the EU Food Security Program donation. 132 Georgia: Public Expenditure Review November 25, 2002 the arrears accumulated prior to that. The local budgets are not current and they owe to the recipients about GEL2 million-one half of the benefit due in 2001. 7.67 Administration. The program is administered by local social protection offices, based upon individual applications. Individual applicants should provide documents proving that they belong to the targeted category and that they have no-one to support them. The benefit is paid to eligible beneficiaries through the social insurance administration (GUSSIF). At the request of the EU Food Security Program that has provided funds for the benefit financing in 2001, the program is monitored quarterly by NGOs. Issues 7.68 Targeting. The poverty benefit is targeted on a categorical principle combined with one proxy that is relatively easy to check (no relatives to support eligible categories). Although found to be well targeted, the program is characterized by both errors of exclusion and inclusion. Table 7.7: The Poverty Benefit Recipients in Georgia: January-November 2001 Month Total number One-member Two and more Biological orphans of households households member households Households Orphans January 44675 37912 6065 698 1013 February 44370 37650 6051 669 980 March 43909 37316 5920 673 960 April 46449 39241 6456 752 1096 May 46225 39079 6441 705 1039 June 45738 38731 6301 706 1040 July 47853 40391 6688 774 1113 August 47766 40319 6662 785 1144 September 47761 40214 6662 785 1139 October 49574 41664 7060 850 1231 November 49574 41664 7060 850 1231 Source: Ministry of Health, Labor and Social Protection. 7.69 Both the recent Poverty Update for Georgia and the results of the quarterly monitoring of the program by NGOs confirm that single pensioners and pensioners' couples with no relatives to support them are among the most vulnerable categories in Georgia. In that respect, the program is well targeted. Still, the program is characterized by both errors of exclusion and inclusion. The World Bank Poverty Update finds other categories of the population that are equally vulnerable to poverty as those targeted by the poverty benefit, but excluded from the program because of budget constraints. Those include families with many children, households with no bread-winner, households with disabled members, etc. 7.70 Indeed, the choice which of equally vulnerable categories to include in the program whose resources are severely constrained is tough and it is often driven by perceptions of vulnerability and/or political priorities. Since the poverty benefit started being delivered regularly, there has been in Georgia a mounting pressure from other vulnerable groups to be included in the program. In that respect, the association of the blind has been particularly vocal and active. 7.71 In difference to error of exclusion, which is primarily caused by budget constraints, the error of inclusion observed through the quarterly NGO performed monitoring is mostly related to poor administration and fraud: some of the benefit recipients are found to actually have children or other relatives who are supporting them. 133 Georgia: Public Expenditure Review November 25, 2002 7.72 Equity. The benefit level is flat for each category of recipient households, thus assuming that they are equally vulnerable, which is mostly not the case: a single pensioner living on a labor pension of GELl 4 is in a different position than a single pensioner receiving a privileged pension of GEL45. Thus, the benefit does not meet a horizontal equity principle. By excluding other equally vulnerable categories, the program does not satisfy a vertical equity principle either. 7.73 Administration. Administration of the poverty benefit is generally found to be poor. Social protection offices tend to be placed in poor premises with no equipment. Benefit administrators are poorly paid, with low morale and motivation. Such conditions coupled with little or no control and auditing create natural environment for fraud. In order to narrow the scope for fraud and abuse of the system, the benefit award decision making could be entrusted to the local community councils with the social protection offices processing applications and preparing decisions. 7.74 Open-ended Character of the Benefit. The poverty benefit is open-ended in two ways. First, once awarded, the benefit does not require any reapplication, which deprives the system of self-targeting. This should be corrected and all recipients should be required to reapply once a year. Second, the benefit is awarded continuously, which is inconsistent with the hard budget constraints reality, puts pressure on the budget and may compromise the program. 7.75 The Benefit Level. The poverty benefit is higher than most of the other types of social transfers, including basic labor pension paid to 90 percent of pensioners. Normally, the situation should be the other way around. Recommendations 7.76 Given the poverty situation in Georgia, in the short term the Government should focus on regular payment of the poverty benefit and improvements in its administration. In the medium term the focus should be on gradual inclusion of other vulnerable categories. The targeting mechanism will need to be re-examined and improved as well (categorical versus other targeting techniques in the context of weak administration and high incentives for rent seeking). 7.77 Regular Payment of the Benefit. The regular payment of the poverty benefit should be among the top priorities for the Government. Also, in order to avoid inequities among recipients caused by the central budget-local budgets split of benefit financing and inability of the later to meet their obligations, the benefit financing should be entirely provided by the central budget. Relative to Georgian standards the current level of benefit is high. For the year 2002, provided that the financial situation allows, instead of increasing the benefit, the Government should consider extending the benefit to some of the other categories of vulnerable population. 7.78 Improvements in Administration. Changes in administration and in particular in employees behavior require substantial changes in incentives, including regular and better wages, better working conditions, clearly defined accountability, stronger supervision and monitoring, and strict disciplinary measures for misconduct. These challenges can only be addressed in the medium term. However some improvements, particularly those related to better office practices, work instructions, better supervision and monitoring, and training are feasible in the short run. Also, given scarce resources, it is extremely important to minimize errors of inclusion. This can be achieved by higher level of involvement of the local community in' the benefit award process. To that end, decision making on awarding the benefit could be entrusted to local communities councils, with the local social protection offices processing applications and preparing the decisions. Quarterly monitoring of the program by local NGOs should be continued. 134 Georgia: Public Expenditure Review November 25, 2002 7.79 Gradual Inclusion of Other Vulnerable Groups in the Program. The benefit should be gradually extended to other groups of vulnerable population. Accordingly, the inclusion should be planned and corresponding resources budgeted. For the year 2002, if the number of beneficiaries remains stable (about 50,000 households), approximately GEL14.3 million is needed to pay the benefit (0.2 percent of GDP). Increasing the budget allocation by GEL 7 million (0.1 percentage point of GDP) would allow inclusion of about 20,000-25,000 new beneficiaries. Doubling the allocation to 0.4 percent of GDP in two years would also double the number of beneficiaries (to about 100,000 beneficiaries) with significant increase in the poverty impact. The program extension is thus not expensive and efficiency gains from improvements in other programs (for instance, cash support to IDPs) would be sufficient to cover its initial extension in the year 2003. 7.80 Improvements in Targeting. Improving the targeting mechanism is likely to become more of an issue with the extension of the program and increase in resources spent. It is also highly relevant for restructuring of other programs, such as support to IDPs, support for electricity payment, etc. Designing an appropriate targeting mechanism is a complex task. The situation in Georgia is even more complicated because individual poverty correlates are not particularly strong (for instance, single pensioners are found to be more vulnerable to poverty than other groups, but their poverty rate is 42 percent), means or income testing is not appropriate because of informal nature of economic activities, and any complex targeting mechanism would be unmanageable for weak administration with high incentives for fraud. More analytical work on vulnerability in Georgia, as well as piloting of proposed mechanism, is recommended prior to its implementation. 135 Georgia: Public Expenditure Review November 25, 2002 136 Georia: Public Expenditure Review - ANNEXES November 25, 2002 ANNEXES  Georgia: Public Expenditure Review - ANNEXES November 25, 2002 Annex 1.1: Gross Domestic Product, Growth and Structure Real Growth Rate (%) 1997 1998 1999 2000 2001 Agriculture, Forestry, Fishing 3.9 -6.6 6.9 -12.2 5.6 Industry 4.3 -1.8 3.7 5.3 -4.2 Processing Agricultural Products by Household 1.5 -2.8 2.0 -2.0 2.6 Construction 51.5 35.2 -21.5 4.0 8.1 Trade 2.2 1.6 3.8 10.8 3.8 Hotels and Restaurants 28.1 6.4 9.7 8.3 34.4 Transport 42.0 51.7 5.4 13.9 3.1 Communications 3.9 36.5 7.0 8.3 14.9 Financial Intermediation 72.5 78.1 58.6 11.3 36.0 Real Estate Operations 5.5 1.5 -2.7 2.9 5.2 General Administration, Defense -1.5 9.2 -6.3 0.4 -1.7 Education 0.3 -3.0 -9.5 8.2 3.0 Health -3.8 -1.0 3.6 3.0 5.0 Other Services, Culture, Sport 3.7 10.0 -5.0 8.8 -2.9 Domestic Services -1.5 -0.2 -1.8 -35.9 -14.1 Adjustment for FISIM 35.0 166.5 33.0 14.0 59.7 Net Taxes 125.2 -13.5 14.5 14.0 15.2 Total GDP 10.6 2.9 3.0 2.0 4.5 Sector Structure as % GDP 1997 1998 1999 2000 2001 Agriculture, Forestry, Fishing 29.0 26.7 24.7 20.1 19.2 Industry 13.3 12.3 13.0 13.7 12.5 Processing Agricultural Products by Household 4.9 4.7 4.5 3.7 4.7 Construction 3.5 4.6 3.7 3.8 3.9 Trade 10.9 10.4 11.5 12.8 12.8 Hotels and Restaurants 1.9 1.9 2.4 2.4 3.1 Transport 6.2 8.9 9.4 11.9 11.0 Communications 1.4 2.0 2.5 2.5 3.3 Financial Intermediation 0.9 1.2 1.6 1.6 1.8 Real Estate Operations 8.9 8.4 7.5 7.1 6.8 General Administration, Defense 3.7 3.9 3.5 3.2 3.5 Education 2.7 2.8 3.3 3.5 3.8 Health 3.5 4.5 4.3 5.2 5.1 Other Services, Culture, Sport 3.3 3.3 3.0 3.1 2.8 Domestic Services 0.4 0.4 0.3 0.3 0.2 Adjustment for FISIM -0.4 -0.9 -0.9 -0.9 -1.3 Net Taxes 5.8 4.8 5.8 6.2 6.8 Total GDP 100.0 100.0 100.0 100.0 100.0 Memo: GDP at current market prices (in millions of Lari) 4638.7 5040.6 5666.0 5970.6 6505.2 GDP at current market prices (in millions of US dollars) 3573.4 3573.4 3573.4 3573.4 3573.4 Source: Georgia State Department Statistics. Georgia: Publc Expenditure Review - ANNEXES November 25, 2002 Annex 2.1: General Government Revenue (million lan) 1997 1998 1999 2000 2001 Total Revenue (exclude grants) 656.2 744.7 837.3 894.8 1032.6 Total Tax Revenue (include special state funds) 589.9 647.6 789.3 852.3 949.1 Taxes on Income 76.9 87.5 101.7 108.4 135.8 Taxes on Profits 39.3 51.5 54.8 80.2 66.5 VAT 195.5 223.1 244.5 289.3 355.8 From Tax Department 125.5 123.0 144.5 164.7 200.7 From Customs 70.0 100.1 100.0 124.5 155.1 Excises 53.3 38.7 118.5 90.2 89.4 From Tax Department 4.8 3.3 14.4 10.5 21.2 From Customs 48.5 42.0 104.1 79.7 68.1 Custom Duties 61.9 67.0 33.3 53.1 55.0 Other Taxes 60.5 62.3 86.0 86.4 90.9 Non-tax Revenue 66.3 97.1 48.0 42.5 83.5 Special State Funds Revenue 102.5 117.5 150.5 144.8 155.9 Road Fund ... ... 33.6 24.5 32.1 O/w. Gasoline Excise (preliminary) 0.0 0.0 12.0 4.3 11.3 Employment Fund 3.2 3.2 4.0 4.9 5.1 Social Security Fund ... ... 112.8 115.4 118.6 Memo: Nominal GDP 4638.7 5040.6 5666.0 5970.6 6505.9 Source: NBG. Annex 2.2: General Government Revenue (percent of GDP) 1997 1998 1999 2000 2001 Total Revenue (exclude grants) 14.1 14.8 14.8 15.0 15.9 Total Tax Revenue (include special state funds) 12.7 12.8 13.9 14.3 14.6 Taxes on Income 1.7 1.7 1.8 1.8 2.1 Taxes on Profits 0.8 1.0 1.0 1.3 1.0 VAT 4.2 4.4 4.3 4.8 5.5 From Tax Department 2.7 2.4 2.5 2.8 3.1 From Customs 1.5 2.0 1.8 2.1 2.4 Excises 1.1 0.8 2.1 1.5 1.4 From Tax Department 0.1 0.1 0.3 0.2 0.3 From Customs 1.0 0.8 1.8 1.3 1.0 Custom Duties 1.3 1.3 0.6 0.9 0.8 Other Taxes 1.3 1.2 1.5 1.4 1.4 Non-tax Revenue 1.4 1.9 0.8 0.7 1.3 Special State Funds Revenue 2.2 2.3 2.7 2.4 2.4 Road Fund ... ... 0.6 0.4 0.5 O/w. Gasoline Excise (preliminary) 0.0 0.0 0.2 0.1 0.2 Employment Fund 0.1 0.1 0.1 0.1 0.1 Social Security Fund ... ... 2.0 1.9 1.8 Source: NBG. Georgia: Public Expenditure Review - ANNEXES November 25, 2002 Annex 2.3: Turnover Limits and Actual Collection of VAT, LTI 2001 Turnover Limits in GEL Total VAT Collected (millions of GEL) From I to 3000 0.0 From 3001 to 6000 0.0 From 6001 to 10000 0.1 From 10001 to 15000 0.1 From 15001 to 20000 0.1 From 20001 to 25000 0.1 From 25001 to 35000 0.5 From 35001 to 50000 0.4 From 50001 to 100000 2.2 From 100001 to 200000 4.8 From 200001 to 500000 10.1 From 500001 to 1000000 11.7 More than 1000000 54.2 Total 84.3 Source: LTI. Georgia: Public Expenditure Review - ANNEXES November 25, 2002 Annex 2.4: Number of Registered Tax Payers by Types of Taxes As of February 1, 2001 Regions VAT Excise Income Revenue Land Tax Tax Tax Abkhazia Aut. Rep. 146 6 378 1489 Adjara Aut. Rep. 5517 34 3696 6119 672 Tbilisi 8288 163 22128 44079 6180 Kakheti 2260 105 2196 8286 10441 Imereti 2588 61 3786 10505 48940 Samegrelo 3227 16 3769 7153 28389 ShidaKartli 878 30 1463 4661 950 KvemoKartli 1868 115 5547 16866 7000 Guria 240 2 265 4217 12905 Samcxe-javaxeti 357 4 695 4412 9244 Mtsxeta-Mtianeti 780 20 699 3976 4615 Racha-Lenchxumi 99 4 88 449 10285 Large Payers Inspection 428 37 514 546 398 Total RepublicofGeorgia 26676 597 45224 112758 140019 Annex 2.5: Number of Active Tax Payers by Types of Taxes As of February 1, 2001 Regions VAT Excise Income Revenue Land Tax Tax Tax Abkhazia Aut. Rep. 146 5 161 1001 Adjara Aut. Rep. 1584 31 2359 3911 447 Tbilisi 7270 118 8437 17320 3979 Kakheti 1880 60 1229 5362 4316 Inereti 2473 59 2365 6378 48621 SamegElo 658 4 1306 3554 26652 ShidaKartli 858 11 882 3626 880 KvemoKartli 992 20 1529 7278 5875 Guria 220 1 215 740 281 Samcxe-javaxeti 282 1 487 1749 7506 Mtsxeta-Mtianeti 356 10 406 1667 4570 Racha-Lenchxumi 92 4 79 424 10275 Large Payers Inspection 428 37 514 546 398 Total Republic of Georgia 17239 361 19969 53556 113800 Georgia- Public Expenditure Review - ANNEXES November 25, 2002 Annex 2.6: Registration and Identification of Tax Payers As of February 1, 2001 Regions Registered Incl ding Number of Number of Number of Number of and have Physical Legal Active Payers Payers Liquidated Identification Person Person Payers Moved to Moved from Payers number another another Region Region Abkhazia Aut. Rep. 2891 2118 773 1380 317 114 Adjara Aut. Rep. 9894 6120 3774 4004 1 1 11 Tbilisi 77111 47216 29895 23331 3301 1673 835 Kakheti 19243 16126 3117 13890 49 8 1047 Imereti 57360 51932 5428 51310 0 0 15 Samegrelo 38127 33705 4422 30613 10 8 8 ShidaKartli 9289 6849 2440 5183 18 5 85 Kvemo Kartli 26011 19532 5481 15040 10 12 25 Guria 15215 13768 1447 13304 0 0 0 Samcxe-javaxeti 13020 11146 1873 9442 1 0 122 Mtsxeta-Mtianeti 8637 7204 1433 4726 21 35 15 Racha-Lenchxumi 11045 10756 289 10850 0 2 3 Large Payers Inspection 546 546 546 1 Total Republic of Georgia 288389 226472 60918 183619 3411 2061 2280 Georgia- Public Expenditure Review - ANNEXES November 25, 2002 Annex 2.7: Stock of Tax Arrears of the State Tax Department and State Funds, 1999 - 2000 14-Ja-9 I-Feb-99 1-ar9 I-Ap-9 I-May-9 .-u-9 iM-9 Ag99 I-cp-99 I-ct9 0 N,-91Me9 I-N-an-_~ l-b0 1~Mar> I-pr i-Myc I-Jun.00 I-Jul.00 I-AUg-0 00 145 4j48 428 Total 1/ 309 6 340 3 344 l 328 8 329 2 353 l 350 4 356 l 374 6 386 8 395 6 424 4 433 9 441 5 450 2 451 9 471 8 491 1 509 9 516 9 538 1 Pripaå19 208 2 208 3 190 4 186 6 198 9 191 7 194 4 197 201 0 205 4 220 3 227 1 222 5 222 1 218 0 220 7 228 8 230 4 235 9 24-6 Penalties 118 2 132 0 135 8 138 4 142 6 154 2 158 6 161 7 174 9 185 8 190 2 204 2 206 8 219 0 228 0 233 9 251 1 262 3 279 4 281 1 2175 SSF 503 554 62 l 531l 538 647 628 670 686 710 733 773 80 l 723 750 769 787 81 4 83 2 767 827 and EF ii i lPnnetoal 1 40 4 44 9 48 9 39 3 38 8 45 2 41 5 44 l 42 0 42 0 l 43 7 45 7 46 1 44 0 44 3 43 6 42 8 44 1 43 7_ 40 8 43 lpenialt,"e l 9 8 10 5 13 l 13 8 15 0 19 5 21 3 22 9 26 7 29 1 29 6 31 7 34 0 28 3 30 7 33 2 36 0 37 3 39 5 35 9 39 4 STD 259 3 284 8 282 0 275 7 275 4 284 27.95 289 l 305 9 315 8 322 3 347 l 353 8 369 3 375 2 375 0 393 1 409 7 426 6 440 2 455 4 totalil Pnneal 1150 9 163 3 159 4 151 1 147 8 153 7 150 2 150 3 57 7 159 0 161 7 174 6 i8] 0 178 5 177 8 174 4 177 9 184 7 186 7 195 0 197 3 Penalties 108 4 121 5 122 6 124 6 127 6 134 7 137 3 138 8 148 2 156 8 160 6 172 5 172 9 190 7 197 3 200 6 215 l 225 0 239 9 245 1 25 I VAT 1025 111 6 1164 jo014 i 001 1043 101 1 1058 1068 1124 1157 1230 1282 1312ý 131 9 1359 143 1 1505 158 j 164 2 1709 Pnepl569 603 621 519 487 503 491 539 50n 530 546 554 629 591i 574 58 1 597 626 642 680 654 UD to 249 247 269 229 195 174 165 185 172 184 23 7 237 312 187 184 204 210 240 253 164 164 6 to 74 1l 1 96 89 88 113 102 11 5 11 5 104 63 63 63 99 1 90 73 l 76 71 76 87 93 1 mora 246 245 256 201 204 215 223 240 213 242 246 254 2541 305 299 304 311t 31 5 313 429 427 Penalties 456 51 4 543 496 514 541 520 l 519 569 594 611l 676 653 720 76 78 34 79 939 962 1025 EXCMs 202 21 2 227 229 226 217 225 180 171 l 72 170 189 153 148 156 i 59 179 172 179 167 170 Prmcal 83 90 96 99 95 87 86 66 64 62 56 70 66 58 66 64 72 67 68 66 63 Upto i 3 1 7 3 2 28 25 24 23 19 14 1 1 27 27 28 17 23 22 27 26 35 25 23 6 to 08 08 08 os 07 07 12 10 1 1 12 09 09 09 09 06 06 04 04 04 03 0.3 more 62 64 56 64 63 57 5 1 38 38 39 20 34 29 32 37 37 41 38 30 38 36 Pentaties 119 1 123 130 129 132 130 139 114 to08 ti0 114 119 87 90 91 95 107 105 tio 101 107 Profit 465 499 522 53 5 539 574 567 591 623 627 638 678 694 705 716 705 745 784 814 845 879 tax iPrm- al1 265 278 287 288 296 l 310_ 294 298 31 5 307 308 329 334 32 1 315 308 307 326 320 33 8 345 Up to 86 94 103 95 104 l 69 56 58 83 46 167 167 167 57 57 120 1 121 137 1181 67 r, 6t0 54 65 65 71 70 89 74 77 49 49 31 31 31 36 32 29 l 23 24 25 51 60 mora 125 119 11 8 122 122 152 163 163 182 213 i09 13 1 136 228 226 159 163 165 114 220 212 Peate 200 22 l 23 5 247 244 263 273 293 309 320 33 1 349 360 385 400 398 438 458 494 507 534 l.cndtax 3l 364 30 35 376 382 389 380 436 1 43 5 468 5327 564 565 578 595 61 l 61 2 1 63 l 642 669 lPrmC.Pal 235 257 262 267 259 258 25 3 240 285 - 274 299 359 364 361 36 l 360 364 36 l 365 369 393 Up to 103 125 13 2 121 i H9 98 7;7 3 79 95 95 95 23 10 07 9 71 5 53 82 6 t 35 31l 389 32 30 50 60 59 46 43 45 45 45 57 63 48 5 1 69 74 66 401 More 96 101 92 114 109 110 120 109 147 153 159 21 9 224 282 188 204 22 l 22 l 234 250 261 Penlh,es _.85 106 1 108 11 7 11 7 125 _ 36_ 140 150 161 169 173 200 205 21 7 236 246 25 l 266 273 2$6 Othe 582 658 538 595 612 668 682 683 76 l 800 791 842 846 962 983 93 1 965 1025 1061 1107 1127 taxes Principal 357 406 328 338 342 379 378 360 41 4 417 408 43 5 418 455 463 43 1 439 468 472 49g8 498 upto 15 3 176 113 13 7 142 164 164 168 159 15 l 153 153 15 3 106 119 124 13 1 148 148 154 1 6 78 75 67 54 54 59 62 63 85 94 86 ¯¯86 69 91 85 82 78 86 78 65 61 Mitre 126 155 147 146 147 156 15 3 129 169 172 169 195 195 258 259 224 23 1 233 246 279 276 Peaus225 22 20 57 270 29 3 -432 347 383 382 407 2 0 520 51 25 55 7 589 609 629 SSF 488 538 559 513 520 63 3 61 6 655 666 681l 700 743 766 695 717 73 5 757 78 1 798 73 4 790l Prnc,pal 393 437 43 5 380 375 43 8 403 426 407 40 1 418 438 441l 423 425 l 41 B 41 1 424 41 8 90 414 Penialties 95 101 124 133 145 195 21 3 229 259 280 282 304 326 272 293 l 31 7 346 357 -380 343 31 EF l15 16 62 l17 1 8 14 i13 1 5 20 29 33 31l 35 28 3 3 3 3 30 33 354 Pr-cpa1 i11 12 5 5 13 13 14 13 15 s 12 18 20 la 21 17 a8 1 7 1 9 l PenMIecs 03 04 07 05 05 08 1 1 l 3 12 14 1 L 14 i15 l 3 16 l15 16 ndum tes Chianageof 307 345 192 196 435 408 465 650 77 860 1149 123 32 40 6 143 62 185 03 273 25 169 170 -109 -48 76 04 30 93 46 141 289 35 7 312 308 266 294 37 39 45 43 Penlt 1138 17 5 201 244 1 360 44 35 567 676 719 859 886 1008 1098 115 7 1329 1440 161 2 1628 19 Source StraeTaxDepiartment 1/ Till July mncludes tar: 5 8 millon subject to rellund Georpia: Public Expenditure Review - ANNEXES November 25, 2002 Annex 2.8: Stock of Tax Arrears, 1999-2001 (millions of lari) Total Tax Arrears to the Bud gets & Extrabudgetary Fund Principal Penalties & Fines Total Jan-99 208.2 132.0 340.3 Feb-99 208.3 135.8 344.1 Mar-99 190.4 138.4 328.8 Apr-99 186.6 142.6 329.2 May-99 198.9 154.2 353.1 Jun-99 191.7 158.6 350.4 Jul-99 194.4 161.7 356.1 Aug-99 199.7 174.9 374.6 Sep-99 201.0 185.8 386.8 Oct-99 205.4 190.2 395.6 Nov-99 220.3 204.2 424.4 Dec-99 227.1 206.8 433.9 Jan-00 229.5 218.9 448.4 Feb-00 230.8 228.0 458.8 Mar-00 226.8 223.9 450.7 Apr-00 225.7 252.0 477.7 May-00 236.6 262.3 498.9 Jun-00 240.4 279.5 519.9 Jul-00 241.9 281.2 523.1 Aug-00 250.6 295.6 546.2 Sep-00 251.2 303.9 555.1 Oct-99 266.3 323.1 589.4 Nov-00 274.6 329.5 604.1 Dec-00 302.3 361.0 663.3 Jan-01 292.2 374.6 666.8 Feb-01 290.8 385.1 675.9 Mar-01 286.5 383.9 670.4 Apr-01 294.2 393.9 688.1 May-01 327.8 444.1 771.9 Jun-01 236.4 465.9 702.3 Jul-01 355.1 477.8 832.9 Aug-01 371.6 502.6 874.2 Sep-01 364.7 546.4 911.1 Oct-01 368.0 564.9 932.9 Nov-01 395.8 596.9 992.7 Dec-01 394.0 630.3 1024.3 Source: State Tax Department, Arrears Department. Georgia: Stock of Tax Arrears, 1999-2001 700.0 - -400.0 200.0 100.0 s.- Principal -U-Penalties & Fines Georgia: Public Expenditure Review - ANNEXES November 25, 2002 Annex 3.1: Georgia - Central Budget - List of Credits Disbursement in 2001 Name of Credit USD IDA - 32630 - Judicial Reform Project 2615275 IDA - 32640 - Structural Reform Support Project 2003202 Kfw - Mikrokredit Bank of Georgia (5) 332708 EBRD - TBILUNIVERSAL BANK 100000 GRANT - 27308 - Institutional Strengthening 105756 IDA - PPF - Q2360 - Education System Strengthening 1001609 IDA - 28520 - HEALTH 3667773 GRANT - 29685 - Health Project II 6302 IDA - 29760 - Municipal Dev. & Decentral. 7449257 IDA - 30200 - Social Investment Fund 5937267 IDA - PPF - Q0790 - WATER 188578 GRANT - 27105 - Japan-Municipal Water Project 33862 IDA - 30400 - Cultural Heritage Credit 1323715 JBIC - POWER REHABILITATION PROJECT (GRG-Pl) 6892336 Kfw - ENERGYSECTOR (14.5) 1011733 GRANT - 25146 - Power & Gas Restructuring Project 18287 EBRD - Enguri Hydra Power Plant Rehabilitation Project 389555 Kfw - ENERGY SECTOR SUPPORT PROGRAM II (13) 11322 Kfw - ENERGY SECTOR SUPPORT PROGRAM 111 (40) 119669 IDA - 29580 - POWER REHABILITATION -10557 Kfw - Agriculture Project 1 (7) 12723 KfW - Land Cadaster and Registration System Establishment (30) 3591173 IDA - 29410 - Agriculture Development 1036594 IFAD - 543-GE - Rural Development Programme for Mountainous and Highland Areas 600000 IDA - 33470 - Agricultural Research, Extension and Training Project - ARET 403637 IDA - PPF - Q2120 - Preparation of the Forests Devel. Project 243332 IFAD - 450 GE - Agriculture Development 194345 GRANT - 25586 - Agriculture Development Pr. II 171803 IDA - PPF - Q1090 - Agricultural Development 2 135952 GRANT - 23752 - GEF2: GEORGIA - AGRICULTURAL (ARET) 125000 GRANT - 25144 - Japanese Grant for Forestry Development Project 53341 GRANT - S103 GE - Rural Development Programme for Mountainous and Highland Areas 36133 GRANT - 20446 - Conservation of Forest EC (GEF 1) 33080 IDA - 33570 - Roads Rehabilitation Project 6309784 EBRD - Trans-Caucasian Rail Link Project 9910741 Kuwait - Road Rehabilitation Project 1788016 IDA - 31290 - Restructuring of the Ministry of Transport Project - LIL 345450 IDA - PPF - Q0780 - Transport Sector Reform -310998 IDA - 31570 - Enterprise Rehabilitation Project - CERMA 2178922 GRANT - 23502- Capital Markets Development Program 304022 IDA - 31560 - Integrated Coastal Management 103563 GRANT - 21598 - Integrated Coastal Management Project 37741 GRANT - 27192 - IDF - Development of Public Procurement Procedures 14507 TOTAL 60516507 Source: Georgian Treasury. Georgia: Public Expenditure Review - ANNEXES November 25, 2002 Annex 3.2: Monthly Revenue Collections Georgia: Central Budget Monthly Tax Revenues (millions of Lari) 50 40- 35 30 - 2001 25 20 -2000 20- 15 - 10 Georgia: Local Budget Monthly Tax Revenues (millions of Lari) 35 - 30 - 20 -+ 2001 - --2000 15 - 0 0 > 30z Georiia Public Expenditure Review - ANNEXES November 25. 2002 Annex 4.1: Expenditures of Georgia Territorial Units, 2000 (Economic Classification) Percent of Tot l Expenditure 0L Territorial Units by m :: w 9 Regions < Tbilisi 153 4.6 0.1 1.7 3.5 2.1 00 0.0 03 8.3 238 11.0 19 27.4 100 Kakheti 50.2 128 07 78 1.7 0.8 01 00 1.0 45 163 00 0.7 32 100 Imereti 461 104 0.7 48 4.3 19 00 0 1 1 6 46 22.7 19 01 09 100 Samegrelo and Zemo 43 1 11 9 1 2 56 1.5 1.7 0.0 02 23 3.2 143 09 14 12.7 100 Svaneti Shida Kartli 535 130 08 55 0.6 0.4 00 01 0.7 2 1 222 00 00 11 100 Kvemo Karth 435 132 0.2 6.0 4.2 10 01 05 1 l 114 95 3 1 27 3 5 100 Guria 536 14.4 1.1 89 06 0.4 00 00 0.5 47 15 1 0.0 00 0.6 100 Samtskhe-Javakheti 560 18.1 12 4.6 25 07 00 00 1.2 47 5 5 22 13 20 100 Mtskheta-Mtianeti 290 85 05 24 3 1 20 00 00 0.8 32 46.5 00 02 3 8 100 Racha-Lechkhumi and 526 159 1 4 63 10 07 00 01 2.2 6 1 5.4 01 02 7.9 100 Kvemo Svane Total 29.7 8.3 0.6 3.5 3.2 1.7 0.0 0.1 0.9 6.7 21.1 6.4 1.5 16.3 100 Descriptive Statistics 560 181 14 8.9 43 21 0.1 05 23 11.4 465 110 27 274 Maximum Value 153 46 0.1 1.7 06 0.4 00 00 03 2 1 54 00 00 0.6 Average Vakue 443 12.3 08 5.4 23 1.2 00 0.1 12 5.3 181 1.9 0.9 63 Standard Deviation 129 3.8 04 22 1.4 07 0.0 01 07 27 120 34 09 83 Co-efficient of 0.29 0.31 0.56 0.41 0.61 0.57 0.80 1.42 0.57 0.52 0.66 1.76 1.07 1.32 Variation l Georzia Public Expemhiture Review --ANNEXES November 25, 2002 Annex 4.2: Disparity of Pe Capita Local Governmert Expendi ures, I197-2001 Descriptive Statistics <The boumläs, 1oor, de~omh~n and." any o ke~mma T URIKEY ~how on th.s m~ d. no m onh part d N heM woldBak A RM ENI\A Grou,~an gmen dn ähgol,t sos d-y ~rny, o r ny AZ ERB A IJAN N) s IRAQ 4r 4SAMC E   22913 GE Typ-: ER