Report No. 24756-RO Romania Building Institutions for Public Expenditure Management: Reforms, Efficiency and Equity A Public Expenditure and Institutions Review August 2002 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank CURRENCY EQUIVALENTS Currency Unit = Romanian Leu (ROL) US$ 1 = ROL 33,070 (as of August 9, 2002) ACRONYMS AND ABBREVIATIONS CGE Computational General Equilibrium DHIFs District Health Insurance Funds EAS European Accounting Standards EU European Union GDP Gross Domestic Product GPs General Practioners IHS Integrated Household Survey MER Ministry of Education and Research MIG Minimum Income Guarantee MLSS Ministry of Labor and Social Solidarity MoHF Ministry of Health and Family MoJ Ministry of Justice MPA Ministry of Public Administration MPF Ministry of Public Finance MTSAB Means Tested Social Assistance Benefit NATO North Atlantic Treaty Organization NGO Non-Government Organization NHIF National Health Insurance Fund NHP National House of Pensions NIM National Institute of Magistrates OECD Organization for Economic Co-operation and Development OSI Open Society Institute PAYG Pay As You Go PSAL Private Sector Adjustment Loan SCM Superior Council of the Magistracy SOEs State-Owned Enterprises Fiscal Year January 1 to December 31 Vice President: Johannes Linn Country Director: Andrew Vorkink Sector Director Cheryl Gray Sector Leader: Kyle Peters Team Leader: Carlos Cavalcanti Ill Contents CURRENCY EQUIVALENTS ................................................................................................................. H ACRONYMS AND ABBREVIATIONS ..................................................................................................11 ACKNOW LEDGM ENTS ...................................................................................................................... VII EXECUTIVE SUM M ARY ...................................................................................................................... IX CHAPTER 1. ECONOMIC REFORMS AND PUBLIC EXPENDITURE MANAGEMENT...........1 Introduction..... .......................................................1 Recent Economic Developments .................................. ......................2 Renewed Actions on Stabilization and Structural Reform ..........................7 The Distributional Implications of the Government's Economic Reform Program ..... ........8 Fiscal Policy for Sustainable Growth ...... ...........................................16 Conclusions..... ......................................................21 CHAPTER 2: STRENGTHENING THE INSTITUTIONS FOR BUDGET MANAGEMENT.......23 Introduction... ........................................................23 Completing Budget Management Reforms ......................................23 Improving Budget Formulation .................................................24 Recommendations on budget preparation...................................................30 Strengthening Budget Execution .............................................31 Conclusions and Recommendations ..............................................36 Recommendations.......................................................41 CHAPTER 3: IMPLEMENTING FISCAL DECENTRALIZATION REFORMS.......................45 Introduction....................................................... ....... 45 Structure and Expenditure Responsibilities of Local Governments in Romania..............................46 Budget Formulation ........ .............................. . ..........50 Recommendation on Local Government Budget Design .....................................54 Budget Execution.......................................................54 Recommendations on Budget Execution ........................................65 Budget Compliance and Evaluation ...........................................66 Policy Recommendations for Budget Compliance and Evaluation ........................68 Conclusions and Recommendations ...........................................69 CHAPTER 4: REFORMING THE BUDGET PROCESS IN THE JUDICIARY.........................71 Introduction... ........................................................71 Current Structure for Court Management in Romania...............................72 Budget Formulation in the Judiciary...........................................75 Budget Execution in the Judiciary.............................................76 Budget Compliance and Evaluation in the Judiciary................................79 Developing the Courts' Administrative Capabilities................................. ..........81 Conclusions and Recommendations ...................8......8.. ...............88 iv CHAPTER 5. IMPROVING RESOURCE ALLOCATION IN THE HEALTH SECTOR.........89 Introduction ...........................................................................................................................................89 O verview ...............................................................................................................................................90 Budget D esign .. .................................................................................................................... 94 Budget Execution.............................................................97 Budget Evaluation .................................................................................101 Conclusions and Recommendations ..........................................107 CHAPTER 6: DELIVERING SOCIAL ASSISTANCE AT THE LOCAL GOVERNMENT LEVEL .....................................................................................................................................................111 Introduction.... ............................ ........................................................111 Poverty and Social Protection in Context ... ..... .............................. ....................................111 The Design of the Minimum Income Guarantee Program .............................115 Implementing the Minimum Income Guarantee Program .... ............................................125 Conclusions and recommendations ...... ..... .................................... ................................130 CHAPTER 7. CONTINUING REFORMS IN THE PENSION SYSTEM.......................................131 Introduction.................................... .................................................................... 131 Structural Imbalances in the Romanian Pension System....................................................................132 The System's Design: The New Pension Law and Subsequent Amendments ...................................135 Implementing the Reforms in the Public Pension System..................................................................136 Policy Recommendations for the Second Phase of Pension Reform..................................................140 Reducing discretion in defining the parameters of the pension system..............................................140 Improving policy-making capacity and the institutional framework... ...............................142 ANNEX 1: DATA AND STRUCTURE OF THE COMPUTATIONAL GENERAL EQUILIBRIUM MODEL .......................................................................................................................145 ANNEX 2: AN ANALYTICAL FRAMEWORK FOR ESTIMATING THE SUSTAINABILITY OF FISCAL POLICY IN ROMANIA................................................................148 ANNEX 3: ASSUMPTIONS AND PROJECTIONS FOR THE ACTUARIAL MODEL FOR THE PENSIONS SYSTEM...........................................................................................................150 STATISCAL APPENDIX.......................................................................................................................152 Tables Table 1.1: Selected Economic Indicators, 1997-2001 ......................................3 Table 1.2: Termoelectrica's Losses, 2000-2001 ................................... 4 Table 1.3: Losses in Production and Distribution of Natural Gas...................................................5 Table 1.4: -Enterprise Arrears to the Government's Consolidated Budget, 1995-2001...................6 Table 1.5: Selected Central European Countries: Statutory and Effective Tax Rates.....................6 Table 1.6: Romania - Real GDP Growth and Key Macroeconomic Balances, 2000-2002.............9 Table 1.7: Simulations from the CGE Model - Sector Reallocation .................................11 Table 1.8: Industrial Output, December 1998 - July 2001 ............................................................14 Table 1.9: Simulations from the Model - Macroeconomic Indicators..........................................16 Table 1.10: Primary Surpluses Required for Stabilizing the Debt.....................................17 Table 1.11 Consolidated Government Revenues and Expenditures, 1995-2001 ............................18 V Table 1.12: Consolidated Government Expenditures by Functions, 1995-2000 . ..............20 Table 2.1: Breakdown of Consolidated Government Expenditure, 2000 and 2002......................27 Table 2.2: Consolidated Public Investment by Main Sources of Funding, 1995 - 2002..............29 Table 2.3: Approval of Technical/Economic Documentation of Investment Projects..................29 Table 3.1: Structure and Population of Local Governments .....................47 Table 3.2: Local Governments Expenditure Responsibilities................ ...............48 Table 3.3: Structure of Local Government Expenditures, 1991-2002 .......... ..........49 Table 3.4: Share of Local Government Expenditures by Type of Government, 2002...................49 Table 3.5: Processing of Equalization Grants by the County (judet) Councils ...... .........53 Table 3.6: Local Public Expenditures Relative to GDP and Consolidated Government Expenditures, 1996 - 2001 .................................... .....55 Table 3.7: Change in Revenues, 1998-2002, by Type of Local Government ...............56 Table 3.8: Distribution of the Population by Type of Local Government According to the Change in Tax Revenues between 1998 and 2000.........................56 Table 3.9: Distribution of Expenditures by Function According to the Change in Tax Revenues between 1998 and 2000 ........................ .. ..........57 Table 3.10: Municipality of lasi - Impact of Earmarks and Mandates ......... ............59 Table 3.11: Funding of Local Capital Expenditures, 1996-2002 ...........................61 Table 3.12: Structure of Capital Expenditures, 1996-2002 ............................61 Table 3.13: Structure of State Education Expenditure, 1998-2000 ........ ...............63 Table 3.14: Changes in the Real Expenditures on Education at the County Level, 1998/2000...........65 Table 3.15: Local Taxes and Fees versus Basic Service Fees..........................68 Table 4.1: Selected Indicators of the Judiciary, 1998-2002............ ................71 Table 4.2: State Budget Allocations to the Judiciary, 2001-2002 .............. .........77 Table 5.1: Health Services Utilization and Resources, Romania and Selected European Countries .....................................................91 Table 5.2: Trends in Health-Care Expenditure in Romania, 1990-2000............... ...............92 Table 5.3; Main Public Sources of Health Financing in Romania, 1990-2000 ...... .........92 Table 5.4: Male Life Expectancy at Birth, Selected European Countries ....... ...........93 Table 5.5: Infant Mortality Rate, Selected European Countries, 1990-2000 ......................93 Table 5.6: Standardized Disease Rates of the Circulatory System, Selected European Countries .....................................................94 Table 5.7: Expenditures for Programs, Ministry of Health and Family, billion ROL....................95 Table 5.8: Expenditure for Health from Public Sources, 2001-2002 .....................96 Table 5.9: Health Insurance Fund Expenditures - Comparison of Framework Contract Provisions, Revised Budget Provisions and Actual Expenditures, 1998-2001.............99 Table 5.10: Functional Allocation of Current Expenditure by Level of Care, 1999-2001.................101 Table 5.11: Use of Health Services by Area of Residence, 1995-2000 .............. ......102 Table 5.12: Choice of Health Care Provider by Insurance Status, 1998-2000 .........................105 Table 5.13: Composition of Private Health Expenditures by Deciles, 1995-2000.............................106 Table 6.1: Percentage of the Population Receiving at Least One Cash Social Transfer..................112 Table 6.2: Outreach of Selected Social Programs in 1999......... .......................112 Table 6.3: Benefit Incidence in 1999, Individuals Ranked by their Household Consumption Per Equivalent Adult, including Pensions but excluding Other Transfers .... .....113 Table 6.4: Cost-Benefit Ratio for Selected Social Cash Transfers..................... .........113 Table 6.5: Ex-ante Analysis of Poverty Reduction under the MIG Scheme....................................114 Table 6.6: Poverty Rates and Poverty Reduction Rates by Different Types of Households............114 Table 6.7: The Dynamics of the Real MTSA/MIG Threshold......................... 117 Table 6.8: Financial Needs for Implementing the MIG - Estimations of the Ministry of Finance versus those Based on the Integrated Household Survey Data ..... ......119 Table 6.9: Estimated Poverty Rates and Expected MIG Eligibility by Sub-Regions ......................120 vi Table 6.10: Means-Tested Residential Heating Subsidies under the Mixed System, January 1 through March 31, 2002............................. ...........124 Table 6.11: Estimation of Additional Costs for a Mixed System of MIG Heating Allowances and Heating Subsidies as per the Former Legislation............ ............. 125 Table 6.12: Expected MIG Eligibility .........................................126 Table 6.13: Number of Full-time Public Employees by Size and Legal Status of the Local - A dm inistrative U nit............................................ ............................................ 127 Table 7.1: Financial Performance of the Pension System, 1995-2001 .............. .....132 Table 7.2: Pension Payroll Taxes............................................................. 135 Table 7.3: Payroll Taxes (Percentage of Gross Wages) ...................................137 Table 7.4: Romania's Aging Population ......................................137 Annex Table 3.1: Macroeconomic Assumptions....................... ...... . ............................ 150 Annex Table 3.2: Compliance Assumptions ....................................... 150 Annex Table 3.3: Projected Changes in Life Expectancies .............................. 150 Annex Table 3.4: Projected Fertility Rates...................... ........................................ 150 Box Box 2.1 Timetable for Budget Preparation ....................................26 Figures Figure 1.1: Real Exchange Rate, 1999-2001 ................................................................................. 15 Figure 5.1: Incidence of Tuberculosis, Selected Central European Countries, 1970-2000.............90 Figure 5.2: Concentration Curves for the Use of Health Services - Primary and Specialized Outpatient Versus Hospital Services .......................................................103 Figure 5.3: Concentration Curves for the Use of Family Doctors, Specialized Outpatient and H ospital Services.....................................................................................................104 Figure 7.1: Pension System - Contributors and Beneficiaries, 1990-2001.......................................133 Figure 7.2: The Evolution of Wages and Old-aged Pension Benefits, 1990-2001............................134 Figure 7.3: Financial Performance of the Public Pillar under Different Scenarios, 2002-2050........140 vii ACKNOWLEDGMENTS The report is based on the research findings and interviews conducted by the World Bank team working in Romania in the fall of 2001 and winter of 2002. The team wishes to thank the Romanian Government for the excellent cooperation received from its senior government officials in the Ministry of Public Finance, the Ministry of Public Administration, the Ministry of Education and Research, the Ministry of Labor and Social Solidarity, the Ministry of Justice, the Ministry of Health and Family, the National Health Insurance Fund, and the National House of Pensions. The team also wishes to thank members of Parliament, officials from the Court of Audit, representatives of local governments and Romanian academics for productive discussions. Carlos Cavalcanti is the principal author of the report, drawing from background papers and contributions prepared by a team composed by Mmes. Amalia Fugaru, Clare Romanik, Cornelia Tesliuc, Esperanza Lasagabaster, Juliana Pigey, Kimberly Nead, Luana Pop, and Yvonne Sin, and Messrs. Catalin Pauna, Francis Conway, Hendrix Koppen, Mihai Nedelcu, Peter Broke, Robert Watt, Silviu Radulescu, Theodore Fetter, Timothy Johnston, Wojtek Maliszewski. The report benefited from constructive criticism from the peer reviewers - Messrs. Bill Dorotinsky and Vinaya Swaroop, as well as comments from Mmes. Irina Kichigina, Maureen Lewis and Randi Ryterman, and from Messrs. Jean-Jacques Dethier, Bernard Funck, Nick Haazen, Ron Hood, Michael Mertaugh, Kyle Peters, and Roy Southworth. The team gratefully acknowledges the assistance of several World Bank colleagues. Ms. Raluca Banioti ably organized the team's visits to Romania. Ms. Cornelia Giurescu provided superb research assistance. Ms. Anita Correa competently assisted in processing the report.  EXECUTIVE SUMMARY 1. This Public Expenditure and Institutions Review was undertaken at the request of the Romanian authorities at a critical juncture for public expenditure management in the country. After three years of economic decline, the Romanian economy began growing again in 2000, reaching a real GDP growth rate of 5.3 percent in 2001. While renewed growth engendered an overall improvement in the outlook for Romania's economy, the way in which it was being financed led to growing macroeconomic imbalances, including a rising current account deficit. To sustain the economic recovery and keep the country on a sustainable growth path, the Government defined a bold economic reform strategy. This strategy aimed at redefining the boundaries of the state by moving forward on banking and enterprise privatization; containing sources of quasi-fiscal deficit, especially in the energy sector; and reducing central government expenditures through better public expenditure management and further fiscal decentralization. These reforms were designed to allow the growth momentum to be sustained by renewed investment and export growth, minimizing the risks of a further build up in macroeconomic imbalances, and opening room to address the country's most pressing concern: rising poverty. 2. This review finds that the challenges for public expenditure management under the Government's economic reform strategy are two-fold: > There is the need to ensure the sustainability of the current economic recovery by containing quasi-fiscal deficits, and keeping the government's consolidated budget deficit on target. The simple reason is that deficits absorb private sector savings (both foreign and domestic) that could otherwise fund productive investments. Lower investment reduces the potential for sustained growth and real increases in living standards. Actions to contain quasi-fiscal deficits include accelerating bank and enterprise privatization, keeping scheduled increases in domestic energy tariffs, eliminating tax exemptions, and better enforcement of tax collection from delinquent state owned enterprises. > There is a need to strengthen budget management to allow a fundamental shift in the manner in which a significant share of public resources is allocated in Romania. At present, large quasi-fiscal deficits (including the deficits of state owned enterprises, tax arrears and tax evasion) imply that public resources are allocated in a non-programmed, discretionary manner, usually in response to political pressures during the budget year. Measures to contain quasi-fiscal deficits need to be accompanied therefore by budget reforms that make the budget more comprehensive and ensure that actual expenditures more closely match the budget approved by Parliament. 3. Actions to strengthen budget management are the focus of this report. These actions include improving the Treasury accounting information to curb the practice of exceeding budget ceilings by accumulating payment arrears; accelerating the elimination of extra-budgetary funds; and subjecting foreign financed expenditures to greater budgetary scrutiny. These actions also include ensuring that the budgeting of local government activities and key public services is x more transparent and that the fimding of these activities is commensurate with actual needs. Measures to achieve these twin objectives range from defining more stable expenditure assignments for local government to improving the allocation criteria for funding of public services. Ensuring the Sustainability of the Current Economic Recovery 4. Romania's economic performance since mid-2000 has been remarkable, with twelve- month industrial output growth averaging around 10 percent since end-2001, and inflation declining to just over 20 percent, down from around 45 percent just two years ago. Faster growth has been matched by a rising current account deficit, however, reaching around 6 percent of GDP in end-2001. The sustainability of the country's economic performance hinges therefore on opening room for private sector investment by limiting the size of the general government deficit and containing sources of quasi-fiscal deficit, such as the losses of the state owned energy utility companies. 5. Containing losses of state owned utility companies is particularly important because these losses reflect a fundamental imbalance in the energy market. Energy tariffs are still set at below market clearing levels, and the tolerance of non-payment of energy bills, especially by state owned enterprises, provides an additional price subsidy. This imbalance in the energy market manifests itself in three principal ways: (i) a worsening of the balance sheets of the energy companies; (ii) an increase in the Government's fiscal deficit as a result both of the non-payment of tax obligations and the calling of debt guarantees on energy enterprise borrowings; and (iii) the deterioration of the external current account deficit, driven in part by the increase in imports of energy inputs (oil and gas). 6. Energy enterprise losses also crowd out private sector investment, reducing the country's growth potential. Since emerging market economies, such as Romania, have limited access to foreign savings, the manner in which these savings are allocated between the financing of the public sector deficit (including state owned enterprises) and private sector investment is important in determining the long term growth prospects of the economy. Any increase in deficit from the public sector implies that the private sector must invest less, reducing productive investments and rate of real GDP growth. 7. Achieving the objective of sustaining the economic recovery will require actions on three fronts: > Lowering the general government's deficit by half of one percent of GDP, bringing the overall deficit to 3.0 percent of GDP in 2002, down from 3.5 percent in 2001. This can be achieved through a combination of tighter tax regulations and targeted increases in support to the poorest groups in the population. The tightening of tax regulations includes the elimination of tax incentives and exemptions, as well as restrictions on rescheduling tax obligations of state owned enterprises. Targeted support to the poor includes (i) the new Minimum Income Guarantee program, involving 0.4 percent of GDP in transfers to mitigate the impact of higher energy tariffs; (ii) an increase in child allowances; and (iii) an increase in the replacement xi rate for the average pensions, compensating pre-1998 retirees for the under indexing of their benefits in the past. > Reducing the losses of the energy sector by 1.7 percent of GDP, which would conservatively improve the savings/investment balance of public sector enterprises by at least 0.75 percent of GDP. This can be achieved by reducing losses in the generation and distribution of electricity and district heating, as well as by eliminating the implicit subsidy provided by domestic producers of natural gas. Also important is better enforcing the collections of energy bills. > Making visible progress on bank and enterprise privatization, reducing the public sector involvement in the economy and improving the business climate. Bank privatization has progressed, with three of the six largest state banks already sold (BRD, Bank Post, and Banca Agricola), and one liquidated (Bancorex). The two remaining banks (BCR and CEC) are slated for privatization in 2002 and 2003, respectively. The agenda for enterprise privatization is more extensive, and progress slower. While many small enterprises were privatized in the early years of the transition, including all the shops and service outlets, almost all large industrial enterprises remain under state ownership. These large enterprises include thestrategic enterprises (Rigies Autonomes) and public utilities, which represent around 20 percent of employment in the economy. The Government needs to accelerate the privatization of large industrial enterprises, to take the first steps towards privatization of some the Rfgies Autonomes under the framework of the Bank's PSAL II operation and to greatly reduce the financial obligations that result from the state ownership of these enterprises. This should be accompanied by measures to improve the business environment, and strengthen public administration, thereby establishing a stable environment for economic development. Strengthening budget management 8. The reform agenda for sustaining the current economic recovery will be left unfinished if it is not accompanied by a strengthening of budget management. The reasons are two-fold. More effective budget management is important to meet lower fiscal deficit targets without compromising quality in the delivery of public services. More effective budget management will also underpin the shift away from allocating public resources through quasi-fiscal channels. 9. The issue of meeting fiscal deficit targets without compromising the delivery of public services arises because, in some instances, fiscal deficit targets have been met by allowing payment arrears to accumulate. There are reports that local authorities, which are not presently subject to ex ante control of budget commitments, have accumulated arrears in the delivery of centrally mandated social benefits. There are also reports that substantial arrears have arisen in payments to hospital suppliers. Finally, there are reports of payment arrears at the central government level when price increases that were neither anticipated in framework contracts, nor captured in a budget rectification, give rise to higher invoices (e.g. for contracted utility services). These reports of payment arrears are a concern because they reflect a disconnect between the budget approved by Parliament and the budget executed by the Government. xii Indeed, expenditure priorities end up being defined during the budget execution, rather than at the time of the budget approval. Changes during the execution of the budget mean that public service providers cannot expect to receive the entire funding defined in the budget. This reduces accountability in the delivery of public services, and undermines the fiscal discipline that strong central controls were designed to achieve in the first place. 10. The actions required by the Romanian authorities to strengthen budget management fall under two broadly defined headings: actions aimed at improving budget design and budget execution; and actions to ensure that budget allocations are commensurate to the needs of the population. 11. Actions aimed at improving budget design and execution include the following: > Improving Treasury accounting information, curbing the practice of exceeding budget ceilings by accumulating payment arrears. Actions in the immediate short run include establishing a central monitoring system for payables and issuing monthly reports on payment arrears. This would increase accountability and minimize the risk of accumulating even larger arrears. This should also be a first step toward developing a modem Treasury system, providing timely information on payment commitments and orders, and reducing the costs of collecting taxes and executing the budget. > Subjecting foreign financed public investment to full budgetary scrutiny. At present, substantial capital expenditures are financed through external credits contracted directly by line ministries. Although the borrowing is covered by Government guarantees, the corresponding investments are not included in the budget. These investments account for one third of total public investment, exceeding investment expenditures in either central or local government budgets. This reliance on guarantees is a non-transparent form of government borrowing, circumventing accountability to Parliament, as well as the scrutiny which applies to foreign funded investments that are included in the state budget such as those funded by multilateral institutions. Also, since there is no priority ranking for projects receiving external funding, available counterpart funding is allocated on a first-come, first-served basis. The scale of this borrowing exacerbates the problem of counterpart fund availability, causing delays in the implementation of ongoing projects and raising the costs of undertaking externally financed projects. > Increasing accountability in the management of extra-budgetary funds. The management of extra-budgetary funds in Romania tends towards two extremes. At one extreme are extra-budgetary funds that come under very little scrutiny regarding the way revenue is raised and expenditure is allocated. For example the Energy Development Fund raises revenues by taxing consumers of electricity, but allocates the funds not always following clear prioritization rules and procedures. This electricity tax adds another burden on electricity consumers at a time when the country is already raising tariffs significantly to reach cost recovery levels. A simple solution would be to eliminate the tax on electricity consumers and fold the xiii investments into the budget, ensuring greater scrutiny over these investments at the time of the budget approval. At the other extreme are the social security funds (pensions, health insurance, unemployment insurance), which rely primarily on payroll taxes, and face excessive interference from the Ministry of Public Finance in defining expenditure ceilings. This happens because the Pension Fund has statutory payment obligations that exceed its revenues and, as a result, has been running large deficits. The Ministry of Public Finance has therefore been imposing spending limits on the other two funds, so that their surpluses' help off set the consolidated budget deficit. The consequence, however, has been to reduce the accountability in the management of these social security funds, as fund managers and service providers under these funds (e.g., hospital and other health care units) accumulate payment arrears to lay claim in advance on these surpluses. 12. Many key programs with an important impact over the well being of the population are presently underfunded. Actions aimed at reducing quasi-fiscal deficits need to translate into immediately tangible benefits for the majority of the population, otherwise progress on reducing quasi-fiscal deficits could be reversed. The following expenditure programs requiring additional attention stand out in the case of Romania: > There is a need to place the health and pension funds on a firmer financial foundation. At present, the very high contribution rates (adding to around 60 percent of gross wages) have encouraged evasion and reduced the share of formal sector employment to around 60 percent of the labor force. There is a need to reduce these payroll taxes and broaden the contribution base for these programs. Further options to improve financing would include better enforcement of mandatory contributions and reductions in non-essential expenditures such as non- insurance related benefits (e.g., two-year maternity leaves), and funding them from general tax revenues. Other options would include merging the collection agencies for the health, pension and unemployment funds, and better executing the obligations of delinquent state owned enterprises. > There is a need to increase funding for primary health care, and for health promotion and prevention. While the health sector reforms instituted in 1997 were designed to ensure more funding for these two activities, in practice funding has been squeezed by hospital overruns, the growing costs of pharmaceutical drugs, and the sharp increase of expenditures on debt service. This situation needs to be reversed or Romania's health indicators will continue to lag other countries in Central and Eastern Europe. Increased funding for primary health care will require greater control over hospital expenditures, and changes in the payment method for family doctors, especially in rural areas. If left unresolved, these problems could threaten the credibility of the health sector reforms. Increased funding for health promotion and prevention will require leadership from the Ministry of Health and Family, as well as greater control over guarantees extended to hospitals and other health care providers. There is also a need to support traditional health education Surplus funds are deposited in the Treasury account, earning annual interest rates of around 10 percent, which is well below the rate of inflation. xiv activities with strengthened regulations (including for tobacco, alcohol, and accident prevention). The latter will require in turn decisive Government involvement. For instance, tobacco control legislation has been developed, but has not yet passed Parliament. There is the need to reconsider the present system of financing of education, moving to a system that does not attach all the weight to the local governments' revenue generating capacity. At present, the financing of local governments comes primarily from two sources: the property taxes (which are kept 100% at the local level); and income taxes. Income taxes are shared between levels of government, with 36.5 percent going to local councils, 25 percent going to county councils, and 38.5 percent going to the central government. There is also an equalization grant designed to address concerns over rising funding disparities across local governments. The funding available through these equalization grants is computed through a formula approved every year with the state budget. The formula, however, allocates additional funding according to the local government's current tax revenues and cannot, therefore, possibly favor the poorest regions. > There is a need to improve the allocation criteria for earmarked transfers to local governments under the minimum income guarantee (MIG) program. This should aim at ensuring an equitable allocation among, as well as within, counties (judets), and at improving the incentives and ability of the local councils to provide the benefit under the program. This might mean reconsidering the requirement that local councils complement earmarked transfers received which is particularly burdensome for poorer localities with a larger share of eligible families. It might also require strengthening local government administrative capacity and developing monitoring and evaluation capacities. While the track record of means- tested social assistance programs indicates that the MIG has the potential to reduce the poverty rate in 2002 by 18 percent, and the severity of poverty by 52 percent, better monitoring and evaluation would increase accountability at all levels, and enable mid-course corrections in the MIG program, in case needed. > There is a broader need to define more stable local government expenditure assignments, clarifying the distinction between local government's own and delegated functions. At present, local government expenditure assignments are revised every year with the state budget law, creating continual uncertainty for financial management and budget practices at the local government level. These repeated changes in expenditure assignments lead to disruption in the delivery of public services. Under clearly defined expenditure assignments, own functions would be those where local governments have broad authority and discretion to determine the service policy and methodology and to decide on service quantity, quality and cost. Local government's own functions should not include national programs, such as education and social assistance. Delegated functions are those that the national government may assign to a local government to perform in a manner that ensures equal access and quality to the entire population, regardless of place of residence. Funding for delegated functions should be clearly earmarked, and defined on the basis of needs rather than through revenue sharing agreements. xv Revenue sharing leads to large discrepancies in the delivery of services, accentuating existing regional disparities. 13. Focusing on the funding and delivery of these public services is important because quasi- fiscal deficits have emerged in part because the Government has been unable to reorganize these public services. State owned enterprises continue to be the main employers in small and medium sized towns, in many instances providing some of these public services. This creates strong pressures to subsidize these enterprises, keeping them operating, even if at a loss. In yielding to these pressures, low energy tariffs, and tax and energy arrears have become the main sources of soft budget constraint, with inevitable implications for the overall budget deficit. Lack of a hard budget constraint weakens cost control and delays improvements in operational efficiency of these state owned enterprises, leading to stagnant productivity and low growth. The dependence on state owned enterprises for maintaining employment leaves governments with very little choice but to accommodate the demands by these enterprises for additional resources, since otherwise either the delivery of key services will suffer (e.g., energy, transport), or a massive number of workers will need to be laid off. Tackling quasi-fiscal deficits requires therefore taking parallel action to improve public service delivery through central and local government agencies, reducing this way the political pressure to subsidize state owned enterprises. 14. Report outline. The rest of this PEIR develops the themes identified in this executive summary. It begins by reviewing Romania's economic reform program, providing the rationale for containing quasi-fiscal deficits and for moving ahead with budget reforms. The subsequent chapters examine budget reforms at the central and local government level, as well as budget reforms in health, social assistance, social security, and the judiciary. The chapters are organized as follows: > Chapter one provides the background against which one can better understand the broader institutional agenda behind the Government's economic reform program. It begins with a brief review of recent economic developments, highlighting some sources of fiscal and quasi-fiscal pressures that are responsible for growing macroeconomic imbalances. It examines next the distributional implication of the Government's reform program, especially as it relates to increasing energy tariffs and reducing subsidies to state owned enterprises. From this section emerges the message that aggregate fiscal discipline, while paramount, is alone not enough to ensure the sustainability of the current economic recovery. Macroeconomic stabilization efforts not accompanied by structural reforms that allow the reallocation of resources to more productive activities will hinder the economic recovery. The economic recovery is in turn needed both to ensure fiscal sustainability, and to mitigate the distributional implications of economic reforms. > Chapter two begins the discussion of the strengthening of the budget process by reviewing recent budget reforms at the central government level. This review highlights how a gap between budget formulation and budget execution has emerged because of strong (and sometimes ad hoc) controls exercised over cash expenditures at the execution stage of the budget. This in turn leads to arrears and reduction in the quality of services. Reforms of the budget process should focus xvi therefore on establishing a modem treasury system in Romania, providing better and more timely accounting information that are critical for the budget execution and evaluation stages. A modem treasury system should help ensure that resources are deployed efficiently, reducing the tendency to accumulate payment arrears, a perennial problem in Romania. Moving away from ad hoc controls is critical to* reaping the benefits of fiscal decentralization. Chapter three therefore assesses the fiscal decentralization process, examining the main issues at every stage of the budget process at the local government level. The findings of this chapter lead to two important conclusions. There is a need for greater stability in expenditure assignmeits and for a clearer distinction between local government's own and delegated expenditure responsibilities include. Funding for delegated expenditures including education and social assistance should be earmarked and defined on the basis of needs instead of through revenue sharing arrangements, since the latter only contribute toward widening existing regional disparities. Chapter four examines another instance where there is a need to move away from strong central controls: the budget process in the judiciary. Greater independence of the judiciary in the budget process is needed to ensure accountability in managing budget resources, and to ensure court performance. The Romanian court system has come under increase demand since the beginning of the transition, leading to considerable delays in resolution of cases, and raising concerns about corruption. The chapter examines reforms therefore aimed at assisting the Romanian authorities manage the scarce resources available to the judiciary system, while increasing the independence of the system, so that it no longer considered an appendix of the executive branch of government. > Chapter five examines the process of resource allocation in the health sector. While the health sector is no longer, strictly speaking, included in the budget process, in practice the allocation of resources within the sector is still very much determined by the central government. The government defines expenditures ceilings for the National Health Insurance Fund, rather than allowing expenditures to be determined by the overall level of contributions by the insured. This interference means in practice that National Health Insurance Fund does not have the ability to develop stable agreements with health care providers. Also, the central government still owns most of the health care facilities, and as a result, has a stake.in how resources are allocated within the system. Not surprisingly, most of the resources still go to hospitals, with inadequate funding available for primary health care and public health prevention programs. Chapter six discusses the challenges involved in implementing the government's newly designed minimum income guarantee program. This program is primarily funded by the central government (75 percent) and entirely delivered by local governments. The program has the potential to reduce significantly the number of poor and the severity of their poverty. Achieving this, however, requires ensuring xvii an equitable allocation among, as well as within, local governments. It also requires improving the incentives and ability of the local governments to provide the benefits that are available under the program. > Chapter seven examines the problems and options for reform in the public pension system. It begins by reviewing where the system now stands, including the system's perennial deficits, the rise in the number of beneficiaries, and the decline in the number of contributors. This provides the background for the needed changes in the management of this public pension system, such as establishing financially viable replacement rates and indexation rules, as well as institutional reforms aimed at reducing the excessive discretion with which the system is currently managed.  CHAPTER 1. ECONOMIC REFORMS AND PUBLIC EXPENDITURE MANAGEMENT Introduction 1.1 Romania began the implementation of its 2002 state budget against a backdrop of renewed economic growth in 2001, fueled in part by lower domestic energy prices, and financed by foreign and domestic borrowing. While renewed growth had created an overall improvement in the outlook for Romania's economy, the way in which it was financed has led to growing macroeconomic imbalances, including a rising current account deficit. In response, the Government's strategy for 2002 was to redefine the boundaries of the state: moving forward on banking and enterprise privatization; containing sources of quasi-fiscal deficit, especially in the energy sector; and reducing central government expenditures through better public expenditure management and further fiscal decentralization. These measures aimed at allowing the growth momentum to be sustained by renewed investment and export growth, minimizing the risks of a further build up in macroeconomic imbalances, and opening room to address the country's most pressing concern: rising poverty. 1.2 This Public Expenditure and Institutions Review (PEIR) examines the institutional agenda behind the implementation of the Government's economic reform strategy. This agenda focuses primarily on improving the three stages of the budget process: (i) policy formulation, forecasting and budget preparation; (ii) budget execution and the operations of the Treasury in the Ministry of Public Finance; and (iii) the institutions for budget compliance, namely internal and external audits and budget evaluation. Improving these three stages of the budget process aims at (i) providing the government with the necessary tools to maintain aggregate fiscal discipline over a longer period of time; (ii) facilitating strategic prioritization across policies, programs and projects for allocative efficiency and equity; and (iii) encouraging better use of resources, achieving outcomes and producing outputs at the lowest possible cost. These three goals are complementary and mutually reinforcing, since judicious budget formulation, with an emphasis on matching priorities to available resources, should facilitate budget implementation. Budget implementation that is transparent and does not rely on ad hoc revision of allocations should facilitate efficiency in budget spending and make assessing budget compliance unproblematic. Accurate information at the budget compliance stage should, in turn, provide those responsible for budget formulation with a better understanding of which programs work and do not work, closing the budget process cycle. 1.3 An improved budget framework should also help the Romanian authorities break the inertia in the current expenditure patterns, shifting resources from old programs to new programs that reflect a more productive use of resources, and doing so by changing the incentives faced by those at the budget execution stage. The link between breaking expenditure inertia and moving beyond cost controls is the following. Breaking expenditure inertia to allow greater expenditure prioritization requires a stronger connection between budget allocations and policy decisions, and this connection is built at the budget formulation stage. Creating the incentives to move beyond cost controls, focusing also on the quality in the delivery of public services, requires making those responsible for executing the budget more accountable for expenditure outcomes. 2 Chapter 1: Economic Reforms and Public Expenditure Management The latter in turn depends on improving budget execution and developing the institutions for budget compliance. 1.4 Improving the government's ability to shift resources across programs should also allow Romania to meet aggregate fiscal targets without compromising the delivery of key public services. Ensuring adequate delivery of public services is important because narrowly focusing on cost controls and postponing capital expenditures may delay improvements in operational efficiency, leading to stagnant productivity, and leaving the government with very little choice but to accommodate the demands for more spending. Otherwise the delivery of services will suffer, compromising aggregate fiscal discipline, the very goal that was trying to be achieved by imposing cost controls. 1.5 The main rationale for focusing on the budget process is that containing quasi-fiscal deficits implies a fundamental shift in the manner in which significant shares of public resources are allocated. At present, quasi-fiscal deficits (including tax arrears and tax evasion) mean that public resources are allocated in an ad hoc, discretionary manner, usually in response to political pressures during the budget year. Measures to contain quasi-fiscal deficits need to be accompanied by budget reforms that strengthen the budget process, allowing public resources to be allocated in an orderly and transparent manner for the general public and policymakers alike. It is to assist the Government in making this shift toward allocating public resources that, after the discussion of the Government's economic reform program in this chapter, this report focuses on strengthening the budget process throughout the public sector. 1.6 This chapter is organized as follows. It begins with a brief review of recent economic developments, highlighting some sources of fiscal and quasi-fiscal pressures that are responsible for growing macroeconomic imbalances. Next it examines the distributional implication of the Government's reform program, especially as it relates to increasing energy tariffs and reducing subsidies to state owned enterprises. From this section emerges the message that aggregate fiscal discipline, while important, is not, by itself, enough to ensure the sustainability of the current economic recovery. Macroeconomic stabilization efforts which are not accompanied by structural reforms that allow the reallocation of resources to more productive activities, will hinder the economic recovery. The recovery in turn is needed both to ensure fiscal sustainability, and to mitigate the distributional implications of economic reforms. The chapter closes with a presentation of what lies ahead, identifying an institutional agenda that would help the government maintain aggregate fiscal discipline and improve the allocation of public resources to achieve equity and efficiency. Recent Economic Developments 1.7 Renewed Growth. After three years of negative real GDP growth rates, the Romanian economy began growing again in 2000, and reached a 5.3 percent real GDP growth rate in 2001 (Table 1.1). The economic recovery has been driven by strong growth in exports, primarily the re-export of processed inputs from EU countries, and a reflation of the economy in the runup to the election and its aftermath. The growth in exports to the EU reflects progress in economic reforms over the last few years, including setting the real exchange rate at a competitive level; achieving some degree of fiscal consolidation, while lowering the tax burden on exports; and Chapter 1: Economic Reforms and Public Expenditure Management 3 initiating bank and enterprise privatization. The reflation of the economy resulted largely from renewed access to international financial markets, which in turn has allowed the Government to delay raising energy tariffs and tolerate increases in arrears to the budget and state owned utility enterprises. Renewed access to international capital markets thus financed the widening of the current account deficit by over one billion dollars from 3.7 percent of GDP in 2000 to 6.2 percent of GDP in 2001. 1.8 The factors driving this current account deterioration are largely found in the trade 2 balance, as imports rose at almost double the rate of growth of exports. This is due in part to higher imports of petroleum, gas and other mineral products reflecting both higher import prices and delays in raising domestic prices.3 While foreign borrowing has allowed Romania to finance this widening current account deficit, and even increase foreign reserves held at the central bank, escalating external indebtedness illustrates the difficulties involved in imposing financial discipline on the operations of state owned enterprises. Almost one third of Romania's US$ 1 billion in new borrowing in 2001 was done directly by two energy sector enterprises (Termoelectrica and Petrom), primarily to purchase imported fuel. Table 1.1: Selected Economic Indicators, 1997-2001 1997 1998 1999 2000 2001 Real GDP growth -6.1 -4.8 -1.2 1.8 5.3 Trade balance (percent of GDP) -5.6 -6.3 -3.6 -4.6 -7.5 CPI (end of period; percent) 151.4 40.6 54.8 40.7 30.3 Unemployment rate (end of period; percent) 8.9 10.4 11.8 10.5 8.6 Consolidated public sector deficit (percent of GDP) -5.3 -5.4 -3.6 -4.0 -3.3 Primary balance of the public sector (percent of GDP) -1.4 -0.7 2.3 0.8 0.6 Exchange rate (Lei/US$, end of period) 7,985 10,951 18,250 25,926 31,597 Interest rates (Interbank rates, end of period) 138.8 105.0 88.7 60.1 39.9 Source: Romanian authorities. World Bank staff calculations. 1.9 Losses of the Energy Enterprises. The need to borrow to finance the losses of energy enterprises can be traced to two factors: the insufficient adjustment in energy tariffs and the worsening collection rates. For instance, in mid-2001, before the Government began raising energy tariffs heating and thermal electricity tariffs were, respectively, 45 and 15 percent below cost recovery levels. Collection rates were also low, averaging around 80 percent for Termoelectrica,4 and around 75 percent for the gas distribution enterprises. These losses in turn led to fiscal pressures when the Government was called to guarantee and honor foreign 2 During this period there was an improvement in the service account and current transfers. The deficit in service shrank to less than a third of the 2000 figure, mostly on account of lower interest payments, while the surplus in current transfers widened by $33 million, or almost 11%. 3 The net import of energy inputs increased by around $300 million in the first 6 months of 2001 relative to the same period last year. This as mentioned above reflects both increased volumes of imports and higher prices. For instance, the volume of crude oil imports increased 23 percent when compared to the first 6 months of last year, while prices have been higher by 8.2 percent. Meanwhile, the volume of imports of natural gas have been lower this year than last year, but this has been more than compensated by higher import prices. 4 Termoelectrica accounts for 50 percent of the country's electricity supply, and 40 percent of the district heating supply. 4 Chapter 1: Economic Reforms and Public Expenditure Management borrowing by these enterprises.5 It also led to fiscal pressures because by signaling tolerance of the non-payment of energy bills, the Government triggered a chain of non-payment throughout the economy that ultimately translated into large tax and social contribution arrears.6 1.10 The magnitude and the factors behind the losses of energy enterprises are better understood by examining the accounts of Termoelectrica under different tariff and bills collection scenarios (Table 1.2). Termoelectrica projected losses for 2001 were conservatively estimated at 1.5 percent of GDP, up from 1.3 percent of GDP in 2000. Most of the increase in losses is attributed to delays in adjusting electricity tariffs. In the absence of the adjustment in tariffs over the summer (57 percent increase), Termoelectrica 's losses would have over $300 million higher. Note that these projections assume constant collection rates. Therefore, a rise in energy tariffs combined with a sustained effort to increase collections is expected to significantly reduce Termoelectrica's losses, helping contain this source of quasi-fiscal deficit. Table 1.2: Termoelectrica's Losses, 2000-2001 2000 2001 2002 W/out W/ Tariff Projections Tariff increase increase Heating Cost recovery tariff ($/Gcal) 22.0 20.0 20.0 20.0 Current tariffs ($/Gcal year average) 10.5 10.4 15.0 19.0 Termoelectrica's heating output (Million Gcal) 22.4 22.4 22.4 22.4 Termoelectrica's losses from heating ($ million) 258 220 132 22 Thermal Electricity Cost recovery tariff ($/Mwh) 34.7 39.0 39.0 39.0 Current tariffs ($/Mwh year average) 34.3 28.5 32.0 38.0 Termoelectrica's electricity output (Million Mwh) 28.6 28.6 28.6 28.6 Termoelectrica's losses from electricity ($ million) 11 294 201 32 Termoelectrica's Joint Operations Total operational losses (% of GDP) 0.7 1.3 0.8 0.1 Collection rate (%) 82 83 83 90 Total losses including non-collection (% of GDP) 1.3 1.9 1.5 0.2 Source: Romanian Electricity Regulatory Agency (ANRE). World Bank and IMF staff estimates. 1.11 Similar factors drive the losses in domestic gas production and distribution (Table 1.3). While losses in gas production and distribution have been a perennial problem in Romania, the s For instance, in April 2001, Termoelectrica issued a EUR 150 million (US$ 132 million) bond guaranteed by the Government. Also, the 2002 budget includes $100 million in external borrowing guarantees for the two gas distribution companies ($50 million each) to cover the difference between these companies' revenues and their operating costs (including purchase of imported fuel). 6 For instance, since Termoelectrica is not being paid by Electrica, it is unable to pay its suppliers (PETROM, gas distribution companies, the coal mines). These suppliers in turn don't pay their tax bills, claiming that it is ultimately the Government's responsibility to ensure payment from Termoelectrica. Assuming that all of Termoelectrica's non-payment to suppliers translates into tax and social contributions arrears, the accumulated arrears to the budget and the social security system in the first three months of this year would have reached $120 million. ' Termoelectrica accounts for 50 percent of the country's electricity supply, and 40 percent of the district heating supply. Chapter 1: Economic Reforms and Public Expenditure Management 5 situation deteriorated further with the increase in the price of imported natural gas in 2000 and 2001 and delays in adjusting domestic prices. For instance, not even the 90 percent tariff increase instituted in September 2001 was enough to bring down the size of the implicit subsidy (as a share of GDP) to the 2000 level. This situation is set to improve this year with the combination of lower international prices, higher domestic tariffs and increased collection. The implicit subsidy however will continue to be an issue, as long as domestic tariffs are set below world market levels. Table 1.3: Losses in Production and Distribution of Natural Gas 1999 2000 2001 2001 w/ 2002 w/out tariff projections tariff increase increase Import price ($/tcm) 65 120 130 139 119 Domestic wellhead price ($/tmc) 35 35 29 40 43 Domestic output (million ton) 14 14 14 14 14 Losses from low prices ($ million) 420 1,190 1,414 1,386 1,064 Collection rate (% of billings) 76 87 81 81 90 Losses from non-collection ($ million) 117.6 63.7 77.14 106.4 60.2 Total loss/implicit subsidy ($ million) 538 1,254 1,491 1,492 1,124 Total loss/implicit subsidy (% of GDP) 1.5 3.4 4.0 4.0 2.8 Source: Romanian Gas Regulatory Agency (ANRGN). World Bank and IMF staff estimates. 1.12 The non-payments to energy producers also constitute an important source of inter- enterprise arrears, imposing an additional burden on the energy sector. Energy intensive sectors, predominantly state owned, run arrears on payments to the electricity distribution company, which in turn defers payments to the electricity producer. The electricity producing company delays payments to its suppliers: oil and gas producers and coal mines, both because it is not paid by its customers and because electricity prices are kept artificially low. The suppliers in turn delay tax payments to the government, arguing that it is the government who fails to ensure timely payments from the state owned electricity producer. In addition, energy sector enterprises borrow abroad to cover their fuel import bill with public guarantees, linking the implicit subsidization through the energy sector to current account imbalances. 1.13 Tax Arrears and Tax Exemptions. The chain of arrears initiated in the energy sector does not account for all the quasi-fiscal pressures arising from the high levels of arrears on taxes and social contribution. Table 1.4 provides estimates of arrears to the consolidated budget (tax and social contribution arrears) over the 1995 to 2001 period, breaking it down between the public and the private sector. The picture that emerges is one where the public sector accounts for around three-quarters of all arrears. This is not surprising, since tax arrears by large state owned enterprises provide a convenient way of subsidizing enterprises that should otherwise be competing in the marketplace. Indeed, the largest non-payers of tax obligations and social security contributions are state owned enterprises (e.g., steel producers, energy producers, coal mines, the railways, oil refineries) producing tradable goods. In the private sector, poor revenue performance can also be credited to high statutory tax rates that provide strong incentives for other, less privileged enterprises to evade taxes and migrate to the informal sector. 6 Chapter 1: Economic Reforms and Public Expenditure Management Table 1.4: Enterprise Arrears to the Government's Consolidated Budget, 1995 - 2001 1995 1996 1997 1998 1999 2000 2001 Total arrears (% of GDP) 5.11 6.89 6.62 8.08 8.29 8.09 8.17 Private Sector (% of Total arrears) 10.2% 19.2% 17.8% 19.4% 24.0% 22.1% 21.0% State Sector (% of Total arrears) 89.8% 80.8% 82.2% 80.6% 76.0% 77.9% 79.0% Source: Romanian authorities. World Bank staff calculations. 1.14 The extent of these two problems (arrears and evasion) can be measured by comparing statutory tax rates and effective tax rates, as reported by Shaffer and Turley (2000). They compared statutory an effective tax rates for several countries in Central and Eastern Europe (Tables 1.5). Their results point to very low tax revenues from the VAT in Romania, reflecting a number of exemptions and a significant degree of tax avoidance, including tax arrears. For instance, VAT revenues in the Baltic countries (Estonia, Latvia and Lithuania) were two times higher as a percentage of GDP, despite having the same statutory rates as Romania. Also, the revenues from the corporate income tax were significantly lower in Romanian relative to other transition countries, again reflecting the exemptions and lax tax collection. Table 1.5: Selected Central European Countries: Statutory and Effective Tax Rates Country Statutory tax rates and revenues collected Effective tax rates VAT Corporate Income VAT Corporate Income Tax (CIT) Tax (CIT) Statutory VAT/GDP Statutory CIT/GDP E/S' NTY2 E/S' NTY2 tax rate tax rate Albania 12.5 4.6 30.0 0.7 0.42 7.0 NA 0.8 Bulgaria 22.0 6.2 36.0 6.4 0.34 5.7 0.31 6.3 Croatia ... ... 35.0 2.0 - - NA 2.0 Estonia 18.0 10.4 26.0 1.9 0.40 6.6 0.23 3.0 Czech 22.0 7.1 39.0 3.4 0.68 11.4 0.22 2.6 Republic FYR ... ... 15.0 0.7 - - NA 1.6 Macedonia Hungary 25.0 7.9 18.0 1.9 0.40 6.6 0.26 3.7 Latvia 18.0 8.8 25.0 2.4 0.58 9.6 0.23 3.4 Lithuania 18.0 8.7 29.0 1.6 0.57 9.6 0.13 2.0 Poland 22.0 8.3 38.0 3.1 0.46 7.7 0.20 2.9 Romania 18.0 4.7 38.0 4.3 0.31 5.1 0.21 4.0 Slovak 23.0 8.4 40.0 3.7 0.45 7.5 0.22 3.3 Republic Slovenia ... ... 25.0 1.2 - - 0.15 1.6 1) E/S = Effective tax rate over Statutory.tax rate. 2) NTY = Normalized Tax Yield. Source: Shaffer and Turley (2001). 1.15 The Government's new economic program reflects concerns with revenue losses due to widespread tax exemptions and arrears, and includes measures to contain these problems. The Government has recently submitted to Parliament a revised corporate income tax law that eliminates corporate tax incentives to small and medium enterprises, and proposed amendments to the VAT legislation aimed at raising rates on exempt and zero-rated products. It has also imposed new limits on the forgiveness of tax arrears by large state owned enterprises, and is Chapter 1: Economic Reforms and Public Expenditure Management 7 working to define new statutory rates for payroll taxes, reducing the incentive for workers and less privileged enterprises, usually in the private sector, to evade taxes and migrate to the informal sector. Renewed Actions on Stabilization and Structural Reform 1.16 To sustain the growth momentum and contain the rising current account deficit, the Government has outlined an economic reform program aimed at addressing these emerging macroeconomic imbalances. The more immediate measures include those already discussed above: raising energy tariffs,8 improving collections of utility companies, making visible progress in state bank and enterprise privatization, and changing the tax legislation to reduce exemptions and increase compliance. Fiscal policy will support the overall program by meeting aggregate fiscal targets, containing sources of quasi-fiscal deficit, and facilitating the implementation of structural reforms. The latter includes more funding to cushion the impact on the population of higher energy tariffs and of layoffs associated with privatization. It also includes reducing Government interference in setting utility tariffs, delegating greater responsibility to local governments, and increasing private sector provision of public services. 1.17 More specifically, the goals of the Government's economic reform program are to keep the external current account deficit within bounds that can be externally financed (around 5.5 percent of GDP); to continue the downward trend in inflation, reaching just over 20 percent by end-2002; and to sustain the economic recovery, maintaining real GDP growth of about 4.5 percent. To meet these goals, the economic reform program envisages: > Lowering the general government deficit by half of one percent of GDP, bringing the overall deficit to 3.0 percent of GDP in 2002. This will be achieved through a combination of tighter tax regulations and lower spending on subsidies to enterprises, as well as targeted increases in support to the poorest groups in the population. The tightening of tax regulations includes the elimination of tax incentives and exemptions, as well as restrictions on rescheduling tax obligations of state owned enterprises. Lower subsidies to enterprises include reduced budget transfers to the railways and to agro-processing enterprises. Targeted support to the poor includes (i) the new Minimum Income Guarantee program, involving 0.4 percent of GDP in transfers to mitigate the impact of higher energy tariffs; (ii) an increase in child allowances; and (iii) an increase in the replacement rate for the average pensions, compensating pre-1998 retirees for the under indexing of their benefits in the past. > Reducing the losses of the energy sector by 1.7 percent of GDP, which would conservatively improve the savings/investment balance of public sector enterprises by at least 0.75 percent of GDP. This will be achieved by reducing losses in the 8 Electricity tariffs, following the 15 percent increase over the summer, are expected to be adjusted by 3.6 percent every month through March 2002, with further 16 percent adjustments scheduled for April 2002. Heating tariffs are also to be adjusted in January and April to keep the price fixed in dollar terms at US 15.40. This comes on top of the 57 percent increased approved in July, and the projected increase to US $ 20 by the Spring. Finally, natural gas tariffs, after the 90 percent increase approved in August, are expected to be adjusted to keep their dollar price in dollar terms constant. 8 Chapter 1: Economic Reforms and Public Expenditure Management generation and distribution of electricity and district heating, as well as eliminating the implicit subsidy provided by domestic producers of natural gas. The first step has been an across-the-board increase in energy tariffs. Electricity tariffs, following the 15 percent increase over the summer, have been adjusted by 3.6 percent every month since October, with a further 14 percent adjustment on April 1, 2002. Heating tariffs were also adjusted in January and April, keeping the price fixed in dollar terms at US 15.40. This increase comes on top of the 57 percent increased approved in July 2001, and the projected increase to US$ 20 by the early summer 2002. Finally, natural gas tariffs, after the 90 percent increase approved in August, have been adjusted to keep their dollar price in dollar terms constant. > Making visible progress on bank and enterprise privatization, reducing the public sector involvement in the economy and improving the business climate. Bank privatization is further ahead, with three of the six largest state banks already sold (BRD, Bank Post, and Banca Agricola), and one liquidated (Bancorex). The two remaining banks (BCR and CEC) are slated for privatization at end-2002 and 2003, respectively. In the meantime, the National Bank of Romania is taking steps to strengthen the regulatory framework and supervision of the financial system by aligning prudential norms and accounting standards with best practices in OECD countries. The agenda for enterprise privatization is more extensive, and is progressing more slowly. While many small enterprises were privatized in the early years of the transition, including all the shops and service outlets, almost all large industrial enterprises remain under state ownership. These large enterprises include the strategic enterprises (Rigies Autonomes) and public utilities, which represent around 20 percent of employment in the economy. The Government is accelerating the privatization of large industrial enterprises, and is taking the first steps towards privatization of some the Rggies Autonomes under the framework of the Bank's PSAL II operation. This is necessary to reduce the financial obligations that result from the state ownership of these enterprises. This should be accompanied by measures to improve the business environment and strengthen public administration, thereby establishing a stable environment for economic development. 1.18 The implementation of this economic reform program will require great political determination, as well as careful consideration of distributional consequences. While the reforms are essential for sustaining the economic recovery, fiscal discipline needs to be complemented with structural reforms that allow resources to flow to productive investments and growing industries. Once the distributional implications of the Government's economic reform program are considered, the chapter turns to the other aspects of the design of fiscal policy - maintaining the primary fiscal balance, breaking expenditure inertia, and improving service delivery. After that is completed, the chapter closes with a summary of the main findings and an outline of the rest of the report. The Distributional Implications of the Government's Economic Reform Program 1.19 This section examines the distributional implications of the Government's economic reform program. It pays particular attention to the distributional implications of dealing with the Chapter 1: Economic Reforms and Public Expenditure Management 9 sources of quasi-fiscal deficit, since reducing these deficits is at the core of the Government's economic program. This reflects two important factors. There is a general belief that the Ministry of Public Finance has a reasonably effective system of expenditure controls and has been able to keep total spending within the agreed overall budget ceiling on a cash basis. In the recent past sudden increases in expenditures have been associated primarily with quasi-fiscal pressures that have unhinged the country's fragile fiscal balance. The focus on quasi-fiscal deficits should avoid this from happening again. Also, closing these sources of quasi-fiscal deficit is critical for creating room for private sector development. Preferential access to public resources through tax arrears, arrears to energy utility companies and lower priced energy inputs gives state owned enterprises an unfair advantage over private enterprises, crowding out private sector investment. 1.20 The last point is important because channeling foreign savings to productive investments is what distinguishes successful from less successful transition economies. The reason is as follows. Foreign transfers can finance either a shortfall in public savings (i.e., a fiscal or a quasi- fiscal deficit), or a shortfall in private savings, where the latter includes the financing private sector investment above and beyond the savings of the private sector. The analytical implications of this choice are summarized in an open economy accounting identity, which is adapted slightly to reflect the issue at stake for Romania, namely the quasi-fiscal deficit in the energy sector: (X-M)=(SP-IP)+(STE-ITE)+(T-G) where X-M is the balance between exports and imports of goods and non-factor prices, Sp-Ip is the balance between savings and investment in the private sector; ST-ITE is the balance between savings and investment in Termoelectrica, the state's main power utility company; and T-G the balance between taxes and expenditures in the public sector (the public sector deficit). 1.21 Table 1.6 presents the figures for this identity for the 2000 to 2002 period. The figures for 2002 are projections, understand why containing the losses in Termoelectrica is important to ensure the sustainability of the current economic recovery. Table 1.6: Romania: Real GDP Growth and Key Macroeconomic Balances, 2000-2002 Real GDP X-M Sp-Ip STE- ITE T-G Growth (%) (% of GDP) (% of GDP) (% of GDP) (% of GDP) 2000 1.6 -3.3 2.0 -1.3 -4.0 2001 5.3 -6.9 1.9 -1.5 -3.5 2002' 4.5 -5.6 2.4 -0.2 -3.0 1) Forecast. Source: National Bank of Romania. 1.22 As can be seen above, Romania's scope for growth is directly related to the amount of savings it can tap from abroad (X-M), and how these savings are allocated between the financing of the public sector deficit and private sector investment. Any increase in deficit from the public sector (including public utilities) implies that the private sector can invest less. This compromises the overall growth of the economy, since there is a direct link between real GDP growth and productive investments. The real GDP growth rate in turn has a positive feedback 10 Chapter 1: Economic Reforms and Public Expenditure Management effect on tax revenues and the overall deficit of the public sector. Consider now Romania's growth performance in 2000 and 2001, and the projections for 2002. In 2000, the relatively low real GDP growth reflects low levels of foreign savings, used mainly to finance the deficit in the public sector and in the state owned electricity and heating utility company (Termoelectrica). The private sector savings and investment balance was actually positive in 2000, meaning that the private savings exceeded investments. In 2001, the private sector was able to increase investment thanks primarily to Romania's greater access to foreign savings and, to a lesser extent, to efforts to contain the public sector deficit. This private sector investment helped propel the economy, leading to a 5.2 percent real GDP growth. The forecast for 2002 is that sustained efforts to reduce the quasi-fiscal deficit in the energy sector, and a lower public sector deficit, will open room for further private sector investment. This is critical to sustaining the current economic recovery, despite the slowdown in the world economy. 1.23 The other important distributional implication of dealing with quasi-fiscal pressures (one that extends beyond the impact of higher energy utility tariffs on the population) is that it affects the profitability of entire industries. This in turn has implications for employment in certain regions, especially among politically vocal groups. These distributional consequences have in the past overwhelmed economic reforms, and explain in part the pattern of stop-and-go reforms that characterized Romania's economic transition in the 1990s. 1.24 The following paragraphs attempt to measure some of these distributional implications of the economic reform program, especially as they relate to indirect subsidies to the enterprise sector, such as higher energy prices and lower tax arrears and exemptions. The main message that emerges from this exercise is that the distributional consequences are quite significant in terms of growth and employment in certain sectors. The only alternative for pushing ahead with these reforms is to ensure that the economy is able to reallocate resources across sectors, allowing resources to move toward sectors with greater growth potential, mitigating macroeconomic imbalances and sustaining the growth momentum. Sustaining the growth momentum in turn has a positive feedback effect, as it is essential for keeping the fiscal balance and minimizing the overall impact on employment and living standards. Understanding these links provides the necessary background for the main theme of the report: the need to shift budget resources toward new and more productive activities within the public sector. This in turn explains the focus on improving the budget process to facilitate this shift in resources. 1.25 To measure the distributional consequences of dealing with these sources of quasi-fiscal pressure, we present the result of a simple simulation model that computes the impact of raising energy tariffs and reducing subsidies to loss making industries. The findings indicate that energy intensive sectors/industries stand to lose the most from economic reform program. This includes many state owned enterprises accounting for a significant share of industrial employment. The findings also confirm that the benefits for the overall economy are significant, especially the higher growth rates experienced by new export oriented sectors/industries. This should both help sustain the current economic recovery and contribute toward the reduction of the current account deficit. Labor and capital market flexibility is critical, however, to reap the benefits of these reforms. Chapter 1: Economic Reforms and Public Expenditure Management 11 1.26 Simulations. Table 1.7 presents the results of the simulation of the impact of raising energy tariffs and removing subsidies on a range of sectors.9 The simulations illustrate two possible policy scenarios: a 20 percent increase in energy tariffs (scenario 1), and the combination of energy tariff increases and the removal of subsidies to loss making enterprises (scenario 2). These simulations have in turn two settings: in the first one assumes that both factors of production (capital and labor) are mobile across sectors and wages are flexible to equilibrate supply and demand for production factors. In the second setting we assume that capital is activity-specific. In this case the price paid for capital services is activity specific and adjusts to equilibrate the (fixed) supply and demand for capital in a given sector. In this setting we also assume that wages are fixed at the current level and the demand for labor at this wage determines employment. Table. 1.7: Simulations from the CGE Model: Sector Reallocation (GDP at factor prices, % change compared to base case equilibrium) Flexible wages, mobile factors Fixed wages, activity specific capital Increase in Increase in Increase in Increase in electricity electric energy electricity electric energy tariffs, plus tariffs energy tariffs, energy tariffs reduction in subsidies plus reduction in subsidies Agriculture -0.9% -1.2% -0.5% -0.7% Coal Mining -2.3% -7.2% -2.3% -7.1% Petroleum Extraction 0.0% 0.0% 0.0% 0.0% Natural Gas 0.0% 0.0% 0.0% 0.0% Other Extracting Activities -3.7% -12.2% -3.3% -11.3% Food Processing & Tobacco -0.9% -1.2% -0.7% -0.9% Textiles 5.7% 10.8% 1.6% 3.9% Wood Processing -0.2% -0.2% -0.3% -0.2% Pulp and Paper -0.5% -0.6% -0.7% -0.9% Refinery Processing -1.0% -2.3% -1.1% -2.2% Chemical Products -1.4% -2.2% -1.4% -2.1% Metallurgy 4.3% -7.8% -3.8% -6.5% Metallic Products -1.1% -1.7% -1.3% -1.7% Machinery and Equipment 2.6% 3.3% 1.3% 1.8% Electric Machinery & Appliances 0.7% 0.8% 0.0% 0.0% Motor Vehicle Production 1.0% 1.5% 0.2% 0.5% Other Transport Equipment -0.6% -2.3% -0.9% -2.3% Electricity, Gas, Thermal Power and Hot Water -1.7% -2.5% -1.9% -2.6% Water and Waste Water Treatment -1.0% -1.4% -1.2% -1.7% Services 0.2% 0.3% -0.2% -0.4% Other Industries -0.3% -0.3% -0.6% -0.6% Total -0.1% -0.2% -0.4% -0.6% Source: Cavalcanti and Maliszewski (2001). 9 As discussed in Appendix I at the end of this chapter, the general equilibrium model used in this chapter is calibrated to the social accounting matrix with only minimal corrections to the gross operating surplus for some branches. The main advantage of using the uncorrected data is that one avoids making controversial assumptions about the scale and direction of subsidization. Although it misses some of the descriptive reality, the coefficients calibrated from these data should still broadly reflect the structure of the Romanian economy. 12 Chapter 1: Economic Reforms and Public Expenditure Management 1.27 The results of these simulations are reflected as changes in the GDP at factor prices in each sector of economic activity. The results are broadly consistent with our expectations, given the structure of the Romanian economy. At present, low energy prices benefit primarily energy production and energy intensive sectors. The former includes coal extraction of coal and the extraction of other minerals (e.g., oil). The latter includes metallurgy, the production of metal for construction, the chemical industry, oil processing, and, to a lesser extend, food processing and agriculture. Thus, in the first scenario, an increase in energy prices leads to a decline in GDP. This happens because higher prices lower the demand for domestic energy production when compared with the current equilibrium with subsidized prices. Also, higher energy prices lead to a lower GDP because this implies a reduction in the implicit subsidy provided under current circumstances to the relatively energy intensive sectors of the economy. After the increase in energy prices, however, resources are shifted to the least energy intensive sectors: first the production of textiles and clothes, next to the production of electric and electronic equipment, machinery and to production of transport equipment.10 The shift of resources also benefits services. 1.28 In the second scenario, the increase in energy prices is combined with the removal of subsidies. Since subsidies are modeled as a negative tax, we assume that after the removal of subsidies, the effective tax rate on any individual sector/industry that was previously subsidized is reduced from a negative number to zero. As in the first scenario, lower explicit subsidies lead to a reallocation of resources towards textile and clothing production, although the magnitude of reallocation is obviously smaller than in the first scenario. Also, energy production declines in this scenario for reasons similar to those in the first scenario: the demand from previously subsidized industries is lower and removal of coal subsidies directly affects energy producers through higher input prices. These two factors reinforce each other, so that the pattern of sectoral reallocation is roughly the same in the two scenarios, although the magnitude is greater in the second. 1.29 Capital and labor mobility. A key message emerging from this simulation is that the reallocation of resources across sectors is indispensable for growth under any scenario of economic reforms. Indeed, more important than the extent of the economic reforms (energy price increases vs. energy price increases plus subsidy removal) is the degree of capital and labor mobility. The GDP drop is higher and unemployment rises when the model assumes that sectors/industries employ specific types of capital, not mobile across sectors, and when wages are fixed at given levels. The differences in the two settings are significant and explained almost entirely by lower GDP growth in the sectors that stand to benefit from the factor reallocation throughout the economy: textiles, machinery and equipment assembly, and services. To understand those factors that could help or hinder capital mobility we turn to three important features of the Romanian economy: its ownership structure; the extent of the reforms since 1998; and the broader macroeconomic implications of these reform scenarios. 1.30 Ownership. The distributional implications of the economic reform program can be better appreciated if one understands the ownership lines along which the Romanian economy is drawn. The private sector is dominant in labor intensive, light industry sectors, such as 1o The increase in production of machines may seem surprising but it reflects a large share of labor intensive outward processing for exports in this sector. Chapter 1: Economic Reforms and Public Expenditure Management 13 production of textiles, clothes and footwear, and the assembly of electric and electronic equipment. It is also partially involved in the chemical industry (e.g., production of plastics and synthetic rubber), and in metal processing. The state sector is dominant in extraction activities, petroleum processing, metallurgy, transport equipment, and energy production. Although the privatization process is gaining momentum, the share of the private sector in the industrial employment and production is still low, at 40 percent. The simulations therefore reflect an adjustment in the industrial sector that pits the emerging private sector against a relatively un- restructured state owned sector. 1.31 Accounting for reforms since 1998. When examining the results of the simulation model, there are several reasons to expect the. outcome of the reforms to be more positive in terms of GDP growth than anticipated in the simulations. The simulations are based on the 1997 input-output matrix for the Romania economy, and since 1998 there have been significant adjustments in the structure of the Romanian economy (Table 1.8). These include a reduction in the share of certain sectors in manufacturing output and an increase in the share of manufacturing output directed toward exports. The most significant adjustment is the reduction of the share of metallurgical production, with the share dropping from around 42 percent in December 1998 to just under 23 percent in July 2001.11 Other important changes include a higher share in manufacturing output in light industries (food, beverages and tobacco, and textiles and textile products). These developments suggest that the results of the simulation overstate the extent of the economic contraction resulting from economic reforms. If one were to adjust the results to account for the economic restructuring since 1998 that has allowed a reduction in the weight of some loss making sectors (e.g., metallurgy), economic reforms would lead to higher GDP growth, while still yielding a similar sector distribution of winners and losers. The usefulness of these simulations is, therefore, primarily in highlighting the distributional impact of these reforms, and in emphasizing the importance of flexibility in reallocating resources. 1.32 What Accounts for the Changes since 1998? In the wake of the 1999 external debt crisis, the Romanian Government undertook several important reforms. Three of these reforms help explain the changes in the structure of the Romanian economy since 1998: (i) the adjustment of the exchange rate, setting it at more competitive levels; (ii) the achievement of some degree of fiscal consolidation; and (iii) the initiation of bank and enterprise privatization. Since December 1998, the Romanian Lei has been adjusted in lockstep with inflation, rising rapidly in the first half of 1999, and preserving its real value thereafter (Figure 1.1). Also, the Government's has reigned in expenditures, reducing the consolidated budget deficit from a high of 5.5 percent of GDP in 1998 to 3.7 percent of GDP in 1999, and it has remained around that level since then. Finally, privatization gained a new momentum after the 1998-99 external crises, with particular emphasis placed on the banking sector. Three of the largest state banks (BRD, Bank Post, and Banca Agricola) were privatized, and one was liquidated (Bancorex), curtailing the practice of channeling funds from the banking sector to loss-making enterprises. " The data from the input-output matrix is not directly comparable with the data from the industrial survey because estimates are centered on different dates. The industrial survey does give nevertheless an accurate indication of the magnitude of the change. 14 Chapter 1: Economic Reforms and Public Expenditure Management Table 1.8: Industrial Output, December 1998-July 2001 Dec. 1998 Dec. 1999 Dec. 2000 July 2001 Total 100 100 100 100 Mining and quarrying 5.2 6.6 10.9 9.5 Petroleum and natural gas extraction 4.1 5.1 8.8 7.7 Other extractive activities 1.2 1.5 2.0 1.8 Manufacturing 82.9 81.6 82.0 87.7 Food, beverages and tobacco 10.8 8.4 15.8 16.2 Textiles and textile products 5.3 5.6 8.7 10.9 Wood processing 1.7 1.8 2.6 2.1 Pulp, paper and cardboard 0.7 0.9 1.2 1.2 Publishing and records reproduction 0.1 0.1 0.3 0.3 Crude oil processing, coal coking and nuclear fuel treatment 4.9 8.5 11.3 10.8 Chemical products, rubber and plastics 5.8 6.5 14.7 11.0 Metallurgy 41.7 39.1 15.1 22.8 Metallic construction and metal products 0.8 0.8 1.1 1.3 Machinery and equipment 6.7 5.7 6.1 6.5 Transport 4.3 4.7 4.5 4.5 Electric and thermal energy, gas and water 11.9 11.8 7.0 2.8 Source: Romanian National Institute of Statistics. 1.33 Macroeconomic implications. Notwithstanding the adjustments since 1998, the implications of the reforms for the country's main macroeconomic aggregates are still significant, and are important to understanding how the distributional impact of these reforms affects growth (Table 1.9). In a setting of flexible wages and mobile factors of production (capital and labor), the reallocation of resources across sectors yields higher growth because of the better performance of exports. This reflects the reallocation of resources from exports in previously subsidized extraction activities and metallurgy to more dynamic export sectors/industries such as textiles, and machinery and equipment. Note that the increase in exports surpasses the rise in imports, which also responds to the reallocation of resources throughout the economy. Imports of investment goods rise, especially in metal construction, machinery and electric and electronic products. Imports of metallurgy products also rise, reflecting the higher costs of domestic production after the removal of energy subsidies. In the case of fixed wages and limited capital mobility between the sectors, while we observe a similar reallocation of resources across sectors, this still leads to a decrease in both exports and imports, coupled with an appreciation of the real exchange rate. The decline in exports reflects lower exports by extraction activities and the metallurgical industry that are not compensated by an increase in exports of textiles and machinery and equipment. The reduction in imports reflects, as before, lower energy imports (crude oil and gas), as well as lower overall income. Economic reforms in a low mobility setting leads therefore to a decline in previously subsidized activities without allowing offsetting growth in labor intensive sectors/industries. Chapter 1: Economic Reforms and Public Expenditure Management 15 Figure 1.1: Real Exchange Rate, 1999-2001 120 115 110 105 100 95- 90 , , 1.34 Another important macroeconomic consequence of reforms is the rise in investment, which in both settings contributes to long run growth. This is due more to higher government savings than to the overall improvement in resource allocation throughout the economy. Higher government savings reflect higher energy prices and increasing utility profits which in turn are, at least in the short run, captured by the government.12 The sectoral reallocation throughout the economy also reinforces the positive effect on government revenues since the sectors benefiting from new prices are subject to a higher effective taxation. Since government consumption is fixed in the model, the increase in government revenues results in the increase in government savings. While lower subsidies entail lower household income, since a large share of these production subsidies are captured by households through employment in loss-making enterprises, the increase in government savings more than compensates for the lower private savings. Since foreign savings are exogenously determined by access to international capital markets, total savings and, as a consequence, total investments rise. 1.35 The results of this simulation suggests that removal of implicit and explicit subsidies will not lead to an immediate increase in GDP from better resource allocation. The removal of subsidies will reduce private consumption, as employees in large, energy intensive sectors that benefit from subsidized employment and higher than average wage industries lose their job. This subsidization policy, however, undermines growth opportunities and, as the evidence from the 1990s indicates, is not sustainable. The removal of the implicit subsidization is a necessary step to improve government finances, increase savings, and place Romania on the sustainable growth path. 12 Given the uncertainty about the initial subsidies to the utility sector, this can be alternatively interpreted as a reduction in subsidies to these utility companies. We also assume with confidence that the profits (or lower subsidies) are larger that a loss in revenues from lower level of activity in some branches. 16 Chapter 1: Economic Reforms and Public Expenditure Management Table 1.9: Simulations from the Model: Macroeconomic Indicators (% change compared to base case equilibrium) Flexible wages, mobile factors Fixed wages, activity specific capital Increase in electric Increase in Increase in electric Increase in energy tariffs electricity energy energy tariffs electricity energy tariffs, plus tariffs, plus reduction in reduction in subsidies subsidies Domestic Absorption -0.1% -0.2% -0.4% -0.6% Private consumption -1.4% -1.9% -1.5% -2.0% Government consumption 0.0% 0.0% 0.0% 0.0% Exports 0.5% 1.0% -0.5% -0.5% Imports 0.4% 0.8% -0.4% -0.4% Fixed investment 5.1% 6.8% 3.6% 4.7% Stock change 0.0% 0.0% 0.0% 0.0% Private savings -0.1% -0.2% -0.1% -0.2% Government savings 1.1% 1.5% 0.9% 1.2% Real Exchange Rate -0.8% -0.9% 0.0% 0.3% GDP at market prices -0.1% -0.2% -0.4% -0.7% Source: Cavalcanti and Maliszewski (2001). 1.36 In the immediate short run, the sustainability of economic recovery will result from the continued increase in exports, in particular exports to the EU countries, and an increase private sector investment, financed by foreign direct investment and retained earnings. The management of fiscal policy can contribute toward this in three principal ways: (i) enhancing the incentives for private savings and investment with prudent and transparent fiscal policies that restrain inflation and instill confidence; (ii) increasing the level of public savings by eliminating unnecessary public expenditures; (iii) increasing overall efficiency of public administration, improving the quality of the delivery of public services. The latter is particularly important because structurally deficient public administration increases the cost of regulatory and tax compliance that are impediments to the development of the Romanian private sector, as well as to attracting needed FDI inflows. 1.37 Since there is little value added in making broad recommendations about pursuing prudent fiscal policies and improving the delivery of public services, the next section attempts to provide more detailed guidance on these recommendations. It examines the primary surplus requirements under different growth scenarios. It also outlines the strategy for improving service delivery defined in the rest of this report. The two messages that emerge are the following. The causality between fiscal policy and growth runs both ways. While the management of fiscal policy is critical for the sustainability of the recovery, sustained growth will make the management of fiscal policy much easier. Also, expenditure reductions need to be implemented in a purposeful manner, aiming at improving efficiency and facilitating factor reallocation, since the latter, as the results of the simulation indicate, is key for sustaining growth. Fiscal Policy for Sustainable Growth 1.38 Prudent Fiscal Management. The previous section highlighted the distributional consequences of economic reforms, and indicating how on factor reallocation affects growth. Chapter 1: Economic Reforms and Public Expenditure Management 17 This section extends this analysis further, establishing the link between growth and the sustainability of fiscal policy, and between prudent fiscal policy and quality in the delivery of public services. We begin by looking at a standard measure of fiscal sustainability: setting the primary surplus at a level that stabilizes the public debt-to-GDP ratio. Table 1.10 presents several scenarios for the primary surplus under different assumptions for growth and interest rates (domestic and foreign). This example aims at shedding some light on sensitivity of fiscal policy outcomes to various macroeconomic assumptions.'3 In this example, we assume that the real GDP growth rate, domestic and foreign real interest rates and real appreciation of the currency remain constant throughout the simulation period (2001-2010). The results are mostly illustrative, and show the increases (reductions) in primary surplus required to stabilize the debt/GDP ratio in 2010 at the level registered in 2001. The 2001 level (30 per cent of GDP) is chosen to indicate what would be required to sustain the current economic conditions, with the constancy of the ratio providing a reasonable indicator of the policy sustainability. Table 1.10: Primary Surpluses Required for Stabilizing the Debt/GDP ratio (% of GDP) Real GDP Interest Real Rate Growth Domestic, Foreign 10%,11% 8%,l1% 6%,1l% 10%,7% 8%,7% 0% 1.67% 1.54% 1.42% 1.21% 1.08% 2% 1.21% 1.08% 0.95% 0.76% 0.63% 4% 0.76% 0.63% 0.50% 0.32% 0.18% 6% 0.33% 0.20% 0.07% -0.11% -0.25% Source: Cavalcanti and Maliszeswki (2001). 1.39 The results reported in Table 1.10 document that some fiscal adjustment will continue to be required, especially if growth slows down and interest rates remain high. The required primary surplus ranges from around 1.5 percent of GDP in the case of no growth and high domestic and foreign interest rates (10 and 11 percent, respectively), to a slight primary deficit in the scenario of high growth (6 percent) and lower domestic and international interest rates (8 and 7 percent, respectively). In most cases, however, the required primary surplus lies in the 0.5 to 1 percent of GDP range, which also happens to be the current level of the primary surplus. This highlights two points made earlier: most of the fiscal adjustment effort lies outside the budget; and the real challenge will be to reallocate resources rather than simply reduce expenditures to sustain the economic recovery. 1.40 A Medium Term Expenditure Framework. Reallocating expenditures toward more productive uses will require a multi-year approach, especially in reducing large expenditure programs. The large expenditure programs that need attention include entitlement programs, public sector wages and salaries, and debt service obligations. The conservative estimate is that these expenditures account for around 50 percent of budget expenditures, and are effectively fixed in the immediate short run (Table 1. 11). A medium term framework would aim at defining a three path for these expenditure programs, increasing the discretionary share of the budget, and reducing the reliance on the current budget practice of compressing capital investments and 13 This sustainability indicator is similar to the one developed in Blanchard et al. (1991). 18 Chapter 1: Economic Reforms and Public Expenditure Management material expenditures.14 The latter can only be compressed for a limited time without seriously reducing the operational effectiveness of the Government. Table 1.11: Consolidated Government Revenues and Expenditures, 1995 - 2001 (% of GDP) 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total Revenues 33.4 31.2 31.4 29.0 28.9 29.9 32.2 32.4 31.9 Total Expenditures 33.8 33.4 34.7 33.8 34.2 35.3 35.9 35.5 35.4 Curren Expenditurest 28.9 27.6 28.9 28.3 28.7 31.1 33.0 32.0 31.9 Goods and services 11.7 11.9 12.6 11.8 10.7 11.5 12.8 12.6 12.6 Wages and salaries 6.7 6.5 6.5 6.0 4.9 5.0 4.9 5.5 5.3 Other goods and services 5.0 5.4 6.1 5.8 5.8 6.5 7.9 7.1 7.3 Interest payments 0.9 1.4 1.4 1.7 3.9 4.7 5.3 4.9 4.1 Subsidies and transfers 16.3 14.4 14.9 14.8 14.1 14.9 14.9 14.6 15.2 Subsidies 6.8 3.8 4.1 4.3 2.5 1.7 1.7 2.2 2.2 Transfers 9.5 10.5 10.8 10.4 11.6 13.2 13.1 12.3 13.0 Child support 0.7 0.6 1.1 1.0 0.7 0.5 0.7 0.6 n.a. State Pensions' 0.5 0.5 0.4 0.5 0.7 0.6 0.5 0.5 n.a. Social insurance pensions 5.9 5.9 5.3 5.9 6.1 5.9 5.9 5.9 n.a. Farmer's pensions 0.4 0.5 0.6 0.6 0.5 0.5 0.4 0.5 n.a. Unemployment benefits 0.3 0.2 0.2 0.4 0.4 0.2 0.3 0.2 n.a. Other social transferS2 3.0 2.7 3.9 4.7 4.6 5.5 3.0 2.7 n.a. Capital 4.2 5.5 5.3 5.2 4.8 3.6 2.8 3.1 3.4 Lending minus repayments 0.7 0.3 0.6 0.3 0.7 0.6 0.1 0.5 0.1 Surplus (+)/Deficit(-) -0.4 -2.2 -3.4 -4.8 -5.3 -5.4 -3.7 -4.0 -3.5 1) Includes pensions for the disabled, orphans, war widows, and the military. 2) Includes allowances for families with many children, and other allowances and indemnities. Source: Romanian Ministry of Public Finance. World Bank staff calculations. 1.41 The expenditure programs that require a three year budget programming include the following: > Public administration. The government sector employment steadily increased from the start of transition until 1996, reflecting a role of the state as an employer of last resort. This increase in employment was only partially reversed after 1997, when the government employment fell by 4 percent after a cumulated increase by about 10 percent in the first six years of transition. Wages and salaries in the public sector, which were reduced below average economy wages between 1991-1993, rebounded in 1994-95, fell again to slightly in 1996-97, and increased sharply thereafter. The result is that the size of the government administration, large even by transition countries' standards, constitutes a burden on economics development, driving up wages and draining resources available to the private sector. A further reduction of employment in the government sector is required not only for improvement of Romania's fiscal sustainability, but also for making room for a slimmer, more efficient and better-paid public administration in the future. 14 In practice, this is what has been happening over the last few years, with public investments usually being delayed, raising the costs of these undertakings further. Chapter 1: Economic Reforms and Public Expenditure Management 19 > Interest expenditure. Notwithstanding Romania's low public debt to GDP ratio (30 percent), interest expenditures account for a relatively large share of budgetary spending (11.6 percent). This reflects two trends. The rapid accumulation of debt obligations to finance large fiscal and quasi-fiscal deficits, despite beginning the transition with essentially no debt. The extension of loan guarantees to the private sector, which have regularly been called. In addition, continuously high levels of inflation have required the National Bank of Romania to keep high positive real interest rates. Fortunately, the overall debt dynamics for the period ahead is not unfavorable, and prudent macroeconomic policies should allow the share of interest expenditures in the budget to decline. This requires, as mentioned above, preserving the current levels of primary surplus and limiting any contingent liabilities coming from state owned banks and enterprises. > Social security system. At the start of the transition process the Romanian pay-as- you-go pension system was already experiencing problems stemming from inadequate individual contributions, extension of benefits to non-contributing employees, and very generous methods of calculating the benefits. The transition and legislative changes brought about a wave of early retirements and further deterioration of the link between the individual wages and retirement benefits. The policy makers' responded to the growing imbalances in the system by increasing the statutory rates and by curtailing some of the benefits. The increase in statutory rates stopped the decline in social security contributions. The high contribution rates, however, substantially increase non-wage costs of employment for enterprises and impose a heavy tax on additional work for individuals. The burden on enterprises led to tax evasion and there is some evidence that the rates are now on the wrong side of the Laffer curve.'5 The implicit tax on additional work favors early retirement and further increases the dependency ratio.16 Reductions in benefits take the form of less-than-full indexation of pensions to inflation. Although quite effective in the short term, the reductions entail high political costs and tend to be quickly reversed, with the benefits often exceeding previous levels in real terms. The growing number and pensioners combined with the falling number of contributors, very high statutory contribution rates but low compliance rates, and the growing deficits of the pension system all require decisive reform steps from policy makers. These reforms are discussed at greater length in Chapter 7 of this report. They include defining a sustainable replacement rate and indexation rule for pension benefits, as well as bringing non-insurance related benefits under greater scrutiny. While the latter account individually for a relatively small share of the overall expenditure of the social security system, combined they account for several percentage points of GDP. 15 The Laffer curve hypothesizes a relationship between the tax rate and the tax revenues collected, where tax revenues collected rise up to an optimum tax rate and then begin declining at tax rates higher then the optimum tax rate. Being on the 'wrong side of the Laffer curve' means that the tax rate has exceeded the rate required to maximize revenues. '6 Defined as the ratio between beneficiaries and contributors to the social security system. 20 Chapter 1: Economic Reforms and Public Expenditure Management 1.42 What happens when we compare expenditures by function? It is important to note the same degree of expenditure inertia emerges when we compare Government expenditures by functions (Table 1.12). With overall expenditures remaining relatively stable during the period, increase in expenditures on public administration, health, social welfare and interest payments on public debt were accommodated by expenditure declines in almost all the other categories. The only exceptions are expenditures on public order, transport and communications. The increase in expenditure on interest payments on public debt exceeded, however, both in absolute and relative terms, the observed increase in these three last categories. Table 1.12: Consolidated Government Expenditures by Functions, 1995-2000 (% of GDP) Change 1995 1996 1997 1998 1999 20001995-2000 Total Expenditures 34.7 33.8 33.9 35.3 35.9 35.6 2.4% Public Administration 1.2 1.1 1.1 1.4 1.3 1.5 24.8% Defense 2.1 1.9 2.3 2.3 1.5 1.7 -21.6% Public Order 1.5 1.5 1.1 1.5 1.4 2.0 31.4% Education 3.4 3.6 3.3 3.3 3.1 3.1 -8.4% Health 2.9 2.8 2.5 3.0 3.6 3.7 29.0% Social Welfare 9.3 8.9 9.6 10.6 10.9 9.9 6.2% Other Social Services 0.3 0.3 0.3 0.2 0.1 0.0 -88.0% Community Services 2.0 1.9 1.8 1.7 1.7 1.9 -6.0% Environment 0.2 0.2 0.1 0.1 0.1 0.1 -19.9% Industry 2.4 2.2 1.1 0.7 0.8 0.8 -66.2% Agriculture and Forestry 2.3 2.4 1.4 1.2 1.0 1.2 -49.7% Transportation and Communications 2.4 2.3 2.3 2.4 2.7 2.7 13.3% Interest Payments on Public Debt 1.3 1.6 3.6 4.5 5.0 4.6 247.3% Source: Romanian Ministry of Public Finance. World Bank staff calculations. 1.43 Improving the delivery of public services. Tackling these large expenditure programs is an important first step toward identifying opportunities for improving public service delivery. This focus on service delivery reflects, in turn, two important concerns. First, the demands for public services by the Romanian population have changed significantly over the last decade with the emergence of open unemployment and poverty. The supply of services has, however, failed to respond adequately to these new demands. Progress has been slow on fiscal decentralization and in setting up the pre-conditions for adequate private delivery of some of these services. Local governments in particular have struggled with challenges ranging from underfunded mandates and unstable expenditure assignments, to problems related to size, geographic location and ethnic composition. 1.44 Second, quasi-fiscal pressures have emerged in part because the Government has been unable to reorganize the delivery of public services to accommodate these new demands. In many instances, large state owned enterprises continue to be the main employers in small to mediuh sized towns. This creates strong pressures to subsidize these enterprises, keeping them operating, even if at a loss. In yielding to these pressures, low energy tariffs, and tax and energy arrears have become the main sources of soft budget constraint, with inevitable implications for the budget. Soft budget constraints fail to encourage cost control and they delay improvements in operational efficiency. The result is stagnant productivity, and an inability to shift resources to new programs that entail a more productive use of resources. This leaves governments with Chapter 1: Economic Reforms and Public Expenditure Management 21 very little choice but to accommodate the demands by public enterprises for additional resources, for fear that the delivery of key services will suffer (e.g., energy, transport), or a massive number of workers will need to be laid off. This in turn compromises the achievement of the fiscal targets, the very goal that is trying to be achieved. Conclusions 1.45 This chapter provides the background against which the 2003 budget will be prepared. It focuses on the design of the Government's economic reform program and its implications for public expenditure management. There are two main findings: > The sustainability of the current economic recovery hinges on containing the quasi- fiscal deficits, and keeping the government's consolidated budget deficit on target. The simple reason is that these deficits absorb private sector savings (both foreign and domestic) that could otherwise fund productive investments. Lower investments reduce the potential for sustained growth and real increases in living standards. > At the budget process level, containing quasi-fiscal deficits implies a fundamental shift in the manner in which significant shares of public resources are allocated. At present, quasi-fiscal deficits (including tax arrears and tax evasion) mean that public resources are allocated in an ad hoc, discretionary manner, usually in response to political pressures during the budget year. Measures to contain quasi-fiscal deficits need to be accompanied therefore by budget reforms that strengthen the budget process, allowing public resources to be allocated in an orderly and transparent manner for the general public and policymakers alike. 1.46 To assist the Government in making this shift toward allocating public resources in a measured and transparent manner the rest of this report focuses on strengthening the budget process throughout the public sector. It begins by reviewing the budget process at the central and the local government level, focusing next on a range of public services that are key in building a fair and democratic society: education, justice, health, social assistance, and social security. Looking forward, these reforms will allow Romania to realign public sector activities to a range of new priorities. This realignment is important both in allowing public expenditures to come into line with available resources, and in improving service delivery. Better service delivery will contribute to overall growth performance and, as a result, to increased tax payment compliance. The latter is important because Romania's integration into the world economy will require increase in expenditures for EU accession and NATO membership.  CHAPTER 2: STRENGTHENING THE INSTITUTIONS FOR BUDGET MANAGEMENT Introduction 2.1 This chapter examines how strengthening the budget process is an integral part of achieving the Government's objectives of meeting aggregate fiscal targets without losing the growth momentum, while still enhancing the quality in the delivery of key public services. Institutional reforms should (i) make the budget document more important, providing the necessary discipline to maintain aggregate fiscal discipline; (ii) facilitate strategic prioritization across policies, programs and projects; and (iii) encourage better use of resources. The most important achievement would be, however, shifting the budget execution away from the present system of strong reliance on management controls, towards a system that gives managers at all levels increased responsibility for the way public resources are used and holds them accountable for results. 2.2 The chapter is organized as follows. It begins by examining the recent changes in the legal framework guiding budget formulation, and the steps still needed to complete these reforms. Next, it reviews the Romanian experience with budget execution under very tight budget constraints, and the role of the Treasury in the Ministry of Public Finance in this process. Finally, it outlines recent efforts to develop the institutions of budget compliance, internal and external auditing, and ex-post evaluation, which aim at building the capacity to learn from the budget process, ensuring that lessons learned at the budget execution stage are incorporated into budget formulation. The chapter closes with a summary of findings and recommendations. Completing Budget Management Reforms 2.3 The Romanian Government has recently initiated a set of far-reaching budget reforms, aiming at preparing more realistic and effective budgets in future. These reforms are reflected in the guidelines for preparing the 2002 budget, and in the draft Public Finance Law recently approved by the Cabinet. In the guidelines for this year's budget, agencies in the central government were asked to (i) formulate expenditure strategies, defining upfront development objectives and policies, as well as expenditure priorities; (ii) provide output/outcome indicators that could be used to monitor performance; and (iii) present expenditure estimates broken down by major program. In addition, all spending agenties at central and local levels, including special funds, had to submit expenditure estimates beyond 2002 for the period 2003-05. These reforms are intended to lead eventually to the preparation of a Medium Term Expenditure Framework (MTEF), providing a strategic framework for the annual budget, and preparing the groundwork for breaking the current expenditure inertia. 2.4 The direction of these reforms has recently been further formalized by the new draft Public Finance Law aimed at bringing budget procedures closer to the practice in EU countries. There are four main modifications proposed in this new draft law: (i) the earmarking of credits targeted to finance multi-annual programs, in order to assure an improved management of funds and the compatibility with EU budget regulations; (ii) the increased transparency in fund 24 Chapter 2: Strengthening the Institutions for Budget Management allocation, by eliminating the out-of-budget retentions and the incorporation in the state budget of the special funds, with a view to unifying the budgetary practices; (iii) the introduction of financial impact assessments to accompany the initiation of new normative acts that can affect budget revenues and expenditures; and (iv) the establishment of clear rules concerning the co- financing of certain programs, which would demonstrate that the resources are directed to a large extent toward the stated priority fields of the reform. 2.5 While these and earlier reforms are extremely important, especially as Romania moves ahead in the EU accession process, most of the attention has focused on refining the legal framework and at introducing budget programming. Less attention has been given to implementing these reforms and to addressing some of the shortcomings at the budget execution stage. The latter is where many of the problems with arrears and quality in the delivery of public services emerge. One of the reasons for giving less attention to the budget execution is that the Ministry of Public Finance has a reasonably effective system of expenditure controls and has been able to keep total spending within the agreed overall budget ceiling on a cash basis. This has lead many to believe that reforming the system of budget management needs to focus primarily on closing the legal gap with EU countries. This perception is further strengthened by the belief that deficiencies in budget preparation have not led to an apparent weakening of aggregate fiscal discipline, since budget revisions in the course of the year have generally been effective in making necessary adjustments. 2.6 The rest of this chapter makes the point that there is a need to address the concerns with budget execution if the government is to meet budget targets without compromising the quality in the delivery of public goods and services. Addressing these concerns requires examining the three stages of budget management in Romania in an integrated fashion. The chapter begins by reviewing recent reforms to the legal framework governing budget management, identifying the accomplishments, and assessing where there is scope for further reforms. Next, the chapter provides a detailed account of budget execution, linking the strong central controls aimed at ensuring that cash expenditures are in line with budget ceilings with some of the problems with arrears and slow execution of the budget. Finally, the chapter outlines the recommendation for further developing the institutions of budget compliance. Many of these institutions have been recently established, and there is a need to invest now in staff resources and data collection. Improving Budget Formulation 2.7 The Romanian government operates an annual budget, with a budget year that is the same as the calendar year. Parliament usually adopts a budget early in the calendar year. Around mid- year there is typically a budget rectification process in which Parliament may alter the budget with either supplemental appropriations or imposed economies if other needs arise or overall revenues are not meeting expectations. All ministries and all other parts of government submit their budget requests to the Ministry of Finance, which compares requests and analyzes cost estimates. The Ministry of Finance naturally measures requests -against anticipated revenueand makes substantial reductions in the requests of many ministries. Then the proposed budget goes to Parliament. The Parliament has committees that examine components of the proposed budget. If a committee wants to increase some element of the budget proposed by the Ministry of Finance, the committee is expected to identify revenue source to fund that increase. Parliament Chapter 2: Strengthening the Institutions for Budget Management 25 adopts a budget, which then goes back to each ministry or other element of government to administer according to the law and according to regulations promulgated by the Ministry of Finance. It is common for many parts of the government to feel themselves under-funded. The budget rectification process is a mid-year review of the entire budget. Transfers can occur as new needs arise and some existing priorities are delayed, or almost all parts of the government can be expected to absorb the effects of a downturn in funds available. 2.8 Legal Framework. As mentioned above, the Romanian legal framework for budget management provides the main elements needed for sound fiscal management,17 and has recently been amended to bring it closer to the budget procedures in EU countries. These amendments aim at (i) making the budget calendar more binding; (ii) increasing transparency by reducing budget fragmentation; (iii) better aligning the responsibilities of the agencies involved in the budget process; and (iv) incorporating new features such as budget programming and multi-year budget planning. The challenge ahead is implementing these modifications in the budget process and making further refinements that bring budget formulation closer to policy priorities. The latter includes adding new steps to the budget calendar and broadening the budget coverage. These changes are proposed below, and should strengthen the budget as an instrument of economic policy management, helping break the budget 'inertia' by focusing expenditure on key priorities, and helping ensure a more efficient use of the funds available. 2.9 Budget Calendar. While the budget calendar defined in the Law on Public Finance (Box 1) was designed to ensure that there were no interruptions in the spending programs of agencies, in practice budgets have not been approved until well into the fiscal year. The 2001 budget became law only in April 2001, with similar delays occurring in previous years. The 2002 budget was the first in about a decade.to be submitted to Parliament on schedule. Delays have usually being associated with difficulties in reaching agreement in the Cabinet on major principles of budget policy. 17 The legal framework is defined in the Public Finance Law, first establish in 1996 and recently modified in draft form by the Government. The Public Finance Law has the following distinct stages of budget preparation are detailed in the Public Finance Law: (1) the Ministry of Public' Finance issues guidelines ('methodology') to all spending agencies (including local authorities and special funds) to help them prepare budget estimates. They include certain 'norms' to be applied, but no expenditure ceilings, and also indicate the format to be followed for presenting the estimates; (2) spending agencies submit estimates for the main economic categories of expenditure for the budget year (in current year's prices) as well as the final budget data for the current year. Local authorities present their estimates of required financing from shared taxes and of transfers from the State budget; (3) counties, municipalities and communes submit budgets to the Ministry of Public Finance; (4) spending agencies and special funds submit detailed budgets to the Ministry of Public Finance. This is followed by negotiations between agencies and the Ministry of Public Finance; (5) expenditure ceilings prepared by the Ministry of Public Finance are issued to each agency following approval by the cabinet; (6) spending agencies re-submit detailed budgets to the Ministry of Public Finance; (7) the Ministry of Public Finance submits a draft budget covering the State budget and all special funds to the Cabinet; (8) the draft budget is presented to Parliament. It is first discussed in the Budget, Finance and Banking Committee and subsequently in the plenary session of Parliament; (9) following parliamentary approval of the budget, the Annual Budget Law is published; (10) the budgets of counties and of municipalities and communes are presented to respective councils for approval. Changes made after budget approval, especially when spending ceilings need to be adjusted to reflect a change in the macroeconomic outlook, are usually introduced through so- called budget rectifications, which have been a regular feature in recent years. 26 Chapter 2: Strengthening the Institutions for Budget Management 2.10 Setting Government Priorities. To avoid these problems in the future, the budget calendar needs to include at least one additional step: a statement by the Cabinet on the Government's priorities for the budget. At present, the Cabinet gets formally involved late in the budget process, only after the initial budget submissions by the line ministries and local governments to the Ministry of Public Finance. Early involvement of the Cabinet is an important first step toward keeping the budget preparation on schedule, and toward breaking the culture of budgetary inertia and compromise, making instead the budget an instrument for achieving well defined policy objectives. The Cabinet's statement should provide an integrated view of the sector strategies, with a priority ranking of ongoing programs and investment projects, along with corresponding funding levels. Based on this statement, and the parameters provided by the Ministry of Public Finance, the line ministries and other credit ordinators would formulate their budget requests. These submissions would then be consolidated by the Ministry of Public Finance in a first draft of the budget document to be reviewed by Cabinet. Box 2.1 Timetable for Budget Preparation March 31 Macroeconomic and social parameters for the next year and the three subsequent years presented to the Government. May 1 The Ministry of Public Finance submits to the Government the objectives of the fiscal policy and the budget, as well as the expenditure ceilings for each line ministry (principal credit ordinators). May 15 The Government informs Parliament of its main assumptions regarding macroeconomic and fiscal policies for the forthcoming and subsequent 3 years. June 1 The Ministry of Public Finance sends the line ministries the guidelines for preparing budget submissions, as well as their expenditure ceilings, as approved by the Government. If macroeconomic circumstances change during this period, the Government can decide, at the proposal of the Ministry of Public Finance, on adjustments to these ceilings. Revised ceilings, if any, should be transmitted to the line ministries by June 15. July 15 The line ministries submit to the Ministry of Public Finance their proposed budgets, including the annexes and estimates for the subsequent 3 years. The local governments also submit their budget proposals. The Ministry of Public Finance initiates discussions with each principal credit ordinator on the draft proposals. In case divergences arise, disputes are taken to Cabinet. August 1 Final budget proposals are submitted to the Ministry of Public Finance. September 30 The Ministry of Public Finance consolidates the proposals and submits the draft budget law to the Government. The submission is accompanied by an analysis of the macroeconomic environment. October 15 The Government submits the budget proposal to Parliament. In case the budget proposal is not approved by the Parliament at least 3 days before the end of the current budget cycle, the standard 1/12 rule applies until approval. Source: Draft Law on Public Finance (February 2002) . The new Law on Public Finance (Law no. 500/2002) was published in the Official Gazette in August 2002. Chapter 2: Strengthening the Institutions for Budget Management 27 2.11 Budget Coverage. While the State and Local Budgets, together with the Consolidated Budget, provide a fairly complete and transparent picture of the Government's financial operations, transparency could be further improved by presenting items related to the financing of the budget deficit 'below the line', following the format found in the reports on European Standard Accounts. In addition, external loans currently being contracted by ministries for which a sovereign guarantee is provided are undistinguishable from direct foreign borrowing by the Government and should be treated as such. This would imply that debt service payments on such loans would no longer appear in the budgets of individual ministries, being budgeted instead as part of the Government's debt service. Similarly, drawings on these loans would have to be presented as expenditure in the state and local budgets instead of being listed as 'External Credits to Ministries' in the Consolidated Budget. Finally, while claims on the budget for guarantees that have been called have been included in the budget ex post, i.e. for past failures, there is a continuous need to provision for outstanding guarantees. Indeed, as the failure rate on guarantees far exceeds the risk premium and thus can be expected to continue to give rise to claims on the Government, comprehensiveness and transparency of the budget would be improved if a provision in the budget were made annually to meet outstanding guarantees. 2.12 What is included in the Consolidated Budget? The Consolidated Budget covers the state (i.e. central government) budget, the budgets of local authorities (counties, municipalities, towns and communes), the Social Security Budget, the Health Insurance Fund, and a number of other special funds financed from earmarked revenues. While overall expenditure amounts as a share of GDP have not changed during the last few years, the distribution of expenditures across budgets has shifted. The central government share in total expenditure has declined, accounting today for less than half of total consolidated public spending, down from over 50 percent in the mid-1990s. Local governments have seen their share in expenditure increase, rising to around one-sixth of total spending (16% in 2002), as the Government further decentralizes expenditure responsibilities. The extra-budgetary funds still account for the other one-third of total public expenditure, although the number of extra-budgetary funds has declined over the years. This reflects that fact that most of the extra-budgetary spending is covered in the social security budget (old age pension and unemployment insurance), and by the health insurance fund (Table 2.1). Table 2.1: Breakdown of Consolidated Government Expenditure, 2000 and 2002 2000 (actual) 2002 (budget) Lei billion % Lei billion % Consolidated expenditure 305.1 100 545.5 100 State budget 149.2 49 238.3 44 Local budgets 33.1 11 88.5 16 Extra-budgetary funds 104.2 34 189.2 35 - Pensions, health insurance, unemployment (90.4) (30) (175.0) (32) - Other funds (13.8) (4) (14.2) (3) External credits to Ministries 18.6 6 29.5 5 Source: Romanian Ministry of Public Finance. 2.13 Budget Fragmentation. Two standard off-budget practices distort the coverage of the Romania budget: the reliance on extra-budgetary funds, and the custom of not subjecting foreign finance public investment to full budgetary scrutiny. While extra-budgetary funds still account 28 Chapter 2: Strengthening the Institutions for Budget Management for a large share of public investments (see below), they have become less important over time, and are planned to be eliminated three years following the approval by Parliament of the new Law on Public Finance9. Already in the 2002 budget the number of extra-budgetary funds had been cut to 720, down from 17 in 2000. This alone has increased the budget's comprehensiveness and reduced fragmentation, since most of the spending through extra-budgetary funds is now dominated by spending on social security (pensions, health insurance, unemployment insurance), with these funds relying primarily on earmarked revenues (payroll taxes). The two only large remaining extra-budgetary funds are the Energy Development Fund and the Roads Fund. 2.14 While the number of extra-budgetary funds has been reduced, two problems in managing them still remain. Some such funds come under very little scrutiny regarding the ways revenues are raised and expenditures are allocated. This is the case of the Energy Development Fund that raises revenues by taxing consumers of electricity, and then allocates the funds to investments not always following clear prioritization rules and procedures. This way of raising revenue places an additional burden on electricity consumers at a time when the country is already increasing tariffs significantly to reach cost recovery levels. A simple solution would be to eliminate the tax on electricity consumers and fold the investments into the budget, ensuring greater scrutiny over these investments at the time of the budget approval. 2.15 The second problem arises, by contrast, with funds such as the social security funds (pensions, health insurance, unemployment insurance), that face excessive interference from the Ministry of Public Finance in defining expenditure ceilings. These funds rely primarily on payroll taxes. However, the Pension Fund has statutory payment obligations that exceed its revenues and, as a result, has been running large deficits. The Ministry of Public Finance has therefore been imposing spending limits on the other two funds, so that their surpluses21 help offset the consolidated budget deficit. The consequence has been a reduction in the accountability of the management of these social security funds, as fund managers and service providers under these funds (e.g., hospital and other health care units) accumulate payment arrears to lay claim in advance on these surpluses. 2.16 Public Investment. Another area where budget coverage is weak is public investment, as the central and local government budgets capture only a fraction of the Government's overall investment effort. Capital expenditures financed through external credits contracted directly by line ministries with Government guarantees are not included in the budget. These investments account for one-third of total public investments, exceeding investment expenditures included in either central or local government budgets (Table 2.2). In addition, some of the special funds undertake capital expenditures, notably the Energy Sector Development Fund. These two sources of off-budget expenditure are significant, since, taken together, public investment outlays have averaged around 4.0 percent of GDP, accounting for just under 10 percent of consolidated government expenditures. Since at present just under 50 percent of public investments are financed directly through external credits and special funds, this excludes a large fraction of the consolidated government expenditure from the budget process - expenditures that do not '9 Law 500/2002. 20 Counting in the Unemployment Fund, and the Fund for Romania (later included in the State budget). 21 Surplus funds are deposited in the Treasury account, earning annual interest rates of around 10 percent, which is well below the rate of inflation. Chapter 2. Strengthening the Institutions for Budget Management 29 compete fairly, openly and simultaneously with other expenditures at the time of the budget preparation. Table 2.2: Consolidated Public Investment by Main Sources of Funding, 1995-2002 (% of Total Public Investment in the Consolidated Budget) 1995 1996 1997 1998 1999 2000 2001 2002 State budget 47.9 42.0 31.2 31.6 23.9 20.4 25.7 29.0 Local budget 25.9 24.0 21.7 25.4 31.2 27.3 25.7 25.1 State Ownership Fund 13.5 22.3 21.3 0.0 0.1 0.0 0.0 0.0 Energy Development Fund 7.1 6.7 6.0 4.6 8.2 9.6 8.0 7.8 Public Roads Fund 1.1 -0.9 1.9 2.2 2.7 1.4 1.9 3.2 Other funds 2.6 2.0 1.5 3.3 2.8 2.9 1.4 1.2 External Credits 1.8 2.1 16.3 32.8 31.0 38.5 37.3 33.5 Public Investment in the Total Consolidated Budget (% of GDP) 5.3 5.2 4.8 3.6 2.8 3.1 3.4 3.4 Source: Romanian Ministry of Public Finance. 2.17 The off-budget treatment of public investments undermines both budget coverage and fiscal discipline. While investment projects considering external funding must have their technical and economic documentation reviewed and approved by various agencies, depending on the size of the project and whether external financing is involved (Table 2.3), approval does not mean that they have either a defined financing plan, or a clear ranking for priority funding. Project approved for external funding competes on equal terms with other projects for funding within the borrowing ceilings approved in the annual budget. Since there is no priority ranking, projects receive funding on a first-come, first-served basis, with line ministries rushing in borrowing requests to outbid each other. In some instances, this has lead to more borrowing for projects than can be accommodated by available counterpart financing in the budget. This in turn has caused delays in the implementation of ongoing projects because of shortages in counterpart funding. Table 2.3: Approval of Technical/Economic Documentation of Investment Projects Cabinet Ministry of County Local Spending Public Finance Council Council Agency State budget Project cost Over Lei 80 billion X X Between Lei 5 and 80 billion X X Under Lei 5 billion X Project with external finance X X Local Government budget Project cost Over Lei 80 billion X X Between Lei 2 and 80 billion X Between Lei 2 and 45 billion X Under Lei 2 billion X Project with external finance X X Source: Law on Public Finance; Law on Local Public Finance 30 Chapter 2: Strengthening the Institutions for Budget Management 2.18 Revenue Forecasting. Another weak link in the budget formulation stage is revenue forecasting. Revenue forecasts for the coming year are the critical stage for setting spending plans and keeping the desired overall. budget balance. Revenue forecasts can also anticipate revenue shortfalls, providing enough time for either provisional measures or a budget amendment. Admittedly, revenue forecasting in the current transition environment is a difficult task. However, both the methodologies used and the frequency of the.forecasts prepared by the Ministry of Development and Prognosis should be improved. The methodology used by the Economic Policy Department for revenue forecasting is based on an outdated version of the Bank's RMSM-X model, which builds on in-house estimates of the national accounts for production, employment and income. Most revenue series are forecast on the basis of a single aggregate variable, with forecasts done on regular intervals, rather than on a continuous basis. Recommendations on budget preparation 2.19 Setting Government Priorities. The Government, in particular the Office of the Prime Minister, should make it a tradition to issue and publish in an accessible way in January of each year a 'Government Policy Priorities' document that outlines the major policy thrusts of government policy for the coming year. This document should become the main vehicle for setting boundary lines between the core public sector and all other activities in the public sector that should be privatized, contracted out or discontinued. The initial allocation of expenditure ceilings by the Ministry of Public Finance to line ministries in the budget circular should reflect these explicitly stated priorities of Government. This statement of overall policy priorities would allow the government to move beyond the item-by-item budget proposals and parliamentary appropriations and, instead, move to using the budget as the overall instrument of economic policy including the management of the macro-economy. 2.20 Defining a Medium Term Expenditure Framework (MTEF). While clearer setting of government priorities is an important first step in reducing expenditure inertia, this should be implemented in the context of a medium term framework. Only over a multi-year period can the share of non-discretionary expenditures increase. A MTEF would indicate broad guidelines for budget preparation, such as the scope for expenditure growth and reallocation, as well as setting out the Government's priorities over the medium term within an agreed financial framework. Introducing a Medium Term Expenditure Framework would require deepening of reforms already introduced in 2002, as well as changes in the current legal framework. A prototype of a Medium Term Expenditure Framework was already prepared as part of the 2002 budget exercise, although it was limited to a set of expenditure projections and emerged at the end of the budget process. The first change, therefore, would be to prepare this expenditure framework at the start of the annual budget cycle and then, following approval by the Cabinet, it would be sent to line Ministries and other spending agents before they start preparing their budgets. 2.21 Strengthening the Underlying Macroeconomic Framework. Whether in preparation for either the annual budget or the Medium Term Expenditure Framework, the Government has to make more explicit the underlying macroeconomic framework. This would provide the rationale for defining the broad policy objectives to be achieved through the budget, and the required reallocating resources to achieve these objectives. While this is what is intended in one of the recent amendments of the Public Finance Law, abiding by the spirit of this amendment will Chapter 2: Strengthening the Institutions for Budget Managem'ent 31 require a significant departure from current practices. For instance, although the final document for the 2002 budget does have a section entitled 'The Main Targets for Macroeconomic Policy for 2002', including a few paragraphs on 'the main targets of budgetary policy', this reflects rather than guides the budget. Also, as mentioned above, a stronger macroeconomic framework would assist in preparing revenue forecasts, one of the critical steps in budget formulation. This would in turn require strengthening of the forecasting capabilities of the Ministry of Development and Prognosis and the Ministry of Public Finance. Forecasting revenues should be a continuous activity. Monthly updates should incorporate what has been learned from the behavior of the revenue series in the previous months, incorporating not only changes in economic variables but also innovations in the tax laws and tax enforcement. The monthly forecast updates should be shared with the Treasury for cash planning purposes. 2.22 Improving budget formulation, while useful in establishing a closer link between resource allocation and policy priorities, should not be seen only as an end in it self. It should lead to more effective budget execution. Budget execution, however, usually receives less attention from policy because it involves several routine steps in implementing decisions that in principle have been taken at the budget formulation stage. Nothing could be further from reality. There is a considerable amount of discretion involved in the budget execution, especially at the level of the local authorities, and understanding and perfecting these routine steps should also be a concern for policymakers. The chapter turns its attention next, therefore, to this process of budget execution. Strengthening Budget Execution 2.23 Budget execution involves a number of steps: matching the pattern of resource needs. over the course of the year with expected revenue flows, establishing credit lines for all agencies, commitment of funds following ex ante control of expenditures, verification of receipt of goods and services followed by submission of payment order, and finally payment. The process is briefly as follows: > Within 20 days of budget approval by Parliament, spending agencies submit a quarterly breakdown of their financial needs to Ministry of Public Finance, which then prepares a quarterly breakdown of aggregate expenditure based on revenue projections; > Each month spending agencies submit a request for funds to the Directorate for Programming in Ministry of Public Finance for approval; > Treasury opens monthly credit lines for each agency, by chapter and category of expenditure, broken down by secondary and tertiary credit ordinator (i.e., spending agency) as well; 32 Chapter 2: Strengthening the Institutions for Budget Management > Agencies initiate the commitment of funds, which requires approval from the Directorate for Financial Control in the Ministry of Public Finance (ex ante control);22 > Once goods and services have been received, the agency sends the invoice and payment order to the Treasury, which checks the purpose of the expenditure and the availability of funds before making the payment. 2.24 The Treasury and the Budget Execution. The Treasury plays a central role in the process of budget execution.23 First, for the purpose of the budget execution, monthly cash projections for expenditure and revenue are produced for the Treasury. The revenue projections are based on an agreed macroeconomic framework, while expenditure projections are produced by the primary agencies and agreed with the Ministry for Public Finance. These projections are essential for cash planning which is vital, inter alia, for debt management purposes. Next, when the projections are agreed, local treasuries are informed of these monthly expenditure limits for each credit ordinator for whom they provide services. These expenditure limits provide the total amount that each agency is permitted to spend based on the allocations approved by Parliament and the agreed monthly projections. 2.25 While at each step of the budget execution there is the strong Treasury control, helping ensure that expenditures are in line with revenues, this comes at a high price. The Romanian Treasury control relies heavily on cash transactions, slowing down the execution of the budget. Also, the focus on cash transactions implies that the Treasury does not monitor expenditure commitments, and, as a result, payment arrears are common throughout the system. This reduces accountability and undermines the objective of fiscal discipline that strong central controls were designed to achieve in the first place. Finally, while, in theory, a strongly centralized system should provide the opportunity to earn a return on unused cash balances, there is little evidence that this is actually taking place. 2.26 The rest of this section elaborates on the key steps for budget execution, and the role of the Romanian Treasury in carrying out this responsibility. It begins with a brief description of the two principal steps in budget execution: opening credit lines for spending agencies and making payments. Next, it discusses the two shortcomings in the current functioning of the Romanian Treasury and its implications for budget execution: weak management of the central 2 What is ex ante control? Ex ante control consists of checking whether commitments comply with budget authorizations. This ex ante control at present extends only to central government agencies and special funds, not to local authorities. To carry out this responsibility the Directorate for Financial Control in the Ministry of Public Finance has representatives in the various spending agencies. All expenditures require a 'visa' from the Directorate before funds can be committed. Expenditures on utilities are covered by a framework contract. Price increases that were not foreseen in the framework contract require a budget rectification or additional resources need to be provided for in next year's budget. 2 Romania has had an independent Treasury since the early 1990s. The State Treasury is a comprehensive system with a Central Branch within the Ministry for Public Finance and a network of Judet level (County level) Treasuries and sub-treasuries throughout the country. There are 42 Judet treasuries and several hundred at a local level operating a cash managed system. This network is responsible for collecting all revenues on behalf of the Government and undertaking all expenditure transactions. Each credit ordinator (i.e. spending agency) must hold an account with a designated Treasury with whom it must undertake all its transactions. Chapter 2: Strengthening the Institutions for Budget Management 33 treasury account and the lack of monitoring of accounting information. The section closes with a summary of the recommendations. 2.27 Opening Credit Lines. In most budgetary systems, spending agencies would have the authority to spend according to the limits set by the budgetary allocations. Commitments could be entered into and Payable Orders issued for services required if they are consistent with the allocations. Monthly returns would be submitted to the Finance Ministry to verify that the approved limits were not exceeded and that the expenditure was in line with the approved appropriation. 2.28 The system that operates in Romania is more complex, however, and involves less delegated autonomy to the individual spending agencies. It is based principally on the opening of monthly credit lines with payment limits. These are the total amount that spending agencies can spend in a given month. Local Treasuries are charged with ensuring that spending agencies do not exceed these limits. 2.29 The procedure for opening credit lines is an important element of the Treasury system. On the 25th of every month Line Ministries submit an assessment of their expenditure needs to the Ministry for Public Finance with a request for the establishment of a credit line for the next month. These requests are prepared in consultation with Secondary and Tertiary credit ordinators. Credit lines for each credit ordinator are requested. These requests are signed by the highest levels in each Line Ministry and are checked by the Department of Economy. Checks are undertaken to ensure that the amounts are within the approved budgets and that the detailed breakdown of expenditure is also in line with the agreed budget. 2.30 Under normal circumstances the requests would be approved. However, a lower than requested credit could be approved if the budget was not evolving as planned. The Ministry for Public Finance has responsibility for monitoring the evolution of the Budget. The revenue position and expenditure patterns are reviewed and circumstances may require an adjustment to the original expenditure allocations. This would oblige agencies to reduce their expenditures plans for the forthcoming period. 2.31 Once credit lines are approved this information is transmitted manually to the local treasuries. The request for opening credit lines is also sent to each Ministry and also to each credit ordinator. Accordingly, a credit line will be opened for each credit ordinator and the relevant Treasury will be informed of the credit limit. Spending in excess of this limit is not possible if the Treasury discharges its functions properly. For each month, total expenditure must be less than the total cumulative credits issued for the year to-date. 2.32 Making Payments. When the credit lines are established the credit ordinators are in a position to make payments. This involves a number of steps. First, each spending agencies has authority to authorize expenditure and enter into commitments. It is the responsibility of the agency to ensure that funds are available before commitments are established. Standard procedures are followed before an order for'payment is prepared. The goods are verified and the invoice is checked against the goods received. If correct, a payment order is prepared. This part of the payment process is the responsibility of the agency. 34 Chapter 2: Strengthening the Institutions for Budget Management 2.33 Next, in order to execute payment, the order must be presented to a local Treasury. Each credit ordinator must carry out all its transactions in this way with a Treasury where it holds an account. When presented with a payment order the local Treasury is responsible for ensuring that it is valid. The Treasury must check that it is within the limit for that ordinator. Payments are not allowed if the limit is already exceeded. In addition, the Treasury verifies the expenditure classification and validates the relevant documentation. This places an audit responsibility on the Treasury, which replicates to an extent the responsibility of the agencies when preparing the order. 2.34 Finally, if approved, the Treasury will account for the transaction by adjusting the account of the relevant credit ordinator as each one holds an account with the relevant Treasury. The credit ordinators must also record the transactions in their own accounts. The treasuries make payments either by cash or by payable order.24 2.35 This reliance on cash payments clearly slows down budget execution and a move to more modem payments systems should be a priority. Payments for goods and services are paid by Payable Orders to the payee. When these are approved these are seit to be cleared through the banking system at Judet level. Pay orders are sent manually from the local treasuries to the Judet Treasury. On receipt of these the Judet Treasury transports the Payable Orders to the Branch of the National Bank. Each Judet Treasury has an account with the Judet Branch of the National Bank. If the proceeds are available the Bank will debit the account of the Judet Treasury. This order will then be processed with the commercial bank where the payee's account is credited. 2.36 The move toward a more modem payment system should increase the reliance on the commercial banking system. In some countries, spending agencies would issue payment orders that would be presented to a commercial bank, but under the Romania system payment orders are presented to a local Treasury, which has responsibility for ensuring that the order is within the limits agreed by the Government. Greater reliance on commercial banks would reduce the need for most payments being made in cash, and strengthen the management of the government central treasury account. This system of commercial payment orders would also improve the accounting information of government revenues and expenditures, minimizing the risk of large payment arrears. 2.37 Central Treasury Account. Steps needed to improve the payment system include stronger management of the central treasury account and upgrading accounting information. One of the positive features of the Romanian Treasury system is the operation of a Central Treasury Account, which records all Government receipts and expenditures, including revenues and expenditures from a number of Government budgets and extra-budgetary funds. Despite the existence of the central and local government budgets, as well as the extra-budgetary funds, all their transactions are accounted for in the central account. Accordingly, there is a single balance 24 The timing of this process is important. The basic Treasury operation will take place at time t. The payable order might arrive at the Judet Treasury at time t if the local and Judet Treasury are located in the same municipality. Generally, the payable order would arrive at the Judet Treasury a day later at t +1. The final stage is the transfer to the Judet Branch of the National Bank. This would normally take a further day so that the transaction would take about two days to complete in total. -Chapter 2: Strengthening the Institutions for Budget Management 35 each day managed by the Romanian National Bank that represents the consolidated position of the Romania Government. 2.38 The consolidation of these balances is an important part of the system. Through its network the Treasury in the Ministry in Public Finance has a daily estimate of the cash position in each Judet Treasury. These balances are then transmitted electronically to the Central Treasury where they are collated and consolidated in one balance at the end of the day. A parallel process is operated through the Romanian National Bank. Each Judet Treasury holds an account with the Judet Branch of the Romanian National Bank. Payable Orders are sent by local Treasuries to the Judet treasuries and then sent to the National Bank for processing through the clearing system. These are debited in the Account of the Treasury. Each day the Romanian National Bank does a sweep of all these Judet level accounts to arrive at a single consolidated balance. This balance should correspond to the balance estimated at the Central Treasury Account. 2.39 While the Central Treasury Account should ensure that there are no unconsolidated cash balances, the management of these balances by the National Bank does not occur on a daily basis. For instance, during the months of October and November 2001 there were no placements on the over-night market. Thus, while, in theory, this centralized system should provide the opportunity to earn a return on unused funds, there is little evidence that this is actually taking place. 2.40 Accounting Information. The other shortcoming of the current process of budget execution in Romania, and its reliance on cash transactions is that there is no monitoring of expenditure commitments. Although the Treasury system in Romania is the principal source of information for budgetary execution and supervision, it relies primarily on monthly reports based on cash estimates of receipts and expenditures. These reports show the total level of expenditure for the ministries and spending agencies by chapter or expenditure heading. While this enables the Ministry for Public Finance to monitor the progression of the budget against the projected profile, it provides no information on unpaid invoices that are due. As a result, there is substantial accumulation of arrears during budget execution. 2.41 The information on budget execution functions in the following manner. The Line Ministries collect monthly information based on payable orders issued rather than payments actually made. When a payable order is issued by an agency this is accounted for in their accounting system. Each agency is responsible for keeping a record of expenditure incurred. This information is transmitted manually to the Line Ministry that collects information on its Secondary and Tertiary agencies. Based on this information monthly reports are sent to the Ministry for Public Finance. This information complements the information produced by the Treasury system but is more detailed. 2.42 Quarterly accounts are also produced by Line Ministries that are sent manually to the Ministry for Public Finance. This provides detailed information on expenditure incurred. It also contains some data on commitments and expenditures due. The system does not provide, however, information on unpaid invoices that are due.. This reflects one of the main deficiencies of the system: the lack of accounting information. Information on expenditures incurred is 36 Chapter 2: Strengthening the Institutions for Budget Management available and presents the key element for budget execution. However, the systems do not collect information on receivables or commitments. This is a particular issue in relation to commitments, as expenditure ceilings that are lower than those approved in the budget lead to payment arrears. There are reports that local authorities, at present not subject to ex ante control of commitments from the Ministry of Finance, have accumulated arrears on the payment of centrally mandated social benefits. Also, there are reports of substantial payment arrears at the level of hospital and other health care facilities. Finally, at the central government level arrears can arise, despite the existence of ex ante controls, when price increases that were neither anticipated in framework contracts, nor captured in a budget rectification give rise to higher invoices (e.g. for contracted utility services). While the following year's budget should cover payment arrears, this is not always done, as the discussion in the next chapter on hospital arrears illustrates. 2.43 For budgetary planning and financial cash planning, an accounting system that provides information on arrears is essential. The core problem appears to be that there are delays in the system due to the manual transfer of information. Improved networking between the Line Ministries and the Ministry for Public Finance should help in this respect, as should also improved links between Line Ministries and Secondary and Tertiary agencies. Improved networking could also free up scarce resources to speed up the recording of transactions and improve basic accounting practices. There is, therefore, clearly a merit in developing a more integrated accounting system that better accounts for arrears. At a very minimum, Treasury and Ministry accounting should be integrated. 2.44 Procurement. The need to improve accounting information is heightened by recent changes in public procurement aimed at bringing Romanian procurement practices closer to the EU. In September 2001, a new law on public sector procurement was adopted by Parliament. The law follows EU directives on procurement and adopts World Bank procurement documentation. The procurement orders are made through the Ministry for Public Works. Ensuring that the new law achieves its stated objective requires, however, regular reporting25 by the contracting authorities on the contracts awarded under the public procurement ordinance. These reports would provide clear information on the amounts involved in these contracts, the number of bids, the technical specification and standards of these contracts, information on the notice for competition, and a brief evaluation. of the contract awarded. The latter should allow the Government to minimize the practice of splitting contracts to avoid the competition thresholds defined in the public procurement law. Accurate monitoring of the amounts involved in these contracts requires, however, that expenditures be accounted on a commitment basis. This in turn requires the development of more timely accounting information. Conclusions and Recommendations 2.45 Timely Budget Proposals. The first recommendation for improving budget execution is submitting timely budget proposals. A timely and well-prepared budget is essential to good budget execution. Unrealistic revenue assumptions, fundamental policy changes within the fiscal year, and changes in the macroeconomic outlook, require budget revisions, complicate 2 Preferably on a quarterly basis. Chapter 2: Strengthening the Institutions for Budget Management 37 expenditure monitoring and can severely disrupt budget execution. The same applies to late approval of the budget, particularly when the inflation rate is high.26 While all these problems have been a regular feature of budget preparation in Romania, the fact that the 2002 budget was prepared on time and that there has been only one budget revision in 2001compared to four in 2000 point to a significant improvement in budget management. 2.46 Avoiding Payment Arrears. The frequent occurrences of payment arrears are one of the least desirable features of Romania's budget execution. There are reports that local authorities, at present not subject to ex ante control of commitments, have accumulated arrears in the delivery of centrally mandated social benefits, and that substantial arrears have arisen in payments to hospitals suppliers.27 There are also reports that even at central government level arrears can arise, despite the existence of ex ante controls. These happen when price increases that were neither anticipated in framework contracts, nor captured in a budget rectification, give rise to higher invoices (e.g. for contracted utility services). While such payment arrears should be covered by next year's budget, this is not always done, as the example of the arrears to hospital -suppliers shows. 2.47 There are several measures that the Romanian authorities should consider when addressing the problem of payment arrears. In the immediate short run there is a need to establish a central monitoring system for payables and issuing monthly reports on payment arrears would increase accountability and minimize the risk of accumulating larger arrears. In the medium term there is a need to develop a more modem treasury system. 2.48 This will require: > Developing an electronic transfer of information. The current reliance on manual transfers of information at all stages in the process, and at all levels, is leading to delays in the consolidation of balances. This implies there is cash in the system that is not being utilized as efficiently as possible. Delays in the flow of information also hinder the efficient recording of budgetary transactions. This reduces the effectiveness of the budgetary control and financial management. Developing an electronic transfer information could begin with the simple exchange of computer diskettes with data inputted on standard forms. Moving toward an integrated management information system will require considerable investment in networking, computerization and modem communication systems. Although some of these investments are already being considered, priority should be given to making better use of the information already available, and putting it together in a 26 At present late budget submissions imply that spending agency's monthly expenditures are restricted to 1/12 of previous year's budget. This has in the past played havoc with its expenditure plans. Expenditures were bunched in the latter part of the year and lost value in real terms due to the effect of substantial price increases. As regards the impact of budget revisions, actual expenditures have consistently been less than the final revised budget, even when the revised budget total was less than the original budget. Available data do not show any consistent effect of the revisions on expenditures of individual agencies. 27 See Chapter 5 28 Currently, the Ministry of Finance monitors monthly a number of 86 public companies with large arrears to the budget, based on a list approved through Government Decision no. 866/2001. However the number of companies with arrears to the budget is much larger, and includes many private firms as well. 38 Chapter 2: Strengthening the Institutions for Budget Management form that would yield better and faster decision-making. This last step should proceed as a matter of priority since it would contribute to on-going cost savings measures in the execution of the budget. > Developing integrated and computerized accounting systems. There is considerable scope for developing an accounting system that provides better accounting of commitments, Payable Orders and other relevant information over the longer run. The lack of timely data makes cash flow forecasting difficult and affects the overall execution of the Budget. Integration of Treasury and Ministry accounting is another area for improvement. This could be made possible by better networking between the Minister for Finance and Line Ministries. Improved links, while enhancing the functioning of the current system, also provide a necessary step in developing a more integrated information management system. In the longer-term, consideration could be given to the development of an Integrated Treasury Financial and Management System. This system has been implemented in a number of countries. 2.49 These steps toward the development of a modem treasury system that allows for the speedy transfer of information and better accounting, should reduces the cost of collecting taxes and executing the budget. This in turn will entail using modem payments and increasing reliance in the banking systems. In doing so, the Treasury should minimize the costs associated with unused cash balances. Given the constraints on resources it is useful to establish clear priorities for action. The first priority should the more active management of unused cash balances. This could lead to an actual cash benefit to the Budget. The second priority is to invest in the development of a computerized inf6rmation system. This would require significant investment over a period of time but planning the development of such a system should commence now. The first step would involve improving the networking between the different elements within the current system and building a more integrated accounting systems. 2.50 The most effective way of strengthening budget execution is to learn from it. While an important first step in this leaming process is building a modem treasury system that can provide faster and more accurate information, internalizing the learning process requires developing the institutions of budget compliance. Therefore, we turn next to these institutions. Developing the Institutions of Budget Compliance and Evaluation 2.51 Three main functions should be carried out at the compliance stage of the budget cycle. First, internal control, including internal audit, should be performed by the ministries and spending agencies themselves, to check the legality of disbursements, and internal ex-post audit should be carried out by the Ministry of Public Finance. The Ministry of Public Finance should provide methodological guidance to these activities, with the spending agencies being responsible for the effective operation of the control system. Second, external ex-post audit should be carried out by an institution that is independent of the executive branch of government to verify the legality and correct use of budgetary funds. This is a fundamental component of good budget systems since the executive branch of government cannot be expected to be strict and objective about its own operations. Third, the evaluation of the performance of budget Chapter 2: Strengthening the Institutions for Budget Management 39 programs should be carried out by several institutions, with emphasis on improving the effectiveness of government expenditures. These three functions of budgetary compliance are complementary and reinforcing. 2.52 How does budget compliance work in Romania? In the Romanian case, in addition to the ex ante control exercised by the Directorate for Financial Control in the Ministry of Public Finance, two other systems of ex post control of the Government's operations exist: internal audit carried out by every spending agency and co-ordinated by the Ministry of Public Finance and external audit performed by the Court of Audit.29 The internal audit function became effective in 2000 (Government Ordinance 119/199930) and its implementation is the responsibility of the General Directorate for Internal Audit and Control (IAC) in the Ministry of Public Finance. Internal audit is not only concerned with financial control of budget execution and detection of fraud and misuse of funds, but also involves an assessment of the effectiveness and efficiency of a unit's operations. Internal auditors are considered 'consultants' to management. To ensure their independence, internal auditors report directly to top managers, and the Court of Audit evaluates their internal audit reports. 2.53 The General Directorate for Internal Audit and Control (IAC) in the Ministry of Public Finance covers all spending agencies at central and local government level as well as the special funds and state enterprises. Each agency employing at least 25 people has an internal audit unit. Smaller. agencies use private accredited auditors. An audit plan is drawn up for each agency. There is a hierarchy in the system, which requires auditors at the higher levels (primary credit ordinators) to supervise those lower down. For instance, the internal audit unit in the Ministry of Health oversees and evaluates audit in hospitals and district health departments and reports on any problems or weaknesses it finds. 2.54 The intention of the Ministry of Public Finance is to move increasingly towards performance audit, giving greater emphasis to the 'consultant' function of auditors. This will require developing performance measures and benchmarks to compare performance of units. Some Ministries have started to do this (e.g. the Ministry of Health), with indicators being developed for hospitals measuring both economic and medical efficiency. The Ministry of Public Finance also plans to establish a separate Harmonization Unit in the Ministry, which would regulate internal audit activities throughout the public sector. This unit, governed by a board that would include outside experts, would, amongst others, (i) develop guidelines on audit methodology, using international standards; (ii) resolve conflicts of interest between managers and auditors; and (iii) follow up on recommendations in audit reports to ensure their implementation. 2.55 Within the General Directorate for Internal Audit and Control there also is an inspection unit, which focuses on inappropriate uses of funds, often following up on complaints. It can ask for a specific audit to be performed. Its report is presented to the head of the agency where the inspection was performed. 29 Previously known as the Court of Accounts. 30 Approved later through Law 301/2002. 31 Ministry of Public Finance, "Policy Paper on Public Internal Financial Control," October, 2001. 40 Chapter 2: Strengthening the Institutions for Budget Management 2.56 Court of Audit. The Court of Audit, previously known as the Court of Accounts, is the external auditor of the Government's financial operations. New legislation (Law 204/1999 and Law 77/2002) introduced substantial reform of its operations and responsibilities. The Court audits the accounts of budget execution and reports on this to Parliament. Until the beginning of 2000, when the Preventive Fiscal Control unit was transferred to the Ministry of Public Finance, it was also responsible for ex ante expenditure control. To assure its independence, the budget of the Court of Audit is submitted directly to, and approved by, Parliament. A Parliamentary Committee examines the budget execution of the Court of Audit. At the county level, external audit is performed by the county Chamber of Audit, which is part of the Court of Audit. 2.57 Having received from the Ministry of Public Finance the budget execution accounts within six months of the end of the fiscal year, the Court has to submit to Parliament within the next six months an annual report on overall budget execution as well as on that of individual spending agencies. The President of the Court of Audit can be questioned by Parliament on the annual report. The Court has 18 membersn, or counselors, and a staff at present of just over 1400, sharply down from around 2200 in 1999 following cuts imposed by Parliament. A reduction in the number of counselors to 18 is also being considered by Parliament. 2.58 To complement their traditional function of ultimate financial controller, the Court of Audit can also act as a court. Following the examination of an agency's accounts by a panel of three counselors, any irregularities may be referred to the Jurisdictional Board in the case of a civil offense or to the Financial Prosecutor General for criminal offenses. Around 70% of the accounts audited are cleared. In the remaining 30% of cases, the bulk of offenses are of a civil nature. Appeal to decisions by the Jurisdictional Board is possible all the way to the Supreme Court. 2.59 The new law gives the Court of Audit responsibility for auditing on a broad basis, including on the effectiveness and efficiency with which public services are being provided. The Court is stressing the importance of performance auditing. With the introduction of a program structure in the 2002 budget, the Court has prepared a report on the methodology for auditing program budgets, which was submitted to Parliament. 2.60 Recent Reforms in Budget Auditing. Significant reforms have been introduced in recent years in the area of financial control and audit. In 1999, new laws were adopted which introduced three major changes: ex ante control of expenditure was separated from the ex post control exercised by the Court of Accounts and transferred to the Ministry of Public Finance; a system of internal audit was established in the Ministry of Public Finance; and the operation of the Court of Accounts was modernized. One result of these reforms has been to open up the way to moving from the heavy emphasis on financial control in the past towards a broader system of control that aims at increasing the effectiveness and efficiency of government operations. 2.61 Romania has been able to keep a tight rein on its finances in terms of maintaining aggregate fiscal discipline while its system of ex ante control has been effective in ensuring that commitments were in line with budget allocations. With the establishment of the internal audit function and the reform in external audit, the Government is taking another step toward moving 32 Law 77/2002 provided for the reduction of the number of members from 25 to 18. Chapter 2: Strengthening the Institutions for Budget Management 41 away from a system that relies on adherence to regulations and procedures towards a more results oriented approach that emphasizes managerial accountability. A similar evolution can be seen in the area of budget preparation, where a program structure of expenditures and the formulation of sector strategies were introduced for the 2002 budget as the beginning of a greater focus on performance. 2.62 These reforms reflect the wish to adopt international, and in particular EU, standards of control and audit. To achieve this goal, however, further steps are needed, especially in terms of institutional strengthening and the development of relevant skills. This includes establishing an appropriate legal basis for internal audit, which currently is based on an ordinance rather than a law. EU standards and concepts have also not yet been fully adopted. 2.63 In addition to setting up the Harmonization Unit in the Ministry of Public Finance, the law should spell out more clearly the role of internal audit, particularly the auditor's role as an advisor to management and the focus of the audit on improving performance. The law should also clarify the appropriate balance between the ex ante control exercised by the Directorate for Financial Control and by the General Directorate for Internal Audit and Control (IAC), which at present is still heavily tilted towards the former. Administrative procedures associated with ex ante control can be a hurdle in achieving efficiency improvements. 2.64 The growing emphasis on using both internal and external audit as a means of improving the efficiency of public institutions that goes beyond checking compliance with the budget and various control mechanisms has implications for staff skills. The skills required of traditional auditors are different from those of management consultants. A considerable investment in staff resources and in data collection and monitoring capacity will be required through recruitment, training programs as well as possibly visits by staff to similar institutions in EU countries to learn from their experience. The sharp cut in staff of the Court of Audit is a concern in view of the broader responsibilities the Court was given in the new law. To allow it to better deal with its new responsibilities within its staffing constraints, the Court might consider involving the private sector in the auditing of state enterprises. Recommendations 2.65 The three functions of budgetary compliance (control, audit, and evaluation) are complementary and mutually reinforcing. Actions to strengthen them should therefore be implemented in parallel. These actions should create an environment where there is continuous leaming from budget formulation and execution, aiming at improving the effectiveness of government expenditures. This can be achieved by improving the legal framework under which these activities are carried out and providing units responsible for audit, control and evaluation with qualified staff. The following paragraphs examine specific actions to be carried out for each institution. 2.66 Internal Audit and Control. To strengthen the internal audit function now carried out by the Internal Audit Unit of the Ministry of Public Finance action is required on at least six fronts. There is a need to: 42 Chapter 2: Strengthening the Institutions for Budget Management > Change the legal basis for internal audit to bring it in line with the international control and audit principles and standards. > Spell out a clearer distinction between ex ante control, and the internal audit unit, where the former aims at ensuring that expenditure commitments are in line with budget allocations, while the latter attempts to ascertain whether actual expenditure complies with the budget and that public resources are efficiently spent. > Create a separate Harmonization Unit in the Ministry of Public Finance to regulate internal audit activities throughout the public sector. This unit, governed by a board that includes outside experts, would, among other things, (i) complete the harmonization with EU standards and definitions, developing guidelines on audit methodology that are consistent with international standards; (ii) resolve conflicts of interest between managers and auditors; and (iii) follow up on recommendations in audit reports to ensure their implementation. > Assist internal auditors by defining clearer rules and regulations for the conduct of public officials in the handling of public funds, characterizing conflict of interests, and imposing administrative sanctions. > Increase the number of qualified staff of the Internal Audit Unit by providing training, and developing further manuals and procedures for internal audit. This would include providing training and resources for data collection and for establishing a monitoring capacity for budget programming. 2.67 External Audit. The Court of Audit's needs to move away from relying on management consultant work, hiring traditional auditors and providing internal and external training programs for well-qualified professions. This need will be heightened by the move toward program auditing, in line with program budgeting, which will require considerable investment in staff resources and in data collection and monitoring capacity. This can be achieved through additional recruitment, training programs as well as possibly visits by staff to similar institutions in EU countries to learn from their experience. Finally, additional resources for the more frequent audit of budgetary institutions could be mobilized if the responsibility for auditing some state enterprises is contracted out to private accounting and audit firms. 2.68 . Budget Evaluation. A budget evaluation function should be developed in the Ministry of Public Finance and in the Court of Audit. The Ministry of Public Finance evaluation unit should specialize in evaluations that are carried out within the budget year, aiming at providing information that is easily fed into future budget decisions, while the Court of Audit would assess performance against strategic objectives. Effective evaluation should benefit from the move toward program budgeting, where expenditure allocations are associated with explicitly stated outcomes. Program budgetting requires each spending unit to develop performance indicators that are submitted along with budget proposals. This evaluation function would feed into broader reforms of expenditure programs and public sector management. Chapter 2: Strengthening the Institutions for Budget Management 43 Conclusions and Recommendations 2.69 This chapter provided a snapshot of the reforms in budget management currently being undertaken, and discusses important next steps. The main conclusions are the following: > Setting Government Priorities. The Government, in particular the Office of the Prime Minister, should make it a tradition to issue and publish in an accessible way a 'Government Policy Priorities' document that outlines the major policy thrusts of government policy for the coming budget year. This document should become the main vehicle for setting boundary lines between the core public sector and all other activities in the public sector that should be privatized, contracted out or discontinued. The initial allocation of expenditure ceilings by the Ministry of Public Finance to line ministries in the budget circular should reflect these explicitly stated priorities of Government. This statement of overall policy priorities would allow the government to move beyond the item-by-item budget proposals and parliamentary appropriations and, instead, move to using the budget as the overall instrument of economic policy including the management of the macro-economy. > Treasury Operations. Action is required in two fronts to strength the Treasury's ability to better execute the budget: strengthening the management of budget information, and improving the accounting system. Strengthening the management of budget information requires tracking of revenue flows, bank balances, and expenditure commitments on a more timely basis. Delays in information flows reduce the effectiveness of the Treasury in financial management and control. Improving the accounting information system will also allow the Treasury to maintain proper records of all transactions (receipts and disbursements) at the time they take place, registering them in a single ledger. > Budget Evaluation. A budget evaluation function should be developed in the Ministry of Public Finance and in the Court of Audit. The Ministry of Public Finance evaluation unit should specialize in evaluations that are carried out in a shorter period of time and include information that can be easily fed into future budget decisions, while the Court of Audit would assess performance against strategic objectives. Effective evaluation should benefit from the move toward program budgeting, where expenditure allocations are associated with explicitly stated outcomes. Program budgetting requires each spending unit to develop performance indicators that are submitted along with budget proposals. This evaluation function would feed into broader reforms of expenditure programs and public sector management. 2.70 These steps to strengthen the institutions for budget management aim at improving expenditure prioritization and breaking expenditure inertia. They will also support the Romanian Government's program of fiscal decentralization, both to central spending agencies and to local administrations. Fiscal decentralization should achieve the second objective associated with better budget management: increasing the effectiveness and the efficiency of public spending, 44 Chapter 2: Strengthening the Institutions for Budget Management since this will allow Romania to move away from the present system of strong reliance on management controls, towards a system that gives managers at all levels increased responsibility for the way public resources are used and holds them accountable for results. 2.71 At present, there are two main obstacles to effectively decentralizing expenditure decisions. There are difficulties involved in measuring performance and hence in holding managers accountable for results. Also, there is limited administrative and analytical capacity at agency and local level. Further reforms in the budget process should be accompanied, therefore, by efforts, possibly in the form of workshops or seminars, to prepare senior officials in spending agencies for the different demands that will be made on managers when a system that relies primarily on management controls makes way for one that emphasizes accountability. Measuring performance also requires data collection on a regular basis and the creation of monitoring capacity. The report turns next therefore to these and other challenges of fiscal decentralization. CHAPTER 3: IMPLEMENTING FISCAL DECENTRALIZATION REFORMS Introduction 3.1 A message that emerges from the previous chapter is that budget reforms are designed to move the public sector away from the present system of reliance on management controls, towards giving managers at all levels increased responsibility for public resources and holding them accountable for results. The fiscal decentralization may be one of the most important steps in this transformation of the budget process. It seeks to improve governance and public service delivery by increasing allocative efficiency through better matching of public services to local needs and productive efficiency through increased accountability of local governments to citizens. These steps should contribute to the Government's twin objectives of keeping overall expenditures within the agreed budget targets without compromising quality in the delivery of public services. 3.2 Achieving the objectives of the fiscal decentralization reform will require close attention to implementation at each of the three stages of budget management at the local government level: budget design, budget implementation, and budget execution. > At the budget design stage, the system of intergovernmental finance has been in flux for the last several years, with constantly evolving procedures for revenue and expenditure assignments. There is particularly a problem with expenditure assignments. The Law on Local Public Administration adopted in 1991 has been amended33, and local government's functional expenditure competencies have been replaced with a list of "tasks" that provides the authority to act. The list is revised each year through an Annex to the Annual Budget Law,34 in a manner that, in the past, provided little stability in expenditure assignments and created uncertainty for financial management and budget practices for local governments3 > At the budget execution stage, the central government's focus on cash expenditures has meant that budget targets have sometimes been met through payment arrears. Arrears are more prevalent at the local government level, where there is no formal ex ante control over expenditures. Also, local governments have less discretion over their expenditures, many of which are centrally mandated, leading them to accumulate arrears as a way to cope with unfunded or under-funded mandates which are particularly acute in rural areas, where the low population density results in higher unit costs of service delivery and own-revenue generating capacity tends to be low. 1 Law no. 215/2001. 34 The problems in the assignment of expenditure responsibilities are further compounded by implementation practices. It is common to assign responsibility to local governments for specific types of expenditures (salaries of social workers assisting the handicapped or education investments) or for specific existing public institutions (centers that care for handicapped children or homes for the aged). There is no sense of a broad local authority to identify local needs and design and implement local strategies to address them. 3 A new Local Public Finances Law, currently under preparation, should address some of these aspects. 46 Chapter 3: Implementing Fiscal Decentralization Reforms > At the budget compliance stage, these shortcomings of the budget process translate into low levels of financial and operational accountability. Because financial accounting and reporting is weak, not every local government can show that funds are raised and spent according to the legal requirements. This precludes them from showing that they are providing services efficiently and effectively, which undermines the achievement of one of the objectives of fiscal decentralization: matching public services to local needs through increased accountability of local governments to citizens. 3.3 The result is that fiscal decentralization presents both opportunities and risks. The opportunities emerge from better matching public services to local needs. The risks are the potential disruptions in the delivery of key services. The challenge is to develop a stable and transparent system of local government financing that matches new expenditure responsibilities. This is particularly evident in the context of the recent decentralization of pre-university education and the renewed responsibility of local governments for the delivery of social assistance through the Minimum Income Guarantee (MIG) program. This chapter focuses on understanding recent developments, reviewing changes in revenue assignments and expenditure responsibilities, and looking at opportunities to strengthen local government's financial and operational accountability. 3.4 The chapter is organized as follows. It begins by placing the decentralization efforts in context by describing the structure and expenditure responsibilities of local governments in Romania. Next, it examines the implementation of fiscal decentralization reforms on the three stages of the budget process: design, implementation and compliance. The second to last section discusses the most recent experience of decentralization in practice: the transfer of the responsibility for pre-university education to local governments. The last section provides a summary of the main findings and recommendations. Structure and Expenditure Responsibilities of Local Governments in Romania 3.5 There are two tiers of local government in Romania. The first includes 2,948 local councils. Local councils include 263 municipalities (municipiu) and towns (oras), and 2,688 communes (comuna) which may include one or more villages or hamlets. Municipality, town and commune are names whose significance is historical and geographic, not legal. Their authority and structure is the same. The second tier of local government consists of 41 county councils (judet). There is no real or presumed hierarchical relationship between these two tiers of local government. They are equal under the law. The geographic limits of local councils are 36 This structure of local governments in Romania reflects principles set out in the Romanian Constitution (1991), which recognizes the principles of local autonomy and decentralization of public services. These principles are elaborated by the Law on Local Public Administration, which provides for the organization of local government on geographically defined administrative subdivisions. Chapter 3: Implementing Fiscal Decentralization Reforms 47 such that they cover the entire territory of Romania. The limits of the county councils overlap with those of the local councils in their area.37 3.6 All local councils (municipalities, towns and communes) are established democratically, with members elected to four-year terms by direct, universal suffrage.38 The number of local councilors is based on the population of the local unit, and varies from 11 to 65, with Bucharest at the higher end of the distribution. Executive authority is held by the mayor, who is directly elected to a four-year term, while the council appoints the deputy mayor. County councils are also elected to a four-year term by direct, universal suffrage. The size of the county council ranges from 37 to 45 members, based on the population of the county. An executive body, consisting of a chair, two vice chairs and four to six councilors, leads the county council. The executive authority of the county council is held by the council chair. 3.7 Population and distribution. Local units range in size from 2 million in Bucharest to an average of just under 4,000 in local communes (Table 3.1). Bucharest municipality has a special status, functioning both as a municipality and a county, and accounting for nine percent of the population. The rest of the population is distributed as follows: 36 percent in the other 83 municipalities, 10 percent in the towns and 45 percent in communes. This distribution would suggest that there is considerable scope for cooperation between the large number of local councils. This would aim at striking a balance between the advantages of a smaller size, such as greater political accountability and overall efficiency, with the benefits that come from cooperation, including economies of scale in the delivery of services, better trained and qualified staff, and easier coordination with the central government. Table 3.1: Structure and Population of Local Governments Number Share of Total Population Share of Total Average Population County Councils 41 22,533,156 549,589 All local councils 2,951 100% 22,533,156 7,636 Bucharest 1 2,027,512 9% 2,027,512 Municipalities 83 3% 8,032,804 36% 96,781 Towns 179 6% 2,320,375 10% 12,963 Communes 2,688 91% 10,152,465 45% 3,777 Source: Ministry of Public Administration. 3.8 Structure of local government expenditure responsibilities. Local government expenditure responsibilities include: local administration; education; social assistance; public " Throughout this chapter we use the term local councils to refer collectively to the first tier, that is, the municipalities, towns and communes. We refer to the second tier as counties or county councils. Finally, the term local governments refer to both local and county councils together. 31 While county and local councils are democratically elected, there is limited participation and community involvement in the budget process. The county councils are the least likely to engage in any kind of consultation, such as hearings, much less in a process of seeking community views on investment needs and priorities. Perhaps this is a function of the large population they represent, on average about 500,000. However, the county councils could make greater use of the knowledge of local needs and priorities of the local councils and of the NGOs and other organizations of civil society. In the local councils there is not all that much continuing involvement by local residents either. In general, mayors tend to be the dominant figure, with the council members often playing a secondary role. The problem in the smaller local councils may be the inverse of that faced by the county councils. In this case the size of population may be so small, in some cases less than 1,000, that it is not evident to the elected officials that there is a need for a formal, structured process of community outreach and participation. 48 Chapter 3: Implementing Fiscal Decentralization Reforms works and housing; and transport and communication. (Table 3.2). The distribution of these responsibilities between the basic level local governments (municipalities, towns and communes) and the counties differ considerably, however. The counties generally are responsible for child protection and care of handicapped children, county public services and county roads, as well as county cultural institutions (museums, libraries, etc.). Municipalities, towns and communes carry out more direct service provision (urban transport, sanitation, public utilities); maintain roads, sidewalks, parks, and green areas; and are responsible for education, other social assistance (such as the minimum income guarantee program), local culture and sports services, local economic development and markets. Table 3.2: Local Governments Expenditure Responsibilities Expenditures Responsibilities Public administration Responsible for operating the local government itself, but not other public institutions subordinated to the local government. Education and culture All pre-university education. Budget support for a wide variety of cultural and sports organizations, such as museums, libraries, operas and houses of culture. Social assistance Counties are responsible for child protection services. Local councils are responsible for partially funding and delivering the transfers under the minimua income guarantee progra, as well as for budget support to social institutions, such as those that provide meals and take care of the aged. Transport and Counties are responsible for county roads, and local councils are responsible for communication local road. Local councils also provide budget support to local public transport enterprises and regional airports. Public works and Local governments own and provide support to public utility companies (water, housing heating). They also support the operation of public markets, and are responsible for street cleaning and repair, housing and solid waste collection and disposal. Source: Law on Local Public Finance (1997); 2002 State Budget Law. 3.9 Trends in local government expenditures. While there has been very little discernable trend in aggregate expenditures at the local government level throughout the 1990s, there have been significant changes in structure of expenditures, revealing the persistent instability of the expenditure assigned to local governments (Table 3.3). For instance, public works and transport accounted for over 80 percent of total expenditures in 1991 and 1992. Starting from 1993, local governments were given expenditure responsibility for health care and some social assistance, with these accounting for close to 20 percent of local expenditures in 1993 and 1994. Local responsibility for health care spending lasted only until 1997, and was taken over by the health insurance system in 1998. In the meantime, education responsibilities were given to the local level in 1995, accounting for 10 to 11 percent of total expenditures from 1995 to 1997. Then education as a share of total expenditures declined to about 8 to 9 percent from 1998 to 2000 due to the addition of other responsibilities. When the responsibility for all pre-university education was given to local governments in the fall of 2001, the share of education in total expenditures rose again, reaching 36 percent of total expenditure in 2001, up from an average ofjust under 10 percent in the 1995 to 2000 period. Chapter 3: Implementing Fiscal Decentralization Reforms 49 Table 3.3: Structure of Local Government Expenditures, 1991-2002 (%) 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001P Public administration 10.0 8.7 7.9 8.5 7.3 6.9 7.9 13.4 14.0 14.4 9.7 Education 0.0 0.0 0.0 0.0 10.4 11.9 10.0 9.0 8.9 7.8 36.6 Health 0.0 0.0 14.6 15.2 12.5 12.4 12.8 0.5 0.4 0.3 0.4 Social assistance 0.1 0.1 4.4 4.5 6.0 6.3 4.6 9.3 8.0 11.5 9.4 Public works and housing 48.2 58.8 45.7 42.6 38.9 37.6 39.0 41.3 40.4 36.0 25.5 rransport 32.0 26.3 22.8 24.2 19.9 19.9 19.6 20.3 12.3 11.7 8.1 Other 9.7 6.1 4.6 5.0 5.0 5.0 6.1 6.2 6.5 8.0 4.5 Special earmarked expenditures 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.5 10.3 5.9 Total expenditures (% of GDP) 4.5 4.1 3.6 3.9 4.2 6.1 p) Preliminary. Source: Ministry of Public Finance, Urban Institute, and World Bank. 3.10 Aggregate measures of local government expenditure, while convenient, conceal a lot of useful information on the real conditions of nearly 3,000 individual local governments. Table 3.4 shows the distribution of local government expenditure across the different types of local governments. There are several interesting patterns. County councils allocate a significant share of their funding to social assistance and culture, especially compared with public works and transport, which account for the greater part of the allocation for all local councils. Within local councils, Bucharest dedicates a higher share of expenditure to transport than other local councils, while local communes spend comparatively less on public works and more on public administration. Indeed, the smaller the unit size, the greater is the share spent on public administration. Table 3.4: Share of Local Government Expenditures by Type of Government, 2000 (%) Public Social Public Administration Education Culture Assistance Works Transport Other Total Counties 10.3 0.0 10.6 24.1 9.6 15.1 30.3 100.0 All local councils 15.0 9.5 2.9 8.1 41.5 10.5 12.5 100.0 Municipalities 9.9 12.3 3.2 6.2 52.4 6.4 9.6 100.0 Bucharest 4.0 4.3 1.7 6.1 46.8 25.9 11.2 100.0 Towns 20.9 11.1 3.9 10.1 38.8 1.0 14.2 100.0 Communes 35.1 9.5 3.5 13.2 15.7 4.0 19.0 100.0 Source: Ministry of Public Administration. 3.11 Is this distribution of expenditures fair? This pattern of expenditures distribution does not always match the distiibution of the population at the local government level. For instance, the municipality of Bucharest, with nine percent of the country's population, accounts for almost 50 percent of all transport-related expenditure, although the latter includes in the case of Bucharest the urban transport system as well as the cost of construction and maintenance of the city's road network. Also, municipalities account for 55 percent of education expenditures, but only for 46 percent of the students. The expenditure disparity at the local government level is even greater in the case of social assistance expenditures and transport. The factors driving these disparities appear to be very different. In the case of social assistance, the needs are fairly evenly 50 Chapter 3: Implementing Fiscal Decentralization Reforms distributed among all local councils. However, as communes have lower per capita revenues on average than municipalities and towns, they have less to spend on these needs. In the case of transport, as mentioned above, the expenditures in Bucharest and other larger jurisdictions include public transport, as well as road maintenance. In the case of the communes, the category only includes maintenance of roads. 3.12 Where are fiscal decentralization reforms heading and what is their impact in the local budget process? Fiscal decentralization in Romania began in earnest in 1998 with the new Law on Local Public Finance.39 The new law placed the local budget process on an equal legislative basis with the state budget. This law aimed at strengthening local fiscal autonomy by clarifying and expanding local control over revenues, expenditures and the budget process. The progress in meeting this objective has been partial at best. The rules on revenue sharing have gone through three amendments. The 1999 state budget law reintroduced dedicated transfers, and the 2002 state budget law added transfers for education, roads, housing and the Minimum Income Guarantee program. All these changes have eroded some of the benefits of the new local finance legislation. It is however the local budget process that continues to have the biggest adverse impact on the autonomy of local and county councils. The next three sections turn therefore to impact of the Local Public Finance Law, and of the changes that followed its adoption on the local budget process. Budget Formulation 3.13 The local budget process begins roughly in June or July of the preceding year with a draft budget that includes estimates of revenues from local taxes and fees and shared tax revenues and a proposal for expenditures for the next year. The individual budget of each local council is consolidated in a single county-level budget that also includes the draft budget of the county council that is forwarded to the Ministry of Public Finance. After reviewing all budgets of local and county councils, the Ministry gives each county and local council a spending limit that they use in finalizing the draft budget for the next year. If at the beginning of the fiscal year the state budget still has, not been approved, the local governments receive authorization from the Ministry of Public Finance to spend each month one-twelfth of their total actual expenditures for the prior year. 3.14 This last feature of the budget process was a particular problem up to 2002, since for most of the 1990s, the state budget was often approved as late as May or June of a fiscal year that runs from January to December. This implied having to adjust the draft budget to the rules set forth in the state budget law. This was a problem because even after 1998 the state budget included changes in the expenditure responsibilities of both local and county councils. It also has included variations in the volume and type of transfers, both general and earmarked. Finally, local councils have to wait for the county councils to allocate to each one of them their corresponding share of the equalization grants received from the state budget. This allocation process follows different rules in each of the 41 counties. While the usual practice involves some quantitative criteria, the process is not based solely on a formula. As a result, local councils never know for sure what amount they will receive from the county council. The 39Law on Local Public Finances, No. 189 of 1998. Chapter 3: Implementing Fiscal Decentralization Reforms 51 process typically involves significant individual communication between the local councils and the county council. 3.15 The allocation of equalization grants emerges as an obstacle for sound management at the local level already at the budget formulation stage. Since each local government's share in the equalization grant is not known, there is no hard budget constraint that compels them to make difficult spending priority decisions within a given level of revenues. Until the state budget has been adopted by Parliament and the equalization grants allocated by the county councils, they do not know with any certainty the full extent of their revenues or of their expenditure responsibilities. 3.16 The adverse impact of this revenue uncertainty is greater in the case of those local councils that are more dependent on the transfers from the state budget. They are much more at the mercy of the changes adopted in the state budget and of the decisions of the county councils on the allocation of equalization grants. It is not surprising therefore that elected officials in many localities-the mayor and the members of the local council-have adopted largely a passive attitude regarding the budget process. Their most important skill is the ability to negotiate with the county council to secure the highest possible equalization grant allocation. 3.17 Why is the allocation of equalization grants important? To the extent feasible, before preparing their budget local and county councils must be able to estimate what level of equalization grants to expect in the coming year. The variation from these estimates should be a function in the size of the overall pool of funds for equalization authorized in the state budget for the specific county, not in the relative share of the pool allocated to each commune. Fixing each local government share in the equalization grant would help create a hard budget constraint, obliging local councils focus on setting spending priorities rather than on negotiating for additional revenues. This would also make local councils more clearly accountable to the local community for the spending priorities included in the budget. 3.18 How does the process of allocating the equalization grant work at present? The process of allocation of the equalization grants begins with the state budget preparation. This consists of three separate steps. In the first step, the budget allocates a sum to each county. Of this sum, the budget allocates one portion to the "own" budget of the County Council and the balance for allocation to each local council in that county. In the second step, the budget allocates specific sums to each county to pay part of the costs of the heating subsidy. Finally, in the third step, the Annual State Budget Law mandates the use by the county councils of a specific formula to allocate equalization grants to the local councils. The formula measures the relative fiscal capacity of county and local councils. It is included as an annex to the state budget. This is a very important step. Any county council that uses some other method to allocate the grants would be subject to a finding by the Court of Accounts that they have failed to follow the law. 3.19 How are county councils actually allocating the equalization grant? Table 3.5 looks at the extent to which four counties adhered to the formula in the actual allocation of equalization grants within their jurisdiction. The attempt to impose on the county councils a single formula for allocating the equalization grants in 2000 did not work in the case of the four counties included in this analysis. Anecdotal evidence also suggests that other county councils failed to 52 Chapter 3: Implementing Fiscal Decentralization Reforms follow the formula. The question is why the county councils are not adhering to the requirement of the Law on Local Public Finances, using a formula-based approach to allocate the distribution grants. 3.20 Methodological factors constraining the use of a formula-based approach. At least one part of the explanation for not following the formula-based approach appears to be the difficulty (real or perceived) of tailoring a formula that reflects the needs of the local councils. When asked what process they followed, the eight county councils referred to one of several factors that required an ad hoc allocation rather than a formula. These factors fall into two main categories: factors related to the revenues of local councils, and factors related to the expenditures of local councils. 3.21 There appear to be two factors related to the revenues of local councils that constrain the use of a formula-base approach: > The county councils appear concerned with the "tax effort" of the local councils. That is, they suspect that the local councils are not doing enough to raise revenues through their own taxes and fees, thus increasing their reliance on equalization grants. This is a common concern when using a formula. There are ways of addressing the problem. > The county councils also appear concerned that some local councils have lost revenues due to factors outside their control. For example the county council of lasi looked at whether one of its local councils had lost revenues from the shared taxes due to the relocation of the headquarters of a company. Since the share of the tax accrues to the jurisdiction where the headquarters is located, the change in location would lead to a loss of revenues of one local council and a gain for another (that might be in an entirely different county). 3.22 There appear to be two factors related to the expenditure of local councils that constrain the use of a formula-base approach: > Almost every county council covered in the analysis feels that it cannot allocate investments using a formula. In part, this reflects the lumpiness of investment expenditures (and hence revenues) that may swell for one or two years while a large project is under construction, then taper off as the project is completed. This is difficult to accommodate in an expenditure formula. > Many county councils also appear concerned with certain expenditure needs that are not shared by all local councils. This includes, for example, institutions that care for the elderly or the handicapped that exist only in a few local councils. This might be a case that cannot be resolved adequately through a formula. It may require an alternative method, such as dedicated transfers. > The county councils appear to feel responsible for helping those local councils that have gotten into problems over time. For example, some local councils have Chapter 3: Implementing Fiscal Decentralization Reforms 53 accumulated unpaid debts for energy or that have deferred maintenance and must make urgent repairs to a school or other such critical facility. This is the expenditure version of the concern with low tax effort discussed above under revenue factors. There are ways to- account for this in a formula-based allocation process. Table 3.5: Processing of Equalization Grants by the County (judet) Councils County (judet) Process used to allocate grants Criteria used to allocate grants council Arad The County Council allocated one The County Council tried to meet the needs of part of the equalization grants by individual local councils in the areas of (i) institutions following the formula included in to care for the elderly; (ii) institutions to care for the the state budget strictly. They handicapped; (iii) school operating and maintenance distributed the balance based on costs; (iv) investments and general repairs other criteria. Brasov The County Council did not refer at The County Council tried to meet the needs of all to the formula in the state budget. individual local councils in the areas of: (i) mainly on They based the distribution on other investments and general repairs; and (ii) to a much criteria. lesser extent, on specific, well-justified operating expenditures, such debts for energy Constanta The County Council initially tried to The County Council tried to meet the needs of apply the formula included in the individual local councils as follows: (i) mainly to state budget. Based on their analysis cover the projected operating deficit, after analyzing of the result, they adjusted these their other revenue sources; and (ii) to a lesser extent, figures based on other criteria. on investments but just for water and pavement of roads in the communes lasi The County Council did not use any The County Council evaluated the revenues and standardized formula. They expenditures of each local council according to the performed an ad hoc analysis of the following criteria. In the case of revenues, whether revenues and expenditures of each the increase or decrease in own local taxes as a result local council. of: (i) decisions by the local council to adjust taxes for inflation (or not); and (ii) changes in the revenues from shared taxes, including those influenced by outside factors, such as the reclassification of the tax payers whose payments accrue to the local council. In the case of expenditures, whether these reflected either changes in operating expenditures, or, to a much lesser extent, investments. Source: Urban Institute (2001). 3.23 Institutional constraints to a formula-based approach. There are also institutional factors that might account for the failure of county councils to implement a formula-based equalization grant allocation process, namely the conflict between the county council's county-wide responsibilities and local council's responsibilities at the commune level. While county councils have legitimate countywide responsibilities by law, they have few instruments with which to fulfill these responsibilities, including coordinating the activities of local councils within the county to address countywide priorities and needs. As a result, county councils have relied almost completely on controlling the allocation of equalization grants across communes. This 54 Chapter 3: Implementing Fiscal Decentralization Reforms creates significant problems already at the budget preparation stage, since neither county, nor local councils are able to estimate what level of equalization grants to expect in the coming year. Recommendation on Local Government Budget Design 3.24 There are several steps that local and national authorities should consider in improving budget formulation at the local government level: > Provide a stable assignment of expenditure responsibilities for local governments in the law, while phasing out as quickly as possible the practice of modifying this assignment annually through the state budget. > Change the sequencing of budget formulation. Since equalization transfers are only defined after local governments present the initial budget proposals, there is a tendency to underestimate local revenues. If the process were reversed, with the Ministry of Public Finance providing an estimate of the amount of the transfer before local governments prepared their budget, there would be no incentive to overestimate expenditures and underestimate own source revenues. This is the approach taken in the Czech Republic, for example, and it has worked well. > Require that all local and county councils develop a draft balanced budget based on their own estimates of the equalization grant they will receive according to the equalization formula. At the same time, reduce significantly the review of draft local budgets by the Ministry of Public Finance. Eventually, this review should be limited to confirming the estimates of equalization grants prepared by the local and county councils. > Indicate that county councils need to learn and apply tools and techniques, other than the subjective allocation of equalization grants, to induce the local councils to consider county-wide priorities and needs when making autonomous decisions on spending priorities in their own budgets. There always will be tension in the relations between the local and county councils. They are two equal levels of local government addressing many of the same needs in the same geographic area. It is unrealistic to expect that they can coexist in full harmony all the time. This suggests a need for a process to channel the tension and to promote constructive solutions to potential conflicts. Budget Execution 3.25 The execution of the local government budget in Romania reflects some of the deficiencies identified at the budget formulation stage, as well significant changes in both the economy and in the overall public finances over the last several years. Table 3.6 summarizes these economy-wide developments, and their impact on local governments. This provides the background against which to discuss the influence of changes in the legal framework on the budget execution of local governments. Chapter 3: Implementing Fiscal Decentralization Reforms 55 3.26 What emerges from the summary in Table 3.6 is that local government revenues and expenditures have seen significant increases and decreases. The reforms in public finance have changed the ways in which the government raises revenues and finances public services, and, as a result, there have been repeated shifts and adjustments in both the sources of revenues of local governments and in their expenditure assignments. Between 1996 and 2001 local government expenditures at first declined relative both to GDP and to consolidated government expenditures, then increased again. The decrease in local expenditures from 1996 to 1998 largely reflects the transfer back to the central level of health care and certain social assistance responsibilities. The increase starting in 1999 reflects both the impact of the 1998 Law on Local Public Finance (that went into effect in 1999), and transfer of additional expenditure responsibilities in the subsequent years, including the responsibility for delivering all pre-university education beginning in mid- 2001 and for partially funding the minimum income guarantee program in 2002. Table 3.6: Local Public Expenditures Relative to GDP and Consolidated Government Expenditures, 1996 - 2001 1996 1997 1998 1999 2000 2001 Real GDP growth 3.9 -6.1 -4.8 -2.3 1.6 4.9 Consolidated expenditures/GDP 33.8 33.9 35.3 35.1 35.1 35.4 Local government expenditures/GDP 4.5 4.1 3.6 3.9 4.2 4.6 Local/consolidated government expenditures 13.3 12.1 10.2 11.1 12.0 13.0 Consolidated government budget deficit/GDP -4.8 -5.3 -5.4 -3.7 -4.0 -3.5 Source: Ministry of Public Finance. World Bank staff calculations. 3.27 The new revenue structure. The most immediate impact of the 1998 Law on Local Public Finances was on the structure of revenues of local governments. The implementation of the law implied both an increase in tax-shared revenues available to local governments, and a redistribution of revenues by regions and by individual local governments. We begin our analysis of the budget execution at the local government level therefore by examining the impact of these changes in revenue structure and in their geographic pattern of distribution on the expenditures of local governments. 3.28 Changes in revenues from 1998 to 2000. There has been much speculation in Romania based on partial and anecdotal data, regarding the impact of the Law on Local Public Finance on the revenues of county and local councils. The accepted wisdom is that the impact was uneven by regions, and that municipalities on average fared best, while communes on average fared worst. This is confirmed by the data presented in Table 3.7 showing how the revenues of the county and local councils changed between 1998 and 2000 in real terms.40 4 In this analysis we skip the intervening year of 1999 to make the presentation easier to follow. 56 Chapter 3: Implementing Fiscal Decentralization Reforms Table 3.7: Change in Revenues, 1998 - 2000, by Type of Local Government (%) Net change from Change in Shared tax General Earmarked 1998 to 2000 local revenues revenues transfers transfers Counties 0.9 20.7 23.1 -37.2 -5.7 Municipalities 32.7 21.4 45.4 -18.9 -15.3 Bucharest 25.8 11.6 74.3 -19.4 -40.7 Town -1.4 16.9 20.6 -20.7 -18.2 Communes 18.7 23.2 10.3 1.9 -16.8 All local governments 19.0 19.2 37.8 -19.1 -18.9 Source: Urban Institute (2001). 3.29 Aggregate measures of revenue distribution, while convenient, conceal a lot of useful information on the real conditions of nearly 3,000 individual local governments.41 Two measures of the impact on expenditures of changes brought about by the Law on Local Public Finance are: (i) the share population served by the county and local councils; and (ii) local needs in key areas. Table 3.8 looks at how the population is distributed among county and local councils by the percent increase in total revenues. The results indicate that revenues were redistributed to local governments accounting for most of the population. Nearly two-thirds of the population lives in the jurisdiction of local governments that had the largest increases in revenues from 1998 to 2000.42 This share of the population rises to 68 percent if Bucharest is included. Table 3.8: Distribution of the Population by Type of Local Government According to the Change in Tax Revenues between 1998 and 2000 (%) Change in revenues Counties Local Councils Municipalities Towns Communes Total Greater than 10 percent 42.1 29.7 4.2 30.6 64.5 Between 10 and 5 percent 2.3 1.0 0.8 1.6 3.5 Between 5 and 0 percent 8.4 1.0 0.6 2.0 3.7 Between 0 and -5 percent 8.5 0.5 0.6 1.4 2.4 Between -5 and -10 percent 6.6 2.9 0.6 1.6 5.1 Greater than -10 percent 32.1 4.0 4.5 12.3 20.7 Total 100.0 39.2 11.3 49.5 100.0 Memo: total share of local governments experiencing declines in revenues 52.8 31.8 5.7 34.3 71.7 Source: Urban Institute (2001). 3.30 The other side of these developments, as shown in Table 3.8, is that counties serving about one-third of the population and local councils serving about 21 percent of the population suffered severe declines in revenues. This is not necessarily a sign of a problem, however. Those localities with the highest per capita total revenues in 1998 were most likely to experience 4 We try to avoid simple comparisons of different types of local governments (counties, municipalities, towns and communes). For example, comparisons of revenue and expenditure data to total population served and per capita figures help overcome the differences in size and volume of operations among different local governments. Where data is available we also have tried to compare expenditures to indicators of local need. For example, we use the number of pre university students per locality, as an indicator of relative need in the area of education expenditures. In a few cases, we also have used examples from specific local governments to illustrate a point. 42 In this table it is important to separate the counties from all the other local councils. The counties and the local councils (municipalities, towns and communes) cover the same areas and serve the same population. Chapter 3: Implementing Fiscal Decentralization Reforms 57 a severe reduction in total revenues between 1998 and 2000. Many of these localities might still have higher per capita total revenues in 2000. The median total revenues per capita are higher in the jurisdictions with social assistance centers in 2000. More than three-quarters of the people served by social assistance centers are in the local councils in the top quintile of total revenues per capita. Almost two-thirds (62 percent) of beneficiaries on means tested income transfers live in the local councils in the top two quintiles of total revenues per capita. 3.31 To provide perspective on this impact of the implementation of the Law on Local Public Finance on expenditure needs, Table 3.9 shows the trends in change in revenue compared to a series of factors that measure the local need in key areas. The Ministry of Public Finance collected these numbers in 1998 as part of its continuing efforts to develop a more equitable formula for allocating transfers to local governments from the state budget. These numbers provide a unique perspective on the distribution of the following key expenditure determinants at the local level: (i) the number of pre-university students by locality in 1998; (ii) the number of persons served by social assistance facilities (homes for the aged, soup kitchens) by locality in 1998; (iii) the average monthly number of social assistance recipients in 1998 by locality; (iv) the extension (in kilometers) of county and rural roads; and (v) the extension of urban streets, measured in kilometers. Table 3.9: Distribution of Expenditures by Function According to the Change in Tax Revenues between 1998 and 2000 (%). Change in revenues Education Social Services Social Assistance Roads Streets Greater than 10 percent 65.4 38.8 57.3 55.7 57.2 Between 10 and 5 percent 3.5 2.3 4.1 2.3 3.6 Between 5 and 0 percent 3.4 5.2 4.0 4.5 3.7 Between 0 and -5 percent 2.3 1.8 3.4 3.6 3.6 Between -5 and -10 percent 5.5 6.1 3.9 3.3 8.0 Greater than -10 percent 19.6 45.8 27.3 30.7 24.0 Total 100.0 100.0 100.0 100.0 100.0 Source: Urban Institute (2001). 3.32 Table 3.9 shows a different pattern than that in Table 3.8. Whereas a large proportion of the population lives in the jurisdictions served by those county and local governments that fared best under the new law, a significant share of the needs is concentrated in the areas served by those that experienced the largest decrease in total revenues under the new law. Around 46 percent of those that benefit from social assistance facilities (such as homes for the aged and soup kitchens), 27 percent of the persons that receive income transfers and 31 percent of the rural roads fall under the jurisdiction of county or local governments that experienced a ten percent or greater reduction in real terms in revenues from 1998 to 2000. 3.33 What was then the impact of implementing fiscal decentralization reforms on local government expenditures? While the new local finance law had a mixed, but generally favorable impact on local revenues, it failed completely to correct a problem in the assignment of local expenditure responsibilities, compromising the execution of the budget. The functional responsibilities (services provided by local governments) are still determined each year through an Annex to the Annual Budget Law. This annual determination of local expenditure responsibilities generates uncertainty and compromises the budget execution. This problem with the assignment of expenditure responsibilities is further compounded by the way the 58 Chapter 3: Implementing Fiscal Decentralization Reforms responsibilities are implemented. It is a common practice to assign responsibility to local governments for specific types of expenditures (salaries of social workers assisting the handicapped or education investments), or for specific existing public institutions (centers that care for handicapped children or homes for the aged). The responsibilities are neither defined, nor implemented in the context of broader local strategies to address them. 3.34 Earmarked funding and expenditure mandates. The most serious constraint to adequate budget execution at the local government level is the practice of earmarking funding and expenditure mandates - where the latter have, at times, taken on the characteristics of unfunded mandates. Although the Law on Local Public Finance eliminated all the earmarked transfers to local governments from the national budget, except for funding for investments financed with loans to the Government, since the initial implementation of this law in 1999, earmarked transfers have reappeared. This happened in parallel with the re-introduction of delegated expenditure. For these new delegated expenditure mandates the national government usually establishes the level of spending, but does not always provide adequate funding to cover the expenditures. This is particularly the case (see para 3.37) with social programs that are widely perceived as unfunded expenditure mandates addressing pressing social needs. For instance, the eligibility criteria and the level of the benefits for means tested income transfers are set nationally. Yet, local governments must fund a share of the cost of the benefits from their own resources. 3.35 The evidence from individual local governments in 1999 and 2000 (Table 3.10) suggests that while earmarked funding declined, these mandates increased to over 50 percent of the operating expenditures of larger, urban local governments. The way in which local governments coped with these mandates was by not funding the full obligation. This is not a good practice, but it is not clear that the alternative in the short run is any better. Should local governments fund the mandates fully, they could only do so by curtailing other important local expenditures.43 3.36 Among the important expenditure mandates that local governments have had to curtail when implementing their budgets three stand out: social assistance, public works and other capital expenditures, and education. The following paragraphs discuss briefly discuss the problems faced in each one of these programs, although the new responsibilities for local government in delivery of the minimum income guarantee program is discussed at greater length in chapters 6 of this report. The message underlying the implementation of all these mandates is the same, however. Unfunded expenditure mandates have squeezed out other expenditure responsibilities, leading local governments, in some instances, to accumulate substantial arrears. This has compromised the quality in the delivery of these services, and accentuated further regional disparities within Romania. 43 The problem with expenditure mandates is not unique to large urban areas such as lasi. The following are some existing mandates that affect all local councils: (i) means tested income support payments (ajutor social); (ii) payroll of the persons assisting the handicapped; (iii) payment per child in institutionalized care (starting in 2001); (iv) payroll of the staff of the veterinary stations; and (v) child protection services. While child protection services are not a formal expenditure mandate, the degree of attention it receives from the central government and foreign donors gives it the characteristics of an expenditure mandate. Chapter 3: Implementing Fiscal Decentralization Reforms 59 Table 3.10: Municipality of lasi: Impact of Earmarks and Mandates (Constant 2000 ROL) 1998 1999 2000 Total operating expenditures 370,090 492,924 356,001 Mandated expenditures, of which 169,991 249,730 126,519 - Social assistance a 0 0 26,596 - Heating b 58,401 201,040 65,006 - Transport 111,590 48,690 34,857 Percent mandated 46% 51% 52% Earmarked revenues, of which 155,755 159,805 113,995 Dedicated local revenues d 0 20,477 41,699 Earmarked operating transfers C 155,755 139,328 72,296 Percent earmarked operating revenues 42% 32% 63%. a. Cost of persons assisting the handicapped. b. Price and social subsidies. c. Price subsidies. d. Local fees with specific purposes and other local revenues with a dedicated use. e. Includes among others transfers from the state budget for heating subsidies. Source: Financial model, Municipality of lasi. 3.37 Social assistance and transfers. To illustrate the broader point made above about expenditure arrears, consider the perennial problems faced by local governments in the delivery of social assistance. Local governments in Romania currently have extensive responsibility for spending that is generally classified under the title of social assistance and transfers. These include the basic income transfer payments, assistance for groups with special needs (handicapped, child protection) and consumption subsidies for public services (heating, public transport). At roughly seven percent of total local government expenditures each, heating subsidies and child protection accounted for the largest share of the total local social assistance expenditures in 2000. Although local governments are responsible for the major income transfer program, expenditures for cash benefits represented just over two percent of total local expenditures. 3.38 A 1999 Bank report on Local Delivery of Social Services" looked in detail at income transfer payments and assistance for groups with special needs managed by local governments in specific localities in five counties of Romania. An Urban Institute (UI) report also looked at funding for child protection services in each of the 41 counties in Romania.45 Both reports found evidence that local governments have not funded the full cost of the social services assigned to them. The problem appeared to be particularly acute in the area of means tested cash benefits. The same is true with consumption subsidies for public services. Evidence from several cities suggested that they are not providing the full funding for heating subsidies required by law. The pattern was not uniform and the debts were not necessarily a problem in each of the 193 local councils that had heating systems. 3.39 The World Bank report also found other problems with local social service delivery. "There is an uneven distribution of service provision across areas due to lower revenues and " World Bank (2001). 4s Urban Institute (2000). 60 Chapter 3: Implementing Fiscal Decentralization Reforms higher costs in some areas. In rural areas unit costs for education and health services tend to be higher because services must be guaranteed in areas where population density (and therefore the number of beneficiaries) is low. This translates, for example, into low ratios of pupil per school. At the same time, because of higher poverty rates, rural areas have low revenues and are the least equipped to cope with these more substantial fiscal obligations. Other factors, such as the concentration of Roma in an area also affect the level and effectiveness of services."46 3.40 Finally, there were other problems with subsidies for consumption of public services, specifically those for heating and public transport. The heating subsidies did not differentiate between those that do and do not need assistance in paying their heating bill. All the population benefited from the general price subsidy under the current system. The report argued that these transfers should not be targeted to heating only. Rather, the concern should be whether a given family can afford to pay for all basic public services. This included water, electricity, heating, gas and solid waste removal. 3.41 Public works and other capital expenditures. To illustrate the broader point made above about unfunded expenditure mandates squeezing out other expenditure responsibilities, consider what has happened with capital expenditures since the introduction of the new law on Local Public Finance.47 One of the objectives of the new law was to allow local governments to shift the funding for capital expenditures from target transfers from the central governments to savings from their own budget.48 This aimed at giving local governments a greater say in determining expenditure priorities and greater stability in making their investment plans. This goal was partially accomplished. The funding of investment through targeted transfers did come down, reaching on 45 percent of all local governments investments in 1999, compared to 82 percent in 1998 (Table 3.11). The greater reliance on own savings to fund investments happened, however, simultaneously with an increase in earmarked expenditures to local governments. This was particularly a problem because earmarked capital expenditures account for about one-half of all earmarked expenditures (Table 3.12). 4 World Bank (2001), page vi. 47 Local governments in Romania are responsible for all investments in key areas including water, waste water, heating, public transport, county and local roads, streets and bridges and public housing. 4 The other dedicated investment transfers exist to promote specific sector priority financing.'They are managed by the line Ministries or by public institutions under the authority of the line Ministries. These include the Road Fund, transfers from Ministry of Public Works for communal road and water works, the housing fund, and other various funds (civil aviation fund, natural disasters, etc.). These funds do not appear, however, to allocated to individual local governments on the basis of specific rules and criteria. These concerns appear to apply in particular to investments by the Communal Road and Water Fund, as well as some other smaller funds. This suggests that either clearer criteria should be defined to allocate these funds, or investment transfers should be made to local governments without setting conditions on areas of investment. Chapter 3: Implementing Fiscal Decentralization Reforms 61 Table 3.11: Funding of Local Capital Expenditures, 1996-2002 (%) 1996 1997 1998 1999 2000 Total capital expenditures' 12,990 9,542 7,590 7,342 7,063 Real % change ... -26.5% -20.5% -3.3% -3.8% Budget savings 13.2 24.2 17.8 55.1 48.8 Targeted transfers 86.8 75.8 82.2 44.9 51.2 Targeted transfers for local investments 76.6 67.1 61.2 0.0 0.0 External loans 0.4 2.9 7.1 12.6 11.5 Communal roads and waterworks 0.0 0.0 8.7 6.4 10.2 Housing fund 0.0 0.0 0.0 18.2 13.2 Other transfers 9.8 5.8 5.2 7.7 16.4 Total capital expenditures/Total expenditures 27.6 25.3 25.7 22.9 20.2 1) In 2000 Billion Lei. 2) Discontinued in 1999 with the implementation of the Law on Local Public Finance. Source: Ministry of Public Administration; Urban Institute (2001); World Bank. Table 3.12: Structure of Capital Expenditures, 1996-2002 (%) 1996 1997 1998 1999 2000 Public administration 3 5 5 6 7 Education 0 0 0 4 4 Public works and housing 68 63 64 51 50 Transport and communications 25 27 26 11 10 Earmarked expenditures' 24 25 1) Beginning in 1999. Source: Ministry of Public Administration; Urban Institute (2001); World Bank. 3.42 The result of this switch away from earmarked transfers for capital expenditure, and a greater reliance on own budget savings, was an overall decline in capital expenditures, both in real terms and relative to other local government expenditures. Most of local government investments are now concentrated in two key sectors-water/wastewater and roads-for which some sources of external financing are still available. The problem is that long-term loan financing is at present available only to the most financially viable municipalities and public enterprises, implying that poorer local governments endured most of the impact of the decline in investments. Also, other areas that have considerable needs in upgrading and modernizing depreciated assets, such as the solid waste disposal, heating and urban transport, are being neglected. For example the heating companies have physical plant so old and in such poor condition that their operating costs are extremely high. With the national price for retail heating set by the central governments, many local district heating companies are unable to pass on their costs of operations to their client base, making them therefore financially unviable. This creates a vicious cycle where lack of investments leads to poor financial performance that makes such companies ineligible for loans thereby further limiting their ability to finance investments. The local governments do not of themselves have sufficient funds to step in to cover the necessary investment costs. 3.43 Decentralization of Education. To illustrate the new challenges of decentralization, consider the recent developments in the decentralization of education. While the decentralization of education was already anticipated in the 1995 Education Law, it has recently 62 Chapter 3: Implementing Fiscal Decentralization Reforms gained speed, making this an important test case of the entire decentralization process. Beginning in 1995 the responsibility for school maintenance was transferred to local governments. This responsibility was extended in 2000 to include investments and current expenditures. Starting with the 2001/2002 school year (September 2001), the payment of teacher salaries, which account for just under three-quarters of all education expenditure, was delegated to local governments, although funding was initially provided through earmarked transfers from the central government. There are now plans to no longer earmark these transfers for teachers' wages and salaries already in 2003, giving local governments full responsibility for education expenditure. This full transfer of responsibility is happening, however, against the backdrop of increasing funding disparities across local governments, and serious underlying problems with the budgeting process in education. These developments raise serious concerns. 3.44 The financing arrangements for the education sector are regulated by the Government Decision (GD) 538/2001 which sets out a capitation principle where 'the budget of the state pre- university education institution is founded on financing in proportion with the number of students or pre-school students, and allocated for staff, textbook, and scholarship costs'. The financing of material and service costs is based on budget norms that are negotiated every budget year to reflect an average annual cost per student/pre-school student according to the level of instruction. 3.45 The budgeting process that underlies these financing arrangements in education follows two different flows: one for wages and salaries, textbooks and scholarships, another for all other expenditures, including material and capital expenditures. The budgeting for wages and salaries, textbooks and scholarships is guided by norms. School units define the number of classes and then, following standard budget norms,49 calculate the expected expenditures based on the number of teachers, the textbook needs and required scholarships. While the budget proposals for these items are scrutinized by several intermediate levels before being submitted by the Ministry of Education and Research to the Ministry of Public Finance, the only issue usually subject to review is the number of classes for each school. The budgeting for all other expenditure units reflects an expenditure prioritization by the local parties involved (the mayor, the local council, the school principal, and, in some instances, the parents). The budget proposal for these items is included and submitted for approval along with the local government's overall budget proposal. While these budget requests are reviewed based on a relationship between the amounts requested and estimated average annual costs for capital and material expenditures, the funding made available reflects overall budget envelopes that may or may not meet the total amount requested. During the budget implementation, at the time of the budget rectification, supplementary funds could be allocated to meet the original request, at least in part. 3.46 This budget process means that, in practice, school units and local governments are not very involved in the budget design and, as a result, can exercise very little discretion at the budget execution stage. While local governments have the responsibility for paying almost all the expenditures, including teacher salaries, they have no control over the factors that determine these expenditures (e.g. human resource policy). This provides little incentive to increase 49 These norms reflect the National Council for Financing State Pre-University Education's basic budget parameters, which are based on estimated average annual cost per student/pre-school student according to the level of instruction. Chapter 3: Implementing Fiscal Decentralization Reforms 63 efficiency by innovating. Indeed, since there is little transparency in the allocation of resources, there is also little incentive to economize on the use of these resources. In most cases, funds are allocated to one line of the budget (e.g., salaries, capital investments), and cannot be used for other purposes. For example, in the case of salaries the allocation rules will be centrally determined according to enrolments and curriculum demands, as well as a clearly defined hierarchy for corresponding salaries. Neither the school principals, nor the school inspectorates will have much opportunity to revise the number of assigned staff to make better use of resources. They can neither pay more to fewer teachers, nor, if fewer teachers are needed, use the savings for something else. 3.47 The outcome of this budgeting process is reflected in Table 3.13. While resources available for education declined 23.6 percent in real terms during the 1998 to 2000 period, staff and teachers wages and salaries continued to command just under three-quarters of total expenditures. Capital and material expenditures bore the brunt of the decline in funding, with declines of 85.3 and 65.1 percent, respectively. The only expenditure category experiencing an increase in real terms were transfers, although most of this increase went to directly' to local governments rather than to students through scholarships. Table 3.13: Structure of State Education Expenditure, 1998-2000 (%) Expenditure category 1998 1999 2000 Total expenditures (Lei 2000 trillion) 31,552 23,665 24,111 Wages and salaries 74.9 74.7 71.9 Materials expenditures 12.9 6.1 5.9 Of which textbooks 0.7 0.6 0.9 Subsidies 1.1 1.2 2.3 Transfers 3.8 15.8 18.0 Of which scholarships 3.7 3.1 3.6 Capital expenditures 7.2 1.8 1.4 External debt service 0.1 0.3 0.5 Source: Fugaru (2002). World Bank staff calculations. 3.48 The increase in the share of transfers in total expenditures reflects concerns that, while decentralization should allow better allocation of limited resources, it will also lead to increased discrepancies among regions and counties. The increasing discrepancies in funding across regions and counties are of particular concern because the financing of education through the state budget is based on the sharing of income tax revenues which tends to favor regions and counties that already raise more taxes and are, therefore, better off. This concern is confirmed when comparing the changes in real expenditures in education across counties between 1998 and 2000. As can be seen in Table 3.14 below, changes in education expenditure at the county level range from a 22 percent increase in Ilfov county (in the Bucharest-Ilfov region) to a 31.6 percent decline in Braila county, in Southeastern Romania. Furthermore, the latter is more common than the former, as twenty-six out of the forty-one counties experienced expenditure declines during this period.50 " The distribution across regions within Romania of the counties experiencing declines does not appear to follow a systematic pattern. Among the last 14 counties experiencing larger declines in education expenditure, four are from the Northwest, three are from Central Romania, three are from the Southeast, two are from the Northeast, one from the Southwest, and one is Bucharest itself. The surprising result is the decline in education expenditure in the 64 Chapter 3: Implementing Fiscal Decentralization Reforms 3.49 How to proceed with the decentralization of education? In 2003 the Romanian authorities plan to stop earmarking the funds currently transferred for teachers wages and salaries. These plans should be reconsidered. This will give local governments excessive discretion in allocating resources across expenditure categories, bringing new responsibilities for which they are not prepared, and heightening ongoing concerns about funding for education. Over the last several years, the overall funding for education has declined, and, as a result, there has been a considerable underfunding of capital and material expenditures. Also, because of revenue sharing arrangements, we have observed a rising inequality in distribution of additional resources across counties. 3.50 Moving ahead with the decentralization of education will require taking actions on two fronts: > There is a need to involve as many local stakeholders as possible, using for instance learning processes, such as the ones where the World Bank trained some 10,000 school principals in budgeting issues. This would prepare local governments to capture efficiency gains associated with decentralization, improving accountability and increasing the responsiveness to specific local needs and/or demand shifts. Other elements of the management of the education program should, however, continue to be managed centrally to ensure quality control and fully exploit economies of scale. This includes the teacher's salaries, the national assessment program, the development of national curricula, as well as the training and supervisory functions. > There is the need to reconsider the present system of financing of education, moving to a system that does not attach all the weight to the local government's revenue generating capacity. At present, the financing of local governments comes primarily from two sources: property taxes (which are kept 100% at the local level); and income taxes. Income taxes are shared between levels of government, with 36.5 percent going to local councils, 25 percent going to county councils, and 38.5 percent going to the central government. There is also an equalization grant designed to address concerns with the rising funding disparities across local governments. The funding available through these equalization grants is computed through a formula approved every year with the state budget. The formula, however, allocates additional funding according to the local government's current tax revenues and cannot, therefore, possibly favor the poorest regions. relatively better off Northwest region. This might reflect a decrease in the number of students enrolled in high- school and vocational education, since many teenagers in this region chose to work abroad in Hungary, Austria, and Germany. Chapter 3: Implementing Fiscal Decentralization Reforms 65 Table 3.14: Changes in the Real Expenditures on Education at the County Level, 1998/2000 (%) County % change County % change Ilfov 22.0 Dimbovita -7.2 Teleonnan 20.0 Hunedoara -8.0 Mehedinti 17.4 Dolj -8.2 Bihor 15.8 Giurgiu -9.1 lasi 15.4 Botosani -9.5 Tulcea 12.7 Alba -9.7 Olt 12.0 Maramures -9.8 Ialomita 9.2 Cluj -10.3 Suceava 8.8 Satu Mare -12.2 Constanta 8.7 Brasov -12.5 Calarasi 6.8 Bucarest -13.5 Sibiu 6.8 Vilcea -15.0 Prahova 4.9 Bistrita-Nasaud -17.2 Galati 3.0 Mures -18.1 Harghita -3.1 Salaj -18.5 Arges -3.3 Neamt -18.7 Caras-Severin -3.4 Buzau -20.2 Bacau -4.3 Vaslui -22.8 Gorj -4.4 Covasna -23.3 Timis -5.8 Vrancea -24.3 Arad -7.0 Braila -30.6 Source: Fugaru (2002). Recommendations on Budget Execution 3.51 The budget execution can be improved primarily by defining more stable local government expenditure assignments, and by clarifying the distinction between local government's own and delegated functions. At present, local government expenditure assignments are revised every year with the state budget law, creating ongoing uncertainty for financial management and budget practices at the local government level. These in turn lead to disruption in the delivery of public services. Under clearly defined expenditure assignments, own functions would be those where local governments have broad authority and discretion to determine the service policy and methodology and to decide on service quantity, quality and cost. Local government's own functions should not include national programs, such as education and social assistance. Delegated functions are those governmental activities that the national government may assign to local government to undertake in a manner that ensures equal access and quality to the entire population, regardless of place of residency. Funding of delegated functions should be clearly earmarked, and defined on the basis of needs, rather than through revenue sharing agreements. The latter leads to large discrepancies in the delivery of services, accentuating existing regional disparities. 3.52 An important step in establishing this distinction between local government's own and delegated expenditure responsibilities is to limit earmarked funding, discontinue under-funded or unfunded expenditure mandates, and define better revenue assignments for delegated functions. Earmarked funding is not appropriate in the case of own functions, and should be limited to certain delegated functions, such as those in the area of social assistance and education. Also, there is a need to revisit the matching of funding and needs in the delivery of these delegated 66 Chapter 3: Implementing Fiscal Decentralization Reforms functions. At present, the funding provided for new delegated functions, such as education, is provided mostly through revenue sharing arrangements. The problem in this case is that, while the aggregate amount of funding might be adequate, the distribution of funds will not match the expenditure needs of each local government. The solution is to increase the share of equalization grants targeted for the delivery of these services, ensuring a better match between funding and needs. Budget Compliance and Evaluation 3.53 How can deficiencies in budget design and budget execution be systematically identified, allowing the development of a stable and transparent system of local government financing that matches existing and new expenditure responsibilities? In principle, the deficiencies should be identified at the budget compliance and evaluation stage, with remedies to correct these deficiencies introduced at the stages of budget design and execution. While Romania has a good basis for adequate the budget compliance and evaluation, there are still several shortcomings that preclude these expectations from being fulfilled. The system still works on a strict cash basis, with no information on expenditure commitments. Each public institution is treated as an independent reporting entity of accounting, so there is no integrated view of the finances of the local government. These problems are compounded by the discrepancies in the capacity of local governments to implement their newfound fiscal and administrative autonomy. 3.54 The Romanian authorities are aware of these problems. They are working to build a professional cadre at the local government level, and to provide them with the necessary tools for to improve financial and operational accountability. This section therefore reviews where the current system of financial accounting and reporting now stands and suggests next steps. 3.55 How does the system offinancial accounting and reporting work in Romania? There is a single, uniform, national system of classification for all public financial data in Romania. It applies both to the national and local governments. This system has clear advantages: > It establishes a single core of knowledge for all public sector accounting and financial administration staff. This facilitates training, and allows staff to move easily from one public institution to another. > It facilitates the aggregation and analysis of public finances. This establishes a uniform basis for comparing the performance of different public sector institutions. This in turn makes it possible to look at all public expenditures in a given functional area, such as education or transport, even when there are multiple institutions involved representing the national and local governmeints. It also is possible to look at public expenditures by level or type of governmental institution, for example, all local governments. The information provided in this chapter draws on this type of information. 3.56 To obtain the full advantage of the existing system, however, it would be necessary to create a single computerized database with the information for each fiscal year. Since the classification creates a unique, numeric code for the different types of revenues and Chapter 3: Implementing Fiscal Decentralization Reforms 67 expenditures, such a database would make it almost effortless to analyze public financial data in Romania. Unfortunately, the data still largely exists in paper form, scattered among a large number of public institutions. 3.57 Also, despite some positive features, the current system of does not meet all the current financial accounting and reporting needs of local governments in Romania. The reasons are three-fold: > The system works on a strict cash basis of accounting. As such it does not produce vital information on outstanding liabilities. This is the case, for example, of the unpaid heating subsidies. The accounting system records only the payments made (actual cash disbursed) to the heating company by the local government. There is no record of the amounts owed for heating subsidies but not paid. > Each public institution is treated as an independent reporting entity, even when various institutions are subject to the authority of the same public officials. As a result, there is no readily available report that provides comprehensive financial data necessary to reflect the full financial accountability of these public officials. This is the case, for example, of data on personnel expenditures. In an alternative approach, the local government would be the reporting entity not only for its own finances but also for those of all its subordinated public institutions. Their combined financial results would appear together in a single statement issued by the local government. In such an approach, it would not be possible to "hide" personnel expenditures in a subordinated public institution. > There are no general-purpose external financial reports. All the financial reports that exist are prepared for internal users, that is, for government managers and elected officials. Indeed, the very notion of a general-purpose external financial report does not exist. Such reports provide essential information needed by all interested parties for a fair assessment of the finances of a government entity. 3.58 These limitations in the rules and practices of financial accounting and reporting represent a serious constraint to the continued development of the system of local government in Romania. Local governments cannot show that their actions have complied with the public decisions on how to raise and spend public funds-financial accountability. They also cannot show that they are providing services efficiently and effectively-operational accountability. 3.59 Public utilities and housing. These limitations in the rules and practice of financial accounting extend beyond those services provided directly by local governments, and also include the delivery of public utility services. These are very important for two reasons. Expenditures on public utilities and housing, as shown in Table 3.4 above, account for around one-third of total local government expenditures. These expenditures are also among the most visible for the population because they include all the major physical plant and structure of urban areas, and have a major impact on the daily quality of life of local residents. Yet, as large as they are, these expenditures do not reflect accurately the extent of local responsibility in the area of public services. As Table 3.15 shows, the amount that households pay for basic public services, 68 Chapter 3: Implementing Fiscal Decentralization Reforms such as water or heating, can be many times greater than what the same households pay in local taxes and fees. This means that for households the entire local budget is but a small part of what they "pay" to the local government in taxes and fees. Table 3.15: Local Taxes and Fees versus Basic Service Fees (Lei billion) Annual Payment for Public Services Constanta' Baia Mare= Targoviste2 Oradea7 Basic service fees Water 1,927,476 66,066 159,433 145,985 Heating 2,227,500 217,128 181,376 376,704 Hot water 1,930,500 201,432 144,00 235,440 Gas 1,200,000 Transport 1,031,250 154,000 104,328 123,750 Solid waste 48,000 29,750 27,000 36,000 Total 8,364,726 668,376 472,281 917,879 Local taxes and fees Property tax 500,000 14,000 14,000 14,000 Car and road tax 140,000 1,400 1,400 1,400 Total 640,000 15,400 15,400 15,400 Taxes/ service fees 7.7% 2.3% 3.3% 1.7% 1) Data for Constanta is for 1999. 2) Data for Baia Mare, Targoviste and Oradea is for 1997. Source: Urban Institute (2001). 3.60 The decisions that local governments take regarding basic public services are as critical as their decisions regarding the local budget. Yet, this large segment of local responsibility comes under very little scrutiny because it is neither entirely reflected in the local budget, nor in the financial statements prepared by local governments. Local governments need to be more transparent and, therefore, be made more accountable for several decisions made in carrying out their legal and regulatory responsibilities in the area of public utilities. This includes responsibility for decisions about service fees, which is currently shared between the local and national governments. It also includes decisions about corporate structure and governance of the local utility companies. Local governments currently appoint the manager of these companies. Eventually, local governments also must decide about the best way to operate the systems, including options for private participation in the management and even ownership of the utilities. This is an area where there is ample room to improve the efficiency-of local public expenditures- broadly defined-without requiring any additional resources. Policy Recommendations for Budget Compliance and Evaluation 3.61 To strengthen locial financial accounting and reporting, the Romanian authorities should institute, in the immediate short run, on a pilot basis: > A process for recording and reporting additional information on local assets and liabilities and net balances on a one-time basis at the end of each budget year Chapter 3: Implementing Fiscal Decentralization Reforms 69 > An expanded definition of local reporting entities so that all the public institutions that are under the authority of the same elected officials appear in a single comprehensive set of financial statements > Simple general purpose external financial statements aimed at external users 3.62 In the longer term, consider a more comprehensive reform of local public financial accounting and reporting standards and practices consistent with EU directives on the subject. Conclusions and Recommendations 3.63 This chapter reviewed the budget process at the local government level, as a way to assess their capacity implement their newfound fiscal and administrative autonomy. While it is misleading to make broad generalization about local government ability to manage their expenditure responsibilities, since there are large discrepancies in the distribution of revenues and administrative capacity across local government, the chapter makes several recommendations that can assist in improving the current budget process. These include: > Allocating the equalization grant at the beginning of the budget formulation stage. The local budget process at present is a major obstacle to any attempt by the local and county councils to manage their finances on a sound basis, consistent with local needs and priorities. There is no hard budget constraint that compels them to make difficult spending priority decisions within a given level of revenues. To a large extent the problems with the local budget process derive from the absence of a transparent and predictable process for allocating equalization grants. Neither county nor local councils are able at present to estimate what level of equalization grants to expect in the coming year, before preparing their budget. There appear to be two major factors that explain the reluctance of most county councils to allocate the equalization grants on a pure formula basis. At least one part of the explanation appears to reflect the difficulty (real or perceived) of tailoring a formula that reflects the needs of the local councils. Another factor appears to be the search by the county councils for tools to coordinate the activities of local councils within the county to address countywide priorities and needs. To date they have relied almost totally on controlling the equalization grant allocation process to this end. > Reconsidering the present system of allocating equalization grants, moving to a system that does not attach all the weight to the local government's revenue generating capacity. At present, the financing of local governments comes primarily from two sources, the property taxes (which are kept 100% at the local level), and income taxes. Income taxes are shared between levels of government, with 36.5 percent going to local councils, 25 percent going to county councils, and 38.5 percent going to the central government. There is also an equalization grant designed to address concerns with the rising funding disparities across local governments. The funding available through these equalization grants is computed through a formula approved every year with the state budget. The formula, however, allocates additional funding according to the local government's current 70 Chapter 3: Implementing Fiscal Decentralization Reforms tax revenues and cannot, therefore, possibly favor the poorest regions. The result is that many delegated responsibilities go unfunded or underfunded. > Establishing a clearer distinction between local government's own and delegated functions at the budget execution stage. Own functions would be those where local governments have broad authority and discretion to determine the service policy and methodology and to decide on service quantity, quality and cost. Delegated functions are those governmental activities that the national government may assign to a local government for performance in a manner and to a degree, which is determined by the central government. Funding in these cases should be defined on the basis of needs instead through revenue sharing agreements. The latter leads to large discrepancies in the delivery of services, accentuating existing regional disparities. > Improving the current practices on financial accounting and reporting at the budget compliance and evaluation stage. The current rules and practices of financial accounting and reporting represent a serious constraint to the continued development of the system of local government in Romania. Local governments cannot show that their actions have complied with the public decisions on how to raise and spend public funds - financial accountability. They also cannot show that they are providing services efficiently and effectively - operational accountability. To strengthen local financial accounting and reporting, the Romanian authorities should institute, in the immediate short run, on a pilot basis: (i) a process for recording and reporting additional information on local assets and liabilities and net balances on a one-time basis at the end of each budget year; (ii) an expanded definition of local reporting entities so that all the public institutions that are under the authority of the same elected officials appear in a single comprehensive set of financial statements; and (iii) a simple general purpose external financial statements aimed at external users. CHAPTER 4: REFORMING THE BUDGET PROCESS IN THE JUDICIARY Introduction 4.1 The judiciary presents one of the most difficult challenges in implementing budget reforms that move the public sector away from the present system of strong reliance on management controls, towards a system that gives managers at all levels increased responsibility for the way public resources are used and holds them accountable for results. There are two main reasons for this change. There is a premium on managing the scarce resources available to the judiciary system, since the demands on the system have steadily increased as the country moves further into the transition process. At present, the system needs to deal with an estimated annual. average of over 400 cases per judge, compared to the European standard of 280 (Table 4.1). In some cases, such as courts in Bucharest, this average can be more than twice as high, raising concerns regarding quality in the delivery of justice. Also, reforming a system that relies heavily on managerial discretion is made more difficult by the heritage of the socialist period, where courts were considered an appendix of the executive branch of government, and enjoyed no independence. For these two reasons, the budget management of almost all the courts in Romania is currently under the control of the Ministry of Justice. While the Ministry has done a reasonable and credible job of administering the courts' budgets, the current structure raises several concerns. The Ministry of Justice's strong central control appears, in practice, to conflict with the Constitutional requirement of an independent Judiciary. The strong central controls of the Ministry of Justice also reduce its ability to effectively measure court performance, and limit transparency and the expression of public opinion. Moving away from centralized budget management of the judiciary is therefore important in building a democratic system of government, and for improving the efficiency and accountability of the judiciary system. Table 4.1: Selected Indicators of the Judiciary, 1998-2002 1998 1999 2000 2001 Number of magistrates 5,357 5,368 5,592 5,647 Judges 3,361 3,383 3,504 3,575 Prosecutors 1,996 1,985 2,088 2,072 Number of specialized clerks n.a. n.a. 6,004' 6,143' Number of court cases filed ('000) 1,298 1,370 1,485 1,2253 Number of cases filed/judge per year 386 405 424 457 1) June 1, 2000. 2) December 1, 2001. 3) First 9 months of 2001. Source: Ministry of Justice. World Bank staff calculations. 4.2 This chapter is organized as follows. It begins with a brief description of the court structure in Romania, providing essential background for the discussions of the chapter. Next, it reviews three stages of the budget process in the judicial system: budget formulation, budget execution, and budget compliance and evaluation. In the section on budget compliance and evaluation the chapter makes the argument for developing indicators of performance measurement and accountability, since this is a key step needed for implementing the 72 Chapter 4: Budget Reforms in the Judiciary decentralization of expenditure responsibilities in the judiciary. Finally, the chapter outlines key steps to refine the structure of the judiciary and its budget process. In outlining these key steps for reform, the chapter develops a paradigm where the more effective operation of the budget system in the courts would allow decentralized management with reasonable oversight. This would imply that local court managers would be empowered to administer their own budgets, once the appropriation had been allocated to them, with as much authority to make decisions as any other decentralized government agency. Local court managers would report to an oversight body in Bucharest that has the interests of the courts fully as its concern. This oversight body would submit budget requests and priorities directly to Parliament, representing the courts directly and fully. The budget process would be transparent at all levels. At the national level, transparency implies clarity on how the funds are being allocated to local courts. At the local level, transparency means that local courts should know what funds are available and how they will be spent. To ensure adequate oversight and transparency, there is a need to develop effective ways to measure court performance and hold them accountable. To ensure accountability, there is the need for a better information system and clearly defined standards to measure and analyze court performance. Current Structure for Court Management in Romania 4.3 Court Structure. The Romanian courts have a four-tiered structure. The highest court is the Supreme Court of Justice, located in Bucharest, with 90 judges. There is an intermediate court of appeal, in fifteen locations around the country, called the Court of Appeal. The upper- level trial court, organized in the 41 judets of Romania, is called the Tribunal. There is a limited jurisdiction trial court, operating in 187 locations, called the Judecatorie. Generally there are two levels of appeal. Cases that begin in the Judecatorie can be appealed to the Tribunal and then to the Court of Appeal, while cases that begin in the Tribunal can be appealed to the Court of Appeal and to the Supreme Court of Justice. 4.4 While this court structure is similar to many judicial organizations around the world, it has a few distinct features. First, there are several other courts that are not seen as part of the Judiciary at all, at least not as "courts of law," as defined in Law 92/1992. The Constitutional Court rules on the constitutionality of legislation and decrees, at least on the constitutionality of laws and decrees adopted under the 1991 Constitution. (The regular courts may adjudicate the constitutionality of earlier laws.) Further, there is a Court of Audit (which reviews government expenditures and collection of government revenue), and a Military Court system (which hears all matters involving police as well as controversies within the military itself). The budget process for these courts is not reviewed in this chapter. 4.5 The "courts of law" also have special features. The establishment and jurisdiction of the Supreme Court of Justice is different from that of the other courts, based on a different law and with key characteristics that are different. The other three levels of court operate under Law 92/1992 and its amendments, while the Supreme Court of Justice is chartered by the Constitution and Law 56/1993. The President of Romania appoints judges at all four levels, and the Constitution (Article 124) states that "judges appointed by the President shall be irremovable, according to the law." While judges in the Judecatorii, Tribunale, and Courts of Appeal serve unlimited terms until a mandatory retirement age (65, 68, and 70, respectively), judges on the Chapter 4: Budget Reforms in the Judiciary 73 Supreme Court of Justice serve for renewable six-year terms. Also, while the Court of Appeal is predominantly a court of cassation, hearing appeals on the record, the Supreme Court of Justice hears many of its appeals de novo, holding trials and taking evidence before it makes the final determination in these cases. 4.6 This court structure is still very much under development. There is tension between the Constitutional Court and the Supreme Court of Justice concerning whether the edicts of the Constitutional Court are binding on the courts of law. Also, there is a strong interest in changing the terms of the Supreme Court judges so that they, too, are unlimited until a mandatory retirement date. Related developments are underway concerning budgeting and administration of the courts as well, for example, a new law in May 2001 privatized more than 350 judgment enforcement officers and several current proposals under review will be discussed later in the paper. 4.7 Management. The management of all these courts is the responsibility of the Court Presidents. Chosen from among the judges of the court, they serve a four-year term and can continue in office for a maximum of two terms. The Court Presidents are assisted by one to three Vice Presidents, depending on the size of the court. Some Court Presidents estimate that they spend as much as 80% of their time on administrative tasks. 4.8 An important institution in the structure of the courts is the Superior Council of the Magistracy, or SCM. The SCM is chartered by Law 92/1992. It has fifteen members, ten judges and five prosecutors, elected by Parliament for four-year terms.51 The Minister of Justice chairs the SCM in most of its work but does not have a vote. The SCM has several roles. On the initial recommendation of the Minister of Justice, it recommends nominees for judgeships, prosecutorial positions, and court leadership positions (Court Presidents and Vice Presidents). These nominees then go to the President for final appointment. The SCM, also on the advice of the Minister, recommends judges for either promotion or transfer. It acts as a disciplinary body for judges and has the authority to discipline or remove judges but, interestingly, not prosecutors, since the prosecutors have a separate disciplinary structure. When acting as a disciplinary body, the President of the Supreme Court of Justice chairs the SCM, not the Minister. Finally, the SCM has a general duty to "watch the observance of the independence of Justice." (Law 92/1992 as amended, Art. 18) 4.9 One other institution deserves separate description in structure of the court system, the National Institute of Magistrates, or NIM. The Law on Judicial Organization, referred to here as Law 92/1992 but amended by Law 142/1997, authorized the creation of NIM to "carry out the specific training of future judges and public prosecutors as well as the improvement of the professional skills of magistrates holding office" (Art.. 70). The Institute trains new candidates for both roles - as judges and prosecutors - and there is an implicit guarantee that successful graduates will receive appointment. There are both one and two year programs for training of graduates of undergraduate-level legal education programs. NIM has been slower to begin 5'The selection of these members is complex. In a secret ballot, Parliament elects four judges from twelve candidates nominated by the Supreme Court of Justice, three prosecutors from nine nominated by the Prosecutor's Office of the Supreme Court, six judges from fifteen nominated by the Courts of Appeal, and two prosecutors from six nominated by the Prosecutor's Office of the Bucharest Court of Appeal. 74 Chapter 4: Budget Reforms in the Judiciary continuing education programs, but some do take place each year. NIM is overseen by a Director and a Council of eleven members. The latter includes judges, prosecutors, a representative of the Ministry of Justice, a representatives of the NIM training staff, and representatives of the students. Admission to NIM is done by examination. 4.10 The Role of the Ministry of Justice. The courts function largely as part of the Ministry of Justice. As mentioned above, the Minister of Justice chairs the Superior Council of the Magistracy and is responsible for recommending candidates for judicial' (and prosecutorial) positions to the SCM. Also, the NIM is located within the Ministry of Justice, and according to Art. 70 of Law 92/1992 the Institute is "exclusively subordinated to the Ministry of Justice".52 The Ministry of Justice represents the court system (apart from the Supreme Court of Justice) in the government, and the Minister is "responsible for the proper organization and functioning of the Judiciary as a public service" (Art. 18, Law 92/1992 as amended). Finally, the Ministry of Justice has a strong role in the Superior Council of the Magistracy. The Minister chairs the SCM, and its actions regarding the appointment, promotion, or discipline of judges are based on the recommendations or requests of the Minister. 4.11 The dominant position of the Ministry of Justice is also felt in the appointment of Economic Directors and chief accountants at each Tribunal. These directors and the accountants are responsible for managing the budget and reporting on its implementation. The dominance of the Ministry of Justice affects the appointment of the Economic Directors and the chief accountants primarily at the selection stage. The examinations for these positions are carried out by the Ministry and occur .in Bucharest only. The process is not entirely closed since a commission reviews the examination results, and the Ministry of Finance and the National Agency for Public Service are involved in the selection of candidates. Also, there is an unofficial local review of references by the examination commission, and the views of the incumbent Court Presidents. can be taken into account at that time. These Economic Directors are then assigned to a particular Tribunal and report both to the Court President and the Ministry of Justice. 4.12 Prosecutors. In the Romanian court system, the prosecutors in Romania are intimately connected to the operation of the courts. Judges and prosecutors together are called "magistrates," and together they make up the "magistracy." Prosecutors along with judges are members of the SCM, and they are trained together in the NIM. The appointment and promotion of prosecutors, or public ministers, as they are also called, occur as the Ministry of Justice recommends candidates to the SCM, which in turn recommends appointment to the President of Romania. Like judges, prosecutors are attached to particular courts, and they work together on the criminal caseloads of that court. 4.13 While the careers of judges and prosecutors seem to go together in many ways, in others the prosecutors have made themselves distinct. Discipline of prosecutors occurs not through the SCM, with its dual judge-prosecutor membership, but through a separate body consisting of .52 it should be noted, however, that the reason for this language is not to challenge an independent Judiciary's control over judicial training programs; on the contrary it is to make clear that other educational oversight organizations in the government do not influence the Institute, and the Institute remains an organization of the legal profession. Chapter 4: Budget Reforms in the Judiciary 75 prosecutors only. Also, the budget for the courts, as we shall see, is shown as part of the Ministry of Justice, while the budget for the prosecutors is shown separately, as the Public Ministry. In at least these two ways, the prosecutors have obtained independent status.53 Budget Formulation in the Judiciary 4.14 Budget Formulation. The construction of budget requests for the courts is a bottom-up process. Each Tribunal submits to the Ministry of Justice a budget request for its organization (itself, a Court of Appeal if there is one in its district, and the Judecatorie that are in that district 4 The budget request has three major components - Personnel, Goods and Services, and Capital. The Tribunale' requests are compiled by the Economic Department of the Ministry of Justice and compared with other needs and with past spending. The Ministry then submits an overall request to the Ministry of Public Finance, similar to other line ministries. The Ministry of Public Finance has the responsibility to review requests from throughout the government to match expected revenues and debt instruments. There are usually significant reductions made to the budget proposals by the Ministry of Public Finance before a budget is submitted to Parliament. At the time of these reductions, there is dialogue between Public Finance and Justice, and the Ministry of Justice has some opportunity to make its case - presumably not only for the courts but also for all the instruments of the Ministry. The Economic Department of the Ministry of Justice indicates that the cuts imposed by Finance are largely the product of two important items of information: the level of spending in the past year, and the priorities set by the whole Government, including both Ministers of Public Finance and Justice. Capital spending has more flexibility than the other two items that largely make up the courts' budgets, since initiatives to build or renovate buildings can more easily be delayed, and capital spending has in fact decreased in the past year as the Government has been forced to limit its spending.5 4.15 Non-budgeted Funds. In addition to the public funds, courts receive some non-budgeted funds. The most significant source of non-budgeted funds is the filing fees, which are collected by the courts and sent to the Ministry of Justice. The Ministry of Justice and the Ministry of Public Finance monitor these revenues jointly, as they know that they need to supplement them with additional funds from the general Treasury.56 The Ministry of Justice allocates these non- budgeted funds, and any revenue not spent at the end of the year is carried forward under the 5 Other commentators have noted several non-budget-related areas in which the boundary between the role of prosecutor and that of judge is evolving. For example, in criminal cases the prosecutor often sits up with the judge in the courtroom, while defense counsel is relegated to lower status. Also, the prosecutor exercises some quasi- judicial functions such as the exclusive authority to issue search warrants and arrest warrants. These arrangements seem not at all stable, however, and there is current pressure to alter the boundary between the two roles. m There are several other budget categories within the appropriation for the Ministry of Justice, as described in Catalin Pauna, Romania: Review of Public Expenditure on the Legal and Judiciary Systems, World Bank, February 2002 draft, but they are very small within the courts' program area. 's See Pauna, op. cit. 56 There was significant difference of opinion on whether the courts were a net cost to the Treasury or a "profit center," with court-generated revenue going to other government needs. While it may have been otherwise in past years, the data for the last few budget years seems to show clearly that the courts generate only about one-third of their overall appropriation. See Pauna, op.cit., pg. 5 76 Chapter 4: Budget Reforms in the Judiciary spending authority of the Ministry of Justice. The estimate is that these funds, however, are not too significant - perhaps only about 6% of the total funds available to the courts. 4.16 Program Budgeting. In line with the introduction of program budgeting, the Ministry of Justice has started to articulate goals with budget estimates applied to each goal. The idea is to better understand the result anticipated from the approval of a given spending item. For instance, if there is a proposal to expand a service or create a new court, that goal can be articulated, the cost for it set forth, and the decision-makers (the Ministry of Public Finance or the Parliament) can rank each goal across ministries in deciding which areas to fund or which to cut. However, notwithstanding the move toward program budgeting, the Ministry of Pubic Finance still attaches considerable weight to past spending patterns in reviewing budget requests. Indeed, if the Ministry of Justice can show that all its funds were well spent, without leaving a surplus of any size, it has a better chance to justify an increase in the following budget cycle. On the other hand, if the Ministry ended a year with surplus or with questions about its spending, it would likely face a cut in appropriations for the following year. 4.17 Other than the Supreme Court of Justice, which has its own budget category, the budgets for the courts, SCM, and NIM are all part of the budget of the Ministry of Justice. The other elements of the Ministry of Justice's budget are the support of the Ministry itself, the operation of the penitentiaries, and a few other small components. A School for Clerks has been authorized and is now in the planning stages; once in operation, it will presumably be a component of the Ministry's budget. The Ministry of Justice is therefore the "principal credit ordinator" for these budgets. The appropriation goes to the Ministry and the Ministry is accountable for the expenditure of those funds. The Supreme Court of Justice is its own principal credit ordinator. Budget Execution in the Judiciary 4.18 Budget Execution. The Ministry of Justice allocates funds to a secondary credit ordinator for purposes of administering the budget and reporting back to the Ministry. The secondary credit ordinator for the courts is not the Court of Appeal (i.e., the next highest ranking court for purposes of legal decision-making) but the Tribunal. The Tribunal has the responsibility for day-to-day budget management for all the courts in its judet, or district. Each Tribunal, then, will receive an allocation from the Ministry for its own operation, the operations of the Judecatorie in that district, and the operations of a Court of Appeal, in case one is located in the district of that particular Tribunal. There is a pending proposal to transfer the responsibility of s7 These filing fees are not paid at the courts themselves but at designated banks. The litigants then come to court with their filing papers and a receipt to show the fees that they have paid. At the time cases are filed, judges review the filings and confirm from the receipt that the proper fee has been paid. Since the accounting for the fee revenue is separated from the courts themselves, very little public money is actually handled within the courts. The fees are established by law (Ordinance 752/1999) on a sliding scale based on the amount of damages sought. They range from 2% of the claim to 10%. For a litigant with a sizable claim (most Civil litigation concerns property rights, especially property confiscated by the Communist government and former owners are asking for return or proper compensation), a 10% fee could be very high. On the other hand, those fees do not seem to deny access to the courts. Caseloads are growing at a fast pace, including civil caseloads with sizable claims, suggesting that these fees are not limiting access to the courts. Chapter 4: Budget Reforms in the Judiciary 77 secondary credit ordinator to the Courts of Appeal, which would make the Tribunale into tertiary credit ordinators. 4.19 It is critical to compare the budget treatment of the courts with that of the Public Ministry (i.e., the prosecutors). The head of the Public Ministry is the Prosecutor General, who is appointed by the Minister of Justice. The Prosecutor General is part of the overall Ministry of Justice, but the Public Ministry's budget is treated separately. The Public Ministry is the principal credit ordinator for the prosecutors throughout the system, and on the staff of the Prosecutor General are the budget analysts and human resources staff that administer the budget of the Public Ministry. 4.20 As mentioned above, the courts participate in a budget rectification process during the course of the budget year. It is an opportunity to respond to new demands not already funded in the primary appropriation process at the beginning of the year, but clearly for the courts it has not been a good opportunity at least over the last year or two. Instead, budget rectification has meant that cutbacks have been imposed, either due to overall budget deficits or to non-court-related (and non-Justice-related) needs that have required the courts to reduce spending. These cutbacks have particularly affected the funding for capital projects, which was reduced considerably in 2001. 4.21 Over recent years, the Ministry of Justice has probably funded the courts as well as it could, given the ability of the Government to meet funding needs. The entire budget of the Ministry of Justice in 2002 is about 2.53% of the total state budget (Table 4.2). Just a little bit less than half of that percentage, about 1.28% of the budget, goes to the three levels of court subordinated to the Ministry, and another 0.07% goes to the Supreme Court of Justice.5 The budget for these courts has therefore generally kept up with inflation over the past few years and then been able to grow, with new courts established every year, and with funds available for capital investment. (Only in late 2001 and 2002 did capital funding begin to slow down considerably.) The indication from the Ministry of Finance for next year, however, is that the total budget for the courts will decline in real terms. A real decline in funding, assuming that salaries will maintain pace, with inflation, means that there will be further reductions in Goods and Services and in Capital expenses. Table 4.2: State Budget Allocations to the Judiciary, 2001-2002 (% of State Budget) 2001 2002 Ministry of Justice 2.65 2.53 Public Ministry 0.57 0.55 Subordinated Courts 1.33 1.28 Supreme Court of Justice 0.07 0.07 Constitutional Court 0.02 0.02 Legislative Council 0.01 0.01 All 3.32 3.18 Source: Ministry of Public Finance. World Bank staff calculations. 58 Pauna, op.cit., pp. 1-3. 78 Chapter 4: Budget Reforms in the Judiciary 4.22 While the budget funding available to the Ministry of Justice and, as a result, to the court system has increased over the years, there are still many needs that are underfunded in the present system. As a result, without action to reduce the significant case overload, there will continue to be a need for more judgeships and more staff. Meeting these needs without due consideration for changes that reduce the number of superfluous cases will eventually come at the expense of maintenance of buildings and other capital expenditures. It will also delay needed investment in information technology and staff training of. 4.23 Example. One example of unmet funding needs of the judicial system comes from the Sixth Sector Judecatori in Bucharest. The court recently moved into new facilities, and they were much more spacious that their previous quarters. There are 30 judges of that limited jurisdiction court and a total of 28 staff. Three judges share an office and a courtroom. They have just one computer for all personnel in the court, but they no longer use it at all, since they are not able to obtain greater memory for it. In this environment of limited funding, very few statistical indicators of court performance, and no automated system to track either caseloads or spending, personalities and existing relationships influence budget allocations. 4.24 Budget Execution for the Courts. Court Presidents are responsible for the administration of their courts. The President of the Tribunal meets with the Vice Presidents, distributes cases to the several panels of the court, and remains aware of the operation of the tribunale and its needs. The President also meets with the Presidents of the judecatorii of that judet or region and learns of the administrative needs, including the budget needs, of the lower courts. At budget time, as the secondary credit ordinator, the President will aggregate the requests from the judecatorii, add to those the requests for the tribunale and (if in the region) the request from the Court of Appeal, and send the combined request to the Ministry of Justice. When the appropriation is approved and comes back to the tribunale , the Court Presidents are responsible for budget management for all three levels of court - the Court of Appeal, the tribunale, and the judecatorii. The Supreme Court of Justice, as a separate budget entity, receives and administers its own budget, as a principal credit ordinator. 4.25 Both the Supreme Court of-Justice and the tribunale have budget managers on the court staff. These Economic Directors, as they are called, have been hired since 1999 to manage the budget. The Court Presidents have generally delegated to the Economic Directors the right to approve expenditures at the local level, within the approved expenditure levels in the budget. This authority to approve expenditures extends, of course, to the judecatorii and Court of Appeal as well, since their budgets are part of the appropriation to the tribunale. They monitor expenditures monthly and report to the Ministry of Justice as well as to the Court President. The Economic Department in the Bucharest tribunale has a total of twelve staff, including an accountant, several bookkeepers, and payroll staff. 4.26 The monitoring of budget execution is done on a monthly basis, though spending reports that go from the courts to the Ministry of Justice in Bucharest. In Bucharest, the Economic Department of the Ministry reviews them and compares them to past spending and the current budget. The Ministry's Economic Department (which has 34 skilled staff) tracks the spending of the courts (and the other components of the Ministry of Justice, such as the penitentiaries), compares it to the budget and to past spending patterns, and offers analysis of any trends or Chapter 4: Budget Reforms in the Judiciary 79 problems it finds. In principle, the tribunal as secondary credit ordinators have the authority to spend on their own within approved budget ceilings. They may spend up to a set maximum in certain areas. For example, the local court may authorize repairs up to 25 million lei, which is the equivalent of one of its judges' monthly salary. However, the goods and services budget is very tight and spending is constrained. As a result, it is common for a tribunal to come to the Ministry's Economic Department for approval to spend. Also, no reallocation among budget categories is permitted without approval of the Ministry. The Ministry itself manages all capital spending projects, most of which involve the construction, purchase, or renovation of buildings and then equipping them. Budget constraints in the last year forced the Ministry to cut back on many capital projects. For instance, of sixteen court projects that were budgeted only five will actually be executed. 4.27 While the tribunale are the locus of budget management for the lower three levels of court, there is anecdotal evidence that there is little communication between these levels of court on budget execution matters. There is no automation in budget tracking, as monthly reports are compiled manually to be compared against the budget and past spending patterns. This makes it impossible to compare spending patterns with changing patterns of incoming cases, backlogs or other workload factor. As a result, there are no real measures in place, either at local courts or at the Ministry of Justice, to evaluate the spending patterns of any court. Budget Compliance and Evaluation in the Judiciary 4.28 The monitoring of budget compliance is done internally through the Ministry of Public Finance and the Courts of Appeal that perform expenditure audits, and externally, through the Court of Audits.59 Judge-inspectors have the responsibility for these audits at the Courts of Appeal. Their role is to review case handling in the tribunal and judecatorie within their region, and reports from the judge-inspectors go to both the Court of Appeal and the Ministry of Justice. Also, aided by auditors, these judge-inspectors look at case processing, financial administration, staffing, and bidding processes. The Ministry of Justice issues the regulations for these reviews and sets the job description of the auditor, and the auditor reports jointly to the President of the Court of Appeal and to the Ministry of Justice's Audit Department. 4.29 Budget evaluation and performance. At present, the performance measurement in the judicial system is done primarily at the level of individuals. The performance of judges is reviewed by Court Presidents and Vice Presidents, with an annual evaluation that reviews a judge's ability to interpret and apply the law, the quality of decisions, and the timeliness of actions. Also, personality factors such as initiative, perseverance, organizational aptitude, ability to supervise auxiliary staff, and relationships with colleagues are included. For Court Presidents and Vice Presidents, management skills are also evaluated. These evaluations are made initially by the Vice President who supervises a judge, then reviewed and confirmed or altered by the Court President, and sent to the Ministry of Justice. Judges may appeal their performance evaluations, and there are proceedings to reconsider what some judges may feel are unfair reviews. 59 Formerly, internal audits were the joint responsibility of the Ministry of Justice and the Ministry of Public Finance that conducted period audits of court expenditures. Beginning in 2001, the Ministry of Justice's audit role was transferred to the fifteen Courts of Appeal. 80 Chapter 4: Budget Reforms in the Judiciary 4.30 There is a need to develop indicators of performance measurement and accountability. This is a key step that needs to be taken before implementing any decentralization of expenditure responsibilities in the judicial system, since otherwise it will be difficult to measure whether a particular court is performing satisfactorily. There are a few clear indicators that need to, be considered, such as corruption or outside influence in decision-making. Also, there is a recognition of the appropriateness of using caseload statistics - cases coming in the court, cases being decided and going out, cases being reversed on appeals -- in order to maintain some review of the quality of decision-making, the time it takes to decide cases, and the amount of backlog that builds up. The problem is that the Romanian courts cannot rely on one or two broadly defined indicators of performance measurement, because the responsibility of the courts is to balance contrary factors. For example, while the fast turnaround of cases is clearly desirable, courts should not rush decisions without listening to the evidence and to the legal argument. Also, courts need to carefully consider the impact of any particular decision on the litigants and on the use of this case as precedent for other decisions. There is a need to reach a balance between having the courts be deliberative, wise, and timely. Similarly, the amount of deference that courts give to the police and to the interests of the government also needs to be balanced. The courts need to validate the work of law enforcement to ensure public safety, while taking care not to merely follow the dictates of the police and the prosecutors. 4.31 These concerns are directly relevant to the courts of Romania. There are real concerns about how to measure workloads, especially considering that only limited resources are available and no measurement of how effectively these resources are being used. Also, public opinion surveys indicate that the courts are not well trusted. It is widely accepted that there is considerable corruption. The Bank's corruption diagnosis indicates that courts are the second lowest-ranking institutions in the country (after customs offices) in the public's level of confidence.6 These concerns need to be addressed if additional funds are going to be allocated to the judicial system, especially since budget proposals have no information on courthouse users' feedback, and only relatively simple measures of workload, when justifying the need for additional resources. 4.32 Fortunately, collecting feedback from court users has become increasingly common, and is now a regular component of performance management systems elsewhere in the world. User surveys can be introduced very easily and at very low cost, providing information that is available from no other source. Options for user surveys include (i) asking courthouse users complete short form asking them which services they received, their level of satisfaction with these services, and whether they were asked to pay a bribe; and (ii) brief surveys of users, judges, prosecutors, and lawyers conducted immediately after a case is heard, asking each party to evaluate the other. These surveys, that could be conducted on a periodic basis, have been used in a number of OECD countries and are currently being piloted in Latvia. 4.33 Also, workload-related performance standards for courts have been developed in other countries, and are now widely available. These standards fall under five broadly defined categories. These are: (i) access to justice, meaning that courts should be open and accessible; (ii) expedition and timeliness, since unnecessary delay is a primary cause of diminished public 6 World Bank (2001), Diagnostic Surveys of Corruption in Romania. Chapter 4: Budget Reforms in the Judiciary 81 trust and confidence in the court system; (iii), equality, fairness and integrity, so that courts ensure equal justice under law, adherence to relevant law, and effectiveness in the results of court orders; (iv) independence and accountability, to ensure that courts protect judges from unwarranted pressure while being able to manage themselves and remain accountable to the public; and (v) public trust and confidence, since justice should not only be done, but should also be seen to be done. 4.34 A review of these workload-related performance standards su gests that nine of the standards across the five categories may be relevant to Romanian courts. These are that (i) the court conducts its proceedings and other public business openly; (ii) the court establishes and complies with recognized guidelines for timely case processing while, at the same time, keeping current with its incoming caseload; (iii) the court disburses funds promptly, provides reports and information according to required schedules, and responds to requests for information on an established schedule that assures its effective use; (iv) the court promptly implements changes in law and procedure; (v) the court procedures faithfully adhere to relevant law, procedural rules, and established policies; (vi) the court gives individual attention to cases, deciding them without undue disparity among the cases and upon legally relevant factors; (vii) the court takes appropriate responsibility for the enforcement of its orders; (viii) the court responsibly seeks, uses, and accounts for its public resources; and (ix) the court informs the public about its programs. 4.35 The priority for implementation of these indicators, meaning collecting the data and establishing benchmarks against which courts should be held accountable, should be the timeliness of case dispositions and the backlogs in cases pending. This is critical for better understanding the strengths and weaknesses in each court's ability to handle its caseload and to administer itself effectively. Other standards should also be developed, such as openness and transparency in operations and ensuring that case decisions apply current law and regulations. These aim at building more public trust and confidence in the court system, and in the budget process. Developing the Courts' Administrative Capabilities 4.36 The Ministry of Justice has been a responsible steward for the Romanian courts. While it has not been able to provide all the resources that the courts need, the Ministry has managed the courts' budgets (other than the budget of the Supreme Court of Justice) with professionalism and with a real effort to fund their legitimate needs. The Ministry is familiar with the need for both operational funds and capital investment, and it effectively manages budget appropriation and the human resources within the existing staffing and resource limitations. The Ministry acknowledges the need for an automated system to track expenditures and to conduct analyses, and it recognizes that it is not able to really identify which courts need resources the most given the information resources available. 4.37 However, as the court system matures and develops, new institutions must develop. The reliance on the Ministry of Justice should be seen as temporary. The Ministry should retain its role as policy-maker, but the subordination of the courts to political interests or to the interests of 61 Fetter (2002), Fiscal operations of the judiciary in Romania, mimeo. 82 Chapter 4: Budget Reforms in the Judiciary the prosecution in criminal cases would be a threat to the independence of the judiciary. This is a very real threat. Judges are acutely conscious of the need to remain in the good graces of the officials in the Ministry. They dare not stray very far from the political priorities and programs of the Minister of Justice, since the Minister has a strong and direct influence over their budget, the potential for capital improvement in their court facilities and equipment, their own promotion to higher courts, and their continued roles as court residents or vice-presidents. These levers of control are much too strong for all but the most fearless of leading judges to ignore. Indeed, several instances have occurred in recent years indicating direct influence by the Ministry of Justice over the decision-making of the courts. Some of the most controversial cases before the courts involve the return of property that had been nationalized under the Communist government, and in 2001 the Ministry offered advice on making decisions in those cases. Also, the Ministry made recommendations that current proceedings in cases involving bankrupt banks be suspended.62 These and related actions raise real concerns about judicial independence, especially when combined with the administrative controls that the Ministry exercises. They are quite distinct from the several avenues of appeal open to the prosecutors and the Minister of Justice, which are protected in the law. 4.38 Taking action to ensure the independence of the Judiciary would greatly reduce these avenues of interference and be entirely consistent with the official policies of Romania. Article 123 of the Constitution of Romania proclaims that "justice shall be rendered in the name of the law. Judges shall be independent and subject only to the law." Article 18 of Law 92/1992 indicates that "the Superior Council of Magistracy and the Minister of Justice safeguard the independence of justice," and Article 1 of Law 56/1993 on the Supreme Court of Justice announces that the Supreme Court of Justice "oversees the correct and uniform enforcement of law by all courts." Assuring that these policies are upheld and protected in future governments will require the courts to develop their own administrative capabilities, especially their budget capabilities and their separate submission of budget requests to the Government. 4.39 There are both short-term and long-term recommendations that can help to accomplish the goal of establishing the courts' own administrative capability. Among the short-term recommendations, at least one key recommendation is already underway. There is a proposal that the Courts of Appeal become the secondary credit ordinator for regional budget administration within the courts, and it should be encouraged. This proposal would necessarily diminish the budget management authority of the tribunale; they would become tertiary credit ordinators and receive an allocation from the Court of Appeal for their own operation and that of the judecatorii within their judet or geographic region. The advantages of this increase far outweigh the effect of this diminution. 4.40 Having only fifteen secondary budget offices instead of forty-one is a significant step in itself. The national-level oversight, now part of the Ministry of Justice, would be much more effective in tracking and evaluating budget management if there were fifteen key managers with whom to interact instead of the unwieldy number of forty-one. Norms and standards could more easily be set, since all fifteen offices would have a certain size of operation, probably with several courts of different sizes within each of them. Oversight might well be more effective in the fight against corruption. There would be only fifteen key offices, and officials of the 62 See Open Society Institute, Judicial Independence in Romania (2001), pp. 365-366. Chapter 4: Budget Reforms in the Judiciary 83 Ministry would become quite familiar with the strengths and weaknesses of that number of offices. The Courts of Appeal, in turn, would have only a handful of subordinate offices among the tribunale and judecatorii in their areas, and they could keep an effective watch for instances of corruption, at least corruption with public money. 4.41 This greater budget authority for the Courts of Appeal would naturally combine with the personnel authority they already have. Many of the auxiliary staff in the lower courts are already hired either directly by the Courts of Appeal or through the competitive examinations that the Courts of Appeal administer. The personnel role that these courts already have leads naturally to a budget role as well. It would naturally still be necessary to develop knowledgeable staff at the higher level to administer the budgets. The position of Economic Director was established in 1999 for each tribunale, and now that same position would need to be created for each Court of Appeal. While such positions would be necessary, it need not involve significant growth in administrative staffing for the courts. First, some of the staff of the Courts of Appeal already monitor their budgets. Second, the staff persons who handle personnel actions for the courts in the area of the Court of Appeal could be trained to do some of the financial administration work. Third, members of the staff of the Economic Directors for the tribunale might be able to move directly up to the Court of Appeal to perform some of the support work at that level, since there is not much new work required, only a shifting of the location from the tribunale to the Court of Appeal. Creating the Economic Director's position itself at the fifteen locations would create more staff, but perhaps that creation would offer good career advancement potential for the better Economic Directors of the tribunale. 4.42 Training the Economic Directors in the Courts ofAppeal. The second measure is to train the Economic Directors thoroughly, both at the new level of the Courts of Appeals and at the tribunale. While these staff persons could emerge as the leading candidates to administer routine court operations, they need to be trained. Most of the training takes place now on the job. Training should allow the Economic Directors to become comfortable with regular budget administration, including how to justify requests, and how to manage major projects like the administration of capital investments (i.e., the local, on-site supervision of renovation or construction of courthouse facilities, recognizing that these projects will always need central coordination and leadership). 4.43 Third, the Ministry of Justice should gradually loosen the degree of oversight in budget matters that it now holds. Dealing with only fifteen secondary credit ordinators, and working with better-trained and more experienced Economic Directors, the officials at the Ministry should be able to loosen the controls and let the courts themselves administer their own budgets to a greater degree. At this stage in the development of budget management, the Ministry would still have an important leadership role. It would present the courts' budget to the Ministry of Finance and support it in Parliament, and would allocate the appropriation and the authorized judgeships and auxiliary staff positions among the areas supervised by the Courts of Appeal, but it would not manage or oversee the day-to-day administration beyond that. The Ministry would represent the courts on budget matters with the rest of Government, and would allocate the available resources (both the state-funded appropriations and any outside-funded projects such as from the European Commission or similar sources). Also, the Ministry would receive audit 84 Chapter 4: Budget Reforms in the Judiciary reports concerning how funds were spent. The regular budget management, once the resources are allocated, however, would be transferred to the Courts of Appeal and the tribunale. 4.44 The fourth measure is designed to increase openness and transparency. The budget and personnel allocations should be transparent, and leaders at all levels should be able to manage within their budgets. The Courts of Appeal should be allocated a certain budget and personnel lines, and they should manage within it, seeking permission from the Ministry for extraordinary changes and new needs, but not seeking approval of regular spending already included in the approved budgets. Likewise, the tribunale should get information on their budget allocation from the Courts of Appeal, with this information being known to all in the system and publicly available if an outsider wanted to know it. The budgets of the judecatorii should be administered at the tribunale level, as they are now, but they should be known to the leaders of the judecatorii and again, available to the public upon request. 4.45 Transparency in budgets has two advantages. The officials at all levels of the court system should know what they have available and should be realistic in their expectations of financial support. They will be able to perform their responsibilities of administration and leadership better. Also, openness is good to increase public confidence in the courts. Citizens should be able to see what the budgets are; if they are low, they will know why court services are not as good as they would like, and if they are high, they will be able to ask good questions to assure that the public funds are being well spent. 4.46 The fifth, and last, short-term recommendation is to improve the management of the information needed for the budget of the courts (and the Ministry of Justice), as well as the information needed for other administrative tasks. At present, there is no effective way to have good oversight of budgets; so the Ministry of Justice considers that it must keep constraints on the Economic Directors and Court Presidents, and even small budget decisions need to be approved by the Ministry. Also, there are few analytical tools that would show how well a particular court is using its budget. All the interviews on this point indicated that the main tool was a comparison from one year to the next in the spending patterns of the courts. Finally, there is a major need for post-audits to identify inappropriate spending as part of the court inspections that the Ministry of Justice carries out. Better managed information can accomplish much in simplifying budget oversight, while still giving local courts considerable flexibility in spending their appropriations. Spending decisions can be made locally, and entered into a common information system. Hierarchical superiors, both in the courts and in the Ministries of Justice and Finance should have access to this information, so that at any point they can examine the spending and review analytical reports on whatever spending patterns they want to study. 4.47 These short-term recommendations would not alter the current structure of the courts or the role of the Ministry of Justice. They take advantage of momentum (i.e., toward a greater administrative role for the Courts of Appeal) that is already underway, but otherwise they build on the foundations of the current operation of the Romanian justice system. There are several longer-term recommendations, however, that do begin to shift the balance of power and to develop further the independence of the Judiciary. .-Chapter 4: Budget Reforms in the Judiciary 85 4.48 The first of these long-term recommendations is to restructure the composition of the Superior Council of Magistracy and to enlarge its role in national leadership of the courts. Currently the Council (the SCM) has fifteen members: three members of the Supreme Court of Justice and seven judges from the Courts of Appeal, and two prosecutors from the Supreme Court of Justice and three from the Courts of Appeal. The Minister of Justice serves as non- voting chair for most of its work, and when it sits as a disciplinary body for judges the President of the Supreme Court of Justice chairs it. The major role of the Minister of Justice and the large percentage of prosecutors as members of the. SCM prevent it from becoming an effective manager of the whole court system. 4.49 There is nothing wrong with the Minister of Justice taking a role with the SCM, and there is no reason not to include a few prosecutors as members. The leadership and the majority of members should, however, be constituted of judges - judges who are not under the influence of the Ministry of Justice. They should be able to take an effective leadership position on the Council, without fear of reprisal. Several more judge-positions should be added to the Council. Also, when the conference of fifteen court presidents of the Court of Appeal takes root, all fifteen could be considered as SCM members. In addition, some leaders of the tribunale could well become SCM members. But if the Minister of Justice and five prosecutors remain on the Council, then there should be perhaps between fifteen and twenty judges. Further, the chair of the SCM should go to a judge, almost certainly the President of the Supreme Court of Justice in order to command respect as the leader for the Judiciary. Such a restructuring of the SCM would make it a body well constituted to take administrative leadership of the court system. 4.50 Representatives of other stakeholders in the justice system could also be considered for SCM membership. Stakeholders to be considered could include representatives from the Romanian Bar Association, respected academics, and leaders from outside the court system. The goal would be to open up the work of the SCM, so that it is better respected by the public, not as a political body following the policies of one political party or another, but as a non-partisan group dedicated to providing an excellent system of justice for Romania. Their recommendations, of course, about judicial appointments and court budgets (see below, next recommendation) would necessarily enter the political arena for decisions, but the public should have a high level of confidence that the work of the Council is based on more than partisan interests. 4.51 A Superior Council of Magistracy of between 10 to 15 permanent members (with an additional 15 and 20 part-time members) would retain its current duties (recomm6nding appointment decisions to the President of Romania and making decisions about discipline of judges), and assume the Ministry of Justice's current role of principal credit ordinator for the Judiciary. The prosecutors have their own body for discipline, and that role can remain outside the SCM's jurisdiction. Also, the prosecutors have a separate principal credit ordinator, the Prosecutor General, and the courts should follow their example. 4.52 The SCM will naturally develop the ability to lead the administration of the Judiciary. Budget management should be one of the first of the tasks that flow to the enlarged and restructured SCM. The Prosecutor General separately submits the budget request for prosecutors to the Ministry of Finance and to Parliament, and it serves as the principal credit ordinator for 86 Chapter 4: Budget Reforms in the Judiciary the Public Ministry or prosecutors. That same model, already established for prosecutors, could operate for the SCM and the courts. The SCM could assemble budget requests from the Courts of Appeal (and eventually from the Supreme Court of Justice as well, see below), negotiate with the Ministry of Finance, and defend the budget request in Parliament. Once the appropriation has been made, the SCM could take responsibility for the allocation decisions within the courts, and it could perform the national-level oversight and management that the Ministry of Justice now does. 4.53 While system-wide budget management would be a new role for the Council, there are precedents in other countries. Countries such as Sweden, Denmark, and Ireland have judicial councils similar in structure to Romania's that not only recommend candidates for judgeships and investigate cases for judicial discipline when necessary, but also represent the court system with the Parliament and oversee administrative functions in budgeting, personnel, and information technology. The Superior Council of Magistracy would need full-time staff support to fulfill these enlarged responsibilities. These responsibilities could initially be fulfilled by some of .the staff already in the Ministry of Justice's Economic Department and Human Resources Department, who would continue performing their current tasks. Also, some of the best Economic Directors at the tribunale level could be considered for assignment in the central office working for the SCM. Over time, there is no doubt that SCM would develop a separate staff entirely, just as the Prosecutor General has for its budget and personnel management, and it is likely of course that there would be some growth of total administrative staff. It is clear, however, that the administrative functions in the Ministry now for the courts are overburdened, and some growth is necessary even without the restructured role for the Council. SCM staff would gradually assume more and more expertise in the operation of the courts themselves, and they would be able to advise local courts on ways to improve their budget management and overall administration. 4.54 There is one concern about making the SCM responsible for system-wide management of the Judiciary's budget, personnel and other administrative work. The members of the SCM must be held fully accountable for the whole court system, not just determined to give their own courts as much of the limited resources available. If that were to occur, then the SCM would not perform an effective leadership role. The best way to avoid such a development is to rotate membership on the Council. While terms should be reasonably long so that expertise and familiarity with administrative issues develops, no court representative should remain a Council member for more than six years. Also, building a close connection between the SCM and the conferences or councils of Court Presidents would help to assure good representation of all courts on the SCM. 4.55 The Superior Council of Magistracy should assume a leadership role with the School for Clerks. The School is just now getting organized and operational, and the SCM can and should exert a strong role. Training of the auxiliary staff is critical for the long-range effectiveness of the courts. Currently judges do almost all of the professional work in the courts because many auxiliary staff are not able to do them; that means that judges oversee case filings and the making of the record and many other tasks that auxiliary staff could easily do if they were properly trained. The result would be economical since the salaries paid to these staff would be less than those paid to judges, and the judges would get greater satisfaction from concentrating on their Chapter 4: Budget Reforms in the Judiciary 87 judicial tasks of decision-making in the cases before them. But in order for this development to occur the clerks must be trained. There seems little need to merge the School for Clerks with the National Institute for Magistrates, since that body must train candidates for both the Judiciary and the Public Ministry. By contrast, the School for Clerks is concerned with auxiliary court staff only. 4.56 Currently Law 92/1992 makes clear that the National Institute for Magistrates (NIM) is exclusively subordinated to the Minister of Justice (Article 70). Such a structure makes sense as long, as NIM trains both judges and prosecutors, and that joint training is efficient and reasonable. But the enlarged and restructured SCM should definitely have a role in overseeing NIM and its curricula. A larger role for SCM could be reflected in NIM's governing council and in some sharing of responsibility with the Minister of Justice for its operation. If the budget for NIM were a joint responsibility of the Ministry and SCM, then there would be a real sharing of authority for its work. 4.57 The next long-range recommendation is to build the Supreme Court of Justice more fully into the leadership of the Romanian court system. Now, of course, it is the final step in the appellate process, but it is hampered by its dependence on six-year terms for its members and by its need to consider many appeals with full new trials rather than as a court of cassation, hearing appeals on the record below. Putting the Supreme Court of Justice more fully into the rest of the court system, with lifetime terms up to a mandatory retirement age and with jurisdiction as a court of cassation would strengthen the independence of the Judiciary. The Supreme Court of Justice would then take its legitimate place as the leader of the system. 4.58 The last long-range recommendation is to build the Performance Standards and accountability measures outlined in paragraphs 4.32 and 4.34 into the management of the court system. Romanian courts have only few measures to judge how well courts are performing. These measures are often limited to the percentage of pending cases, and the ratio between the number of decisions issued and the number of cases filed. These measures are easily distorted by pressuring judges to rush decisions, sometimes without due consideration to the legal arguments involved. Court leaders, as well as their hierarchical superiors in Bucharest, Parliament and the public, need to have reasonable measures to judge the efficiency and the effectiveness of the court process. The adoption of these Court Performance Standards could also help' protect the court against corruption, something that every country needs to guard against. 4.59 The nine standards outlined here and drawn from the Trial Court Performance Standards are currently in use in many courts in the United States. The Superior Council of Magistracy could adapt these standards to the courts in Romania and set priorities on which ones could be implemented first. It is too time-consuming to implement all nine standards at the same time in the same court or group of courts. But two or three standards could be undertaken. They involve measures of case management performance, examination of records for handling funds, surveys of users of the court, and other techniques to determine how well the courts are performing. These standards, or ones very similar to them, could be institutionalized. They could be tested in a small number of courts and then implemented throughout the system in a phased way, such as three per year over three years, so that all nine are used within a reasonable amount of time. 88 Chapter 4: Budget Reforms in the Judiciary 4.60 Measuring court performance should be done in a transparent way so that the public confidence in the courts grows. If the public were more convinced of the integrity and effectiveness of the courts, the ability of the courts to carry out their legitimate functions would be enhanced immeasurably. In short, measuring performance, and making the results public, benefits both the courts and the Romanian population. Conclusions and Recommendations 4.61 Romania has made improvements over the last decade in how it manages its court system. It is struggling now with large caseloads, considerable delays in resolution of cases, concerns about corruption, and a budget system that is inflexible. The seeds of improvement are, however, already planted. The system itself is evolving, and the institutions are young. Establishing budget management at the level of the Courts of Appeal, which is already under consideration, opens up a number of improvements in operation. Restructuring the Superior Council of Magistracy to become a real leader of the administration of the court system opens up many more. These major steps, which are natural developments from the foundation that the Ministry of Justice and the Parliament have already laid in Romania, will lead to a substantial improvement in the effectiveness of the court system. 4.62 Other recommendations aimed both at increasing efficiency in resource allocation and at enhancing the independence of the judiciary include the following: > Training the Economic Directors at the Courts of Appeal (to be established) and at the existing ones at the Tribunale. These Economic Directors should be fully trained in their duties, including the authority they have for budget administration. This step will allow these Economic Directors to take full responsibility for their budgets, and train them on how to prepare justifications for future budgets. > Delegating most of the budget administration to the Courts ofAppeal, which in turn should delegate some responsibilities to the Tribunale. The Ministry of Justice would retain its external responsibility for presenting the budget for the time being, and it should continue to allocate resources, delegating the routine budget administration to the courts. > Increasing budget transparency. All the institutions involved in the budget of the courts should be familiar with the structure of the budget and how resources are allocated. Also, these budgets should be available to the public. >. Implementing performance standards and measures in the courts. There are a number of performance standards and measures that would assure the accountability and the effective performance of the courts at all levels. The priority for implementation of these indicators, meaning collecting the data and establishing benchmarks against which courts should be accountable, should be courthouse users' feedback, the timeliness of case dispositions, the number of reversals at the appeal stage, and the backlogs in case pending. This is critical for better understanding on the strengths and weaknesses in each court's ability to handle its caseload and to administer itself effectively. CHAPTER 5. IMPROVING RESOURCE ALLOCATION IN THE HEALTH SECTOR Introduction 5.1 In 1997 Romania adopted a Law on Social Health Insurance that initiated the transformation of the health system from a state financed model to a social insurance-based model. The key provisions of this law regulate the revenues of the health sector, and the allocation of funds across health services and health care providers.6 Through this social insurance-based system, health care providers are contracted by the District Health Insurance Funds (DHIFs) to deliver health services to their insured. This system is designed to allow a better matching between patients needs and the services provided by making the allocation of health insurance funds responsive to the demands of the insured. This matching of needs to services is designed to improve the efficiency in the allocation of resources, and increase service provider accountability to the insured. In practice, however, the system has operated as a hybrid between social insurance and a publicly managed system, where revenues are collected as health insurance contributions, while the Government still exercises a considerable amount of discretion in allocating funds. 5.2 The Government, through the budget approval stage, defines overall expenditure limits for the National Health Insurance Fund (NHIF). This aims at preventing the health system from spending all of the revenue raised from health insurance contributions, so that the resultant surpluses can be used to offset deficits in other social insurance budgets, namely the pension fund. This has created tensions within the system, weakening accountability and financial control, as the practice of having expenditures still defined by the central government is inconsistent with the contributory social insurance model adopted by Romania. The excessive control exercised through the budget process from outside the health insurance funds reduces the ability of the funds to develop stable, rational funding agreements with providers. This in turn leads many providers to develop mechanisms to deal with the budget uncertainty (including substantial arrears), with the perverse effect that overall level of financial control of the health system is reduced and weakened. 5.3 The result is that the Romanian reformed health care system has not yet fulfilled the expectations created in 1997. Although health insurance did succeed in increasing the revenues available to the sector, one third of the users of the health care system still report having to make informal out-of-pocket payments for services. Also, life expectancy and other health status 63 With this 1997 law, insurance membership was made mandatory and linked with employment. Contributions reflected employees' income and are paid in even shares by the insured and the employer. Children and young people, disabled persons, and war veterans, as well as dependents without income have free access to health insurance. For conscripted soldiers and people serving prison sentences, insurance contributions are paid from the budgets of the Ministry of Defense and Ministry of Justice. "These include hospitals and their outpatient units, diagnosis and treatment centers, health centers, dispensaries, and medical offices). 65 Although the surplus is nominally retained by the NHIF, it is not authorized to use it and, as a result, the value is eroded by inflation over time. 90 ChapterS: Improving Resources Allocation in the Health Sector indicators are still well below EU and regional averages. While this is in part due the country's low starting point at the beginning of the transition, and the health financing reforms cannot be held entirely responsible for these trends, the new system has not succeed in reversing some of the trends. Morbidity indicators are still very high, and the incidence of preventable diseases - tuberculosis66 and cardiovascular67 - has increased since 1995 (Figure 5.1). Since many of these persistent health problems can be dealt by controlling risks factors, and by health promotion and prevention measures, there is a concern that the system has been allocating resources primarily according to the interests of the providers, and failing to adequately fund public health programs. Figure 5.1. Incidence of Tuberculosis, Selected Central European Countries, 1970-2000 (Rates per 100,000) 120 100 80 60 40 20 1970 1975 1980 1985 1990 1995 2000 Bulgaria - - Poland Czech Republic 0 Romania - - - Hungary Source: WHO Health for all database. 5.4 This chapter reviews the evidence available to date regarding impact of financing reforms in the health sector, with a special focus on expenditure allocation processes and their relationship with the efficiency and effectiveness of using public funds in achieving the desired outcomes. The chapter is organized as follows. It begins with a brief overview of the structure of the sector and recent trends in health status indicators. Next, it examines the budget process: design, implementation and evaluation. There is a particular emphasis placed on evaluation, since the reforms have been in place for several years. This analysis relies on both administrative and survey data, assessing the efficiency and equity impact of the reforms. The chapter closes with a summary of the policy recommendations. Overview 5.5 The provision of health care services in Romania includes primary health care delivered by family doctors that are independent practitioners contracted by the (public) health insurance 66The incidence of tuberculosis declined rapidly during the 1970s, stagnated during the 1980s and has increased dramatically during transition taking Romania back to the high levels achieved during the mid 1970s. 67 As in other countries in the region, Romania's morbidity and mortality patterns are characterized by a disproportionate burden of heart and circulatory diseases stemming from preventable lifestyle factors, particularly, tobacco consumption, alcohol abuse, and poor dietary habits. Chapter 5: Improving Resource Allocation in the Health Sector 91 fund but operating from their own offices. It also includes ambulatory secondary care delivered by the network of hospital outpatients departments, centers for diagnosis and treatment and office-based specialists. Finally, there is the tertiary care that is provided in teaching hospitals and specialized hospitals, most of which are publicly owned and are under state administration. 5.6 These facilities combined provide nearly 167,000 hospital beds, including short-term acute care and long-term care beds, representing an average of 7.4 beds per 1000 people. The regional distribution of these beds range from 10.5 beds in western Romania and in Bucharest to 6.9 in southern Romania. The number of hospital admissions is higher than in most European countries (Table 5.1), with a marked increase in admissions since 1997 and an associated decrease in average length of stay. This reflects the payment arrangements introduced by the health insurance scheme in 1998, where admissions are an important indicator for service contracts, and hospitals are not reimbursed for stays beyond the average for a range of medical treatments. Table 5.1: Health Services Utilization and Resources, Romania and Selected European Countries 1997 or most Out-patient Inpatient Physicians Hospital Occupancy Average recent available visits (per Admission (per 1000 Beds (per rate (%) Length of year capita) Rate (per population) 1000 Stay 100 population) (days) population) Romania 1997 4.2 20.9 1.8 7.4 78.6 10.1 Romania 2000 3.6 22.4 1.9 7.4 75.3 8.8 Austria 6.5 25 3.5 9.3 74 11 Bulgaria 5.5 14.8 3.4 7.6 64 11 Czech Republic 14.5 19 3.0 6.7 70.8 12 Germany 6.5 21 3.4 9.7 76.6 14 Hungary 13.7 24 3.4 9.0 75.8 11 Italy 6.6 16 5.5 6.4 76 10 Poland 5.3 14 2.3 6.2 .. 11 Slovak Republic 16.4 20 3.0 7.5 77.9 11 Slovenia 7.1 16 2.1 5.7 75.4 11 Sweden 2.8 18 3.1 6.3 77.5 8 European Union 6.1 18 3.2 7.9 75.5 13 Source: Radulescu (2002). 5.7 The public health expenditure during the 1990 to 2000 period ranged between 2.8 to 3.8 percent of GDP (Table 5.2), which is equivalent to $28 to 58 per capita. Adding another 0.7 to 0.8% of GDP in private expenditure on health, brings the estimated total expenditure on health services (recurrent and capital, public and private) up to 4.6 percent of GDP. This level of spending on health service places Romania at the lower end of the spending distribution among countries with a similar per capita GDP, as well as among most other countries in Central and Eastern Europe. For instance, spending by the public sector (the only comparable spending data available across countries in the region) is significantly higher in other Central Eastern European countries, such as Hungary and Croatia, where public spending on health is respectively over 6 and 8 percent of GDP. Similarly, in countries that established health insurance schemes during the 1990s, such as the Czech Republic and Slovakia, health spending increased by about two percentage points to over 7 percent of GDP. The latter is important because there was a shift in 92 ChapterS: Improving Resources Allocation in the Health Sector revenue source for the health sector with the introduction of the health insurance scheme, moving from a diversified pool of taxes to a heavy reliance on payroll taxes (Table 5.3). Table 5.2: Trends in Health-Care Expenditure in Romania, 1990-2000 (by source of funds, in current billions of Lei unless otherwise specified in the table) Source of Funds 1990 1991 1992 1993 1994 1995 1996. 1997 1998 1999 2000 State budget 24 62 153 366 1,003 1,324 2,009 4,419 3,184 3,518 3,710 External credits 6 10 86 119 74 123 483 890 2,390 Local budgets - - - 102 263 408 616 1,325 69 95 105 Special health fund - - 32 126 283 414 530 1,197 462 492 765 Health insurance - - - - - - - 7,404 15,482 24,817 Total public expenditure 69 24 62 191 604 1,636 2,265 3,228 7,064 11,458 20,067 30,395 Total public expenditure 1,092 813 621 794 989 1,114 1,047 985 1,291 1,309 (millions of US$) Public expenditure as share 2.9 2.8 3.2 3.0 3.3 3.1 3.0 2.8 3.1 3.8 3.8 of GDP (percent) Private expenditure on 7.5 16.5 na na Na 461 782 1,792 3,073 4,580 6,479 health, value in current prices (billions of Lei)70 Total expenditure as-share 3.7 3.6 ... ... ... 3.7 3.7 3.5 3.9 4.6 4.6 of GDP (percent) Source: Ministry of Public Finance and National Commission for Statistics. Table 5.3: Main Public Sources of Health Financing in Romania, 1990-2000 (%) Source of finance 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Taxes 100 100 98.3 94.7 94.7 97.7 98.3 32.4 20.5 15.1 - general 100 100 80.1 60.6 61.3 58.5 62.2 62.6 27.8 17.5 12.2 - local - - - 16.9 16.1 18.0 19.1 0.6 0.5 0.3 - earmarked - - 16.8 20.9 17.3 18.3 16.4 16.9 4.0 2.5 2.5 Statutory Insurance - - - - - - - - 64.6 75.1 77.0 External loans - - 3.1 1.7 5.3 5.3 2.3 1.7 4.2 4.4 7.9 Source: Ministry of Public Finance. 5.8 Are the low levels of public expenditure on health reflected in the health profile of the population? The statistics on life expectancy at birth for males provides an illustrative answer. Male life expectancy in Romania, at 67.8 years in 2000, is the lowest in Central and Eastern Europe other than the Newly Independent states of the Former Soviet Union (Table 5.4). There are positive developments, however. In 1998, life expectancy for males began rising, in a pattern similar to other Central European countries, such as the Czech Republic, Poland, Bulgaria, Hungary. " Includes amounts used for partial reimbursement of prescription drugs from 1992 to 1997. 69 Consolidated expenses, including interest and loan amortization payments for past investment loans in the health sector. 70 Based on survey data Chapter 5: Improving Resource Allocation in the Health Sector 93 Table 5.4: Male Life Expectancy at Birth, Selected European Countries (years) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Albania 69.6 n.a. 69.8 70.3 71.9 71.5 71.3 70.8 71.6 n.a. n.a. Bulgaria 68.3 68.4 67.8 67.6 67.2 67.4 67.5 67.0 67.4 68.4 68.5 Croatia 68.7 66.1 67.1 68.5 69.2 69.3 68.7 68.6 68.6 68.9 69.1 Czech Republic 67.6 68.3 68.8 69.3 69.5 69.8 70.4 70.6 71.2 71.5 71.8 Hungary 65.2 65.1 64.6 64.5 64.9 65.3 66.1 66.4 66.1 66.4 67.2 Latvia 64.3 63.7 62.7 60.5 59.2 60.2 63.0 63.9 63.5 64.8 64.9 Lithuania 66.6 65.3 64.9 63.3 62.8 63.6 65.1 66.0 66.7 67.1 67.6 Moldova 65.1 64.2 63.9 64.0 62.3 62.0 63.0 63.3 64.2 63.8 64.0 Poland 66.6 66.2 66.7 67.4 67.5 67.7 68.2 68.5 68.9 68.9 69.8 Romania 66.6 66.9 66.1 65.9 65.7 65.5 65.1 65.3 66.3 67.2 67.8 Russia 63.8 63.4 62.0 58.9 57.6 58.3 59.8 61.0 61.4 60.0 59.2 Slovak Republic 66.8 66.9 67.7 68.4 68.4 68.4 68.9 68.9 68.6 69.1 69.3 Slovenia 70.0 69.6 69.7 69.4 70.2 70.9 71.2 71.2 71.3 71.8 n.a. Ukraine 65.7 64.7 63.9 63.2 62.4 61.3 61.7 62.4 63.3 62.8 62.3 EU average 73.1 73.2 73.5 73.6 74.1 74.2 74.4 74.9 75.0 CEE average 67.1 67.0 67.0 67.4 67.5 67.6 67.8 68.0 68.4 68.7 69.3 Source: WHO Health for All Database. 5.9 Another illustrative indicator is the infant mortality rate. Despite recent declines, it is still the highest in Central and Eastern Europe, including the Newly Independent states of the Former Soviet Union (Table 5.5). Again, the positive development is that beginning in 1998 infant mortality starts declining, attaining its lowest level of 18.6 in 1999 and 2000. While infant mortality is still high, this is an encouraging development because after a visible reduction in 1991, this indicator had stagnated at the same level for six years. Table 5.5: Infant Mortality Rate, Selected European Countries, 1990-2000 (death per 10000) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Albania 33.8 32.9 25.2 23.3 20.2 15.6 14.7 Bulgaria 14.8 16.9 15.9 15.5 16.3 14.8 15.6 17.5 14.4 14.6 13.3 Croatia 10.7 11.1 11.6 9.9 10.2 8.9 8.1 8.2 8.2 7.8 7.4 Czech Republic 10.8 10.4 9.9 8.5 7.9 7.7 6.1 5.9 5.2 4.6 4.1 Estonia 12.4 13.4 15.8 15.8 14.5 14.8 10.4 10.1 9.3 9.5 Hungary 14.8 15.6 14.1 12.5 11.6 10.7 10.9 9.9 9.7 8.4 9.2 Latvia 13.7 15.7 17.5 16.6 15.7 18.9 15.9 15.4 15.0 11.3 10.4 Lithuania 10.2 14.3 16.5 16.0 14.1 12.5 10.1 10.3 9.3 8.6 8.6 Moldova 19.2 19.3 18.3 21.9 22.9 21.5 20.6 19.6 17.9 18.5 18.4 Poland 15.9 15.0 14.4 13.4 15.1 13.6 12.2 10.2 9.5 8.9 8.1 Romania 26.9 22.7 23.4 23.3 23.9 21.2 22.3 22.0 20.5 18.6 18.6 Russia 17.6 18.1 18.4 20.3 18.6 18.2 17.5 17.3 16.4 17.1 15.2 Slovak Republic 12.0 13.2 12.6 10.6 11.2 11.0 10.2 8.7 8.8 8.3 8.6 Slovenia 8.3 8.2 8.9 6.8 6.5 5.6 4.8 5.2 5.2 4.6 Ukraine 13.0 14.0 14.1 15.6 15.1 14.8 14.5 14.2 12.9 13.0 12.0 EU average 7.6 7.4 6.9 6.5 6.1 5.6 5.5 5.2 5.1 CEE average 18.2 17.1 16.9 15.8 16.2 14.8 14.0 13.0 12.3 11.4 11.1 Source: WHO health for All Database 94 ChapterS: Improving Resources Allocation in the Health Sector 5.10 These positive developments come with another concern, however. While there are reasons to believe that the health care system is belatedly starting to address some of the worst health indicators for Romania, funding appears to go disproportionately to health care providers, with a smaller share of the funding going to health programs aimed at reducing the incidence of preventable diseases. This is reflected in the morbidity indicators, where cardiovascular diseases ranked among the leading causes of death among the population below 64 years of age. The incidence of cardiovascular disease in Romania rose throughout the 1990s, placing the country among the three highest.in Central and Eastern Europe, other than the Newly Independent states of the Former Soviet Union (Table 5.6). Table 5.6: Standardized Disease Rates of the Circulatory System, Selected European Countries (0-64 years/100,000, male) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Albania 108 100 81 94 100 87 90 ... ... Bulgaria 225 227 237 256 253 262 257 276 270 243 259 Croatia 163 160 148 151 144 143 152 149 152 151 142 Czech Republic 225 208 201 185 177 172 168 162 146 142 132 Estonia 278 273 288 315 342 330 270 253 269 253 ... Hungary 266 264 271 275 256 250 240 229 235 229 210 Latvia 290 287 327 405 454 419 330 292 299 266 262 Lithuania 231 235 247 286 276 259 230 202 196 201 191 Moldova 200 210 194 213 258 271 250 253 221 246 243 Poland 241 252 238 221 208 207 196 ... ... 179 ... Romania 201 209 229 240 246 254 264 263 240 222 211 Russia 265 267 294 380 450 412 372 338 329 371 406 Slovak Republic 243 238 217 204 204 201 193 193 190 185 178 Slovenia 127 127 126 128 120 105 97 103 102 94 Ukraine 229 249 269 291 312 347 341 328 301 318 336 EU average 96 95 90 88 83 82 79 75 73 ... CEE average 218 222 221 217 210 211 207 204 196 187 183 Source: WHO health for All Database. Budget Design 5.11 The annual budget for the National Health Insurance Fund (NHIF) is based on a framework contract, agreed upon annually by the NHIF and the College of Physicians. This framework contract is approved by the Government, and defines the benefits package, conditions for service delivery, and payment mechanisms. There are co-payments required for drugs and allowed for other services. Family physicians play a gatekeeper role. The framework contract sets the general terms for provision of health care services and is the basis of the contracts between the District Health Insurance Funds (DHIFs) and health-care providers (hospital and their outpatient units, diagnosis and treatment centers, health centers, and medical offices). The budget for the NHIF, and its forty-one DHIFs, is designed to cover ambulatory (primary and specialist), inpatient, and dental care, including clinical preventive services and drugs. The insured are entitled to prescription drugs, health-care materials needed to correct eyesight and hearing, and prosthesis of the limbs is also either partially reimbursed by the insurance funds or free. The Ministry of Health and Family and the NHIF, using recommendations from the College of Physicians and the College of Pharmacists, compile a list of prescription drugs on a Chapter 5: Improving Resource Allocation in the Health Sector 95 yearly basis with reference prices. Pharmacists must sell the cheapest available drug, if only the generic name is on the prescription, and must mention potential substitutes. 5.12 Most of the other public expenditures on public health are included in the budget for the Ministry of Health and Family (MoHF), which is prepared in the context of the state budget. The 2002 budget provides a good example of the principles and practices governing the preparation of the MoHF's budget. While it introduces the principles of program budgeting, including programs for allocating funds by broad expenditure categories, with objectives, outputs and cost effectiveness indicators for 2002 and estimates until 2005, there is still a need for further development of this program budgeting, including the realism of cost estimates, assumptions about links between outputs and outcomes, and the need for a public health strategy with quantitative health status improvement indicators. Also, there is a worrying development, where the resources allocated for some of the programs are decreasing even in nominal terms, a decrease reflected in reduction of some of the planned outputs of those programs (Table 5.7). These programs include control of non-communicable diseases and community public health. Table 5.7: Expenditure for Programs, Ministry of Health and Family, billion ROL Nominal% Real% 2001 2002 Change Change program program 2002/2001 2002/2001' Ministry of Health and Family, programs total 3,879 4,451 15% -6% Control of non-communicable diseases 331 233 -30% -42% Community public health 1,249 1,243 0% -18% Children and family health 323 513 59% 30% Implementation of strategy for protection of disabled persons 25 27 8% -11% Reform, health policy and health administration 1,952 2,435 25% 2% 1) Assumes 22% average inflation rate, according to the draft budget law. Source: Romanian Authorities and World Bank staff calculations. 5.13 The shift in the composition of health expenditures raises important concerns. There is a sharp increase of expenditure on debt service (mainly for servicing commercial borrowing with state guarantees in 1999 and 2000), implying that more than 7.1% of total public expenditure on health (around $125 million) will be directed toward the repayment of credits and interests in 2002. This is a 37 percent real increase from the $91 million budgeted for 2001. The increase in debt service places, as a result, pressure on other expenditure categories, such as the national health programs, as well as leaving limited resources available for completion of investments, some of which started over 10 years ago. 5.14 The limited resources available for the completion of investments are a symptom of the ambiguities surrounding the decision-making for capital investments, and the ownership of assets in the health system. Major capital investments remain the responsibility of the Ministry of Health, and are financed from the state budget. Although the law gives local governments some discretion over local hospital expenditures, they generally lack sufficient funds for new investments. To circumvent restrictions on investments, the health service providers usually finance new equipment by leasing, disguising actual investments as current expenditure items. More importantly, the reliance on guarantees for supplier credit becomes a non-transparent form of government borrowing, circumventing accountability to Parliament, as well as the scrutiny 96 ChapterS: Improving Resources Allocation in the Health Sector given to investments that are included in the state budget such as those funded by multilateral institutions. This problem is compounded by the ambiguities and delays involved in the decentralization of state health facilities, as defined in the Law on Public Ownership approved in 1999. The law creates additional gaps in responsibility for public sector investment, since it mandates the decentralization of hospital ownership to local governments. Local governments however are likely to face greater difficulties in keeping up with the needed investments, while facing greater political constraints in rationalizing the hospital structure. Table 5.8: Expenditure for Health from Public Sources, 2001-2002 (Lei billion) Nominal % Real % 2001 2002 change Change program program 2002/2001 2002/2001' Health expenditure, public, total 47,458 61,273 129% 106% % GDP 4.31% 4.26% Total expenditure on health in state budget 10,237 13,477 132% 108% Health expenditure from state budget 7,260 9,812 135% 111% Capital investments 921 1,199 130% 107% Debt service 2,648 4,362 165% 135% Credit disbursement (mainly for capital expenditures) 2,977 3,562 120% 98% Expenditure funded from extra-budgetary revenue 103 Health Insurance Fund Revenue 4,1434 50,863 123% 101% Expenditure 37221 47,796 128% 105% Reserve 502 636 127% 104% Surplus 3712 2,431 65% 54% 1) Assumes 22% inflation rate, according to the draft budget law. Source: Romanian Authorities. World Bank staff calculations. 5.15 There is a need to review the process for approval of borrowings with state guarantee for investments in the health sector. This review should aim at increasing accountability, ensuring investment projects approved have adequate appraisal and implementation arrangements, and at improving resource planning to guarantee timely deployment and efficient operation. Also, the legal provisions requiring the approval of investment in high performance equipment by a joint commission of the Ministry of Health and Family, the Health Insurance Fund and the College of Physicians should be implemented in practice. This review should be based on a systematic evaluation of local needs, and an assessment of the capacity to finance investments and operations. Finally, charges for health services should include the cost of capital investment in the total cost of services covered from contracts with the health insurance funds. This inclusion of capital cost in the total cost of services would increase accountability in capital expenditure decisions by better matching investments made with health services rendered. Chapter 5: Improving Resource Allocation in the Health Sector 97 5.16 The final issue in the budget design is the manner in which budget surpluses are 71 allocated. In 2002 expenditures (including reserves) were set at 95 percent of revenues. Although the amount of the surplus will later be transferred to next year's insurance fund budget, there is a significant loss of revenue in real terms, because of the very low nominal interest rates earned while funds are deposited in the state treasury account. This creates tensions within the system, weakening accountability and financial control, as the practice of having expenditures still defined by the central government is inconsistent with the contributory social insurance model adopted by Romania. The excessive control over expenditures exercised through the budget process (coming from outside the health insurance funds) reduces the ability of the funds to develop stable, rational funding agreements with providers. This existence of this large surplus in the health insurance system leads many providers to develop mechanisms to deal with claims on this surplus, including building payment arrears, with the perverse effect that overall level of financial control of the health system is reduced and weakened. Budget Execution 5.17 Budget execution is strongly influenced by the incentives embedded in the budget design. Public hospitals receive global budgets negotiated with the health insurance funds and based on a mix of historical and performance-related criteria. The contracts that the District Health Insurance Funds sign with hospitals specify target number of admissions, hospitalization days and average costs per day (specific by specialty department and category of hospital). The trends in hospital utilization since 1998 already identified above (the shortening of hospital stays and the increased number of admissions) are likely to be related to the new payment mechanism for hospitals. The actual payments made to hospitals are based on reported hospital days, with a cut- off threshold labeled "optimal" length of stay. The hospital stays in excess of the average length of stay (measured in number of days) in the previous year for the given type of medical treatment are not reimbursed. The hospitals have, therefore, an incentive to shorten length of stay and increase the number of admissions to avoid a decrease in hospital revenues. 5.18 The manner in which physicians are paid for medical services provided also influences budget execution. The doctors in primary health-care are paid using a combination of weighted capitation and fee-for-service. Seventy percent of their income is determined by the number of patients that register on their lists, and the rest is allocated on a fee-for-service basis for preventive and health promotion services such as immunizations or cancer screening. The payment of physicians in ambulatory secondary care is based on a fee-for-service schedule that relies on a points system. Both categories of physicians also receive an allowance to cover administrative expenses related to their practice and pay for other staff with whom they work. Former dispensaries and smaller polyclinics now operate as autonomous entities, and have a self- employed status, although the largest part of their income still comes from contracts with public financing bodies. 5.19 The early experience with the new payment system for family doctors is mixed. There are signs that patients have become more demanding and less inclined to make informal payments. In many instances, the additional financial autonomy was used creatively to improve the range of services provided or the physical quality of medical offices. There are reports of 71 This is an increase from the 91 percent of revenues defined in 2001. 98 ChapterS: Improving Resources Allocation in the Health Sector providers that regard capitation payments as an entitlement based solely on the fact that patients are registered, without offering in return an adequate level of services, and abusing the right to refer patients to other providers. In the absence of adequate controls, the fee-for-service element is opening the way towards fraud through over-reporting. Also, most family doctors are reluctant to start any major investment to improve their offices as long as ownership issues are not clarified. 5.20 These trends show that, while the separation between health purchaser and provider has changed the behavior of health care units, it has not lead to expected changes in the cost of services. The favorable impact on improving technical efficiency of hospital services delivery has not translated into an increase in allocative efficiency and, therefore, has not released resources for other cost-effective services, such as primary health care. To accomplish this shift toward more cost effective services there is a need for a systematic assessment of service capacity and costs. Based on this assessment, changes need to be implemented using the authority and regulatory powers awarded to the central government, rather than relying exclusively on the payment mechanism to accomplish this shift. 5.21 Health sector arrears. The distortions in the budget design stage have also meant that budget execution are marked by the accumulation of substantial arrears and the need to reduce funding for key services (e.g. primary care, ambulatory specialist care, emergency services). For instance, in early 2001 late payments by state owned hospital reached close to Lei 4 trillion, even though the accumulated surplus of the National Health Insurance Fund amounted to Lei 3.6 trillion, and there was an additional Lei 1.4 trillion set aside in a mandatory reserve fund. This placed tremendous pressures in the implementation of the 2001 budget, leading to a reduction of expenditures in key services, as well as to a rapid build-up of new arrears to suppliers by hospitals. The latter is estimated by the National Health Insurance Fund to have reached Lei 4.6 trillion in end-September 2001. Furthermore, the District Insurance Funds had more than Lei 8 trillion deposited in their accounts from carrying forward the surplus from previous years and from the difference between revenue collected in first nine months of 2001 and the expenditure made in accordance to the budget program. 5.22 While there are better prospects for greater financial discipline in 2002,72 there is a need to ensure that the arrears accumulated in 2001 are financed from the surplus of the 2001 budget for the National Health Insurance Fund, not the 2002 budget. This can be achieved by reviewing the cost per hospital day in accordance with actual evolution of input costs. The resources to pay these arrears are, in principle, available, and would avoid having the burden for cost of services provided in 2001 being passed on to the 2002 budget. Indeed, in 2002 the participants of the health system (the Ministry of Health and Family, the insurance funds, and the hospitals) should 72 The reasons for this optimism are that the budget law and the framework contract were approved before the beginning of the 2002 calendar year; and the hospital law amended through Ordinance 68/2001 included several important provisions. These were (i) the approval of hospital budgets by Ministry of Health and Family authorities in agreement with insurance funds and based on actual revenue resulting from contracts of the hospitals with insurance funds; (ii) the introduction of administration contracts for hospital managers, including specific performance indicators; (iii) allowing hospitals to use part of their surplus to pay bonuses to staff, provided they don't register any arrears; and (iv) the draft framework contract for 2002 allows implmentation on a pilot basis of a prospective payment mechanism similar to Diagnostic Related Groups in 24 hospitals and payment mechanisms to all hospitals will be in 2002 a mixture between payment by hospital days and presumptive costs. Chapter 5: Improving Resource Allocation in the Health Sector 99 focus on (i) making use of the opportunities for better control of hospital expenditures; (ii) continuing preparatory activities for moving to a prospective payment mechanism, with payments made on the basis of presumptive costs; (iii) clarifying the roles of Ministry of Health and Family and insurance funds to identify and deal with hospitals facing structural deficits and develop effective mechanisms for restructuring. 5.23 Although the current financial situation of the National Health Insurance Fund also allows it to deal with the arrears accumulated during 2001, this should not detract from the broader issue of acknowledging that there is increasing evidence of a disconnect between resource allocation at the budget design and execution stages. The policy objectives and financial plans defined in the framework contract and approved in budget do not match the actual resource allocation of the executed budgets of the health insurance funds. This is can be seen by comparing the actual budget provisions with revised budget and the actual expenditure by type of service during the period 1998 to 2000 (Table 5.9). The actual expenditure for primary care, outpatient specialists, drugs, dentistry, medical aids and pre-hospital emergencies were systematically lower than planned, whereas for hospital services actual expenditure was always higher than approved in the framework contract and budgeted. Indeed, despite the stated objective of increasing expenditure in primary health care, the actual expenditures remained relatively constant as a proportion in total expenditures. Meanwhile, expenditures on hospitailizations continued accounting for around two-thirds of total expenditures, with this share rising to just under three quarters if one takes into account that almost all insurance payments for drugs are made to hospitals. Table 5.9: Health Insurance Fund Expenditures: Comparison of Framework Contract Provisions, Revised Budget Provisions and Actual Expenditures, 1998-2001 (%) Type of Service Actual Framework Revised Actual Framework Revised Actual Framework Expenditure Contract Budget Expenditure Contract Budget Expenditure Contract 1998 1999 1999 1999 2000 2000 2000 2001 Primary care 9.0 15.5 9.5 9.1 14.5-15 9.8 8.8 14.5-15 Outpatient 5.8 11.7 6.6 6.1 8.75 7.8 7.5 8.75 specialists Hospitals 67.2 40.0 61.2 64.2 59-61 64.0 65.5 50-53 Outpatient 6.8 20.0 9.3 8.0 10-11 12.8 11.3 10 drugs Dentistry 2.7 4.2 2.8 2.4 2.5-3 1.6 1.4 3 Rehabilitation 0.8 1.0 1.2 1.1 1 0.6 0.6 1-1.2 Medical aids, 3.2 3.0 0.6 0.3 1 0.3 0.3 1 prostheses Pre-hospital 4.3 4.5 3.8 3.7 3-4 3.0 3.0 3 emergencies Programs 0.1 0.0 5.0 5.2 0.1-1 0.0 7.0 8 Total 100 100 100 100 101 100 100 100 Source: National Health Insurance Fund. World Bank staff calculations. 5.24 To deal with this disconnect between budget design and execution action is required on three fronts: > There is a need for the Ministry of Health and Family, the insurance funds, and the representatives of hospitals to develop a framework for better control of hospital 100 ChapterS: Improving Resources Allocation in the Health Sector expenditure. This framework will require improved approval process of hospital budgets by Ministry of Health and Family and the Health Insurance Funds. The approval process should be based on actual revenue resulting from contracts of the hospitals with the insurance funds, and include performance contracts for hospital managers. The latter in turn should include specific performance indicators and allow hospitals to use part of their surplus (in the absence of payments arrears) to pay bonuses to staff. To ensure clarity in the compensation provided to health care professionals, the Ministry of Health and Family and the Insurance Fund should proceed with plans to generalize the adoption of prospective payment arrangements, such as Diagnostic Related Groups. Also, there is a need to clarify the role of the Ministry of Health and Family and the Health Insurance Funds in identifying hospitals facing structural deficits and in developing effective mechanisms for restructuring. > There is a need to improve the presentation formats of the budget data, to be used throughout the phases of the budget process (from planning to execution). At present, budget information is mainly presented in the economic classification for state budget expenditure and administrative expenditure of insurance houses, and is over-aggregated for the largest item of insurance fund expenditure (medical services). A better presentation format would combine the economic with a functional classification, by level of care (primary, specialist-outpatient, hospital and others), facilitating policy design, analysis and formulation. In addition, fragmentation of health care expenditures across institutional lines (with separate budgets presented by the Ministry of Health, National Health Insurance House, special health fund), the use of different presentation formats by different institutions, and the existence of transfers between institutions, limit the opportunity for a comprehensive external review of consolidated health care expenditures. To overcome this problem, National Health Accounts should be systematically prepared and published. > There is a need to review the process for approval of credits with state guarantees for investments in the health sector. This review would aim at increasing accountability, ensuring that guarantees are based on adequate appraisal and implementation arrangements, and allowing resources to be efficiently and timely deployed. The legal provisions regarding approval of investment in high performance equipment by a joint commission of the Ministry of Health and Family, the Health Insurance Fund and the College of Physicians should be implemented in practice, based on systematic review of local needs and an assessment of the capacity to finance investment and operation. The overall regulatory framework regarding roles for financing investments in the health sector should be reviewed, aiming at inclusion of the cost of capital in the total cost of services covered from contracts with insurance funds. Chapter 5: Improving Resource Allocation in the Health Sector 101 Budget Evaluation 5.25 What has been the impact of these problems at the budget design and execution stage on health outcomes? This section attempts to answer this question. In doing so, this section also assesses whether the budget reforms undertaken in the health sector in Romania in 1998 succeeded in shifting the demand from secondary and specialist care toward primary care - meaning that resources are allocated to services where there is greater demand (allocative efficiency); and in maintaining or improving access to health care services for the poor and vulnerable - equity. Since these reforms were introduced five years ago, the section relies on both administrative and household budget survey data to make this assessment. This allows the analysis to overcome some of the limitation of the administrative data.73 5.26 Have the budget reforms undertaken in the health sector in Romania in 1998 succeeded in shifting the demand for (and use of) health services from secondary and specialist care toward primary care? The short answer to this question is no. Most of the public resources in the health system continues to be spent on hospitals (almost 70%), with an increasing share now allocated to ambulatory specialized outpatient treatment (Table 5.10). As a result, there is a decline in spending allocated to primary care (family doctors) that would likely be magnified if capital expenditure were added to these figures.74 Table 5.10: Functional Allocation of Current Expenditure by Level of Care, 1999-2001. Current Expenditure on Health services and drugs 15,482 24,817 36,620 100.0 100.0 100.0 Primary care 1,409 2,186 2,673 9.1 8.8 7.3 Specialized outpatient 958 1,860 1,447 6.2 7.5 4.0 Para clinic services 1,018 0.0 0.0 2.8 Dental services 390 340 468 2.5 1.4 1.3 Hospitals 9,889 14,925 20,912 63.9 60.1 57.1 Drugs & medical supplies (chronic hospitals) 806 1,746 3,359 5.2 7.0 9.2 Drugs for ambulatory treatment 1,251 2,810 5,261 8.1 11.3 14.4 - drugs requiring co-payment 705 4.6 0.0 0.0 - free drugs 546 3.5 0.0 0.0 Medical equipment (prosthesis, aid equip) 42 72 174 0.3 0.3 0.5 Pre- hospital emergency 565 740 1,076 3.6 3.0 2.9 Physical therapy/ rehabilitation/ home care 172 137 233 1.1 0.6 0.6 Source: National Health Insurance Fund. 5.27 This allocation of resources within the public health care system appears to go against the trends in the use of services. Household level data on use of services point toward a sizable " Although the administrative data is informative, limitations because of disclosure requirements imply the information is not always useful. For instance, there are no details on main expenditure items, such as payment of services for physician by type or specialty. 4 Most of the capital expenditures are financed from the state budget revenues and are within the budget administered by the Ministry of Health, not the Health Insurance House. 102 ChapterS: Improving Resources Allocation in the Health Sector increase in the relative use of primary care (which includes specialized outpatient in this case) and a decrease in the utilization of hospitals.75 Furthermore, the pattern of use of health services is different for urban and rural residents (Table 5.11). A larger share of urban residents uses private services. While the use of primary care services has increased in both areas after 1998, the rate of growth differs. For the rural residents, primary care is (and was) the most accessible service, for urban residents, only after 1998 have these services become more attractive than going first to the hospital. Table 5.11: Use of Health Services by Area of Residence, 1995-2000 (%) Private Primary Hospital Other Total Urban 1995 8.37 70.67 17.21 3.76 100 1996 19.33 61.89 16.2 2.57 100 1997 22.63 66.91 9.21 1.24 100 1998 22.85 66.13 9.49 1.53 100 1999 18.95 73.57 6.4 1.08 100 2000 10.97 83.23 5.16 0.65 100 Rural 1995 5.15 72.09 20.33 2.43 100 1996 8.36 68.98 20.71 1.95 100 1997 11.29 74.74 13.18 0.79 100 1998 11.19 75.82 12.34 0.65 100 1999 12.41 78.02 9.07 0.5 100 2000 8.01 84.64 6.68 0.66 100 Source: Tesliuc (2002) 5.28 Have the budget reforms undertaken in the health sector in Romania in 1998 succeeded in maintaining or improving access to health care services for the poor and vulnerable? To answer this question, we look at the distribution of health services across individuals ranked by consumption. This is shown in Figure 5.2 which depicts three curves: the distribution of health services across income levels (concentration curves), the distribution of consumption across income levels, and the equality line where consumption is equally distributed across income levels. The position of these so-called concentration curves indicates whether the health program (primary, outpatient, hospital), is progressive or not. If the concentration curve of a program lies below the consumption line, that program is regressive. The poor get a share of benefits that is lower than their share in consumption. If the concentration curves lies between the consumption line and the line of equality, that program is mildly progressive. The poor receive a share of benefits that is larger than their share in consumption, but smaller than their share in population (e.g.: if the consumption of the poorer 20% is 5% of total, and they receive 10% of the public spending on primary care services, that program is mildly progressive). The last case is where the concentration curve lies above the line of equality, and that transfer is highly progressive. The poor receive a larger share of benefits than their share in total population. Such a program 7s Note that information about utilization of health care facilities during 1995-1998 does not discern between primary health care (dispensaries) and specialized outpatient (polyclinics), making it difficult to ascertain what services are more popular. While the data in 1999 and 2000 do allow one to distinguish between primary care (family doctors) and specialized outpatient care, for comparison reasons data on utilization of primary and specialized care has been aggregated also for 1999 and 2000. Chapter 5: Improving Resource Allocation in the Health Sector 103 will distribute a relatively large share of health services expenditure to the poor. We will argue, however, that a correct benchmark against to judge the distributional outcomes of the publicly financed health services is the line of equality. In its spirit and letter, publicly financed health services should guarantee equal access for all patients, irrespective of their economic status. Thus, concentration curves that lie above the line of equality should be termed progressive, the ones that lies between the line of equality and the line of consumption should be termed mildly regressive, while curves that lie below the consumption line are strongly regressive. Figure 5.2. Concentration Curves for the Use of Health Services: Primary and Specialized Outpatient Versus Hospital Services 995 S98 09 ------------- -- ------09---- ----- 08 ----. ---------------- -- ------------- 0.8 ------- o.e ----- ----- -- --- j ----- --- ---. .-.. - - --. .... ..-. --.. .. . --- -- -- ---------- 05 - - - - - - - - --- ---- --- - - --- --- --- -- --- 0 57 --- ------------ - - --r - -- - - - - -I Z 0.6 -- - -- - - - - - - - - - -- - - - - -4- - - - - --- - -0 - - - - - - - - -- - - - - - - - - - - --- 0.4 ---------------- ----- - - - -- --- ---- -- 04 o-- 0 83 --- - -- ----- -- --- ---- E ---- ------- -- ---- --0------- - - -- - - - - - - t - -- - - - - - S 0 6 - - - -- - -- - - - - - - - - - --063 ----- ----- - - --- - - -I- - - --- - ---- --- - 02 -- - ---- - - ---- 02 -------- --- - ---- -02-- ----- ------ 01 0---- ------ - I --- 01 ---- - ---- ----------- -------- ----- ---------- ON 0 0 01 02 03 04 05 06 07 08 09 1 0 01 02 03 04 05 06 07 08 09 1 crrulativeshar of popubtio cundiveshtsf o d puaion ---eqalty pinrsy-Z-hoeplal- -cormurptim -+-equaly --prinay -- hospltal misurpon 0 8 - - -- - - -- - - - - - -- - .- - -1-- --- -08 - - - - - - - - -- - - - - - - - - g 0.7 -- - -- - -- - --- - -- - - - -0 7 - - -- - - . 05 .... --- ----0--L5-- 0 2 -- - - -- - . . .0.2 - - - - - - - - -- .. . . .. . .. . . . 01- - -- - . .3 - -- - 1- - -- i- - - -I - - - 01 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 01 0.2 03 0.4 0.5 0.6 07 0.8 0.9 1 anidahve davm of popuatio cumiaveshare of pDpulabn --equalty ---pinary -g-hosplII41corsurpiUn F;:-euIly -4-- pairnary -I- hospta-40--om w-PODn 104 ChapterS: Improving Resources Allocation in the Health Sector 5.29 The results from Figure 5.2 indicate that from 1995 to 1998 the distributional incidence of publicly financed health services worsened. In 1995, both primary and hospital care were situated close to the line of equality (mildly regressive). From 1996 to 1998, both concentration curves shifted downward, signaling a less equitable distribution of benefits. Surprisingly, the largest increase in inequality occurred in primary care (included here are dispensaries and polyclinics), and to a lesser extent in hospital care. These findings are confirmed when analyzing the information available for 1999 and 2000, when there is there is information on health care providers disaggregated among dispensary and polyclinic care. Figure 5.3 illustrate the concentration curves for the use of dispensary/family doctors, polyclinics and hospital care for this period. The most equitable outcome in terms of care is again associated with hospital care, while access to polyclinic services turns out to be highly regressive. This suggests that the well-off strata of the population can capture specialist services in polyclinics with much greater success than the poor. The relatively better access of the poor into hospital care may be explained by low and worsening quality of care in hospitals. In addition, the poor affected by illness seems to report only severe episodes of illness and seek care only in cases that strongly threaten their health. In contrast, the middle class and the well-off are investing more in preventive care, seeking care earlier in the disease cycle, even for milder episodes. Figure 5.3 Concentration Curves for the Use of Family Doctors, Specialized Outpatient and Hospital Services 1999 2000 0 1 02004 0 6 4 8 0 1 2 0 4 4 6 0 8 0 5.0 Another-imprtant-equiy-issue-tht-emerges-rom-the-bu08 refrm o the health J 04 ----- 03 ----------- --------------- 0o3-------- 02------------- - ------ 01----- ------ ......o - ---------- I---I------ 0 a1 02 03 04 05 0e 0.7 09 0s I 0 01 2 03 0'4 05 08 07 00 0g I 5.30 Another important equity issue that emerges from the budget reforms of the health system is the access by non-insured to services. The 1998 reform changed the rule of access to health care, from universal access to access conditional on contributions to the health insurance system. Some population groups, like employees and pensioners, make compulsory contributions to the health insurance system. The Unemployment Fund pays the contributions for registered unemployed. The state budget covers the health contributions for children, students, and welfare beneficiaries. Other categories, notably self-employed, farmers and employers, contribute voluntarily. Anecdotal evidence suggests that a large proportion of the non-mandatory groups do not contribute to the health insurance system. Chapter 5: Improving Resource Allocation in the Health Sector 105 5.31 Table 5.12 presents comparative information on the use of different types of health care providers over 1998-2000. In 1998, there are no significant differences in the choice of health care providers by the insured and the non-insured status76 (we focus on the two main categories, choosing no treatment (nowhere) and choosing primary care). Starting 1999, there is a statistically significant difference between insured and non-insured with respect to the choice of primary care and no treatment. The non-insured are less likely to seek primary health care when ill, and more likely to forgo treatment when needed. 5.32 If we conjecture that non-insured individuals are usually poorer, the findings of Table 5.12 are consistent with the more general observation that poorer individuals seek treatment less often than better off individuals, everything else held constant. Furthermore, this inequality in utilization of a formal health care provider appears to have risen over time. In 1995, the chances of seeking treatment when ill were similar for the poor, the middle class and the rich. This situation changed in 1996-97, when the odds of seeking treatment were between one-quarter to one-third higher for richer individuals with, otherwise, identical characteristics. In 1999 and 2000 this difference in odds rose to three-quarters. This result is somewhat surprising and raises concerns, since there does not appear to be a significant difference in seeking health care between the poor insured and the poor non-insured. The concern is that access to health care may be conditional on under-the-table payments, limiting all the poor (insured or not) from visiting health care providers. Table 5.12: Choice of Health Care Provider by Insurance Stats 1998-2000 % Nowhere 38 37 37 37 31 32 32 29 29 Private-Clinic 10 12 11 11 11 11 5 7 7 Primary_Care 43 44 44 42 48 47 50 54 54 Specialist_Outpatien na na na 3 4 4 5 5 5 Hospital 9 6 7 6 5 5 7 4 4 Other 1 1 1 0 0 0 1 0 0 _ _ _ _ _g _ _ _i _ WIgbi 1i l ie W i ei l r ofot h *) In 1998, primary and specialist outpatient care are lumped together. Source: Tesliuc (2002). 5.33 Out-of-pocket Payments. This last result is important because there is considerable reliance on out-of-pocket payments for services delivered in Romania. For instance, private spending on health-care in 2000 was estimated at Lei 6,479 billion ($ 250 million), or about one- third of total health expenditures. Around one-quarter of these payments go directly or indirectly to public providers and their staff through charges for services and/or under-the-table payments (that is illegal payments to providers for health and dental services that are nominally free). About three-quarters of these out of pocket payments, however, are household expenditure on 76 As the household budget survey does not indicate if a person contributes voluntarily, we group the individuals who are not mandatory required to contribute to the health insurance system into a non-insured category. 106 ChapterS: Improving Resources Allocation in the Health Sector pharmaceutical drugs.77 Thus, while out of pocket payments is one of the features of the Romanian health system, and should be expected, they need to be monitored to ensure that there is no abuse and that access to health services is not denied to certain groups. The latter is important because the utilization of health services is positively associated with the prevalence of illness and the supply of publicly financed services, but negatively correlated with the out-of- pocket payments. Table 5.13: Composition of Private Health Expenditures b (Deciles, 1995-2000 () 7Wed 3 - WISMkf~AR Year Medicines 1995 84% 81% 81% 81% 78% 75% 70% 67% 60% 49% 65% 1996 87% 78% 79% 81% 76% 76% 71% 69% 63% 49% 65% 1997 85% 87% 83% 79% 81% 79% 73% 72% 69% 53% 69% 1998 88% 86% 80% 79% 81% 77% 76% 72% 64% 51% 68% 1999 83% 86% 85% 81% 77% 77% 78% 75% 68% 58% 710/6 2000 80% 84% 86% 85% 81% 82% 77% 77% 68% 63% 73% Health Services 1995 11% 12% 11% 10% 11% 14% 14% 16% 17% 22% 16% 1996 5% 14% 12% 10% 14% 14% 14% 15% 20% 24% 18% 1997 8% 6% 10% 12% 9% 10% 14% 11% 15% 21% 15% 1998 6% 7% 10% 12% 11% 13% 13% 13% 21% 30% 18% 1999 9% 7% 9% 11% 12% 12% 10% 13% 16% 22% 15% 2000 12% 9% 4% 7% 9% 8% 11% 11% 15% 20% 13% Dental Services 1995 0% 1% 2% 3% 4% 5% 9% 9% 14% 18% 10% 1996 2% 2% 1% 2% 4% 4% 7% 7% 10% 19% 10% 1997 2% 1% 1% 3% 2% 4% 4% 7% 7% 16% 8% 1998 0% 1% 4% 1% 3% 3% 4% 8% 7% 11% 6% 1999 0% 1% 1% 1% 3% 4% 5% 5% 8% 12% 7% 2000 1% 2% 1% 2% 4% 4% 5% 4% 8% 10% 6% 1) The share devoted to medicines, health and dental services do not add up to 100 in any year, as other expenditures items (such as optical and medical equipment) are not reported. Source: Tesliuc (2002). 5.34 An analysis of private expenditures on health services and pharmaceutical drugs indicates that this concern is warranted for two reasons. First, while the share of income usually devoted to pharmaceutical drugs tends to decline as income rises, making room for health and dental services, the very poor in Romania (the poorest 10 percent of the population) spend a higher proportion of their out-of-pocket expenditure on health services (Table 5.13). This signals that the very poor are limited in their free access to public facilities. Second, the fact that most of the population faces a trade-off between paying for pharmaceutical drugs and paying for access for health services is a particular public policy concerns in the case of dental services. Dental' services in Romania were privatized in 1992-93, and access is now based on market prices. Table 5.13 reveals that access to dental services by the poor worsened throughout the 1995-2000 period. The poor not only spent a smaller amount on private care, but they spent a smaller 7 The level of out-of-pocket payments in Romania is higher than the average level of private expenditure in the OECD countries, although,lower than in countries such as Australia and the United States. Also, comparqd to other countries in the region, Romania is in the middle of the distribution in terms of amounts and frequency of out of pocket payments. Chapter 5: Improving Resource Allocation in the Health Sector 107 fraction of their budget too. This is an important alert for the Romania authorities because recent reforms in health care financing do not address this issue of access to dental services. Conclusions and Recommendations 5.35 As in many transition countries, the Romanian health system was highly centralized, with excess hospital capacity, limited focus on primary care and prevention, and little client responsiveness. The entire system suffered from extreme neglect during the latter years of the Ceasescu period. Despite initial optimism, health reform stalled during the mid-1990s-the government was interested in 7physical investments in the health sector, but showed limited commitment to major reforms." Following a change of government in 1996, Romania enacted several significant health reforms, aimed at shifting the health system from the model of centralized government financing and delivery of services to a more decentralized and pluralistic approach. The major elements of the reform program are similar to those being undertaken in many other Eastern European countries. Key reforms enacted by Parliament have included establishment of a National Health Insurance Fund and 41 District Health Insurance Funds, paid for through an earmarked wage tax. These reforms also envisioned the decentralization of service delivery; an increase in hospital autonomy; and the establishment of general practice doctors (GPs) as private practitioners, paid on a mixed capitation and fee-for-service basis. The health insurance law was approved in 1997 and came into full force in 1999. While still in their early stages, reforms remain fragile, and have not yet led to significant improvements in health system quality, efficiency, or public satisfaction. 5.36 The Government recognizes the need for further refinement of health sector reforms. This includes developing regulations and policies, improving incentives within the health system, and building capacity at all levels. While implementing this agenda of reform presents significant challenges - technically, institutionally, and politically, this chapter identifies seven steps for moving forward: > Improving health financing to ensure system sustainability. The financial sustainability of the new health insurance system faces a number of risks, in light of the relatively small proportion of formal sector workers, high rates of tax evasion, rapid escalation of pharmaceutical costs, and an inefficient hospital sector that consumes a disproportionate share of sector resources. To address these risks, parallel action will be necessary on a variety of fronts. First, the payment system will need to be further refined to better align provider incentives with desired quality and efficiency objectives. This will require technical support and capacity building in the health insurers. Second, the "separation of purchasing from provision" is unlikely, on its own, to result in substantial efficiency gains or rationalizing of the hospital system. The Insurance Funds do not currently have the regulatory authority or political clout to force significant rationalization in hospitals. Hospital downsizing and restructuring will be necessary, although politically challenging to implement. Third, the capacity 78 Proposals to establish a national health insurance scheme, paid for with a dedicated wage tax, did attract support both within Parliament and among some senior MOH officials. The latter appeared to have viewed this initially as a means to obtain additional resources for the health system, rather than to improve efficiency or promote wider reforms in service delivery. 108 ChapterS: Improving Resources Allocation in the Health Sector and governance structure of health insurers will need to be strengthened. The current system of appointing the boards for health insurance houses by the government has led to administrative instability and politicization of senior positions. > Clarifying the legal framework Despite the passage of key legislation, a number of aspects of the legal and regulatory framework remain to be clarified. Most important is to clarify the relative roles and relations of the Health Insurance Funds and the Ministry of Health and Family (MoHF)-which still owns most secondary and tertiary facilities. Accountability for health insurance revenues also needs to be clarified, given that the MoHF has asserted control over "surplus" revenues to help close the government's chronic fiscal deficit. Clarification is also necessary regarding the ownership of rural health centers and related equipment, given that these are now staffed by GPs. The government's strategy highlights the need for progress in this area; drafting and passing effective legislation will require both technical and political skill, however. > Increasing funding for primary health care. Establishing general practice as a distinct specialty, paid through a mix of capitation (70 percent) and fee-for-service was intended to strengthen primary service delivery, improve client responsiveness through competition among GPs, and improve the referral system through giving GPs a "gatekeeper" function. These reforms were tested initially through the district pilots and have now been implemented nationwide. The reforms create the potential for more client-responsive primary care, but also create additional risks. First, the major risk is that financing for primary care and prevention through the health insurance funds is being increasingly "squeezed" as a result of cost overruns in the hospital sector. This was a major problem in 2000, resulting in wide press coverage of inadequate funding for drugs, and underpayment for GPs. Left unresolved, this could threaten the credibility of the reforms. Second, most GPs are former pediatricians or adult medicine specialists with little additional training to help them adapt either to their new roles as family doctors or to the new demands of running a private practice. Some training is being provided through NGOs and other donors, but it is inadequate. Third, the reforms may have initially weakened the referral system by separating primary care from hospitals. Further policy development and training is necessary to strengthen referral and clarify new roles. Fourth, financing equipment and other capital investments for general practitioners remains an unresolved issue. A supplementary "practice allowance" was to help cover these costs but has since been reduced significantly. Fifth, encouraging group practice-particularly in urban areas-holds promise for addressing a number of issues, including pooled use of equipment and administrative assistance, and improved coverage for after-hours care, which could in turn help reduce reliance on emergency services for primary health care. Group practice remains rare, however. Also, while the current approach to primary care has the potential to succeed in urban areas, it may prove inadequate for rural primary health care. Nearly half of the population lives in rural areas, where "competition" among general practice doctors is unlikely to bring about quality improvements, since there is limited choice of alternative providers. Capitation payments create some incentive for providers to locate to rural areas, and the government's strategy calls for further strengthening such incentives. Chapter 5: Improving Resource Allocation in the Health Sector 109 > Placing greater emphasis on equity in the delivery of health services. Access to and quality of health services in Romania remain inequitable among geographic areas, among income groups, and for ethnic minorities. Several concerns deserve highlighting. First, because most of the rural population is engaged in subsistence agriculture, many rural households are at risk of losing access to health services. This would further reduce the incentives for GPs to work in rural areas and significantly undermine equity. Second, the practice of informal payments for doctor remains widespread; this distorts provider incentives and constitutes a significant burden on the poor. Third, a transparent criterion for redistributing health insurance income among districts needs to be established and consistently implemented. Fourth, minority ethnic groups, particularly the Roma, are often marginalized or excluded from care. Different approaches may be needed to reach these groups, and moving forward will require making this a political priority. > Giving greater attention to health promotion and prevention. Recent reforms create a significant opportunity for the MoHF to take a leadership role on health promotion and prevention. The challenge is to establish a comprehensive approach to health promotion, which incorporates traditional health education activities with strengthened regulations (including for tobacco, alcohol, and accident prevention), and increased community participation. For example, tobacco control legislation has been developed, but has not yet passed Parliament. Improving monitoring and evaluation on the causes of ill health-and wide dissemination of findings among decision makers and the public-will be important to building support for health promotion. > Containing the escalation of pharmaceutical costs. With support from WHO and other partners, the government has taken steps to improve the policy framework for the pharmaceutical sector. Effective cost controls, however, have yet to be implemented-both in terms of establishing an essential drug policy and strengthening incentives in the reimbursement system to promote cost containment. Continued cost escalation could threaten the financial sustainability of social health insurance and undermine equity-as poor patients are forced to pay for high-cost drugs or go without. Various vested interests oppose these reforms, however. The Bank would like to remain engage in the policy dialogue on pharmaceutical policy development - and could consider the possibility of incorporating key pharmaceutical sector reforms into future programmatic lending. > Developing health information, monitoring, and evaluation. The lack of adequate information systems for monitoring and evaluation in the health system remains a major concern. The government's strategy calls for strengthening health information systems, but design, responsibilities, and financing of such a system remain unclear. Moreover, and data collection from the health insurance funds, MoHF, and other statistical bodies are not standardized or integrated in ways to guide management or policymaking. While funds have been spent in purchasing the hardware required for this task, there is a need to make a bottom up investment now on those collecting and using the information.  CHAPTER 6: DELIVERING SOCIAL ASSISTANCE AT THE LOCAL GOVERNMENT LEVEL Introduction 6.1 Chapter 3 examined Romania's experience to date with implementing fiscal decentralization reforms, and suggested ways to strengthen local governments' financial and operational accountability. One of the findings was that local governments' revenue and expenditure responsibilities (including responsibilities for the delivering social assistance) have undergone significant and frequent changes in recent years. In some instances, the way in which revenue and expenditure responsibilities were shifted hindered the effective delivery of social assistance services and benefits. This was particularly true in the case of the means-tested social assistance benefit (MTSAB), introduced in April 1995 as Romania's principal income support mechanism for households with incomes below an established threshold. Originally financed by the central government, the MTSAB lost its effectiveness as a poverty alleviation tool after the government shifted responsibility for funding the benefit to the local councils. The result was that the program coverage shrank and benefits transferred to households contracted sharply. 6.2 In recognition of the near total collapse of the MTSAB program, Romania passed the Law on the Minimum Income Guarantee in late 2001. The law aims at strengthening the country's social safety net by providing an improved administrative and financial framework for the delivery of four social assistance benefits. These included: (i) the MIG benefit, ensuring a minimum income for all those falling below a certain threshold; (ii) the benefit for the wives of soldiers in compulsory military service, who either have preschool children or children that are disabled; (iii) a lump-sum benefit to all eligible households at the time of child birth, up to the fourth child; and (iv) a heating allowance to help defray residential heating costs during the winter season (November 1 to March 31). The most important of these benefits is the Minimum Income Guarantee (MIG) that was designed to replace the MTSAB. 6.3 A part of the financing, and. all of the delivery of MIG program, is the responsibility of local governments. This chapter examines the challenges local governments face in carrying out these responsibilities. The chapter is organized as follows. It begins with a snapshot of the poverty profile in Romania, and a short description of the country's social protection programs. The chapter then examines the Minimum Income Guarantee program along three dimensions: design, implementation, and monitoring and evaluation. The chapter closes with recommendations for increasing the odds of the Minimum Income Guarantee program becoming an effective and efficient tool for poverty alleviation. Poverty and Social Protection in Context 6.4 Poverty in Romania is widespread but mostly shallow, with some deep pockets among households with many children and households headed by farmers, the unemployed, and ethnic Roma. The poverty incidence is also high in the Northeastern region of the country, where there is a greater concentration of farmers. Farmers account for 16 percent of the poor, with a poverty incidence of almost 60 percent. Households headed by unemployed account for about 11 percent 112 Chapter 6: Delivering Social Assistance at the Local Government Level of the poor, with a poverty incidence of almost 70 percent. Roma-headed households account for just under 7 percent of the poor,t but the poverty incidence among these households reaches almost 90 percent. Among pensioner-headed households, poverty is higher for those earning lower pensions (e.g., farmers, survivors) and pensioners who retired before 1997, when benefits were under-indexed to inflation. Among employee-headed households, the risk of poverty is highly correlated with the number of children in the household, although poverty is also high for households where the main income earner has a seasonal job, lower educational attainment, or a low paying job. 6.5 It is against this backdrop of widespread but shallow poverty that social protection measures have relied primarily on categorical benefits, with a very small share of the funding going toward means-tested social assistance. Means-tested social assistance accounts for between one and two percent of total social spending. While social protection programs reach around 80 percent of the population (Table 6.1), most receive old age pensions and child allowances (Table 6.2). This has done little to mitigate hardship during a period of rising poverty,79 as large categorical benefits are not necessarily well targeted toward the poor.80 The Government's renewed funding commitment for means-tested social assistance is expected, therefore, to make a positive contribution toward poverty reduction. Table 6.1: Percentage of the Population Receiving at Least One Cash Social Transfer Total Urban Rural Non- Non- Non- Year Recipient Recipient Recipient Recipient Recipient Recipient 1995 20.3 79.7 24.1 76.0 15.7 84.4 1996 21.3 78.7 25.4 74.6 16.1 83.9 1997 17.8 82.2 20.7 79.3 14.3 85.8 1998 17.8 82.2 20.6 79.4 14.4 85.6 1999 19.3 80.7 23.0 77.0 14.8 85.3 Source: Tesliuc (2001). Estimations based on Integrated Household Survey 1995-1999 data. Table 6.2: Outreach of Selected Social Programs in 1999 % of Households benefiting: Total Urban Rural Old-Age Pension 30.8 31.5 29.9 Farmer Pension 14.1 1.3 30.0 Unemployment Benefit 9.0 9.6 8.3 Child Allowance Benefit 37.0 39.0 35.1 Scholarship Benefit 0.4 0.6 0.2 Social Assistance Benefits 0.6 0.6 0.7 Source: Tesliuc (2001). Estimations based on Integrated Household Survey 1999 data. 6.6 The optimism generated by introduction of the MIG program is based on three factors. First, the record of Romania's previous social assistance program - the means-tested social assistance benefit (MTSAB) - is encouraging. A 1999 analysis of the benefit incidence of various social transfers in Romania indicated that benefits under the MTSAB were consistently 7 Estimates suggest an increase in incidence of poverty from 20 percent in 1996 to 41 percent in 1999. Tesliuc (2001). so Pension benefits are not designed for poverty alleviation Chapter 6:-Delivering Social Assistance at the Local Government Level 113 well-targeted, with the poorest 20 percent of the Romanian population receiving almost three- quarters of the benefits (Table 6.3). Also, the MTSAB was more cost effective than unemployment benefits and child allowances (Table 6.4), with every 1.7 leu spent on this program leading to a 1 leu reduction in the poverty gap (Table 6.5). Table 63: Benefit Incidence in 1999, Individuals Ranked by their Household Consumption per Equivalent Adult, including Pensions but excluding Other Transfers Deciles 1 2 3 4 5 6 7 8 9 10 % per decile Share ofpopulation 10% 10% 10% 10% 10% 10% 10% 10% 10% 10% Current consumption per equivalent adult 3% 4% 6% 7% 8% 9% 11% 12% 15% 25% Scholarships 4% 14% 6% 9% 14% 16% 8% 14% 6% 8% Remittances received 5% 6% 5% 5% 6% 7% 9% 7% 13% 36% VWW aid 11% 9% 4% 11% 8% 9% 9% 10% 13% 15% PPP aid 19% 2% 10% 4% 10% 17% 11% 13% 2% 14% Child allowances 21% 16% 13% 11% 10% 9% 7% 6% 4% 3% WOW pension 23% 3% 7% 8% 3% 5% 4% 17% 13% 16% Handicapped aid 27% 18% 15% 14% 8% 6%. 5% 3% 3% 2% Unemployment benefit 34% 17% 14% 10% 7% 6% 5% 4% 2% 2% Social pension 37% 23% 7% 4% 8% 5% 6% 2% 7% 2% Social assistance (MTSAB) 43% 21% 6% 3% 5% 12% 1% 2% 1% 7% Source: Tesliuc (2002). Estimation based on Integrated Household Survey 1999 data. Table 6.4: Cost-Benefit Ratio for Selected Social Cash Transfers Reduction in Lei spent per No. of Poor Cumulative the Income 1 Leu Number of lifted out by Average Income Gap Gap Reduction in Scenario Poor the Program Income Gap (Lei Jan 95) (Lei Jan 95) Income Gap If household consumption includes all 57,047,025,81 transfers 4,983,504 0.2419 9 If child allowances are excluded from HH 78,912,718,92 consumption 5,919,173 935669 0.2818 1 21865693102 2.63 If unemployment benefits are excluded 66,773,357,13 from HH consumption 5,364,215 380711 0.2631 3 9726331314 2.56 If social assistance (MTSAB) is excluded 57,503,035,77 from HH consumption 5,003,507 20003 0.2429 0 456009951 1.72 Explanatory Note: The poverty line used is 50% of mean consumption per adult equivalent. PO is the poverty level based on current consumption. Poverty levels given the absence of a particular cash benefit were then computed by subtracting the transfers from PO, one at a time. Source: Tesliuc (2002). Estimation based on Integrated Household Survey 1998 data. 81 Unfortunately, the available data do not allow an estimation of the full cost-benefit ratio of the programs, since the administrative cost data were not available for all programs. One would expect the administrative costs of the social aid program, a means tested program, to be higher than the costs of providing the unemployment benefit and the child allowance. However, even such implausibly high administrative cost-benefit ratios of 25 percent, or even 50 percent, would not challenge the MTSAB's position as the most efficient program from a poverty alleviation perspective. 114 Chapter 6: Delivering Social Assistance at the Local Government Level 6.7 The second reason for optimism about the poverty reduction potential of the MIG is that the results of more recent simulations, which account for the design of the MIG and are based on the 2000 Integrated Household Survey, continue to indicate that the program has a strong potential to reduce poverty. According to these simulations, the MIG scheme could reduce the poverty rate in 2002 by about 18%, the poverty gap index by 36%, and the severity of poverty by about 52% (Table 6.5). These figures are based on a poverty line of 60% of the median consumption per equivalent adult and take into account the impact of inflation during the year on the number of eligible households, under the assumption that the income thresholds probably won't be indexed during the year. Looking only at the static January picture, which doesn't account for the likely decline in eligible households, the poverty impact is higher. Table 6.5: Ex-ante Analysis of Poverty Reduction under the MIG Scheme Before MIG After MIG, After MIG and Change Rate (1) Change Rate (2) without Heating Heating Subsidies (1) Subsidies (2) 2002 - January Poverty rate 13.9% 11.1% 10.1% -20.14% -27.34% Poverty gap 2.91% 1.77% 1.49% -39.18% -48.80% Poverty severity 0.95% 0.42% 0.34% -55.79% -64.21% 2002 - overall Poverty rate 13.9% 11.7% 11.4% -15.83% -17.99% Poverty gap 2.91% 1.99% 1.86% -31.62% -36.08% Poverty severity 0.95% 0.51% 0.46% -46.32% -51.58% Source: Pop (2002). Calculations based on the Integrated Household Survey 2000, NIS. 6.8 Table 6.6 reports the ex ante evaluation's results regarding the MIG program's potential impact on different types of households. The estimated poverty reduction is greatest for single- parent households, persons living alone, and two-adult households containing many children (at least three). The poverty rate for single-parent households decreases nearly 27 percent as a result of the MIG program, while it drops around 20 percent for persons living alone and large families. Table 6.6: Poverty Rates and Poverty Reduction Rates by Different Types of Households Poverty Rates (ea.) by Single Households Single Households Households Children Total Household Types (%) Person without Parent with at Least with at Least Households Children and Households two Adults two Adults at Least Two and 1-2 and 3 or Adults Children More Children Poverty Rate before 8.8% 7.5% 25.8% 13.3% 33.7% 20.1% 13.9% MIG January 2002 Poverty rate after MIG 6.2% 5.8% 16.4% 10.0% 23.3% 14.3% 10.1% - January 2002 Poverty rate after MIG 7.0% 6.4% 18.9% 11.1% 26.9% 16.2% 11.4% - 2002, allover the year Poverty reduction rate - 29.5% 22.7% 36.4% 24.8% 30.9% 29.0% 27.3% January 2002 Poverty reduction rate - 20.5% 14.7% 26.7% 16.5% 20.2% 19.2% 18.0% 2002, allover the year Source: Pop (2002). Calculations based on the Integrated Household Survey 2000, NIS. Chapter 6: Delivering Social Assistance at the Local Government Level 115 6.9 The third reason for optimism is that the MIG law defines a minimum level of income per household. This should provide the resources needed to reach many of those eligible, avoiding the main shortcoming of the MTSAB, which was the lack of funding, and the resulting limited coverage. The MTSAB was first introduced in April 1995 as a centrally funded, and locally administered, program. When the financing responsibility was transferred to the local councils in 1996, program coverage fell quickly, since reduced funding allowed the income threshold for eligibility to erode. By 1997 the number of program participants had fallen to just 33 percent of its level in 1995. This lack of funding was compounded by competing demands on local government budgetary resources, due in part to the constantly shifting expenditure responsibilities. 6.10 The rest of the chapter examines what is new under the MIG program, and what is being done to avoid the mistakes of the past. The main positive feature, as mentioned just above, is that the MIG has a strong poverty reduction potential. This strength is drawn from the guaranteed funding from the central government, and the careful design of the program, such as clearer eligibility criteria and higher income thresholds. The main concern is how the funding is being allocated between local governments. Preliminary estimates suggest that some counties (judels) might not receive their "fair share" of the earmarked transfers for MIG implementation in the State Budget Law for 2002. The most disadvantaged judefs are those whose potential beneficiary population has been underestimated and whose poor population has larger average income gaps compared to other judels. The concern is that those jude,ts that receive relatively larger earmarked transfer amounts compared to the needs that emerge will tend to increase the number of program beneficiaries in order to use the available funds, and/or decrease the amount of their own revenues spent on implementing the program (at least the amount in excess of the legally required percentage). This main concern is compounded by other concerns about the administrative capacity of local governments to implement this program. However, since this is the first year of the MIG, the program includes an increased attention to monitoring, aiming at raising accountability at all levels and enabling mid-course corrections in case needed. The Design of the Minimum Income Guarantee Program 6.11 The Minimum Guaranteed Income legislation. The Law on the Minimum Guaranteed Income, adopted in late 2001 and effective starting January 1, 2002, underlines the renewed government commitment to ensuring the right of all Romanian families to a minimum income. It introduced a new means-tested cash transfer, the MIG benefit, to replace the former Means- Tested Social Assistance Benefit (MTSAB). Like its predecessor, the MIG program provides a monthly cash benefit that bridges the gap between actual family income and the minimum income threshold established for a family of a given size. The law also entitles MIG beneficiaries to heating allowances during the five months of the cold season, and introduces higher benefit levels for two other social assistance benefits, the "birth allowance" and assistance to the wives of military conscripts. 6.12 Institutional and legal framework for the MIG program. The institutional and legal framework for the delivery of the MIG benefits is fairly similar to the arrangements that existed for provision of the former MTSAB. The local governments (the mayor's offices) are responsible for delivering the MIG benefits, and the local councils are required to finance the 116 Chapter 6: Delivering Social Assistance at the Local Government Level benefit payments from the local budget. The MIG law places the responsibility for ensuring that the local councils and mayors comply with the legislation in the hands of the county (judet) offices of the Ministry of Labor and Social Solidarity (MMSS). Also, it gives these offices the authority to levy fines against the mayors and local councils for failing to perform their duties under the MIG legislation. Finally, the law lays the groundwork for national monitoring and evaluation of the program by requiring the mayors to ensure that the MMSS county (judet) offices receive updates on the number. of beneficiary families, the status of payments and the amount of cash benefits approved. In doing this, the Romanian government is aiming at strengthening the capacities of the these county offices and at building a social assistance network that is required to improve accountability, coordination, and consistency in the implementation of decentralized social assistance programs, such as the MIG. 6.13 The law heavily emphasizes the mayor's legal responsibility to provide the benefit according to the MIG legislation, and the local council's duty to provide funds for the payment of the benefit through the approved local budget. The law also states that applicants can challenge the mayor's decision in the administrative courts. However, it is worth noting that the MTSA legislation also contained this clause, with little apparent effect on the likelihood of city/town halls providing the benefit per the legislation. Following the shift of financing responsibility to the local councils, most city halls either stopped providing the benefit entirely, delayed benefit payments, stopped accepting new applications, or introduced additional, unauthorized eligibility criteria in an attempt to target the neediest families and limit benefit payments, due to local budget constraints. 6.14 Eligibility determination under the MIG program. As in the former MTSA program, those interested in receiving the MIG benefit apply at the social assistance office at the city or town hall, fill out an application, make a personal statement as to all family members' incomes, and submit supporting documentation. A "social investigation" conducted at the applicant's home to better assess the full income and assets of the family constitute a mandatory part of the eligibility and benefit determination process. This "investigation" is based on methodological norms that specify what income and assets local authorities must take into account when calculating the family's net monthly income. Although the specifications are similar to those under the MTSA benefit program, local authorities have somewhat less discretion than they did under the former program. The MIG law stipulates that net monthly income embraces all earnings (including social cash transfers, such as pensions, unemployment benefits and child allowances) as well as rents, land lease and other revenue generated from the ownership of land and buildings. The MTSA law specified additional asset categories (livestock, paintings, luxury objects) for local authorities to take into account when determining program eligibility. The MIG law has simplified the eligibility determination process, and to some extent decreased the scope for excessive discretion and local variability in the application of eligibility criteria, by emphasizing housing standards, land and real estate properties over other assets. 6.15 Income thresholds under the MIG program. The income thresholds established in the MIG legislation represent a substantial increase, in real terms, over the MTSA thresholds in effect during the period from 1998 through 2001, when the real value of the threshold was allowed to erode drastically. Still, the thresholds specified by the MIG legislation are less generous than those against which MTSA eligibility was tested in 1995, 1996 and even 1997. Chapter 6: Delivering Social Assistance at the Local Government Level 117 The MIG program targets the ultra-poor, about 10 percent of the population in 2002. As Table 6.7 shows, the real value of the income threshold for a single person family, for example, is just 66 percent of the 1995 MTSA threshold, which itself represented just 92 percent of the World Bank poverty line for Romania. The law mandates annual indexation of the MIG levels to account for inflation and maintain the real value of the thresholds and benefits. This is an important stipulation, since lack of income threshold adjustment was one reason MTSAB program coverage was so low. Table 6.7: The Dynamics of the Real MTSA/MIG Threshold (1995=100%) 1996 1997 1998 1999 2000 2001 2002 Single person family 70.75% 76.70% 53.89% 41.07% 29.58% 22.11% 66.39% For every additional person in the 70.87% 76.83% 54.04% 41.18% 29.65% 22.16% 29.88% family, for families with more than 5 persons 1) Data for 2002 refer to the MIG threshold. Source: Pop (2002). 6.16 Work incentives in the MIG program. The MIG legislation contains several provisions designed to prevent program participation from becoming a poverty trap for the beneficiary families. First, families whose monthly net income falls below the MIG threshold have their monthly benefit amount increased by 15 percent if at least one family member can demonstrate that he/she has an employment contract. Second, "work-fit" persons who are unemployed, or have no other paying job, must demonstrate that they are registered with the local employment and training office, and have not refused any job offers or training programs, to count as family members for the income eligibility calculation. Third, work-fit family members must perform community service activities, as assigned by the mayor each month that they receive MIG benefits. Conditioning benefit receipt on community work is not a program innovation, but a formalization of a de facto practice that most local councils had already adopted on their own initiative to increase the poverty targeting of the MIG's predecessor, the MTSAB. Pensioners, adults caring for children under the age of 7, students and persons with a documented work disability are exempt from fulfilling the workfare and employment office registration requirements. 6.17 Financing arrangements for the MIG program. The MIG law makes the local governments ultimately responsible for financing the benefits, by stipulating that the local council budgets shall support the MIG payments and the associated heating allowances. However, the legislation also specifies a source of funding from the state budget-an earmarked transfer from the funds set aside for the equalization grant. This is an important and positive development. For the first time since 1998, fiscal decentralization of a social program has been accompanied by the indication of a revenue source. It is worth noting, though, that the law does not suggest that the central government will fully fund the MIG program. The State Budget Law for 2002 earmarks funds for implementing the MIG program, by judet, but also stipulates for each judet an additional amount of funds that must come from the local budgets to support the MIG program. These vary by judet, but average out to about 25 percent of the earmarked transfers. 118 Chapter 6: Delivering Social Assistance at the Local Government Level 6.18 Are the amounts budgeted in 2002 for earmarked transfers to the judet councils adequate for financing the MIG program? The State Budget Law for 2002 dedicates 4,374.8 billion lei, or one-half of the total "equalization grant" funds budgeted for 2002, for the provision of earmarked transfers to local governments to finance benefit payments made under the MIG program. The revenue source for these funds is the wage (income) tax. Annex 5 of the state Budget Law specifies an earmarked amount for each judet, based on estimates of the number of potential MIG beneficiaries in each judef multiplied by an average benefit amount. These transfer amounts are projections; the actual budget execution could differ. For example, local councils may not be able to spend all of the earmarked transfer funds because of administrative bottlenecks, low program take-up, or difficulty meeting the State Budget Law's requirement that the local councils complement the transfers with an additional amount from their budget revenues. Still, an examination of the proposed amounts and distribution sheds light on the extent to which MIG program design and budget formulation support effective implementation of the MIG program by the local governments. 6.19 In September 2001 the MMSS estimated the number of families eligible for the MIG benefits and their distribution across the judels, based on the MTSA beneficiary distribution. The estimated number of potential beneficiary families equaled 758,000. The amount of funds needed to support the MIG program (including heating allowances per the MIG law) was then calculated by imputing an average income gap per family of 630,000 lei per month.82 The MMSS determined that 5,700 billion lei would be required to finance the MIG program in 2002. The Ministry of Finance adopted these estimates for the purpose of preparing the state budget for 2002. Earmarked state transfers actually budgeted for 2002 (4,374.8 billion lei) amount to 77 percent of this estimated cost. 6.20 As part of the Public Expenditure and Institutions Review, the World Bank commissioned a more precise estimation of the costs of implementing the MIG program in 2002, using Integrated Household Survey (IHS) data from 2000. 3 Two calculations were made. The first assumes that the number of persons eligible in January 2002, and the benefit amounts calculated for that month, carry over to the rest of the year. The second calculation accounts for the decline in the number of eligible families likely to occur as family incomes rise with inflation, while the income thresholds for the MIG remain the same. 6.21 Table 6.8 presents the estimates of the Ministry of Finance and MMSS, and those based on the IHS data. The IHS-based calculations indicate a much smaller number of potential beneficiaries than the MMSS estimates (683,038 families or about 75,000 fewer beneficiary families) and a somewhat higher level of financing requirements (6,762 billion lei for providing the MIG benefits and MIG heating allowances). Figures incorporating the potential impact of inflation indicate lower total financing needs since the number of eligible households declines in the course of the year. Depending on which estimates are used, the amount budgeted for 82TiS amount approximates the income threshold set in the MIG law for a single-person family. * See Pop (2002). The Integrated Household Survey is a family budget survey conducted annually by the NIS. The simulation inflated the income and consumption data to reflect the growth in wages, pensions, and prices that occurred between the survey date in 2000 and January 1, 2002, when the income thresholds established in the MIG legislation were scheduled to take effect. Chapter 6: Delivering Social Assistance at the Local Government Level 119 earmarked transfers to local governments for implementing the MIG covers between 65 and 81 percent of the required sum. 6.22 These initial estimates of the number of families eligible for the MIG benefit and the amount of payments they will be eligible for in 2002, compared with the amount of earmarked state transfers budgeted for 2002, suggest that the MIG program may operate as an underfunded mandate, at least in 2002. These estimates also completely ignore the administrative costs of implementing the MIG program, and merely estimate the monetary value of the cash transfers to all eligible families. This is completely unrealistic, of course, for a program that requires a means test, mandates conducting a social investigation at the residence of every applicant, and has a workfare component. Finally, the requirement in the State Budget Law that local councils fund a portion of the MIG expenditures ensures that the program will operate as an underfunded mandate, regardless of the theoretical coverage of the benefit payments, since the local governments must fund part of the delegated function (providing a nationally designed and mandated means-tested income transfer to eligible citizens within their jurisdiction) from their own resources. Table 6.8: Financial Needs for Implementing the MIG - Estimations of the Ministry of Finance versus those based on the Integrated Household Survey Data (Billion Average No. of lei) Beneficiary Families Estimations by Ministry of Finance and MMSS 758, 000 Estimations of total necessary amount by MMSS and MoF 5,700 Earmarked state transfers for upholding the MIG scheme 4,375 Estimated needs based on the Integrated Household Survey data, if 683, 038 replicating the January 2002 picture for the entire year IHS-based estimations of the amount necessary for upholding the 5,666 MIG benefit, excluding heating allowances IHS-based estimations of the amount necessary for upholding the 6,762 MIG benefit, including heating allowances per the MIG legislation Estimated needs based on the Integrated Household Survey data, 569,714 taking inflation during the year 2002 into account IHS-based estimations of the amount necessary for upholding the 4,449 MIG benefit, excluding heating allowances IHS-based estimations of the amount necessary for upholding. the 5,403 MIG benefit, including heating allowances per the MIG legislation Source: Pop (2002). Calculations based on Integrated Household Survey 2000 data, NIS and Ministry of Finance. 6.23 The transfer allocation process. The State Budget Law for 2002 specifies the amount of earmarked transfers each judel council will receive for distribution to the local councils to help cover the costs of the MIG program. MMSS estimations of the share of all potential MIG beneficiaries in each judef reportedly form the basis for this allocation. The jude; councils are responsible for distributing the funds to the local councils on a monthly basis, in proportion to the number of approved beneficiaries, as submitted by the mayor's offices. As noted above, the State Budget Law requires local councils to complement the funds transferred to their budget for the MIG program by allocating additional local budget funds for MIG implementation equal to a percentage of the amount transferred (this percentage varies by county (judet), but averages 25 percent). This allocation process for distributing the earmarked transfers has several 120 Chapter 6: Delivering Social Assistance at the Local Government Level shortcomings likely to present obstacles to the MIG program reaching its full potential as an effective and efficient poverty alleviation program: 6.24 Allocating the earmarked transfers to judets according to the historical beneficiary distribution of the MTSA program may perpetuate horizontal inequities. This distribution does not necessarily reflect the distribution of the poor or of the MIG-eligible population across the jude,ts. For example, the Local Social Services Delivery Study recently completed by the World Bank found no association in 1998 between a jude,t's poverty rate and the percentage of MTSAB beneficiaries in the jude,t population.84 The distribution of beneficiaries may reflect greater fiscal or administrative capacity rather than greater needs. 6.25 Table 6.9 compares the MMSS estimations used to allocate the earmarked transfers in the budget with the 2000 Integrated Household Survey-based estimations of the percent of eligible households in each subregion (each containing between two and four judets). The number of beneficiary households is underestimated by the MMSS in five subregions. Three of these subregions have high poverty rates and high poverty gaps relative to other subregions. Conversely, the MMSS calculations overestimate the percentage of eligible households in several subregions, notably Bucharest, which has the lowest poverty rate and poverty gap of all the subregions. The MMSS estimates the percentage of eligible households at 8.8 percent, while the IHS-based estimates indicate that only 4 percent of the population is eligible for the MIG. The IHS-based estimates correlate more closely with poverty rates than do the estimates the Ministry of Finance used to allocate the MIG transfers to the county (judet) councils. Table 6.9: Estimated Poverty Rates and Expected MIG Eligibility by Sub-Regions Sub-Regions Poverty MIG % of MIG Difference Rate Eligibility Eligible between IHS- Rate per IHS Households based Estimations data (% of per MMSS and the MMSS Households) Estimation Estimations 1 Botosani lasi Vaslui 24.0% 12.0% 12.3% -0.3% 2 Bacau Neamt Suceava 19.0% 12.0% 10.5% 14 3 Braila Tulcea Constanta Galati 16.0% 12.0% 10.3%1 4 Buzau Vrancea 15.0% 9.0% 9.8% -0.8% 5 Prahova Arges Dambovita 8.0% 6.0% 8.9% -2.9% 6 Calarasi Giurgiu Ialomita Teleorman 20.0% 11.0% 10.5%f 7 7 Olt Dolj Mehedinti 12.0% 8.0% 9.9% -1.9% 8 Gorj Valcea 15.0% 8.0% 9.1% -1.1% 9 Timis Arad 11.0% 8.0% 7.1% 9 10 Caras-Severin Hunedoara 14.0% 8.0% 10.4% -2.4% 11 Bihor Cluj 9.0% 5.0% 8.4% -3.4% 12 Maramures Salaj Satu Mare Bistrita- 16.0% 10.0% 10.9% -0.9% Nasaud 3 Brasov Sibiu 13.0% 11.0% 8.6% K 14 Harghita Covasna Alba Mures 12.0% 9.0% 9.1% -0.1% 15 Municipi llfov 5.0% 4.0% 8.8% -4.8% Bucuresti 13.93% 8.9% 9.7% -0.8% Source: Pop (2002). Calculations based on Integrated Household Survey 2002 data, NIS. 8 Although the study did find a correlation for 1997, indicating a decline in the poverty targeting of the MTSA program in 1998. Chapter 6: Delivering Social Assistance at the Local Government Level 121 6.26 In addition, the MMSS estimation of financial needs for implementing the MIG in each judef, which determined the final allocation of the earmarked transfers in the State Budget Law for 2002, does not account for differences among the jude,ts in the depth of poverty. By imputing an average income gap to all beneficiary families, the MMSS analysis assumes that the distribution of income among the poor is the same in all the judefs. This is not the case; given the same number of approved beneficiaries, two judets will have different costs associated with their MIG payments. 6.27 To sum up, some judels do not seem to receive their "fair share" of the earmarked transfers for MIG implementation in the State Budget Law for 2002. The most disadvantaged judefs are those whose potential beneficiary population has been underestimated and whose poor population has larger average income gaps compared to other jude,ts. It is highly unlikely that the funds will be redistributed across the jude,ts after the true needs become apparent. Local councils in those jude,ts that receive relatively larger earmarked transfer amounts compared to the needs that emerge will tend to increase the number of program beneficiaries in order to use the available funds, and/or decrease the amount of their own revenues spent on implementing the program (at least the amount in excess of the legally required percentage). 6.28 Glaring errors of inclusion will not necessarily result from this misallocation, since the funds are not unlimited and may well prove inadequate to cover benefit payments even for judets that receive a disproportionate share of the transfer relative to the distribution of the eligible population and the magnitude of the beneficiaries' average income shortfall in the judef. The more relevant case is probably the opposite one, in which local councils in poorer judels shortchanged by the allocation process, which would have to fund a larger share of the MIG payments from their own resources given the gap between needs and earmarked transfer amounts, may choose to cover a smaller share of their eligible population, and may apply more rigorous means testing or use other tactics to limit the amount of benefit payments. This could lead to horizontal inequities in program coverage and increase welfare disparities acrossjudets. 6.29 Conditioning local council receipt of transfers on proportional contributions from the local budgets may increase horizontal inequities and create perverse incentives. Although it is not clear whether, or how, the requirement will be enforced, if local councils are obligated to devote additional budgetary funds for implementing the MIG program in proportion to the amount of state transfers received, the MIG's ability to strengthen the Romanian safety net may be compromised. Poorer localities, which are likely to have both more limited fiscal capacity and a higher number of eligible households per capita, may have difficulty finding the budgetary funds needed to enable them to rbeceive the state transfers to which they are entitled given the number of eligible families in their jurisdiction. To fully implement the MIG program and provide benefits to all eligible families, local councils in this situation would have to spend more per capita, and a much larger share of their budget, on the MIG program. 6.30 One outcome of the proportional contribution requirement may be that local councils faced with a large potential beneficiary population, especially councils with limited fiscal capacity, will have perverse incentives to delay application processing, apply eligibility criteria in a more restrictive manner, enforce more rigorous work requirements to discourage program participation, or otherwise try to restrict the number of approved beneficiaries and payments. In 122 Chapter 6: Delivering Social Assistance at the Local Government Level short, those locations with the potential to benefit the most from earmarked transfers in per capita amounts, given the share of the poor in their total population, could end up being the locations that receive the fewest resources per capita from the state budget. Poor people in worse-off localities might be much less likely to receive the MIG benefit than poor people in better-off locations facing less daunting budget constraints. 6.31 Allocating earmarked transfers to local councils in proportion to the number of approved beneficiaries could create horizontal inequities. Just as judet-level variations in the depth of poverty require taking approved benefit amounts, and not just the number of approved beneficiaries, into account when aiming for an equitable allocation of the earmarked transfers among thejudets, geographic variations in the depth of poverty withinjudets suggest that judets should take approved benefit amounts into account when allocating the earmarked transfers among the local councils. Accounting only for the number of approved beneficiary families unfairly disadvantages poorer localities where the average benefit payments are higher because of deeper poverty (since the MIG payment depends on the shortfall between net family income and the minimum income threshold). 6.32 The process by which judets are to allocate the earmarked transfers to the local councils has not been clearly elaborated. The State Budget Law for 2002 indicates that the judets should distribute the funds to the local councils in proportion to the number of approved beneficiaries reported by the mayor. Many aspects of this allocation process are less than clear. One such component is the method by which the total amounts budgeted for earmarked transfers will be divided into monthly amounts. Will each month's distribution budget simply amount to one- twelfth of the total amount budgeted for the year, or will the monthly amounts available for distribution by the judets vary based on seasonal variations, inflation, application processing backlogs at the start of the program, and other variables that tend to affect the number of approved beneficiaries in a given month? This question becomes more important for judets where the amounts allocated are insufficient to cover the eligible population. 6.33 A second, major area of uncertainty is the method by which the judets will allocate the monthly earmarked transfers to the local councils. While the State Budget Law indicates they should be distributed proportional to the number of approved beneficiaries in each locality, the procedure for doing so is not clearly spelled out. How will the judets deal with the problem of local councils unable (or unwilling) to match the earmarked transfers with funds from their budgets? Will all local councils be expected to contribute an amount exactly proportional to their transfer receipts? Or will judets vary the requirements depending on the fiscal capacity of the local council, the extent of other competing demands on the local budget, or the locality's level of "need" measured in some other manner? Similarly, if the total amount available for distribution to the local councils in a given month falls far short of the total approved MIG payments for the judet, how will the judet allocate the available funds to the local councils? If the allocation occurs in strict proportion to the number of approved beneficiaries, and does not account for differences in fiscal capacity, horizontal inequities could result. 6.34 Given the lack of a well-defined allocation procedure, the process of allocating the monthly sums to the local councils could involve considerable discretion. The final allocation could depend to some extent on the outcome of negotiations between the judet councils and the Chapter 6: Delivering Social Assistance at the Local Government Level 123 local councils. Chapter 3 described howjudets allocated the equalization grants to local councils in 2000, using formulas or criteria that differed significantly from the formula mandated in the State Budget Law. The process varied considerably by judet, and often entailed a lot of discussion between judet councils and the local councils. The allocation process for the MIG funds is much less clearly elaborated than that for the equalization grant. Judets will probably adopt different approaches to allocation, as was true for the allocation of the equalization grant. 6.35 To conclude, considerable scope exists for elaborating the allocation process and for improving the guidelines so that the maximum poverty alleviation is achieved by the MIG program. 6.36 Heating subsidy. The other important entitlement under the MIG legislation is the heating subsidy. This subsidy entitles families enrolled in the MIG program to receive an additional amount to help defray residential heating costs during the cold season, from November 1 to March 31. The monthly heating supplement was set at 500,000 lei for families using thermal energy supplied by central heating systems, and 250,000 lei for families using other energy sources (such as wood, natural gas, kerosene, petroleum and coal) to heat their homes. The MIG heating allowance was intended to replace the means-tested heating subsidy introduced by the Romanian government in November 1999 and in operation, with periodic modifications to the eligibility criteria and benefit levels, through December 31, 2001. This subsidy represented part of the government's response to the difficulties families faced when the price of thermal energy was liberalized startinp in 1997, and heating prices for families connected to the central heating grid rose sharply.8 6.37 Initially the means-tested heating subsidies went only to families living in buildings connected to the central heating grid, but the government extended the heating subsidy program in September 2001 to include residential users of natural gas, who had also faced price increases. The heating subsidies were not well targeted to the poor. The governing legislation set such generous income thresholds that a large share of the heating subsidies went to the non-poor; the lowest income threshold was set above the poverty line. At the same time, the subsidy program left out many poor families, due to its urban bias and the exclusion of coal, wood, kerosene and petroleum users from the beneficiary population. Over half of all Romanian families, and a disproportionate share of the poor, use these energy sources for residential heating. 6.38 Before the MIG legislation went into effect on January 1, 2002, the Romanian government passed an emergency ordinance that instituted, at least for the period January through March 2002, a mixed system of targeted heating subsidies. For families using wood, coal, kerosene or petroleum to heat their home, eligibility for heating subsidies is conditioned on eligibility for the MIG benefit, and the heating allowance is provided per that legislation. Users of central heating and natural gas are eligible for heating subsidies based on the former legislation that regulated the November and December 2001 benefits. The emergency ordinance aimed to prevent the number of heating subsidy beneficiaries from decreasing, at least for the first part of 2002. As was the case under the previous legislation, the program differentiates 85 The government's response also included a general thermal energy subsidy to bridge the gap between the local cost of provision and a national reference price, partially funded from earmarked State transfers and paid by the local councils to the local energy provider. 124 Chapter 6: Delivering Social Assistance at the Local Government Level between central heating and natural gas users; the former are eligible for larger subsidy amounts relative to natural gas users with the same income level. The program also differentiates between lower and higher income beneficiaries by providing smaller heating subsidies to better- off families. 6.39 Table 6.10 summarizes the eligibility requirements and benefits provided under the mixed system of heating subsidies implemented January through March 2002. Institutional arrangements for heating subsidies for central heating and natural gas users remain as they were under the former program. Applicants submit income declarations, along with documentation from their employer or the unemployment office, to the administration of the flat block, where the beneficiary pays the heating bill, less the subsidy amount. The block administration submits the paperwork to the city hall, and the local authorities then transfer the entire amount of the subsidy for the block residents to the heating provider. 6.40 To support the additional costs of paying heating subsidies to central heating and natural gas users based on the former, much more inclusive eligibility criteria and larger benefit amounts, the Annual State Budget Law for 2002 provided for the transfer of an additional 2,151.0 billion lei from the state budget to the local councils whose jurisdictions have central heating grids or natural gas distribution networks. Table 6.10: Means-Tested Residential Heating Subsidies under the Mixed System, January 1 through March 31, 2002 MIG Heating Allowances for Wood, Subsidies for Central Heating Users Subsidies for Natural Gas Coal and Oil Users - Users' Subsidies for Natural Amount of Monthly per capita Amount of Monthly Amount of Gas Users monthly income (ROL) monthly per capita monthly heating heating income heating allowance subsidy (ROL) subsidy (ROL) (ROL)' (ROL)' < 630,000 - 1 person 250,000 <900,000 <900,000* family 700,000 400,000 < 1,134,000 - 2 person 900,001 to 1,100,000 420,000 family 900,001 to 240,000 < 1,575,000 - 3 1,100,000 person family 1,100,001 to < 1,953,000 - 4 person 1,400,000 210,000 1,100,001 120,000 family to < 2,331,000 - 5 person 1,400,000 family 1,400,001 to 1,800,000 110,000 1) These amounts represent maximum subsidy, which can not exceed the actual monthly heating bill. Source: Pop (2002) and MIG legislation. 6.41 The additional costs of providing heating subsidies from January 1 through March 31, 2002 based on the mixed system were estimated using Integrated Household Survey data for 2000. Table 6.11 presents the results of the simulation. The estimated additional cost is 1,748.1 billion lei. Thus, the 2,151.0 billion lei transferred by the state budget to the local council budgets should adequately cover the payments required under the emergency ordinance, assuming that local authorities successfully enforce income eligibility requirements. However, Chapter 6: Delivering Social Assistance at the Local Government Level 125 the earmarked transfer would not prove adequate to cover heating subsidy payments in the event the program is extended for November and December 2002. Table 6.11: Estimation of Additional Costs for a Mixed System of MIG Heating Allowances and Heating Subsidies as per the Former Legislation Estimated subsidy costs for January through March 2002 (in billion ROL) Necessary amount for providing all heating subsidies per the MIG legislation 657.9 Necessary amount for providing heating subsidies as in Nov./Dec. 2001 for 2076 central heating system and natural gas users Necessary amount for providing heating subsidies per the MIG legislation 330 only for those categories not covered by the Nov./Dec. 2001 heating subsidies (i.e., users of energy sources other than thermal energy or natural gas) Additional cost of implementing the mixed system of subsidies 1748.1 Source: Pop (2002). Calculations based on Integrated Household Survey 2000 data, NIS. 6.42 Other benefits included in the MIG Legislation. The Law on the Guaranteed Minimum Income also reaffirms the Romanian population's right to two additional social assistance benefits: the allowance for newborn children, a lump-sum cash benefit paid to mothers after the birth of a child, and the social assistance allowance for wives of military conscripts, a monthly income supplement paid to wives earning less than the minimum wage who are disabled, pregnant or caring for children under the age of 7. The legislation increased the "birth allowance" to 1,400,000 lei and emphasized that the local council budgets are responsible for funding the benefit. It also revised the eligibility criteria so that the benefit, formerly paid to mothers having at least one child, is now paid for each of the first four children. The eligibility criteria and revenue source (the state budget) for the military conscripts wives' benefits did not change under the legislation, although monthly benefit levels also increased to 1,400,000 lei. The birth allowance and monthly allowance for military conscripts' wives represent relatively minor social assistance programs in terms of the size of the budget attached to them and potential program coverage. Implementing the Minimum Income Guarantee Program 6.43 Local governments will face a number of challenges implementing the Minimum Income Guarantee program. The obstacles to the MIG -becoming an effective and efficient poverty alleviation tool include administrative capacity constraints, inadequate local incentives for implementing the MIG, and possible barriers to program participation by the poor. Some of these difficulties relate back to the design of the program, particularly its nature as a delegated function for which local councils have little decision making authority but will incur costs when implementing. The obstacles to effective implementation should be considered in the context they occur--dramatically increased loads on the administrative and financial resources of the local councils, and unstable expenditure assignments and considerable program uncertainty, due to the way fiscal decentralization is occurring and frequent changes in social policy and programs. 126 Chapter 6: Delivering Social Assistance at the Local Government Level 6.44 Administrative capacity issues in implementing the MIG. Full coverage of the MIG's target population implies growth in the number of families receiving means-tested social assistance on the order of 360 to 480 percent. Upholding the spirit of the law, which proclaims the right of all Romanian families to a minimum guaranteed income as a form of social assistance, also implies a large increase in the percentage of city halls providing the means-tested benefit. In most judets no more than 5 percent of city halls granted the benefit at all in 1999. Scaling up the program to this extent presents a major challenge to local councils. Especially during the first year of MIG implementation, the city halls' efficiency and skill in processing applications and conducting the social investigation may leave room for improvement, due to both limited experience implementing means-tested social assistance and limited human resources. Mayor's offices in small communes are likely to face a particularly challenging scenario, as they tend to have the most limited resources, while having a larger number of potential beneficiaries per capita. As Table 6.7 shows, a slightly higher percentage of rural households are eligible for the MIG. About 45 percent of the anticipated eligible population lives in rural areas. Table 6.12: Expected MIG Eligibility January 2002 % of Households Eligible for the MIG Total 8.7% Urban 8.3% Rural 9.2% % ofpopulation living in eligible households Total 10.3% Urban 9.3% Rural 11.6% Calculations based on the Integrated Household Survey 2000, NIS. Source: Pop (2002). 6.45 Many city halls in Romania have few employees, because they serve a small population and have limited resources. Table 6.13 shows the average number of full-time positions in local government administrations. Local administrations serving areas designated as communes (rural, consisting of one or more villages) have an average of 7.1 staff positions, including the mayor, the accountant and the secretary. More than 91 percent of all mayors' offices are located in communes. Seventy-one percent of all mayors' offices serve populations under 5,000, with an average number of staff positions between 4.6 and 7. One-quarter of all city halls do not have any employees with a university degree. Chapter 6: Delivering Social Assistance at the Local Government Level 127 Table 6.13: Number of Full-time Public Employees by Size and Legal Status of the Local Administrative Unit 2001 Total no. of Positions for Full Proportion of Vacancies time Public Employees (Non-auxiliary Staff) Average median Average median Total 11.7 7 5.3 0 Size of the locality Under 1,000 inhabitants 4.6 4 1 0 Between 1,000 and 2,000 inhabitants 5.6 5 5.3 0 Between 2,000 and 5,000 inhabitants 7 6 5.3 0 Between 5,000 and 10,000 inhabitants 10.3 9 4.5 0 Between 10,000 and 5000 inhabitants 35.7 30 8 3.9 Over 50,000 inhabitants 186 151 8.7 9 Legal status of the locality Commune (rural) 7.1 6 5.1 0 Town 26.2 24 6.8 5.4 Municipal capital city 63.4 59 6.9 3.6 Judel capital city 205 172 9.1 9.1 Data source: Democracy of Local Government Survey - Tocqueville Research Center Budapest, 2001 Source: Pop (2002). 6.46 These relatively few public servants have taken on an expanded range of responsibilities in recent years as a result of fiscal decentralization. Most notably, starting in 1999 each city hall became responsible for collecting its own revenues usingilocal resources. This put a strain on local administrations, especially those having fewer than 10 employees. Some city halls had to hire additional personnel to carry out their new responsibilities for tax collection, lowering the opportunities to increase the quantity and quality of human resources available for the implementation of other programs. 6.47 Given the small number of staff in the numerous small city halls, financial and human resources are, in most cases, too limited to permit functional specialization and professionalization of staff, although the newly transferred responsibilities regarding tax collection, social assistance provision and financing/provision of other public services to some extent require these. Of relevance for implementing the MIG, the smallest communes sometimes have no designated employees for social assistance. The mayor, the accountant and the city hall secretary register and screen beneficiaries for means-tested benefits. The small number of city hall staff having professional training in social assistance provision is compounded by Romania's generally limited experience with means-tested social assistance, since the first such program was only introduced in 1995. 6.48 These administrative capacity constraints indicate that local authorities will face challenges effectively screening MIG applicants via the means test. However, aspects of the MIG's design complement the means-testing procedure in helping target benefits on the poor. First, by keeping benefits low, the program is less likely to attract non-needy persons. Second, conditioning benefit receipt on performing community work, if the work requirement entails a significant time commitment and is effectively monitored, provides an additional targeting method, since the requirement discourages those who don't really need the assistance from applying. The effectiveness of this targeting method has already been proved, to some extent, by 128 Chapter 6: Delivering Social Assistance at the Local Government Level the fact that many of the mayor's offices instituted work requirements for the MTSA on their own initiative as an additional eligibility requirement to limit the number of beneficiaries. 6.49 Based on the documented experience with the MTSAB, the MIG has tremendous potential to become Romania's most efficient and effective poverty alleviation tool. The largest stumbling block may not be implementation capacity, per se, but rather the local councils' ability and incentives to implement the MIG program in the face of competing demands on local budgetary resources and constantly shifting expenditure responsibilities, particularly since not all local officials appear convinced of the merit of means-tested social assistance. 6.50 Local Council Incentives to Implement the MIG. As Chapter 3 discussed, the process of fiscal decentralization has proceeded in a manner that has created considerable uncertainty and challenges for local governments. Expenditure responsibilities have shifted almost continuously, often in the annual State Budget Law but sometimes even during the middle of the fiscal year. At times local councils have had unfunded mandates thrust upon them, particularly in the area of social assistance. Obviously this situation has made it difficult for local councils to plan their activities and budgets in any meaningful way, and has put pressure on local resources. 6.51 As the Romania Local Social Service Delivery Study (World Bank, 2002) noted: * "The transfers of additional financing responsibilities, often following the start of the fiscal year - and in an environment of tight fiscal constraints - has led local government officials to perceive decentralization as an increase in imposed obligations, rather than an opportunity to develop locally responsive policies. As a result, local government decision-making in social assistance and education has been largely focused on developing coping strategies to address the tight fiscal environment." 6.52 In this regard, Romania's recent experience with the Means-Tested Social Assistance program is relevant for assessing the obstacles to effective implementation of the MIG, in that many of the factors that contributed to the local councils' failure to provide the benefit as directed by law still exist. The introduction of earmarked transfers to fund a large portion of the anticipated needs certainly is an improvement over the former situation, but that innovation has not solved all of the fiscal issues that negatively affect the local councils' ability and inclination to implement a means-tested social assistance program imposed from "above." In part this issue goes back to program design and budget formulation. These budgetary pressures are only a factor because the program functions as an underfunded mandate by not covering all the costs local councils will incur implementing the MIG, such as administrative costs and the required local budget contribution in proportion to earmarked transfer receipts. 6.53 Pressures on local budgets have increased, mostly due to growing responsibilities for funding social assistance institutions and services. In 1999 expenditure responsibility for all social assistance institutions, other than those for abandoned children, shifted to local councils. Because most of these institutions, especially those for the handicapped, have been "hidden" in rural areas, the pressures have been enormous on most local councils with such institutions located in their jurisdiction. In 2000, the Social Solidarity Fund supported some of these costs, Chapter 6: Delivering Social Assistance at the Local Government Level 129 but at the same time, all of the costs relating to salaries for personal attendants for handicapped persons were transferred to the local budgets. An additional social assistance expenditure imposed on local councils in 2000 was a contribution for the support of abandoned children in social assistance institutions, payable to the judet council. Expenditure responsibility for a host of services for the aged also shifted to the local councils, with the result that these services are no longer being provided. 6.54 These represent mere financial obligations; the local councils have no control over the organization or provision of the services, which are under the authority of deconcentrated judet level line agencies. As a consequence of the increase in inelastic expenditures in the field of social assistance, local governments further decreased the only elastic expenditure for which they had control regarding its provision: the MTSAB. These budgetary pressures could affect the local councils' motivation and ability to provide the MIG benefits as well. Moreover, the increase in earmarked transfers to local councils in 2002-for the MIG program, educational salaries and decentralized cultural institutions-has been accompanied by a decrease in equalization grants, both in real terms and as a share of total state transfers, leaving fewer resources available for local councils to address other pressing needs. 6.55 The frequent changes in local councils' expenditure responsibilities, as well as in national social policy and programs, present an additional challenge to effective implementation of the MIG. Local councils have little incentive to invest in the MIG program, say by enhancing their human resources for effectively screening applicants and administering the program, if its future is uncertain. Romania's policies in this area change so frequently that it is hard to keep up with them. For example, a key part of the MIG legislation's design entailed the replacement of the former means-tested heating subsidies with a much more narrowly targeted heating allowance. Before the MIG legislation could go into effect, it was in some ways superceded by the emergency ordinance on heating subsidies. The MTSAB's history is also relevant in this regard. Soon after the MTSAB was introduced as a centrally financed, but locally implemented, program, the local councils became fully responsible for funding the program. There is little stability in the policy arena. 6.56 Another factor that may influence the degree to which local councils actively promote the MIG program is the attitude many local officials take toward means-tested social assistance. Interviews conducted with local officials as part of the Local Social Service Delivery Study (World Bank, 2002) found local authorities tend to view means-tested social assistance as their last priority when ranking their social assistance cash benefits; birth grants, scholarships and emergency assistance were viewed as greater priorities. Similarly, many local officials express a preference for in-kind social assistance such as the provision of free meals through soup kitchens. There is a stigma attached to the receipt of means-tested cash benefits. 130 Chapter 6: Delivering Social Assistance at the Local Government Level Conclusions and recommendations 6.57 This chapter examines the opportunities and challenges faced by local government in implementing the newly designed Minimum Income Guarantee program. The main opportunity is to adequately address concerns with extreme poverty. The MIG has a strong poverty reduction potential, with simulations prepared for this report indicating that the MIG scheme could reduce the poverty rate in 2002 by about 18%, the poverty gap index by 36%, and the severity of poverty by about 52%. This strength is drawn from the guaranteed funding from the central government, and the careful design of the program, including clearer eligibility criteria and higher income thresholds. The main challenge is to get the funding for this program to be better allocated between local governments. Preliminary estimates suggest that some counties (ude,ts) might not receive their "fair share" of the earmarked transfers for MIG implementation in the State Budget Law for 2002. The most disadvantaged judefs are those whose potential beneficiary population has been underestimated and whose poor population has larger average income gaps compared to other judefs. The concern is that those judels that receive relatively larger earmarked transfer amounts compared to the needs that emerge will tend to increase the number of program beneficiaries in order to use the available funds, and/or decrease the amount of their own revenues spent on implementing the program (at least the amount in excess of the legally required percentage). This concern is compounded by other concerns about the administrative capacity of local governments to implement this program. However, since this is the first year of the MIG, the program includes an increased attention to monitoring, aiming at raising accountability at all levels and enabling mid-course corrections in case needed. 6.58 To capitalize on the potential of the MIG and smooth out the shift in expenditure responsibilities to the local councils in the area of social assistance, actions are required on the following fronts: > Elaborating comprehensive allocation criteria for the earmarked transfers for the MIG, to ensure an equitable allocation among, as well as within, judets and to improve the incentives and ability of the local councils to provide the benefit. > Reconsidering the requirement that local councils complement the amount of earmarked transfers received with a proportional amount from their own budget, particularly for poorer localities with a larger share of eligible families. > Improving the administrative capacity of the local administrations to implement the program, particularly the means test. > Strengthening the monitoring and evaluation provisions related to the MIG program at all administrative levels, increasing accountability and enable mid-course corrections in the MIG program, in case needed. CHAPTER 7. CONTINUING REFORMS IN THE PENSION SYSTEM Introduction 7.1 Romanian public pension system constitutes the single largest expenditure in the Governments consolidated budget, accounting for between 7 and 8 percent of GDP. It is also a perennial source of fiscal pressure, as the system has been running persistent deficits since early 1990s. Deficits in the operations of pension systems are a common feature for transition economies in Central and Eastern European economies, where pension systems have been strained by the changes in the labor market and the socialist legacy of relatively generous provisions for certain professional categories. Misguided policies to deal with the restructuring of the labor market in Romania have, however, exacerbated these imbalances. As a result, by the end of the first decade of transition, the Romanian pension system had one of the most unfavorable dependency ratios in the region, reaching 98 in 2001,86 and making reforms more urgent and more difficult. 7.2 The urgency and the difficulties in reforming the Romanian pension system reflect very different pressures. The urgency reflects the need to bridge a recurring deficit in the system without resorting to further increases in contributions. The contribution rate at 35 percent of the gross wage is already quite high (especially when compared to the average benefit), and has encouraged evasion through early retirement and informal sector employment. This has only worsened the system's dependency ratio, and the sustainability of its finances. The difficulties in implementing pension reforms in Romania reflect the obstacles in establishing an institutional setting that provides transparency and stability to the system. The reforms introduced in April 2001 illustrate the nature of this problem. The reforms aimed at placing the system on a sounder financial footing by expanding contributions to the entire labor force, raising the retirement age, reducing benefits to privileged categories, and defining a benefit formula that made the system actuarially fairer. The managers of the system retained, however, a considerable amount of discretion, allowing the actual benefit to be set within a very wide range. Also, soon after the system became effective, amendment were introduced reversing some of the gains. These included the lowering of the penalties for non-compliance, and the increase in the maximum ceiling for pension benefits. The results of the reforms were, therefore, largely undone by an institutional setting that is not accountable to either the contributors to the pension system, or the great majority of the beneficiaries. 7.3 This chapter examines how this lack of accountability has worsened the pension system's structural problems, and how this can be resolved to allow needed reforms to be implemented. The chapter is organized as follows. It begins with a brief description of where the system now stands. This includes the financial performance of the pension system, the number of beneficiaries and contributors, and their contribution rates. It turns next to recent reforms that define the current design of the system, and to an assessment of how these reforms are being implemented. The chapter closes with a summary of the policy recommendations, covering both 86 If one includes the farner's pensioners fund, the dependency ratio reaches 134 percent in 2001. 132 Chapter 7: Continuing Reforms in the Pension System the reform of the public pension system and the options for moving toward a system that encourages greater accountability by providing greater choice to pension system participants. Structural Imbalances in the Romanian Pension System 7.4 Romania's mandatory pension insurance scheme provides old-age, disability and survivor pensions to 6.2 million beneficiaries.87 It also provides other non-pension benefits, such as sickness benefits, birth allowances, and childcare, although these only constitute about 6 percent of total expenditures. The system accounts for a large share of consolidated government expenditures, reaching an average of 7.2 percent of GDP during the period 1995-2001 and more than 8 percent in 2001 (Table 7.1). The present system, based on a Defined Benefit (DB) Pay- as-you-go (PAYG), results from the merger of the state and the farmers' pension schemes under the new pension law that has been effective since April 2001. The State Pension System, that was already based on a PAYG DB plan, was mandatory for salaried workers, and voluntary for the self-employed and workers with civil contracts, although participation from the last two groups was negligible. The Farmers' scheme was largely non-contributory, and special taxes were imposed on companies producing and/or commercializing agricultural products to finance the scheme. Table 7.1: Financial Performance of the Pension System, 1995-2001 (% of GDP) Revenues State Farmers' Short-term Other Balance Pensions Pensions Benefits 1995 6.1 5.9 0.4 0.4 0.1 -0.8 1996 5.8 5.8 0.5 0.4 0.1 -1.1 1997 5.0 5.2 0.6 0.4 0.1 -1.3 1998 5.5 5.9 0.6 0.5 0.1 -1.6 1999 6.0 6.1 0.5 0.5 0.1 -1.2 2000 6.8 6.4 0.5 0.5 0.1 -0.8 2001 7.2 7.0 0.6 0.5 0.2 -0.9 Source: National House of Pensions. World Bank staff calculations 7.5 One of the reasons underlying the financial difficulties of the pension system has been the rapid increase in beneficiaries since the early 1990s. Between 1990 and 2002, the number of beneficiaries in the state and farmers' pension system surged from 3.4 to 6.2 million, reaching a cumulative increase of over 70 percent (Figure 7.1). While the largest one-time increase took place during the early years of the transition, the growth in the number of beneficiaries still remained high (close to 4 percent) during the second half of the decade, outpacing the growth rate of the elderly population." Indeed, while the legal retirement age is 62 for men and 57 for women, the average retirement ages for men and women are 56 and 51, respectively. This rapid rise in the number of beneficiaries is also reflected in the increase in the number of disabled, 7In 2001, old-age pensioners comprised 70 percent of beneficiaries; disabled and survivors constituted the remaining amount in approximately equal terms. " This reflects the fact that the system implicitly subsidized early retirement because of its high tax on additional work. This implicit tax is the higher payroll tax plus the reduction in the present value of total lifetime pension due to the loss of a year's benefits. Prior to the recent pension reform, the actuarial yield dropped from 1.8 percent at age 55 to only 0.7 percent at age 60 for a man and from 3.1 percent at age 52 to 1.6 percent at age 60 for a woman. Chapter 7: Continuing Reforms in the Pension System 133 which rose significantly during the 1990s because of lax rules allowing workers to claim disability to cope with unemployment.89 7.6 The rapid growth in the number of pensioners continued into 2001. Although the new Pension Law tightened retirement provisions and increased the retirement age, it granted a one- year transition period that allowed a sudden increase in the number of pensioners in 2001. This rise in the number of pensioners should decelerate only over the next few years and the rest of the decade, as the new provisions take effect. Figure 7.1. Pension System: Contributors and Beneficiaries, 1990-2001 (in thousand) 10,000.0 8,000.0 6,000.0- 4,000.0 2,000.0 0.0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 I IState Pensioners =Farmers' Pensioners 6 Employees Source: National House of Pensions (NHP) 7.7 This increase in the number of pensioners, while the number of contributors dwindled, meant that the real value of pensions fell sharply. The impact on the real value of pensions was further affected by the decline in wages, which have not recovered from the sharp drop experienced in the early 1990s (Figure 7.2). The end result has been that pensions have declined faster than wages, leading to a fall in the net replacement rate (net old-age pension to net wage) from 51 percent in 1990 to 47 percent in 2001 (The gross replacement rate was 36 percent in 2001).91 While the current net replacement rate is adequate for a mandatory insurance scheme, it is low relative to the high contribution rate, as will be discussed further below. 89 While the number of beneficiaries rose almost two-fold during this period, the share of disabled as a proportion of total beneficiaries within the State Pension System still grew from 9 percent in 1989 to 15 percent in 2001. 9 As of 2000, the methodology for calculating net pensions deducts both taxes and the share of social security contributions paid by the employee. The previous methodology only deducted taxes. 9' The continued fluctuation in pensions from 1995 through 2001 responds to ad hoc indexation of benefits, which frequently lagged inflation and was later followed by special plans to recalculate (re-correlate) benefits. Since 1996, five re-correlation plans have been issued for the State Pension System and two re-correlation plans for the Farmers' Pension System. (See Annex 1). Indexation rules and re-correlation plans have not been uniform across different groups of beneficiaries--old-age, disability, and survivors-nor among members within a group. As a result, pension benefits have undergone periods of compression and periods of subsequent decompression. All in all, the system remains fairly compressed. 134 Chapter 7: Continuing Reforms in the Pension System Figure 7.2. The Evolution of Wages and Old-aged Pension Benefits, 1990-2001 (1990=100) 120 100 80 X 60 20 0 1990 1993 1995 1997 1999 2001 Net Fnsion index 1990 100 Net Wage index 1990 =100 Net Replacement Rate (pension average) Source: NHP and World Bank estimates. 7.8 The Romanian authorities responded to these developments by raising contribution rates. The rate rose from 28 percent of gross wages in 1992 to 30 percent in 1998, and 35 percent in 1999 (Table 7.2). While the contribution rate of contributors in categories I and II (i.e., with access to special early retirement provisions) was 10 and 5 percentage points higher, the number of individuals within these special categories was allowed to increase to about 30 percent of all participants, further increasing the system's future liabilities. The result was that the system revenues fell as percent of GDP during the early transition years, only modestly recovering in the second half of the decade. 7.9 Meanwhile, the contribution base shrank from 8.2 million insured in 1989 to 6.2 in 1995, and further to 4.5 million in 2000. This reflected the rise in unemployment, and the growth employment in the informal sector. For instance, the number of self-employed stood at 1.8 million in 2000, with less than 70,000 participating in the system. Also, while the new Pension Law expanded mandatory coverage to self-employed, farmers and workers under civil contracts, the impact in terms of increasing the number of contributors to the system has been negligible. The sum of self-employed, workers under civil-contracts, and agricultural workers participating in the system barely surpassed 110,000 at the end of 2001, well below the 2.7 million individuals estimated to have this employment status. Chapter 7: Continuing Reforms in the Pension System 135 Table 7.2: Pension Payroll Taxes (1992-2001) 5/1992-11/1998 12/1998-4/1999 5/1999-3/2001 Category I 38 40 45 Category II 33 35 40 Category III 28 30 35 Others* 18 20 20 Farmers 7 7 7 *Includes self-employed and civil contracts Source: NHP The System's Design: The New Pension Law and Subsequent Amendments 7.10 ' The new Public Pension Law, that became effective April 2001, was the first attempt to introduce to place the public pension scheme on a sounder financial balance. The reforms included: > Expanding the coverage of the public scheme to the entire the labor force, including farmers, self-employed, and unemployed; 0 Increasing the retirement age from 57 to 60 for female, and from 62 to 65 for males by 2014, and a concomitant rise in the minimum length of service; > Defining a new benefit formula (based on a point system) accounting for the entire contribution history and treating early retirement on a near actuarially fair basis; and > Rationalizing the retirement benefits for special and hazardous occupations that benefited from early retirement provisions with an increase in their minimum retirement age and required length of service. 7.11 While the new law constitutes a step in the right direction, it still leaves a significant amount of discretion in the hands of the Romanian authorities. For instance, the point value for estimating benefits can vary within a very wide range. Also, several Emergency Ordinances to amend the Public Pension Law were introduced soon after the new law became effective, reversing some of the provisions of the new law. These included lowering the penalties for non- compliance with payment obligations and raising the maximum ceiling in the point value from 45 percent to 50 percent.92 These amendments were, furthermore, introduced with no public discussion and no known analysis of their implications for the financial stability of the pension system. 7.12 The financial impact of these amendments to the new law was compounded by the decision to re-correlate the pension benefits of individuals that had retired before 1999. The decision taken in late 2001 aimed at correcting some of the past inequities that resulted in the erosion of the real value of pensions for the pre-1999 cohort and subsequent ad hoc changes in 92 The lower ceiling is set at 30 percent of the average gross wage. 136 Chapter 7: Continuing Reforms in the Pension System pension benefits.93 While the decision to redress past inequities is understandable, they way in which it is being financed raises concerns about the transparency and stability of the system over the next few years. The re-correlation plan envisions that the point value will lag behind inflation for the period 2002-04, meaning that the value of all other pensioners will fall in real terms. Also, there is no clarity on the future value of points-and therefore benefits-after 2004. This might open room for pressures coming from several groups of beneficiaries. For instance, the disabled who were not included in the re-correlation plan might later claim compensation. New pensioners with low benefits might also claim to be part of the re-correlation, particularly when they realize that the new indexation provisions will not fully maintain the real value of benefits. In view of budget constraints and limitations on correcting all past inequities, it would have been preferable to apply less arbitrary rules and focus on some social priorities, such as raising the pension level of all beneficiaries who fall below a minimum threshold, while maintaining the real value of other benefits. Implementing the Reforms in the Public Pension System 7.13 While improvements were introduced with the new law in April 2001, important structural problems and institutional weaknesses still hamper the operation of the system. The structural problems include a disincentive for participation because of the very high contribution rate (especially when compared to benefits provided) and the impact of worsening demographic trends. The institutional weaknesses include outdated collection and contribution enforcement structures, limited policy making capacity, and low levels of accountability because of excessive discretion left in the hands of policymakers. 7.14 Very high contribution rates. As indicated above, a significant part of the financial problems of the pension system reflect the continued narrowing of the system's revenue base. The contributions rates to the pension fund, at 35 percent of gross wages, have reached very high levels, and they are even higher for workers in special and hazardous occupations. After accounting for contributions to other social insurance funds (health and unemployment), the sum of payroll taxes adds to 55 percent of gross wages (Table 7.3). Moreover, there is another 3 percent payroll tax that is planned to be introduced to finance the new occupational risk insurance. This would bring payroll taxes to 58 percent of the gross wages, making the Romanian workforce one of the most heavily taxed workforces in Central and Eastern European.94 9 These re-correlation plans will revalue the pensions of old-age pensioners (with full length of service) and survivors that retired prior to 1999. Since existing pensions were converted into the point system as of April 2001, the benefits of existing pensioners will be revalued by increasing the number of points. The recalculation will be implemented in phases from 2002 through 2004. Those with lowest pensions will receive higher increases up front. 9 By comparison, payroll taxes in Bulgaria and Poland, some of the highest in the Central and Eastern European Region, amount to 42.7 percent and 46.8 percent, respectively, including pensions, short-term benefits, work injury, unemployment insurance and health insurance. Among OECD countries, Italy has one of the highest payroll tax rates, close to 40 percent. While burdensome compared to international standards, the Italian payroll tax rate is significantly below Romanian levels. Chapter 7: Continuing Reforms in the Pension System 137 Table 7.3: Payroll Taxes (Percentage of Gross Wages) Fund Employer Employee Total Pension and Short-term Benefits 23 12 35' Unemployment .5 1 6 Health 7 7 14 Occupational Risk2 3 -- 3' All sectors 38 20 58 ) Mil taxes for special and hazardous conditions are 5 and 10 percent higher. 2) Occupational risk insurance has not yet come into effect. Source: World Bank Staff calculations. 7.15 These very high payroll taxes have raised non-wage labor costs, with adverse consequences for employment in the formal sector and for participation the pension system. These problems of low levels of formal sector employment, and, as a result, low participation in the pension system, are compounded by the degree of discretion exerted in managing the pension system. This high degree of discretion in managing the pension system has reduces the credibility in the eyes of contributors. In an environment of high uncertainty about the future value of pension benefits, and an already very low rate of return on current contributions, employers and employees have been evading the system either by contributing the minimum amount possible, or by migrating to the informal sector. The system is therefore trapped in a vicious circle, where the high contribution rate continuously narrows the contribution base. 7.16 Aging demographics. Romania's demographic profile is expected to undergo an important transformation over coming decades. The trend is similarly to other Central and Eastern European countries. According to World Bank population projections, life expectancy at 60 for males and females is expected to rise, respectively, from 75.8 and 79.2 in 2000 to 80.2 and 83.1 in 2050 (Table 7.4). Fertility rates are also projected to recover from the slump suffered during the last decade, reflecting renewed economic growth and enhanced living standards. Overall, the Romania population is expected to age rapidly as the rise in longevity outweighs the recovery infertility rates, and the old-age dependency ratio grows from 35.4 in 2000 to 50.9 in 2050. This aging process will impose an additional retirement burden on the working age population, as each worker will have to support am even larger number of retirees.95 Table 7.4: Romania's Aging Population 2000 2025 2050 Total population (millions) 22.4 20.9 19.4 Old age dependency ratio (%) 35.4 31.6 50.9 Life expectancy at 60 Males 15.8 18.0 20.2 Females 19.2 21.1 23.1 Source: World Bank. 95 The old-age dependency ratio measures the share of the population above the current retirement age (62 and 57 for male and female, respectively) to working age individuals (age 15 to retirement). 138 Chapter 7: Continuing Reforms in the Pension System 7.17 Collection and contribution enforcement. The administration of the pension fund- including collection and payment of benefits-has traditionally been the purview of the Ministry of Labor and Social Protection (MLSP), and has undergone very little modernization over time. The collection and enforcement methodology is outdated and ill-suited to the challenges of operating in a market economy. The pension fund had no registry of contributors, and there is no systematic methodology for planning, auditing and control activities.96 Most of the agency's efforts are focused on benefit payment. Services also have fallen behind international standards with delays in the processing of claims and provision of information to affiliates and pensioners. 7.18 As part of a broader effort to improve pension administration, the new pension law established a new independent agency for administering the system, the National House of Pensions and Other Social Insurance Rights (NHP). The separation of the NHP from the MLSP was designed to promote greater operational flexibility. However, the NHP's independence was reversed already in 2001, becoming a quasi-department of the MLSP. While important institutional changes are now underway at the NHP, progress has been slow. There have been extensive delays in the nomination of the NHP Chairman and the formation of a more permanent management team. The management team is extremely thin, limiting its ability to respond to day-to-day activities and plan institutional changes. Further, salary constraints hinder the institution's capacity to attract staff for highly technical positions. As a result, the establishment of the registry of contributors is not yet complete. The development of information technology systems (IT) to support the new collection methodology is also delayed. The latter is a problem because new information technology system is not only about the acquisition of the needed hardware, but a first step in the overall reengineering of key business processes. 7.19 Besides concerns on the slow progress achieved within the NHP, the broader institutional framework for collecting social security contributions remains fragmented and highly inefficient. At present, there is a separate department (or quasi-government agency) for collecting unemployment insurance, also under the MLSP, which duplicates the NHP collection, audit and enforcement arrangements. The National Health Insurance Fund (NHIF), a quasi-government agency within the Ministry of Health and Family administering the health fund, also has its own collection structure. Information exchange between these agencies and the tax administration is extremely limited. 7.20 Policy making capacity. These institutional weaknesses are exacerbated by the short- term focus of policy-making process surrounding the pension system. Most of the attention goes toward managing the day-to-day operations of the system and toward accommodating political pressures for increasing benefits in the immediate term, with no regard to the medium- and long- term implications. The yearly preparation of the budget for the pension fund illustrates these problems well. The budget for 2002 assumed that an estimated 2.7 million self-employed farmers and workers under civil contracts would shift to part time employment contracts that require social contributions. These workers under part time contracts would generate an additional Lei 6.0 trillion (0.4 percent of projected GDP) in revenues, up from around Lei 0.5 9 Until April 2001, contributions were collected on the basis of the total wage bill of an enterprise, and no disaggregated information on each worker was submitted to the MLSP. Contributions were recorded by the employer on a workbook. On retirement, the workbook was presented to the MLSP to calculate benefits, easily lending itself to fraud. Chapter 7: Continuing Reforms in the Pension System 139 trillion in 2001. The budget was approved with this twelve-fold increase in revenues, notwithstanding the low likelihood of cash-constrained farmers signing on for social security contributions, and the even lower probability that civil contracts could be easily changed in time to meet the budget's revenue projections. Also, the budget was approved with an estimated collection of 85 percent, up from 77 percent in 2001. Where Next? Evaluating the future of the pension system 7.21 To evaluate the future of the pension system, the World Bank has simulated the impact of these recent reforms on the financial accounts of the system during the 2002 to 2050 period. These simulations are based on three main assumptions. The point value will be maintained at 35 percent of the gross wage, close to its current value, and benefits will be indexed to inflation. Also, there will be a progressive improvement in compliance, reaching 87 percent and 89 percent in 2030 and 2050, respectively, due to a gradual expansion of the system to farmers, workers under civil contracts, and the self-employed. Finally, the number of workers under special and hazardous occupations benefiting from early retirement is maintained at 6 percent (its current level) throughout the projection period.97 7.22 Figure 7.3 presents the simulation results. The results suggest that the public pension system can only achieve financial sustainability by increasing compliance rates, while setting prudent replacement rates and indexation rules. One concern is that the resulting replacement rate is disproportionately low compared to the contribution rate, making the mandatory scheme unattractive for many of the younger participants. Another concern is that the results are extremely sensitive to the assumptions of the simulation. For instance, lower real GDP growth rate (positive but in the range of one to two percent) would shift the pension system back into deficit, even under prudent management with low replacement rates and price indexation rules. Also, indexing benefits to the average wage (rather than the consumer price index as in the optimistic scenario) would wreak havoc with the system, even under a strong macroeconomic growth scenario. Wage indexation leads to a rapid rise in expenditures, broadening the wedge between revenues and costs, and producing ever-growing deficits, approaching 5 percent of GDP by the end of the projection period. 97 Other macroeconomic and demographic assumptions are included in Annex 3. 140 Chapter 7: Continuing Reforms in the Pension System Figure 7.3 Financial Performance of the Public Pillar Under Different Scenarios, 2002-2050 (percent of GDP) 1.5% 0mmeam -3.0%- -4.5%- -6.0% - Base case - - - Wage Indexation Pessimistic Policy Recommendations for the Second Phase of Pension Reform 7.23 Pension reform is far from complete in Romania. To move forward, the next phase of reform needs to set forth a clear long-term direction for the pension system, not only focusing on current beneficiaries but also on the future security of younger workers during their old age. The design of the next phase of reform needs to address three key concerns: restoring confidence in the stability of the system by increasing transparency; achieving financial stability by rebuilding the system's contribution base, and providing adequate income during retirement for current and future pensioners. This in turn requires considering the following measures. Reducing discretion in defining the parameters of the pension system. 7.24 Legislating a viable replacement rate and indexation rule: The new pension law did not stipulate the provision for determining benefits for a given level of accumulated contributions to the system. In practice, this open point system offers extremely weak links between contributions and benefits. Indexation of benefits is also open to discretion. There is no pension system in the European landscape that enjoys this wide level of discretion.98 Other European transition economies that endured important pension crisis during the 1990s and exercised discretionary policies to cope with it have by now turned around. The reformed pension systems are rooted in more transparent and financially viable rules in order to re-establish confidence and rebuild their diminished contributory bases. The legal framework of the Romanian pension system needs to be amended, therefore, to stipulate the long-term replacement (i.e., point value) and a reasonable post-retirement indexation policy. These parameters should be set at financially viable levels over the long-term, so participants can count on their pension promises and their trust in the system can be rebuilt. Pre-defining benefit policies will lead to better governance by isolating the system from political maneuvering and misguided short-term policies. At most, pension systems grant some modest level of discretion on benefit indexation. Chapter 7: Continuing Reforms in the Pension System 141 7.25 Preserving the real value ofpension benefits: The last re-correlation issued in late 2001 should close the cycle of policies that compress benefits through ad hoc ceilings and indexation provisions. These re-correlations create distortions in the pension structure that subsequently lead to further recalculation of benefits. It is strongly recommended, therefore, that the Government define soon a sustainable point value and indexation provision, allowing the real value of benefits outside the re-correlation to be preserved, minimizing the need for one more re- correlation plan a few years from now.99 7.26 Setting a replacement rates for pensions at a level that preserves the real value of benefits requires actions on three fronts. Two actions are on the expenditure side - bringing non-pension benefits under greater scrutiny and limiting the obligations of the pension fund toward farmers' pensions. The other action is on the revenue side, and involves raising compliance and enhancing the institutional framework for collecting payroll taxes. The details of the actions on these three fronts are as follows: > Increasing scrutiny over short term (non-pension) benefits in the pension fund. Pension fund expenditures include several non-pension benefits, such as two-year maternity leaves, and sanatorium leaves for the elderly. Vhile these account for a relatively small share of total expenditures (6 percent), together they amount to around 0.5% of the projected GDP for 2002. Including these benefits in the pension fund constitutes poor budgeting practice, since these should be covered by general taxation, bringing them under full budgetary scrutiny. The practice of having these benefits covered by the payroll tax also limits the opportunities for reducing payroll taxes, which today represent one of the main barriers for formal employment in Romania. > Explicitly financing of part of existing pensioners' debt As described above, the need for a high contribution rate is due in part to the system's past liabilities, such as those that resulted from encouraging early retirement to deal with unemployment and the decision to include farmers with no contribution record under the pension insurance fund." Explicitly acknowledging these liabilities, covering them with a separate source of revenue, such as proceeds of privatization, or an alternative tax base, is critical for stabilizing the finances of the pension system. It would allow the costs of distortionary labor and pension policies established in the 1990s to be shared more broadly within society. This in turn would spread transition costs more equitably, lowering the incentive for workers to remain in the informal sector. > Raising compliance and enhancing the institutional framework for collecting payroll taxes. Arrears from state enterprises should not be permitted, and efforts to increase collection should receive a forceful political endorsement. Also, the 9 The last re-correlation plan tries to address the inequity created by the disproportionately higher pensions issued from 1999 through early 2001, due to a problem with the calculation of the underlying reference wage. In the process, it creates new inequities that concern those that retired after April 2001 and those that retired prior to 1998 and were not included in the re-correlation plan. ' The benefits to farmers amount to around 0.6 percent of GDP annually, and are projected to decrease by 2025. 142 Chapter 7: Continuing Reforms in the Pension System Government's decision to rationalize the fragmented structure for collecting social security contributions under the NHP will bring about important economies of scale and will lower taxpayer compliance costs. These institutional changes will need to occur without disrupting revenue collection. Their positive impact will only be realized in the medium term, which implies that the process will need to receive special impetus. Consideration also needs to given to whether introducing one more social insurance program, namely occupational risk insurance, would not be counterproductive to the efforts to reduce the payroll taxes and improve payment compliance. 7.27 The slow progress attained at the NHP during the last two years reflects the complexity of institutional change. To avoid a recurrence of these problems, it is paramount that the unification of social security contributions is supported by a strong political mandate, especially since it will entail cooperation among various agencies. To this end, the Government should form a task force led by the NHP and with representation from the National Health Insurance Fund, the Unemployment Fund and the tax administration authority. The participation of the tax administration will help identify areas of common collaboration in the future (e.g. exchanging information and conducting joint audits). The task force should elaborate a plan for implementing the unification and report on progress to the Cabinet of Ministers. It is particularly important that the plan ensures adequate staffing, a major shortcoming of the current NHP structure. The task force would greatly benefit from experts who have managed similar institutional changes in other countries. Improving policy-making capacity and the institutional framework 7.28 The absence of adequate policy making capacity has been a major cause of the prevailing reliance on short-term measures. The MLSP has adopted the first steps to remedy this major shortcoming by instituting a new pension policy group. The group should receive a clear mandate to lead policy work in the pension field: evaluating the current system and elaborating reform proposals. The policy group should also assess the proposed yearly budgets for the pension fund and their mid-term reviews. 7.29 The policy group should rely on more rigorous and systematic analytical tools (inter alia actuarial modeling to simulate the longer-term effect of policies) and should construct a supporting database. Indeed, the lack of sufficient and reliable data has been a major deficiency in policy analysis and the yearly budgetary process.101 The completion of the contributors' and employers' registries at the NHP should provide the foundations for elaborating an adequate database, helping improve the system's analysis. The urgency behind the need to develop this 1o'Planning for funding future pension liabilities requires an accurate estimate of the number of existing pensioners, the number of employees, a timeline for the retirement of these employees, and the projected exit of pensioners through death. Presently, there is no up-to-date information on the distribution of current contributors and pensioners by age groups, necessary to estimate the entry and exist from the pension pool. It seems that pension payments from the previous year are used to budget for the next year, usually by applying some rule of thumb (called natural rate of increase). This method reflects neither the underlying demographic shifts, nor the probable changes in behavior resulting from new policies. Chapter 7: Continuing Reforms in the Pension System 143 database and undertake the related policy analysis is reflected in the wide differences between the budgeted and executed pension payouts and collected revenues over the years.'02 7.30 Enhancing the budgetary review process: The Ministry of Finance should also contribute to safeguarding the system's financial viability by placing proposed budgets for the pension fund under greater scrutiny. Special attention should be given to the robustness of the underlying assumptions, and the impact of policy reforms on the system's overall liabilities, not only their immediate, short-term effects. 7.31 Creating an independent National Actuary:. In the medium term, the Government should also consider the creation of a National Actuary, a small independent unit that would report to Parliament on the actuarial performance of all social security funds (pension, unemployment and health). Such an institution already operates in the United Kingdom. Poland is presently working on the establishment of a similar institution to promote good governance of social security funds and consolidate the reform of pension and health systems. A comparable institution in Romania would minimize the risk of continued discretionary management of the public pension system, and reduce the threat of further re-correlations. '02 The wedge on the pension payout may partly respond to changes in retirement behavior in anticipation of the new pension law. An appropriate database would have facilitated the analysis of potential behavioral changes.  145 ANNEX 1: DATA AND STRUCTURE OF THE COMPUTATIONAL GENERAL EQUILIBRIUM MODEL The empirical analysis of resource reallocation presented in this chapter is based on the computable general equilibrium model of the Romanian economy calibrated for 1998 data. This appendix presents the data and structure of the model, beginning with the 1998 social accounting matrix used in the analysis, moving next to the structure of the static general equilibrium model.'03 The Social Accounting Matrix for Romania The National Commission for Statistics' 1998 input-output table and sectoral accounts are the latest ones available for Romania. These data form the base for constructing the social accounting matrix (SAM) used for building this static computational general equilibrium model. In the basic version of the model we use the SAM with the input-output part aggregated to twenty-one activities to calibrate the parameters of the model. There are some peculiarities of the Romanian economy that require a great degree of caution in using the SAM constructed from these sources in general equilibrium modeling. In particular, since some sectors generate very low or negative gross operating surplus, these values cannot be used as a measure of capital input in the calibration of the production function. One can assume that the low values of gross operating surplus indicate that the sectors are subsidized implicitly by suppliers or banks, or directly by the government. Subsidization by suppliers takes a form of arrears to producers of intermediate goods. As discussed in Chapter 1, the non-payment for energy inputs is at the core of the arrears problem. Another source of implicit subsidies is the banking sector which provides subsidies if either the ailing industry/sector does not service its debt obligations, or transfers unrecoverable debt obligations to the Government owned debt recovery agency (AVAB). Finally, part of these subsidies is provided by the government in the form of tax arrears. Another peculiarity of the Romanian input-output tables is the very high rates of the value added tax imposed on some branches of the economy. A closer examination of the numbers reveals that some enterprises were registering negative value added in the industries with high VAT rates. Our approach to solve this problem is to assume that the negative value added in these industries is also financed by the some form of subsidies from the same sources as above. In the present version of the model, we work with the social accounting matrix with only minimal corrections compared to the original data. In order to calibrate the parameters of production functions, zero replaces the figures for sectors registering negative gross operating surplus. The amount of the negative gross operating surplus is assumed to be covered by government subsidies to the industry/sector. For simplicity, it is assumed that the subsidies are financed exclusively by taxes. 103 The static general equilibrium model used here is based on the modeling framework proposed by Lfgren, Harris and Robinson (2001). 146 Static General Equilibrium Model The static general equilibrium model is a system of equations explaining transactions in the social accounting matrix. The equations describe the behavior of twenty-one producers (activities), two factors of production (capital and labor), and three institutions (households, government and rest of the world). In this model producers maximize profits subject to production technology, which is identical for all sectors and based on nested CES (constant elasticity of substitution) and Leontief functions. At the top of the production technology value added and aggregate intermediate input are used in fixed proportions (Leontief function). Value added is in turn a CES function of capital and labor and aggregate intermediate input is a Leontief function of intermediate inputs. Profit maximization requires that producers do not waste inputs and employ primary factors up to the point where factors' marginal revenue products are equal to their wages. In the basic version of the model the factor market is closed by assuming that the wages are flexible and supply of labor is fixed at the current level. In other simulations we assume that wages are fixed at the observed level (equal to the average economy wide wage plus sector specific wage variations). The quantity demanded at this wage determines the level of employment, allowing for unemployment in the short run. The behavior of electric energy and gas producers is based on costs minimization, although prices are controlled by the government. The quantity produced by the electric energy sector adjusts to meet demand. The quantity of gas production is fixed at the current level, reflecting the constraints of extraction capabilities. The government automatically taxes all profits or covers all losses of the producers given the given price. Moreover, given that both prices and quantities of domestic natural gas production are fixed, and the prices of imported gas are fixed at the international level, the government sets subsidies to the imported gas in order to equilibrate domestic supply and demand. The price of crude oil is set at the international level and the quantity produced is fixed, reflecting constraints on extraction from limited oil reserves. Households in the model receive income from the factors and production and transfers from other institutions. They use this income to pay taxes, consume, save and make transfers to other institutions. The direct taxes and transfers are defined in the model as fixed shares of income. Households' savings are flexible and equilibrate the macroeconomic savings-investment balance. The remaining income is spent on consumption of commodities purchased on the market at prices which include commercial margins and commodity taxes. The choice of consumption basket is determined by the linear expenditure system (LES) demand function. The government collects taxes, modeled as ad valorem rates, and receives transfers from other institutions. The government spends its income on consumption and transfers to other institutions. The consumption is fixed in real terms and transfers are indexed to the Consumer Price Index. Profits from, or a transfer to, energy producers constitutes another source of profits/expenditures for the government. Government savings are endogenous and defined as a difference between revenues and expenditures. 147 The rest of the world pays and receives transfers to and from other institutions, fixed in foreign currency. In addition, it exports and imports commodities, as described below. The resulting difference between foreign currency spending and receipts defines the current account deficit. Commodity markets distribute domestic production between domestic consumption and exports and determines proportions of domestic production and imports in the composite commodity bundles. The domestic output is allocated between exports and domestic sales by assuming that suppliers maximize sales revenue subject to a Constant Elasticity of Transformation (CET) function describing imperfect transformability between sales in the two markets. The domestic demand in turn is met by a mixture of domestic production and imports, which are assumed to be imperfect substitutes and aggregated in the CES function (Armington function). It is also assumed that both export demand and import supply are infinitely elastic at given world prices, determining quantities of exports and imports. There are three macroeconomic balances in the model: the government balance, the external balance and the savings-investment balance. The government balance is closed by assuming that the government deficit is a flexible residual between government revenues and expenditures. For the current account balance we assume that foreign savings are fixed and the flexible exchange rate adjusts to equilibrate demand and supply for foreign currency by changing the commodity trade balance. Finally, we assume that total investment is flexible in real terms and adjust to total savings available in the economy. 148 ANNEX 2: AN ANALYTICAL FRAMEWORK FOR ESTIMATING THE SUSTAINABILITY OF FISCAL POLICY IN ROMANIA The analysis of fiscal sustainability in this note is based on the medium- to long-term fiscal projection from simple macroeconomic model'0 linking the government budget identity with economy's resource constraint and some basic assumptions about behavior of macroeconomic aggregates. We start from some basic assumptions about the GDP growth rate, investment consistent with the projected growth, inflation targets, exchange rate depreciation, and interest rates. With additional assumptions, we project main elements of the current account and general government revenues and expenditures. The general government deficit can be covered by domestic borrowing, foreign borrowing and grants, and privatization revenues. We make assumptions about the shares of domestic and foreign borrowing in the deficit financing, given the projected privatization revenues and grants. Finally, given our assumptions about the total GDP, investments and government consumption, the investment-savings gap in the model is closed by adjustments in private consumption. The foreign exchange gap is closed by private borrowing from abroad, given the projected current account deficit/surplus, public sector borrowing requirements, the increase in central bank's reserves and other elements of the capital account. We consider two macroeconomic scenarios and assess fiscal policy sustainability by evaluating the financial requirements of the budget and available sources of financing. We also conduct a sensitivity analysis based on the fiscal sustainability indicator proposed by Blanchard (1990). The starting point for the long run fiscal sustainability analysis is a general government budget identity given by: 3t +itBt-1 = ABt where fD is primary deficit of the general government, B is general government debt and i denotes interest rate. By iterating this equation forward we arrive at the following equation describing the government intertemporal budget constraint: Bt =- 0i(t,t+j)-1 Dt+j+1im lim i(t,t+T)-1 Bt+T+1 j=0 T-+oo where i(t,t+j)= Ij it+k is the discount factor. The most common definition of fiscal k=O sustainability corresponds to the 'no Ponzi game' condition: under this condition, setting the second element of the equation to zero shows that at any point the current debt must be equal to the present value of expected government surpluses. The second definition derived from the same budget identity requires the time path of the debt/GDP ratio to be bounded. The boundedness condition is stronger that the present value constraint: if the real interest rate is '1 A mini-RMSM-X model developed in ECA for Russian economy. 149 higher that the growth rate, the present value constraint may be consistent with unbounded debt/GDP ratio as long as the debt growth rate is somewhere between the real interest rate and the GDP growth rate. If the real interest rate is higher that the GDP growth rate and the debt/GDP ratio are bounded, the real interest rate is higher than the growth rate of debt and the present value condition is satisfied. Several econometric procedures were proposed to test if the above conditions hold in the data. The econometric tests, however, require long time series and are not appropriate for developing or transition countries. The simpler method, proposed by Blanchard (1990) among others, compares a current primary surplus generated by the fiscal policy with the constant surplus required to keep the GDP/debt ratio constant at the specified time horizon. The current policy is classified as sustainable if the required primary surplus is lower or equal to the current one. Yet another approach, introduced by Buiter (1993) and Anand and Wijenbergen (1992), is based on a simple accounting framework for checking the consistency of policy targets. After a consolidation of government and central bank's budget identities, and given initial values of the debt to GDP ratios and projections for the real growth rate, real exchange rate depreciation and real interest rates at home and abroad, there are five endogenous variables in the public sector budget constraint: change in domestic debt, change in foreign debt, change in reserves, seignorage revenues and primary deficit. The internal consistency of the planned fiscal and monetary policies can be checked by assigning numerical values to four of these variables and calculating the value of the fifth one from budget identity equation. 150 ANNEX 3: ASSUMPTIONS AND PROJECTIONS FOR THE ACTUARIAL MODEL FOR THE PENSIONS SYSTEM Annex Table 3.1: Macroeconomic Assumptions Base case 2001 2002 2003 2004 2005 2010 2035 2050 Real GDP 4.5% 3.5% 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% Real wage growth 15.0% 4.5% 4.0% 5.0% 4.0% 3.5% 4.3% 3.4% Pessimistic 2001 2002 2003 2004 2005 2010 2035 2050 Real GDP 4.5% 2.0% 1.5% 1.5% 1.5% 2.0% 2.0% 2.0% Real wage growth 15.0% 3.0% 1.0% 1.0% 1.0% 1.5% 3.3% 2.4% Source: World Bank estimates Annex Table 3.2: Compliance Assumptions' Base and Pessimistic Case. Salaried Employees Collection Rate --in 2001 81% -in 2010 81% - in 2030 87% --in 2050 89% 1) Figures for 2001-2010 assume an 80% recovery of past debt, which translates to an additional 4% of contribution revenue Annex Table 3.3: Projected Changes in Life Expectancies 2001 2010 2020 2030 2040 2050 Male Life Expectancy: At Birth 66.3 67.4 68.7 70.2 72.0 72.5 At Age 20 48.8 49.7 50.9 52.3 53.8 54.4 At Age 60 15.8 16.6 17.5 18.5 19.8 20.2 At Age 65 12.9 13.5 14.4 15.4 16.6 17.0 Female Life Expectancy: At Birth 73.8 74.7 75.7 76.8 78.1 78.6 At Age 20 55.9 56.6 57.6 58.6 59.7 60.2 At Age 60 19.3 19.9 20.7 21.6 22.6 23.1 At Age 65 15.4 16.0 16.8 17.6 18.6 19.1 Source: World Bank estimates Annex Table 3.4: Projected Fertility Rates 2001 1 2010 1 2020 230 2040 2050 Total Ferility Rate 141.8% 159.9% 179.9% 200.0% 200.0% 200.0 Source: World Bank estimates References Blanchard, Olivier, J. Chouraqui, R. Hagemann, and N. Sartor, 1991, The Sustainability of Fiscal Policy: New Answers to an Old Question, NBER Working Paper No. R1547, March. Cavalcanti, Carlos and Wojtek Malizewski, 2001, Transformation, restructuring and macroeconomic policy in Romania, mimeo, August. Conway, Francis, Juliana Pigey, Clare Romanik, Mihai Nedelcu, 2001, Romania: Local Expenditure Review, processed, Urban Institute, Washington DC. Fetter, Theodore, 2002, Fiscal operations of the judiciary in Romania, mimeo, March. Kehoe, T.J. and J. Serra-Puche, 1991, A General Equilibrium Appraisal of Energy Policy in Mexico, Empirical Economics 16, 71-93. L6fgren, H., R. L. Harris and S. Robinson, 2001, A Standard Computable General Equilibrium (CGE) Model in GAMS, Trade and Macroeconomics Division Discussion Paper No. 75, International Food Policy Research Institute, Washington DC. Pauna, Catalin, 2002, Romania: Review of Public Expenditure on the Legal and Judiciary Systems, mimeo, World Bank, February. Pop, Luana, 2002, The Impact of the Romanian Minimum Income Guarantee Law in the framework of the state budget project for 2002, mimeo, January. Radulescu, Silviu, 2002, The impact of the introduction of insurance-based financing on the performance of the Romanian health sector, mimeo, March. Schaffer, M and Gerard Turley (2000). Effective versus Statutory Taxation: Measuring Effective Tax Administration in Transition Economies, William Davidson Institute Working Papers Series 347, William Davidson Institute at the University of Michigan Business School, November. Tesliuc, Cornelia, 2001, Romania: distributional and poverty alleviation impact of social and fiscal programs, mimeo, July. Tesliuc, Cornelia, 2002, Romania: Has the 1998 health sector reform achieved its efficiency and equity objectives?, mimeo, February. World Bank, 2002, Romania: Local Social Service Delivery Study, Report No. 23492-RO, January 2002, Washington, DC. STATISTICAL APPENDIX Statistical Appendix Population, Vital and Social Statistics S1 Population by gender S2 Population by Area S3 Vital Statistics S4 Yearly Average Consumption per Capita for Main Food Products S5 Population Endowment with Durable Goods S6 Daily Average food consumption, in calories Employment and Unemployment S7 Employment by Activities of the National Economy (end of year) S8 Average Number of Employees by Activities of the National Economy S9 Registered Unemployed and Unemployment Rate (end of year) S1O Employment, by Type of Ownership, by Jan 1, 2000 National accounts Sli Real GDP Growth and Structure, 1998 - 2001 S12 Gross Domestic Product, in billion ROL S13 Indices of Gross Domestic product, 1990 = 100 S14 GDP by Region, in Year 1998 Real Sector Statistics S15 Industrial Production Indices, by Activities S16 Structure of the Industrial Production by Activities (%) S17 Output of Main Agricultural Products S18 Average Net Nominal Monthly Salary Earning, by Activities of the National Economy S19 Exports (FOB) and Imports (CIF) by Main Sections of the Combined Nomenclature S20 Geographical Orientation of the Foreign Trade S21 Consumer Price Index , per Month, Oct. 1990 = 100 S22 Consumer Price Index , per Month, previous Month = 100 S23 Consumer Price Index (yearly evolution) S24 Industrial Production Price Index per Month S25 Indices of Investment by Activities S26 Investment by Sector S27 Employment, Unemployment, Average Monthly Net Nominal Salary Earnings, by Region, in 2000 S28 Energy Balance S29 Production, Domestic Consumption, Export and Import of Oil products (1985-1999) S30 Evolution of the Privatization Process S31 Private Sector Share of GDP General Government Consolidated Revenues and Expenditure S32 (IMF data) Monetary Statistics, BOP, External Debt S33 Monetary Developments S34 Interest Rates on ROL Government Securities (new and rolled over issues) S35 Interest Rates to the Non-bank Sector, 1991 - 2000 S36 Loans Classification Credits and Risk provisions S37 Interbank Foreign Exchange Market S38 Balance of Payments (in million US$, NBR format, non-convertible currencies included) S39 External Debt by Creditor S40 Reserve Assets and External Debt by Debtor  SI: Population by Gender million inhabitants as % of total: Total Male Female Male Female July 1, 1985 22.725 11.214 11.511 49.3 50.7 July 1, 1986 22.823 11.261 11.562 49.3 50.7 July 1, 1987 22.940 11.319 11.621 49.3 50.7 July 1, 1988 23.054 11.375 11.679 49.3 50.7 July 1, 1989 23.152 11.422 11.729 49.3 50.7 July 1, 1990 23.207 11.449 11.758 49.3 50.7 July 1, 1991 23.185 11.435 11.750 49.3 50.7 January 7, 1992 22.810 11.214 11.596 49.2 50.8 July 1, 1992 22.789 11.201 11.588 49.1 50.9 July 1, 1993 22.755 11.176 11.579 49.1 50.9 July 1, 1994 22.731 11.157 11.574 49.1 50.9 July 1, 1995 22.681 11.124 11.557 49.0 51.0 July 1, 1996 22.608 11.081 11.527 49.0 51.0 July 1, 1997 22.546 11.041 11.505 49.0 51.0 July 1, 1998 22.503 11.012 11.491 48.9 51.1 July 1, 1999 22.458 10.985 11.473 48.9 51.1 July 1, 2000 22.435 10.969 11.466 48.9 51.1 1) Census Source: Statistical Yearbook of Romania 2001 and NIS S2: Population by Area million inhabitants as % of total Total Urban Rural Urban Rural July 1, 1985 22.725 11.370 11.355 50.0 50.0 July 1, 1986 22.823 11.540 11.283 50.6 49.4 July 1, 1987 22.940 11.771 11.170 51.3 48.7 July 1, 1988 23.054 11.962 11.092 51.9 48.1 July 1, 1989 23.152 12.312 10.840 53.2 46.8 July 1, 1990 23.207 12.609 10.598 54.3 45.7 July 1, 1991 23.185 12.552 10.633 54.1 45.9 January 7, 1992 22.810 12.392 10.418 54.3 45.7 July 1, 1992 22.789 12.367 10.422 54.3 45.7 July 1, 1993 22.755 12.406 10.349 54.5 45.5 July 1, 1994 22.731 12.428 10.303 54.7 45.3 July 1, 1995 22.681 12.457 10.224 54.9 45.1 July 1, 1996 22.608 12.411 10.196 54.9 45.1 July 1, 1997 22.546 12.405 10.141 55.0 45.0 July 1, 1998 22.503 12.348 10.155 54.9 45.1 July 1, 1999 22.458 12.303 10.155 54.8 45.2 July 1, 2000 22.435 12.244598 10.191 54.6 45.4 1) Census Source: Statistical Yearbook ofRomania 20001  S3: Vital Statistics Absolute number Rates (per 1000 inhabitants) Late fetal deaths per Infant Natural Infant Natural 1000 births deaths per Increase Deaths Increase (live births 1000 liv and fetal births deaths) 1985 358797 246670 112127 9191 15.8 10.9 4.9 7.8 25.6 1986 376896 242330 134566 8746 16.5 10.6 5.9 8.0 23.2 1987 383199 254286 128913 11077 16.7 11.1 5.6 7.9 28.9 1988 380043 253370 126673 9643 16.5 11.0 5.5 7.6 25.4 1989 369544 247306 122238 9940 16.0 10.7 5.3 7.6 26.9 1990 314746 247086 67660 8471 13.6 10.6 3.0 7.0 26.9 1991 275275 251760 23515 6258 11.9 10.9 1.0 6.9 22.7 1992 260393 263855 -3462 6080 11.4 11.6 -0.2 6.5 23.3 1993 249994 263323 -13329 5822 11.0 11.6 -0.6 6.3 23.3 1994 246736 266101 -19365 5894 10.9 11.7 -0.8 6.5 23.9 1995 236640 271672 -35032 5027 10.4 12.0 -1.6 6.2 21.2 1996 231348 286158 -54810 5158 10.2 12.7 -2.5 6.1 22.3 1997 236891 279315 -42424 5209 10.5 12.4 -1.9 6.2 22.0 1998 237297 269166 -31869 4868 10.5 12.0 -1.5 6.3 20.5 1999 234600 265194 -30594 4360 10.4 11.8 -1.4 6.2 18.6 2000 234521 255820 -21299 4370 10.5 11.4 -1.9 6.1 18.6 Source: Statistical Yearbook ofRomania 2001  S4: Yearly Average Consumption per Capita for Main Food Products 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Meat and meat products (equivalent fresh meat), Kg 54.4 45.7 47.7 45.5 47.8 47.2 45.1 48.0 45.2 45.2 Fats - total (gross weight), Kg 16.5 14.4 14.1 12.9 14.2 14.3 13 13.4 14.5 na Milk and dairy products in equivalent milk (butter excluded), liters 163.3 163.7 176.9 179.5 188.6 192.7 192.4 194.4 194.0 193.0 Eggs, pcs. 241 196 190 194 197 199 186 201 206 208 Sugar and confectioneries (equivalent sugar), Kg 26.4 24.4 23.7 24.5 23.5 24.8 19.9 20.5 20.9 23.0 Cereal products (equivalent flour), Kg 145.3 146.5 159.6 158.6 162.4 160.6 169.8 166.7 166.3 165.8 Potatoes, Kg 48 60 73.9 66.7 71 73.4 81.7 84.1 86.1 86.5 Vegetables and vegetable products (equivalent vegetables), Kg 88.5 100.7 112.7 110.3 115.6 115.4 111.6 120.5 130 na Fruit and fruit products (equivalent fruit), Kg 45.7 47.1 64.3 47.8 45.8 50.5 44.5 45.8 43.4 44.5 Source: National Institute for Statistics  S5: Population Endowment with Durable Goods End of Year, pcs./1000 Inhabitants 1991 1992 1993 1994 1995 1996 1997 1998 1999 1999 Radio sets 298.9 302.9 304.0 304.5 306.0 318.4 326.9 356.1 358.0 362.6 TV sets 195.9 200.4 201.6 201.8 224.7 231.6 233.7 247.8 259.4 270.8 Refrigerators 173.3 175.3 176.6 177.6 196.4 201.4 204.3 211.3 212.7 216.2 Gas stoves 141.2 139.4 141.7 146.0 160.6 170.2 174.3 182.6 185.3 189.1 Electric washing machines 150.8 149.5 148.5 153.3 147.7 150.1 149.6 151.9 150.5 152.7 Vacuum cleaners 79.1 811.0 84.8 85.1 86.8 91.3 95.2 98.4 97.7 100.4 Cars 61.4 67.9 76.2 85.5 92.8 101.0 109.9 118.8 125.5 131.5 Source: National Institute for Statistics S6: Daily Average Food Consumption, in Calories and Nutrients per Capita 1991 1992 1993 '1994 1995 1996 1997 1998 1999 2000 Calories, number 2846 2771 2972 2886 2933 2953 2933 2959 2981 3020 of which: Animal-products, number 736 684 712 703 717 721 698 716 700 691 Proteins, g 92.9 90.1 96.3 95.1 95.9 94.7 95.8 97.5 96.6 94.7 of which: Animal products 44.6 40.1 41.5 41.2 43.2 43.5 42.2 44.1 42.9 42.0 Fats, g 89.4 80.8 83.1 78.7 79.3 80.5 76.1 78.4 80.6 85.1 of which: Animal products '50.3 46.9 49.2 47.7 47.0 47.1 45.2 46.0 45.1 44.4 Carbohydrate, g 399.4 403.3 440.8 431.8 441.4 444.2 448 446.6 448.7 449.7 Note: Sarting 1990 data were recalculated in accordance with EUROSTAT and FAQ suggestions Source: National Institute for Statistics  S7: Employment ' by Activities of the National Economy (end of year) thousands of persons 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total 10786 10458 10062 10011 9493 9379 9023 8813 8420 8629 Agriculture 3116 3362 3537 3561 3187 3249 3322 3296 3419 3523 Sylviculture, forestry and hunting 89 81 77 86 78 71 62 53 47 47 Industry 3803 3301 3030 2882 2714 2741 2450 2317 2054 2004 Mining and quarrying 277 272 259 256 250 250 184 169 146 140 Manufacturing 3372 2865 2606 2456 2293 2302 2079 1964 1734 1691 Electric and thermal energy, gas and water 154 164 165 170 171 189 187 184 174 173 Construction 501 579 574 563 479 475 439 391 338 353 Trade 699 754 585 636 865 772 802 835 756 776 Hotels and restaurants 213 175 131 136 123 116 130 98 100 93 Transport and storage 594 556 497 462 458 448 405 361 310 326 Post and telecommunications 96 93 95 94 98 99 100 100 95 93 Financial, banking and insurance activities 44 57 66 59 71 71 73 76 69 74 Real estate and other services 421 441 417 438 324 257 199 243 238 271 Public administration 99 113 117 125 131 125 130 134 141 147 Education 426 432 432 437 437 441 426 426 429 421 Health and social assistance 311 306 308 333 333 337 315 317 277 341 Other activities of the national economy 374 208 196 199 195 177 170 166 147 160 1) Employment: includes, in accordance with NIS methodology used for the labor force balance, all persons who, during the reference year, carried out a socio-economic profitable activity, excepting military staff and similar, political and community organizations employees, and convicts. Source: Statistical Yearbook of Romania 1997 and 2001. Employment' by Activities of the National Economy (end of year) as % of total 1991 1992 1993 . 1994 1995 1996 1997 1998 1999 2000 Total 100 100 100 100 100 100 100 100 100 100 Agriculture 28.9 32.1 35.2 35.6 33.6 34.6 36.8 37.4 40.6 40.8 Sylviculture, forestry and hunting 0.8 0.8 0.8 0.9 0.8 0.8 0.7 0.6 0.6 0.5 Industry 35.3 31.6 30.1 28.8 28.6 29.2 27.2 26.3 24.4 23.2 Mining and quarrying 2.6 2.6 2.6 2.6 2.6 2.7 2.0 1.9 1.7 1.6 Manufacturing 31.3 27.4 25.9 24.5 24.2 24.5 23.0 22.3 20.6 19.6 Electric and thermal energy, gas and water 1.4 1.6 1.6 1.7 1.8 2.0 2.1 2.1 2.1 2.0 Construction 4.6 5.5 5.7 5.6 5.0 5.1 4.9 4.4 4.0 4.1 Trade 6.5 7.2 5.8 6.4 9.1 8.2 8.9 9.5 9.0 9.0 Hotels and restaurants 2.0 1.7 1.3 1.4 1.3 1.2 1.4 1.1 1.2 1.1 Transport and storage 5.5 5.3 4.9 4.6 4.8 . 4.8 4.5 4.1 3.7 3.8 Post and telecommunications 0.9 0.9 0.9 0.9 1.0 1.1 1.1 1.1 1.1 1.1 Financial, banking and insurance activities 0.4 0.5 0.7 0.6 0.7 0.8 0.8 0.9 0.8 0.9 Real estate and other services 3.9 4.2 4.1 4.4 3.4 2.7 2.2 2.8 2.8 3.1 Public administration 0.9 1.1 1.2 1.2 1.4 1.3 1.4 1.5 1.7 1.7 Education 3.9 4.1 4.3 4.4 4.6 4.7 4.7 4.8 5.1 4.9 Health and social assistance 2.9 2.9 3.1 3.3 3.5 3.6 3.5 3.6 3.3 4.0 Other activities of the national economy 3.5 2.0 1.9 2.0 2.1 1.9 1.9 1.9 1.7 1.9  S8: Average Number of EmployeesI by Activities of the National Economy thousand persons 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total 7574 6888 6672 6438 6160 5939 5597 5369 4761 4623 Agriculture 609 561 560 484 420 364 283 250 187 146 Sylviculture, forestry and hunting 99 93 88 91 83 78 69 61 53 50 Industry 2 3643 3245 3017 2856 2615 2586 2443 2272 1991 1873 Construction 486 459 536 515 443 431 387 378 309 316 Trade 496 471 468 495 660 587 613 654 579 568 Hotels and restaurants 172 122 112 124 115 109 118 94 91 84 Transport and storage 580 514 478 448 420 398 365 340 279 277 Post and telecommunications 96 93 94 93 96 97 97 100 95 93 Financial, banking and insurance activities 41 51 62 56 67 69 74 76 69 71 Real estate and other services 297 257 232 233 195 185 144 164 167 177 Public administration 93 106 110 123 130 125 130 134 141 148 Education 440 428 427 428 427 432 423 414 415 407 Health and social assistance 315 304 306 329 328 335 321 311 283 305 Other activities of the national economy 207 184 182 163 161 143 130 121 102 108 1) Employment: all persons aged 15 and over who carried out an economic or social activity producing goods or services for at least on hours during the reference period, in order to get income as salaries, payment in kind or other benefits. Employee: person who carries out his activity based on an employment contract into an economic or social unit- no matter of its ownership type - or for private persons, for a remuneration as salary, in cash or in kind, as commission, a.s.o. 2) Includes electric and thermal energy, gas and water Source: The Statistical Yearbook of Romania 1997 and 2001 Average Number of Employees1 by Activities of the National Economy as % of total number of employees 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total 100 100 100 100 100 100 100 100 100 100 Agriculture 8.0 8.1 8.4 7.5 6.8 6.1 5.1 4.7 3.9 3.2 Sylviculture, forestry and hunting 1.3 1.4 1.3 1.4 1.3 1.3 1.2 1.1 1.1 1.1 Industry 48.1 47.1 45.2 44.4 42.5 43.5 43.6 42.3 41.8 40.5 Construction 6.4 6.7 8.0 8.0 7.2 7.3 6.9 7.0 6.5 6.8 Trade 6.5 6.8 7.0 7.7 10.7 9.9 11.0 12.2 12.2 12.3 Hotels and restaurants 2.3 1.8 1.7 1.9 1.9 1.8 2.1 1.8 1.9 1.8 Transport and storage 7.7 7.5 7.2 7.0 6.8 6.7 6.5 6.3 5.9 6.0 Post and telecommunications 1.3 1.4 1.4 1.4 1.6 1.6 1.7 1.9 2.0 2.0 Financial, banking and insurance activities 0.5 0.7 0.9 0.9 1.1 1.2 1.3 1.4 1.4 1.5 Real estate and other services 3.9 3.7 3.5 3.6 3.2 3.1 2.6 3.1 3.5 3.8 Public administration 1.2 1.5 1.6 1.9 2.1 2.1 2.3 2.5 3.0 3.2 Education 5.8 6.2 6.4 6.6 6.9 7.3 7.6 7.7 8.7 8.8 Health and social assistance 4.2 4.4 4.6 5.1 5.3 5.6 5.7 5.8 5.9 6.6 Other activities of the national economy 2.7 2.7 2.7 2.5 2.6 2.4 2.3 2.3 2.1 2.3 1) Employment: all persons aged 15 and over who carried out an economic or social activity producing goods or services for at least on  S9: Registered Unemployed and Unemployment Rate (end of year) number of persons Total Unemployment 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total 337440 929019 1164705 1223925 998432 657564 881435 1025056 1130296 1007131 826932 Unemployment rate' (%) 3 8.2 10.4 10.9 9.5 6.6 8.9 10.4 11.8 10.5 8.6 Recipients of unemployment benefits and of allowance for vocational integration 265978 602957 549785 -504284 317142 202233 438044 402980 386517 .307065 316413 Workers 223776 526305 470324 413951 247490 152655 327926 287194 267121 204674 na High school graduates 34198 65468 70260 79781 62429 44188 96696 102573 104421 86806 na University graduates 8004 11184 9201 10552 7223 5390 13422 13213 14975 15585 na Recipients of support allowance 239642 516059 564066 457079 259457 217959 390038 445992 391932 286214 Workers 206850 441595 477595 380754 214973 169935 304979 341116 289742 na High school graduates 28268 66491 79903 72014 42475 45764 78239 96844 94048 na University graduates 4524 7973 6568 4311 2009 2260 6820 8032 8142 na Beneficiaries of Compensation Payment (according to Government ordinance 98/1999) 39042 52914 5063 Workers .. .. .. 27687 38168 na High school graduates .. .. 8877 11882 na University graduates .. .. .. 2478 2864 na Unpaid unemployed 71462 86420 98861 155575 224211 195874 225432 232038 258745 255220 219242 1) Registered Unemployment Rate: ratio between the number of registered unemployed and active civil population (unemployed + employment) Note: Support allowance is paid following expiry of a 270 days period. Source: National Agency for Employment and Professional Training of which: women 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total 208457 563065 685496 693342 551492 355435 428620 485181 530119 471608 Unemployment rate' (%) 4 10.3 12,9 12,9 11,4 7,5 9,3 10,4 11.6 10.1 Recipients of unemployment benefits and of allowance for vocational integration 158145 364032 313981 280726 174345 112547 200336 182865 173609 143042 Workers 130590 313589 265599 228710 134164 81978 135967 115453 104454 81439 High school graduates 23398 44208 43494 46385 36099 27608 57402 60689 61345 53003 University graduates 4082 2961 6967 6723 7810 8600 Recipients of support allowance 142152 309190 318834 247820 134925 108240 185015 209953 182138 Workers 120658 260545 264189 201797 106579 79068 134101 145832 119657 High school graduates 19051 44639 51142 43292 27190 27922 47264 59865 57934 University graduates 2443 4006 3503 2731 1156 1250 3650 4256 4547 Beneficiaries of Compensation Payment (according to Government ordinance 98/1999) 16464 23172 Workers .. 11097 15667 High school graduates .. .. .. 4383 6243 University graduates .. .. 984 1262 Unpaid unemployed 50312 56881 62325 93782 129327 107963 120044 117301 130093 123256 1) Registered Unemployment Rate: ratio between the number of registered unemployed and active population (unemployed + employment) Source: National Agency for Employment and Professional Training  S10: Employment, by Type of Ownership, End 2000 thousand of persons Total, of which: Majority: Total o.w: gender Public Private Total 8629 4181 2154 6475 of which: women 4181 1016 3165 Agriculture .3523 1829 60 3463 Sylviculture, forestry and hunting 47 5 29 18 Industry 2004 871 671 1333 Mining and quarrying 140 20 130 10 Manufacturing 1691 808 373 1318 Electric and thermal energy, 173 43 168 5 gas and water Construction 353 46 53 300 Trade 776 394 35 741 Hotels and restaurants 93 58 11 82 Transport and storage 326 60 190 136 Post and telecommunications 93 50 78 15 Financial, banking and insurance activities 74 53 42 32 Real estate and other services 271 107 83 188 Public administration 147 86 147 - Education 421 284 402 19 Health and social assistance 341 265 300 41 Other activities of the national economy 160 73 53 107 percent of total employment' Total, of which: Majority: Total o.w: gender Public Private Total 100 48 25.0 75.0 of which: women 48 Agriculture 41 21 0.7 40.1 Sylviculture, forestry and hunting 1 0 0.3 0.2 Industry 23. 10 7.8 15.4 Mining and quarrying 2 0 1.5 0.1 Manufacturing 20 9 4.3 15.3 Electric and thermal energy, 2 0 1.9 0.1 Gas and water Construction 4 1 0.6 3.5 Trade 9 5 0.4 8.6 Hotels and restaurants 1 1 0.1 1.0 Transport and storage 4 1 2.2 1.6 Post and telecommunications 1 1 0.9 0.2 Financial, banking and insurance activities 1 1 0.5 0.4 Real estate and other services 3 1 1.0 2.2 Public administration 2 1 1.7 - Education 5 3 4.7 0.2 Health and social assistance 4 3 3.5 0.5 Other activities of the national economy 2 1 0.6 1.2 1) Employment: includes, in accordance with NIS methodology used for the labor force balance, all persons who, during the reference year, carried out a socio-economic profitable activity, excepting military staff and similar, political and community organizations employees, and convicts. Source: Statistical Yearbook of Romania 2001  Si1: Real GDP Growth and Structure, 1998 - 2001 (provisional data, ESA 1995 methodology) 1998 1999 2000 2001 1998 1999 2000 2001 1999 2000 2001 GDP Structure - Supply side nominal bn ROL volume indices % change over previous year Agriculture' 53773 72805 88537 154311 100.0 103.3 81.8 121.2 3.3 -18.2 21.2 Industry 98213 135344 201953 297246 100.0 98.5 106.2 107.9 -1.5 6.2 7.9 Construction 19029 27376 39822 57977 100.0 97.7 106.3 105.2 -2.3 6.3 5.2 Services 166097 253356 387110 535449 100.0 99.4 103.8 101.7 -0.6 3.8 1.7 FISIM2 -5565 -8407 -8579 -16983 100.0 107.9. 74.8 146.3 7.9 -25.2 46.3 GVA 331548 480475 708842 1027999 100.0 99.5 101.8 105.5 -0.5 1.8 5.5 Taxes, net 42251 65255 91466 126127 100.0 93.7 101.5 103.1 -6.3 1.5 3.1 GDP 373798 545730 800308 1154126 100.0 98.8 101.8 105.3 -1.2 1.8 5.3 as% of GDP as% of GVA Agriculture' 14.4 13.3 11.1 13.4 16.2 15.2 12.5 15.0 Industry 26.3 24.8 25.2 25.8 29.6 28.2 28.5 28.9 Construction 5.1 5.0 5.0 5.0 5.7 5.7 5.6 5.6 Services 44.4 46.4 48.4 46.4 50.1 52.7 54.6 52.1 IFSID -1.5 -1.5 -1.1 -1.5 -1.7 -1.7 -1.2 -1.7 GVA 88.7 88.0 88.6 89.1 Taxes, net 11.3 12.0 11.4 10.9 12.7 13.6 12.9 12.3 1998 1999 2000 2001 1998 1999 2000 2001 1999 2000 GDP Structure - Demand side nominal bn ROL volume indices % change over previous year Final consumption, o.w: 337469 484361 687879 995018 100.0 97.5 101.1 105.7 -2.5 1.1 5.7 -households3 310923 453308 632378 921805 100.0 98.9 100.1 106.4 -1.1 0.1 6.4 -public administration' 26546 31054 55501 73213 100.0 80.9 115.9 98.1 -19.1 15.9 -1.9 Gross Capital Formation, o.w: 66334 87741 157680 252650 100.0 - - - - - - -gross fixed capital formation 67920 96630 151486 219289 100.0 95.2 104.6 106.6 -4.8 4.6 6.6 Exports goods and services 84559 152903 264187 386832 100.0 110.5 124.1 110.6 10.5 24.1 10.6 Imports goods and services 114563 179275 309438 480373 100.0 98.5 128.2 117.5 -1.5 28.2 17.5 Net Export -30004 -26373 -45251 -93542 as % of GDP Final consumption, o.w: 90.3 88.8 86.0 86.2 -households' 83.2 83.1 79.0 79.9 -public administration' 7.1 5.7 6.9 6.3 Gross Capital Formation, o.w: 17.7 16.1 19.7 21.9 -gross fixed capital formation 18.2 17.7 18.9 19.0 Exports 22.6 28.0 33.0 33.5 Imports 30.6 32.9 38.7 41.6, 1) Includes forestry and fishing; 2) Financial Intermediation Services, Indirectly Measured 3) actual final consumption Source: Monthly Bulletin of Statistics Nol/January 2002, page 2  S12: Gross Domestic Product, In Billion ROL 1991 1992 1993 1994 1995 1996 1997 1998 1998 1999 2000 2001 ESA 1979 methodology ESA 1995 methodology Industry' 835 2311 6781 18018 23711 36182 78094 103054 98213 135344 201953 297246 Agriculture 404 1120 4124 9654 13941 20460 44583 52212 52212 70456 85075 154311 s Forestry, fishing, hunting 12 28 82 244 328 490 950 1552 1561 2349 3462 57977 Construction 96 290 1040 3251 4755 7067 13230 18731 19029 27377 39822 na Transport 125 457 1759 3712 4568 7659 17271 22891 23862 33983 51359 na Post and Telecommunication 22 57 255 642 1009 2146 5395 11581 11665 20307 32453 na Trade, hotels and restaurants 297 860 2058 4076 7570 12722 28768 49494 50403 74599 115154 na Real estate and other services 86 261 840 2158 3453 5871 25860 42971 42454 69557 108685 na General government 68 203 620 1611 2692 3355 6763 13222 12834 18275 31405 na Education 61 160 492 1260 1854 2777 5153 9975 10016 16343 21628 na Health and Social Assistance 50 127 362 944 1305 2106 3709 7917 8240 10793 14111 na Financial, banking and insurance activities 58 322 1015 2376 3588 3243 4299 6529 6623 9500 12314 na Gross Value Added 2066 5915 18579 45955 66599 101854 232818 334606 331548 480475 708842 1027999 Taxes by product (including VAT) 193 485 1858 3848 5579 7459 18569 32952 39807 61606 87430 na Customs Duties 25 87 304 649 1189 1852 3808 5741 5859 7243 8702 na Subsidies by product -80 -458 -705 -679 -1231 -2246 -2269 -2106 -3415 -3593 -4667 na GDP 2204 6029 20036 49773 72136 108920 252926 371194 373798 545730 800308 1154126 Final consumption 1673 4643 15236 38452 58662 89939 218620 334672 337469 484361 687879 995018 a. Households 1324 3751 12670 31442 48545 75289 186238 278626 - - - - b. General Government 334 861 2473 6852 9877 14274 31000 52741 - - - c. Private Non-Profit Sector is 31 92 159 240 377 1382 3305 - - - - Final consumption' 337469 484361 687879 995018 Households actual individual final consumption 310923 453308 632378 921805 Governments actual individual final consumption 26546 31054 55501 73213 Gross fixed capital formation 317 1157 3584 10096 15425 24999 53540 68112 67920 96630 151486 219289 Change in stocks 301 737 2212 2253 2085 3161 -1369 -1586 -1586 -8890 6194 na Exports of Goods and Services 388 1676 4612 12394 19921 30651 73796 87105 84559 152903 264187 386832 Imports of Goods and Services 475 2182 5608 13422 23958 39831 91661 117108 114563 179275 309438 480373 GDP per capita, based on PPP' 5087 5068 5259 5628 6095 6595 6422 6153 6000 na na Gross Value Added Rate (GVA / Output, %) 37 39 43 46 44 42 43 46 46 na na na Investment rate ( GFCF 4' GVA,%) 15 20 19 22 23 25 23 20 20 20 21 21 1) InIcudes electric and thennal Energy, gas and water 2) Apply to years 1993 to 1998 (ESA 1979 methodology) 3) Apply to revized serie 1998-2001 (ESA 1995 methodology) 4) GFCF: Gross Fix Capital Formation 5) Incluedes foresty, fishing, hunting; 6) Human Development Report. UNDP. 2000  S13: Indices of Gross Domestic Product, 1990 = 100 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 1990=100 1998= 100 GDP' 87.1 79.4 80.6 83.8 89.8 93.4 87.7 83.5 98.8 100.6 Final consumption' 88.2 83.3 84.3 87.5 97.0 103.7 99.3 100.4 a. Households 83.8 77.5 78.2 80.2 90.7 98.0 94.4 94.9 - - b. General Government 110.6 113.0 116.0 128.8 130.1 132.0 120.9 123.0 - - Final consumption 97.5 98.6 Households actual individual final consumption - - - - - - - - 98.9 99.0 Government's actual individual final consumption - - - - - - - - 80.9 93.7 Gross fixed capital formation 68.4 75.9 82.2 99.2 106.1 112.1 114.0 107.5 95.2 99.6 Change in stocks Exports ofGoods and Services 82.1 102.9 111.1 119.0 117.0 102.0 111.4 98.3 Imports of Goods and Services 70.4 107.5 104.4 102.8 116.3 108.7 107.5 111.3 Industry4 87.2 75.2 76.0 78.6 83.0 88.7 81.6 77.3 98.5 104.6 Agriculture 88.1 76.5 87.4 89.9 94.2 90.2 89.1 79.6 103.3 83.8 Forestry, fishing, hunting 73.5 69.2 63.3 64.9 66.9 66.9 63.1 63.0 100.8 103.1 Construction 80.6 76.1 94.9 120.8 128.9 129.9 104.9 100.2 97.7 103.9 Transport and storage 83.2 78.1 78.5 78.1 78.7 79.3 70.0 60.4 95.1 98.1 Post and Telecommunication 138.3 149.7 181.3 199.3 211.9 279.2 274.7 279.7 100.1 101.4 Trade, hotels and restaurants 74.4 67.6 61.5 60.9 74.2 83.4 74.4 77.5 100.4 105.2 Financial, banking and insurance activities 99.7 139.3 140.0 142.3 144.3 123.7 95.8 97.6 100.6 93.4 Real estate and other services 98.2 117.0 114.7 122.4 136.0 156.3 145.6 145.3 103.8 112.0 General government 92.0 101.9 103.9 113.9 119.0 110.5 106.9 103.3 95.9 102.5 Education 116.5 115.6 119.6 120.0 124.2 136.6 123.8 123.3 103.7 97.8 Health and Social Assistance 98.0 95.2 97.3 106.2 109.1 119.4 105.2 121.7 81.0 78.2 Gross Value Added 88.2 80.3 83.0 86.5 92.4 96.0 89.0 84.2 99.5 101.3 Taxes by product (including VAT) 77.1 69.0 66.0 62.1 68.0 71.6 67.6 61.9 94.2 95.9 Customs Duties 102.8 109.8 102.8 117.7 129.7 138.8 137.8 165.0 91.1 85.1 Subsidies by product 84.6 73.3 85.4 77.6 77.1 85.4 38.6 17.9 94.3 86.2 GDP per capita 87.2 80.9 82.3 85.6 91.9 95.8 90.2 86.1 99.0 100.9 1) Apply to years 1993 to 1998 (ESA 1979 methodology) 2) Apply to revized serie 1998-2001 (ESA 1995 methodology) 3) GFCF: Gross Fix Capital Formation 4) Includes Electric and thermal energy, gas and water  S14: GDP by Region , in Year 1998 billions lei, current prices Total North-East South-East South South-West West North-West Center Bucharest Other Regional Gross Value Added 331548 44779 43271 43929 32246 30590 40281 41380 54624 450 of which: Agriculture 52212 9381 7422 9086 6587 5783 6973 6167 814 - Sylviculture,forestry,fishery 1561 368 123 171 148 221 210 306 13 - Industry1 98213 13576 12996 14084 10423 6589 11630 14629 13925 362 Construction 19029 2192 2891 2245 1928 1801 1964 2164 3845 - Trade2 50403 5770 6476 5134 3815 4634 5267 6185 13121 Transport and storage 23162 2122 3126 3084 2122 2625 3174 2876 4034 Post and telecommunications 11665 1091 1140 977 690 906 1107 1123 4631 Financial, banking and insurance activities 6623 644 647 665 700 484 679 650 2155 Real estate and other services 42454 4577 5511 4903 3112 5130 5584 4265 9372 - Generalgovernment 12834 2152 1667 1993 1373 1176 1630 1516 1233 95 Education 10016 1676 1062 1227 1073 886 1569 1229 1295 - Health and social assistance 8240 1283 937 1099 815 868 1170 966 1103 - Regional GDP 373798 50385 48959 49675 36102 34378 45320 46683 61785 511 Regional GDP per capita (ROL, NIS data) 16611185 13252775 16635893 14245353 14951802 16761724 15859722 17584213 26945539 - Regional GDP percapita(US$)3 1872 1493 1874 1605 1685 1889 1787 1981 3036 Exchange Rate ROIUS$, ave 1998 8876 1) Including electric and thennal energy, gas and water 2) Including hotels and restaurants 3) Calculated using the ROL data of the NIS and the average annual exchange rate in 1998 Source: Statistical Yearbook ofRomania 2001. page 724 - 725  S15: Industrial Production Indices, by Activities (1990=100) (1995-100) (1998=100) 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 1999 2000 INDUSTRY 77.2 60.3 61.1 63.1 69.0 75.8 1063 98.6 85.0 83.1 89.4 97.6 104.5 Mining and quarrying 81.7 81.0 79.9 81.3 80.8 80.2 101.3 94.9 80.7 75.3 79.7 92.6 97.2 Coal mining and preparation 83.4 95.6 96.8 99.5 99.7 102.4 105.7 85.7 65.0 56.8 69.0 84.6 104.0 Petroleum and gas extraction 85.4 82.1 82.4 80.8 80.2 78.3 100.1 98.1 91.2 87.4 88.0 96.0 96.1 Mettaliferous ores quarrying and preparation 77.2 72.1 72.7 83.0 80.3 76.0 97.4 87.5 77.3 59.4 57.8 92.0 91.0 Other extractive activities 74.5 49.5 44.8 40.9 42.2 41.4 91.5 122.0 87.0 86.3 88.5 87.0 91.1 Manufacturing industry 76.2 57.9 58.3 60.5 67.8 76.3 107.9 100.6 86.6 85.6 93.0 98.6 106.6 Food and beverages 80.9 66.3 57.1 64.6 66.0 64.2 101.8 85.5 84.2 98.5 114.3 105.4 117.7 Tobacco 97.0 96.6 84.5 81.0 84.8 92.7 107.9 111.8 124.5 127.4 119.8 102.8 95.6 Textile and textile products 83.3 57.9 56.6 57.1 56.1 46.5 95.6 83.9 73.5 55.2 55.1 85.7 86.3 Textile, furs and leather wearing apparel 88.9 61.8 59.0 85.6 105.1 144.8 125.6 137.7 103.4 109.1 147.9 126.3 149.5 Leather and footwear 83.9 59.5 58.5 65.6 70.8 64.4 103.4 100.9 94.3 95.9 112.4 105.8 119.9 Wood processing industry (excluding furniture) 76.5 62.2 56.5 48.1 47.3 49.1 109.5 99.3 88.3 86.9 100.9 104.5 111.1 Pulp, paper and cardboard 68.9 51.1 46.6 46.6 61.2 56.5 104.8 89.6 77.0 80.0 85.0 112.9 117.1 Crude oil processing, coal coking and treatment of 64.1 57.0 56.5 66.8 67.3 61.4 92.1 89.5 87.5 65.1 67.5 84.3 87.9 Chemical industry and synthetic and artificial fibres 73.5 56.2 58.3 53.6 58.6 55.4 87.8 67.3 51.8 52.9 63.6 104.1 126.9 Rubber and plastics processing 72.2 50.6 53.5 43.8 46.0 47.8 . 100.3 79.8 66.5 52.0 50.4 91.9 93.7 Other products of non metallic minerals 77.2 62.6 62.6 57.8 64.2 67.7 110.2 104.8 91.5 80.9 88.0 94.4 95.4 Metallurgy 71.2 50.1 54.1 56.5 66.4 65.8 92.2 93.1 92.6 58.6 74.2 68.7 87.5 Metallic constructions and metal products 83.4 58.5 54.0 46.2 55.3 56.4 62.0 87.2 51.0 59.4 61.2 105.3 107.1 Machines and equipments 65.1 50.9 52.4 49.4 60.5 78.8 124.4 109.2 72.5 60.1 77.7 78.0 79.9 Computers 61.5 44.9 32.6 35.9 38.0 50.4 na na na na na na na Electric machines and equipments 70.9 47.0 50.4 78.1 97.6 124.0 116.8 128.0 110.2 97.5 100.3 102.2 106.8 Radio, TV sets and communications equipment 88.5 57.9 69.1 148.3 185.2 338.0 134.1 138.6 116.8 107.7 28.8 83.2 43.3 Medical, precision, optical and watchmaking instruments and apparatus 81.0 59.8 60.1 51.7 69.5 98.1 127.1 123.1 85.8 118.0 129.3 111.0 114.0 Road transports means 75.3 53.3 67.1 61.0 61.4 76.4 169.2 161.7 181.1 185.4 137.4 108.2 78.0 Other transport means 74.6 63.8 71.7 56.4 66.8 80.1 105.6 115.8 130.4 118.4 131.5 96.0 104.6 Furniture and other nonclassified activities 94.3 76.7 99.4 112.3 144.9 218.1 143.4 137.7 102.9 108.8 114.1 110.8 121.4 Electric and thermal energy, gas and water 90.6 84.1 89.9 89.8 92.8 94.4 100.8 88.6 78.3 74.6 75.4 94.7 94.3 Electric and thermal energy, gas and hot water production, transport and distribution 83.3 78.0 79.8 79.6 81.6 83.2 103.0 89.3 78.3 75.1 76.8 95.9 98.4 Source: Statistical Yearbooks of Romania 1996, 1999, 2001  S16: Structure of the Industrial Production by Activities (%) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total 100 100 100 100 100 100 100 100 100 100 100 Mining and quarrying 9.0 8.9 7.5 6.8 7.1 6.9 6.7 7.9 6.8 5.3 5.6 of which: Coal mining and preparation 1.0 1.1 2.2 2.1 2.4 2.2 2.2 1.8 1.5 1.3 1.4 Petroleum and natural gas extraction 6.8 6.7 4.4 4.0 3.7 3.6 3.6 5.1 4.4 3.1 3.5 Other extractive activities ' 1.1 0.9 0.8 0.7 0.9 1.1 0.9 1.0 0.9 0.9 0.7 Manufacturing 85.8 83.5 81.6 84.6 79.6 80.5 81.9 78.9 78.8 74.4 79.4 Food and beverages 14.4 14.8 14.7 19.4 15.9 16.3 17.1 16.1 18.2 13.9 16.1 Tobacco industry 0.5 0.5 0.4 0.6 0.5 0.5 0.7 1.1 1.5 1.8 1.5 Textiles and textile products 6.8 6.7 4.9 4.3 3.5 3.2 2.8 2.6 2.4 2.2 2.1 Textile, fur and leather wearing apparel 3.6 2.7 2.1 2.6 2.2 2.6 2.9 2.4 3.2 3.5 3.3 Leather goods and footwear 1.9 1.9 1.7 1.6 1.2 1.5 1.4 1.2 1.4 1.3 1.3 Wood processing (excluding furniture) 1.6 1.8 2.0 1.8 1.8 1.9 1.9 1.8 2.0 2.6 2.5 Pulp, paper and cardboard 1.2 1.5 1.4 1.0 0.9 1.2 1.2 1.0 1.0 1.1 1.2 Publishing houses, polygraphy and records reproduction on supports 0.3 0.3 0.3 1.5 1.1 1.1 1.3 1.1 1.5 1.5 1.3 Crude oil processing, coal coking and nuclear fuel treatment 6.9 6.5 6.7 8.6 7.7 7.7 6.4 8.3 6.3 7.9 10.1 Chemistry and synthetic and man made fibres 7.3 7.8 8.9 7.4 7.5 8.7 7.8 7.2 5.8 5.8 7.0 Rubber and plastics processing 2.6 2.6 2.2 2.3 1.6 2.1 2.1 1.7 1.7 1.8 1.7 Other nonmetallic mineral products 3.5 3.6 3.9 2.9 3.7 3.8 3.9 4.2 3.8 3.5 3.3 Metallurgy 8.5 9.6 10.6 8.8 9.5 10.4 10.4 11.6 9.7 9.0 11.4 Metallic construction and metal products 4.1 3.3 3.0 2.5 4.4 2.8 3.0 2.5 3.1 2.7 2.5 Machinery and equipment 9.3 7.8 7.5 6.2 5.4 5.8 5.6 4.6 4.4 3.8 3.6 Computers and office means 0.5 0.1 0.1 0.3 0.1 0.2 0.4 0.4 0.4 0.3 0.2 Electric machinery and appliances 2.5 2.4 2.2 1.7 1.8 2.0 2.3 2.1 2.3 2.0 1.9 Radio, TV and communications equipment and apparatus 0.8 0.7 0.4 4.1 3.2 0.9 1.3 1.0 0.9 0.8 0.7 Medical, precision, optical, watchmaking instruments and apparatus 1.1 0.9 0.6 0.3 0.3 0.4 0.5 0.4 0.4 0.4 0.4 Means of road transport 3.7 3.5 3.1 2.7 2.9 3.0 4.2 3.3 3.9 3.5 2.5 Other means oftransport 2.3 2.1 2.6 1.5 1.3 1.5 1.7 1.7 2.1 2.2 1.9 Furniture and other nonclassified activities 2.2 2.3 2.2 2.2 2.8 2.6 2.7 2.4 2.5 2.4 2.3 Waste recovering 0.2 0.1 0.1 0.3 0.3 0.3 0.4 0.2 0.3 0.4 0.6 Electric and thermal energy, gas and water 5.2 7.6 10.9 8.6 13.3 12.6 11.3 13.2 14.5 20.3 15.0 Electric and thermal energy, gas and hot water production, transport and distribution 5.0 7.3 10.3 7.9 12.4 11.6 10.3 12.2 13.0 19.1 14.0 Water collection, treatment and distribution 0.2 0.3 0.6 0.7 0.9 1.0 1.0 1.0 1.4 1.2 1.0 1) Including extraction and preparation of radioactive and of metalliferous ores Source: Statistical Yearbook ofRomania 1996 and 2001  S17: Output of Main Agricultural Products 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 thousands of tons, unless otherwise indicated Grains 19307 12289 15493 18184 19883 14200 22107 15453 17037 10478 18871 of which: wheat and rye 5559 3228 5355 6187 7709 3164 7186 5208 4683 4456 7764 maize 10497 6828 7987 9343 9923 9608 12680 8623 10935 4898 9119 Sunflower seeds 612 774 696 764 933 1096 858 1073 1301 721 824 Sugar beet 4703 2897 1776 2764 2655 2848 2726 2361 1415 667 876 Potatoes 1873 2602 3709 2947 3020 3591 3206 3319 3957 3470 3997 Field vegetables' 2112 2518 2766 2476 2783 2647 2354 2754 2996 2478 2826 Fruit 1165 1167 2183 980 917 1632 1416 1036 936 1301 1353 Grapes 849 905 1339 1033 1314 1431 1179 874 1117 1295 1121 Livestock production meat (live weight)2 2023 1895 1935 1852 1846 1868 1705 1672 1521 1414 milk (in million hectoliters) 46 45 47 54 57 57 56 54 53 52 eggs(inmillions) 7177 6140 5633 5407 5567 5783 5271 5331 5668 5711 wool (in tons) 32537 28020 26011 25141 24323 23165 22120 19967 18983 17997 honey(intones) 8279 10414 9936 9820 10435 11157 10543 10198 11153 11746 fish (tons) 25383 23857 19381 17925 13528 17099 Memo items: Agricultural area Total (tho hectars) 14793 14798 14797 14789 14794 14746 14781 14857 of which: Private 10336 10371 10694 10694 10431 12342 12561 14218 Per capita output wheatandrye(inkg) 240 142 235 272 340 140 319 231 209 199 potatoes(inkg) 81 114 163 130 133 159 142 148 176 155 meat(in kg) 87 83 85 82 81 83 76 74 68 63 milk (in liters) 199 198 208 236 250 253 249 242 234 230 1) 2001: includes solaria production 2) represents the livestock lroduction Source: Stattstical Yearbook ofRomania 1997 and 2001 and NIS Statistical Bulletin 1/2002 Employment in Agriculture (including self employed) (thousand of persons, end of year) 1993 1994 1995 1996 1997 1998 Total employment in agriculture 3537 3561 3187 3249 3322 3296 as % of total employment 35.2 35.6 33.6 34.6 36.8 37.4 Prvate farms 3139 3242 2926 3000 3156 3143 State farms (public and mixed) 398 318 261 249 166 153 Agroprocessing (ave) 255 244 231 219 213 214 % of total employment in agriculture Private farms 88.7 91.0 91.8 92.3 95.0 95.4 State farms (public and mixed) 11.3 8.9 8.2 7.7 5.0 4.6 Agroprocessing (aye) 7.2 6.9 7.2 6.7 6.4 6.5 Source: IMF (data provided by the Romanian authorities)  S18: Average Net' Nominal Monthly Salary Earning, by Activities of the National Economy leilemployee 1991 1992 1993 1994 1995 1995 1997 1998 1999 9 2000 Total 7460 20140 59717 141951 211373 321169 632086 1042274 1522878 2139138 Agriculture 6752 17151 49053 112922 171148 254194 469680 764800 1164405 1532934 Sylviculture, forestry and hunting 6689 18551 50166 116344 179319 277082 573426 964280 1480857, 1956305 Industry 7499 20508 62388 146999 227469 350039 693437 1094757 1554217 2229496 Construction 7981 21438 64189 164945 224855 332082 617101 986083 1399927c 1861422 Trade 6857 18075 52751 124038 168777 250282 459497 717877 1066958!4 1502294 Hotels and restaurants 5881 14770 42023 98078 145403 216496 412334 663357 9414551 1381068 Transport and storage 8850 25229 71548 166573 253344 397682 772902 1222723 1805850 2590861 Post and telecommunications 7269 19590 65459 162605 265877 385872 905341 1664782 2533361 3583359 Financial, banking and insurance activities 8526 29785 85571 234287 389521 659092 1482926 2763051 3999188 5258061 Real estate and other services 7431 21052 59196 138667 226271 340445 681983 1062108 1520096 2159136 Public administration 8206 20973 62663 149888 225914 304649 608716 1373164 2143292 3044988 Education 7408 18735 54337 132361 194772 275597 539919 1051738 1415535 2046107 Health and social assistance 7983 19760 55067 126183 161252 229743 463440 850351 1506768 1768105 Other activities of national economy 5747 14703 45011 113796 156041 251638 519982 857994 1314496- 1884899 1) According to NIS methodology, applied until year 2000, net salary earnings exclude tax but include social security contributions. Starting year 2000, the net salary earnings are calculated by subtracting from the gross nominal salary earnings the tax and the employee's contributions for unemployment, supplementary pension and health insurance Source. The National Institute for Statistics and staffcalculations USS per employee 1991 1992 1993 1994 1995 1995 1997 1998 1999 2000 Memo: exch rate (ROL/US$) 76.47 307.95 760.05 1655.09 2033.26 3082.60 7167.94 8875.55 15332.93 21692.74 Total 98 65 79 86 104 104 88 117 99 99 Agriculture 88 56 65 68 84 82 66 86 76 71 Sylviculture, forestry and hunting 87 60 66 70 88 90 80 109 97 90 Industry 98 67 82 89 112 114 97 123 101 103 Construction 104 70 84 100 111 108 86 111 91 86 Trade 90 59 69 75 83 81 64 81 70 69 Hotels and restaurants 77 48 55 59 72 70 58 75 61 64 Transport and storage 116 82 94 101 125 129 108 138 118 119 Post and telecommunications 95 64 86 98 131 125 126 188 165 165 Financial, banking and insurance activities Ill 97 113 142 192 214 207 311 261 242 Real estate and other services 97 68 78 84 Ill 110 95 120 99 100 Public administration 107 68 82 91 111 99 85 155 140 140 Education 97 61 71 80 96 89 75 118 92 94 Health and social assistance 104 64 72 76 79 75 65 96 98 82 Other activities of national economy 75 48 59 69 77 82 73 97 86 87  519: Exports by Main Sections (Combined Nomenclature) Section TOTAL FOB EXPORT 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 million USS Total 5775 4266 4363 4892 6151 7910 8084 8431 8302 8487 10367 11385 I Live animals and animal products 23 148 185 162 219 170 155 199 94 116 126 131 11 Vegetable products 38 70 50 58 64 209 374 164 181 250 119 181 Ill Animal or vegetable oils and fats 0 6 7 61 46 80 67 126 70 54 21 25 IV Food, beverages and tobacco 21 36 48 47 68 72 108 106 91 64 72 97 V Mineral products 1136 621 572 574 714 731 692 638 508 502 822 784 VI Chemical products 301 285 423 345 489 724 691 560 336 327 518 503 VII Plastics and rubber 106 54 85 82 143 202 193 186 174 182 230 231 VIl Hides and skins, dressed leather, furs 36 25 25 35 57 53 53 66 62 65 99 143 IX Wooden products, excludng furniture 180 115 155 177 222 258 296 338 385 495 564 530 X Paper and paper articles 33 22 17 19 34 85 58 64 46 48 81 105 XI Textiles and textile articles 573 404 455 785 1156 1570 1734 1942 2162 2197 2506 2979 XII Footwear, headger, umbrellas, others 129 87 73 160 311 428 505 545 609 683 789 982 XIII Articles ofstone,plaster,cement 78 75 88 100 Ill 152 150 151 157 159 161 172 ceramics, glass and similar materials XV Basic metals and articles 949 634 735 959 1065 1437 1268 1556 1583 1310 1658 1516 XVI Machinery, appliances and electric 947 692 508 439 519 656 673 737 789 968 1451 1680 equipments; audio and video XVII Transportation means and materials 770 492 473 405 392 428 441 448 427 450 509 598 XX Miscellaneous goods and products 413 404 366 402 436 574 552 526 519 502 533 613 Others(XIV+XVIII+XIX,XXII) 42 96 99 85 107 84 75 79 109 115 108 115 Source: Statistical Yearbook of Romania 1997and 2000 for 2001. NIS Monthly Bulletinfor Foreign Trade 1/2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 percent of otal FOB Export 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 I Live animals and animal products 0.4 3.5 4.2 3.3 3.6 2.1 1.9 2.4 1.1 1.4 1.2 1.1 II Vegetable products 0.7 1.6 1.1 1.2 1.0 2.6 4.6 1.9 2.2 2.9 1.2 1.6 III Animal or vegetable oils and fats 0.0 0.1 0.2 1.2 0.7 1.0 0.8 1.5 0.8 0.6 0.2 0.2 IV Food, beverages and tobacco 0.4 0.9 1.1 1.0 1.1 0.9 1.3 1.3 1.1 0.8 0.7 0.9 V Mineralproducts 19.7 14.5 13.1 11.7 11.6 9.2 8.6 7.6 6.1 5.9 7.9 6.9 VI Chemical products 5.2 6.7 9.7 7.1 7.9 9.2 8.5 6.6 4.0 3.9 5.0 4.4 VII Plastics and rubber 1.8 1.3 2.0 1.7 2.3 2.6 2.4 2.2 2.1 2.1 2.2 2.0 VIII Hides and skins, dressed leather, furs 0.6 0.6 0.6 0.7 0.9 0.7 0.7 0.8 0.7 0.8 1.0 1.3 IX Wooden products, excluding furniture 3.1 2.7 3.5 3.6 3.6 3.3 3.7 4.0 4.6 5.8 5.4 4.7 X Paper and paper articles 0.6 0.5 0.4 0.4 0.5 1.1 0.7 0.8 0.6 0.6 0.8 0.9 XI Textiles and textile articles 9.9 9.5 10.4 16.0 18.8 19.8 21.4 23.0 26.0 25.9 24.2 26.2 XII Footwear, headger, umbrellas, others 2.2 2.0 1.7 3.3 5.1 5.4 6.2 6.5 7.3 8.0 7.6 8.6 XIII Articles of stone,plaster,cement 1.4 1.8 2.0 2.0 1.8 1.9 1.9 1.8 1.9 1.9 1.6 1.5 ceramics, glass and similar materials XV Basic metals and articles 16.4 14.9 16.8 19.6 17.3 18.2 15.7 18.5 19.1 15.4 16.0 13.3 XVI Machinery, appliances and electric 16.4 16.2 11.6 9.0 8.4 8.3 8.3 8.7 9.5 11.4 14.0 14.8 equipments; audio and video 0.0 0.0 XVII Transportation means and materials 13.3 11.5 10.8 8.3 6.4 5.4 5.5 5.3 5.1 5.3 4.9 5.2 XX Miscellaneous goods and products 7.2 9.5 8.4 8.2 7.1 7.3 6.8 6.2 6.3 5.9 5.1 5.4 Others (XIV+XVIII+XIX, XXII) 0.7 2.3 2.3 1.7 1.7 1.1 0.9 0.9 1.3 1.4 1.0 1.0  Import by Main Sections (Combined Nomenclature) TOTAL CIF IMPORT 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 million USS Total 9202 5793 6260 6522 7109 10278 11435 11280 11338 10557 13055 15551.6 1 Live animals and animal products 352 42 86 68 99 129 74 65 208 123 148 280 II Vegetable products 424 369 429 475 140 154 167 170 226 224 273 335 III Animal or vegetable oils and fats 46 14 39 21 30 28 21 33 56 32 35 34 IV Food, beverages and tobacco 389 360 442 400 392 580 605 426 523 417 476 558 V Mineral products 3,896 2626 2028 1872 1906 2484 2689 2408 1687 1251 1893 2237 VI Chemical products 655 410 412 514 563 932 990 940 1028 974 1077 1218 VII Plastics and rubber 92 76 180 204 227 390 448 444 512 477 582 758 VIII Hides and skins, dressed leather, furs 92 18 76 I14 149 213 264 278 306 302 362 509 IX Wooden products, excluding furniture 75 35 33 30 27 56 52 53 60 68 83 110 X Paper and paper articles 44 39 91 96 135 235 274 255 31I 264 291 345 XI Textiles and textile articles 264 246 562 659 810 1209 1343 1564 1825 1937 2131 2500 XII Footwear, headger, unbrellas, others I5 45 42 48 67 106 124 157 198 190 223 254 XIII Articles of stone, plaster, cement 114 74 55 63 69 14 145 143 163 155 179 230 ceramics, glass and similar materials XV Basic metals and articles 522 243 272 278 352 549 715 670 790 687 891 1138 XVI Machinery, appliances and electric equipments; audio and video 1,321 838 914 1149 1450 2113 2506 2593 2723 2483 3216 3527 XVII Transportation means and materials 637 142 294 283 332 397 415 384 485 420 552 799 XX Miscellaneous goods and products 72 31 52 77 113 181 202 209 227 207 242 275 Others (XIV+XVIII+XIX, XXII) 192 185 253 171 249 410 404 489 510 348 402 445 Source: Staristial Yearbook of Romania 1997 and 2000: for 2001: NIS Monthly Bullen for Foreign Trade 1/000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 percent of total CIF Import 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 I Live animals and animal products 3.8 0.7 1.4 1.0 1.4 1.3 0.6 0.6 1.8 1.2 1.1 1.8 II Vegetable products 4.6 6.4 6.9 7.3 2.0 1.5 1.5 1.5 1.9 2.1 2.1 2.2 III Animal or vegetable oils and fats 0.5 0.2 0.6 0.4 0.5 0.3 0.2 0.3 0.5 0.3 0.3 0.2 IV Food, beverages and tobacco 4.2 6.2 7.1 6.1 5.5 5.7 5.3 3.8 4.4 3.9 3.6 3.6 V Mineral products 42.3 45.3 32.4 28.7 26.8 24.2 23.5 21.3 14.3 11.9 14.5 14.4 VI Chemical products 7.1 7.1 6.6 7.8 7.9 9.0 8.6 8.3 8.7 9.2 8.3 7.8 VII Plastics and rubber 1.0 1.3 2.9 3.1 3.2 3.8 3.9 3.9 4.3 4.5 4.5 4.9 VIII Hides and skins, dressed leather, firs 1.0 0.3 1.2 1.7 2.1 2.1 2.3 2.5 2.6 2.9 2.8 3.3 IX Wooden products, excluding furniture 0.8 0.6 0.5 0.5 0.4 0.5 0.5 0.5 0.5 0.6 0.6 0.7 X Paper and paper articles 0.5 0.7 1.4 1.5 1.9 2.3 2.4 2.3 2.6 2.5 2.2 2.2 XI Textiles and textile articles 2.9 4.2 9.0 10.1 11.4 11.8 11.7 13.9 15.4 18.3 16.3 16.1 XH1 Footwear, headger, umbrellas, others 0.2 0.8 0.7 0.7 0.9 1.0 1.1 1.4 1.7 1.8 1.7 1.6 XIII Articles ofstone,plaster,cement 1.2 1.3 0.9 1.0 1.0 1.1 1.3 1.3 1.4 1.5 1.4 1.5 ceramics, glass and similar materials 0.0 0.0 XV Basic metals and articles 5.7 4.2 4.3 4.3 4.9 5.3 6.3 5.9 6.7 6.5 6.8 7.3 XVI Machinery, appliances and electric 14.4 14.5 14.6 17.6 20.4 20.6 21.9 23.0 23.0 23.5 24.6 22.7 equipments; audio and video 0.0 0.0 XVII Transportation means and materials 6.9 2.5 4.7 4.3 4.7 3.9 3.6 3.4 4.1 4.0 4.2 5.1 XX Miscellaneous goods and products 0.8 0.5 0.8 1.2 1.6 1.8 1.8 1.9 1.9 2.0 1.9 1.8 Others (XIV+XVIII+XIX, XXII) 2.1 3.2 4.0 2.6 3.5 4.0 3.5 4.3 4.3 3.3 3.1 2.9  S20: Geographical Orientation of the Foreign Trade million USD as% of FOB export/ CIF import 1993 1994 1995 1996 1997 1998 1999 2000 2001 1993 1994 1995 1996 1997 1998 1999 2000 2001 FOREIGN TRADE Exports - total 4892 6151 7910 8084 8431 8302 8487 10367 11385 100 100 100 100 100 100 100 100 100 Imports - total 6522 7109 10278 11435 11280 11838 10557 13055 15552 100 100 100 100 100 100 100 100 100 Europe Exports 2906 4163 5615 5984 6389 6744 7134 8798 9691 59.4 67.7 71.0 74.0 75.8 81.2 84.1 84.9 85.1 Imports 4499 5389 7878 8843 8756 9775 8722 10541 12697 69.0 75.8 76.6 77.3 77.6 82.6 82.6 80.7 81.6 European Union 1) Exports 2023 2965 4283 4569 4768 5358 5562 6618 7720 41.4 48.2 54.1 56.5 56.6 64.5 65.5 63.8 67.8 Imports 2955 3427 5186 5986 5922 6829 6405 7388 8918 45.3 48.2 50.5 52.3 52.5 57.7 60.7 56.6 57.3 Austria Exports 76 97 155 169 178 249 243 251 342 1.6 1.6 2.0 2.1 2.1 3.0 2.9 2.4 3.0 Imports 166 195 315 349 304 347 311 332 441 2.5 2.7 3.1 3.1 2.7 2.9 2.9 2.5 2.8 France Exports 219 316 458 460 465 491 530 722 919 4.5 5.1 5.8 5.7 55 5.9 6.2 7.0 8.1 Imports 507 362 536 561 648 815 706 799 979 7.8 5.1 5.2 4.9 5.7 6.9 6.7 6.1 6.3 Germany Exports 700 988 1435 1486 1419 1629 1510 1627 1781 14.3 16.1 18.1 18.4 16.8 19.6 17.8 15.7 15.6 Imports 1031 1278 1803 2008 1851 2064 1841 1923 2365 15.8 18.0 17.5 17.6 16.4 17.4 17.4 14.7 15.2 Italy Exports 406 795 1243 1385 1643 1827 1980 2319 2838 8.3 12.9 15.7 17.1 19.5 22.0 233 22.4 24.9 Imports 614 841 1363 1751 1784 2059 2062 2443 3100 9.4 11.8 13.3 15.3 15.8 17.4 19.5 18.7 19.9 Netherlands Exports 212 216 239 343 264 313 327 329 386 4.3 3.5 3.0 4.2 3.1 3.8 3.9 3.2 3.4 Imports 157 178 276 266 229 278 243 284 322 2.4 2.5 2.7 2.3 2.0 2.3 2.3 2.2 2.1 United Kingdom of Great Britain Exports 185 200 236 248 299 303 412 546 587 3.8 3.3 3.0 3.1 3.5 3.6 4.9 5.3 5.2 Imports 171 223 303 330 388 397 443 536 539 2.6 3.1 2.9 2.9 3.4 3.4 4.2 4.1 3.5 Other Countries and/or Continents Switzerland Exports 119 45 63 37 44 S 57 58 54 2.4 0.7 0.8 0.5 0.5 0.6 0.7 0.6 0.5 Imports 148 147 199 193 149 135 127 153 169 2.3 2.1 1.9 1.7 1.3 1.1 1.2 1.2 1.1 Poland Exports 20 14 39 53 102 82 120 102 100 0.4 0.2 0.5 0.7 1.2 1.0 1.4 1.0 0.9 Imports 33 30 57 86 90 146 160 193 275 0.5 0.4 0.6 0.8 0.8 1.2 1.5 1.5 1.8 Hungary Exports 117 161 174 171 183 219 271 355 372 2.4 2.6 22 2.1 2.2 2.6 3.2 3.4 3.3 Imports 168 167 315 289 347 547 414 513 599 2.6 2.3 3.1 2.5 3.1 4.6 3.9 3.9 3.9 Bulgaria Exports 103 103 71 71 57 78 136 290 202 2.1 1.7 0.9 0.9 0.7 0.9 1.6 2.8 1.8 Imports 71 65 77 73 56 47 55 88 149 1.1 0.9 0.7 0.6 0.5 0.4 0.5 0.7 1.0 Yugoslavia Exports 8 20 31 137 140 119 86 138 155 0.2 0.3 0.4 1.7 1.7 1.4 1.0 1.3 1.4 Imports * * 31 54 61 56 65 34 * * 0.3 0.5 0.5 0.5 0.5 0.2 Republic of Moldova Exports 94 60 82 101 129 129 101 142 112 1.9 1.0 1.0 1.2 1.5 1.6 1.2 1.4 1.0 Imports 87 95 105 80 62 62 40 41 39 1.3 1.3 1.0 0.7 0.5 0.5 0.4 03 0.2 Russian Federation Exports 220 207 158 164 250 81 47 89 83 4.5 3.4 2.0 2.0 3.0 1.0 0.6 0.9 0.7 Imports 764 984 1229 1432 1356 1062 704 1120 1184 11.7 13.8 12.0 12.5 12.0 9.0 6.7 8.6 7.6 Ukraine Exports 105 95 131 65 92 53 63 90 45 2.1 1.5 1.7 0.8 1.1 0.6 0.7 0.9 0.4 Imports 133 150 194 182 186 168 109 196 320 2.0 2.1 1.9 1.6 1.6 1.4 1.0 1.5 2.1 Turkey Exports 277 251 348 390 354 323 468 627 451 5.7 4.1 44 4.8 4.2 3.9 5.5 6.0 4.0 Imports 147 151 251 219 .213 272 237 271 377 2.3 2.1 2.4 1.9 1.9 2.3 2.2 2.1 2.4 Asia Exports 1521 1207 1310 1147 1087 665 590 678 846 31.1 19.6 16.6 14.2 12.9 8.0 7.0 6.5 7.4 Imports 1245 823 1153 1323 1358 1083 1065 1248 1388 19.1 11.6 11.2 11.6 12.0 9.1 10.1 9.6 8.9 China Exports 419 276 179 93 44 23 36 85 89 8.6 4.5 23 1.2 0.5 0.3 0.4 0.8 0.8 Imports 93 60 90 115 120 178 145 174 253 1.4 0.8 0.9 1.0 1.1 1.5 1.4 13 1.6 Africa Exports 257 418 571 551 457 367 327 378 300 53 6.8 7.2 6.8 5.4 4.4 3.9 3.6 2.6 Imports 184 170 483 312 189 86 82 77 123 2.8 2.4 4.7 2.7 1.7 0.7 0.8 0.6 0.8 America Exports 201 355 407 395 485 523 426 501 523 4.1 5.8 5.1 4.9 5.8 6.3 5.0 4.8 4.6 Imports 566 672 690 835 853 818 610 743 912 8.7 9.5 6.7 7.3 7.6 6.9 5.8 5.7 5.9 United States of America Exports 69 193 201 192 320 319 317 380 357 1.4 3.1 2.5 2.4 3.8 3.8 3.7 3.7 3.1 Imports 373 465 419 431 462 499 370 391 491 5.7 6.5 4.1 3.8 4.1 4.2 3.5 3.0 - 3.2 Source: Statistical Yearbook ofRomania 1997 and 2001; NIS Foreign Trade Bulletin 1/2002; staffcalculations  S21: Consumer Price Index per Month, Oct 1990 = 100 (Oct. 1990= 100) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total Estimated' January 158 531 1483 5518 8678 11000 19386 44960 62079 97318 136135 February 169 597 1605 5843 8803 11207 23025 48193 63863 99427 139267 March 180 657 1752 6328 8885 11401 30097 50002 67925 101209 142052 April 228 688 1927 6708 9028 11622 32174 51365 71222 106051 145887 May 240 771 2514 7042 9125 12243 33545 52536 75006 107982 148367 June 245 804 2651 7223 9245 12370 34316 53196 78827 111042 150741 July 268 830 3001 7338 9483 13300 34553 53908 80129 115791 152701 August 298 858 3326 7468 9575 13805 35768 54251 81105 117918 156060 September 319 945 3689 7762 9727 14136 36952 55717 83691 121230 159025 October 353 1035 4291 8103 10074 14615 39346 57878 87174 124579 162842 November 123 391 1175 4899 8334 10484 15458 41026 58984 90651 128115 167239 December 138 445 1330 5260 8507 10868 17052 42872 60265 93297 131278 170918 Food January 137 597 1677 6047 9851 12138 20609 47814 59312 84476 119718 February 143 671 1773 6465 10002 12370 25804 51494 60782 87056 123430 March 161 724 2043 7182 10027 12549 33914 52780 63630 89137 126515 April 256 769 2285 7605 10169 12855 35775 53892 66942 91178 130690 May 258 894 2977 8025 10245 13744 36770 54281 69043 92919 133173 June 254 934 2989 8112 10258 13829 37156 54259 69356 96353 135837 July 282 943 3292 8160 10574 14571 36981 53737 68557 101386 135973 August 330 958 3612 8192 10575 14850 38408 53206 68580 102580 136925 September 353 1074 4044 8604 10701 15199 39215 54625 71058 105619 138842 October 387 1190 4736 9030 11072 15745 41548 55499 73489 108847 141341 November 120 421 1364 5517 9316 11488 16655 43264 56492 75990 111996 143037 December 127 491 1568 5820 9561 12009 18634 45769 57888 79127 115336 146756  Consumer Price Index per Month, Oct. 1990 = 100 (Oct. 1990= 100) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Non - Food January 182 501 1425 5490 8147 10345 19121 40729 59583 95329 130910 February 198 571 1594 5748 8250 10529 21482 43573 61601 96595 132612 March 205 639 1643 6030 8328 10690 27392 45843 67043 97769 134999 April 215 661 1782 6414. 8502 10834 29556 46889 69549 102993 138239 May 234 725 2362 6677 8587 11255 31442 48823 71909 104968 140727 June 246 755 2571 6889 8751 11414 32187 50087 77130 108115 142416 July 269 789 2931 7074 8919 12553 32747 51390 79107 112334 145264 August 285 828 3312 7274 9056 13185 33637 52357 80880 114672 150929 September 306 894 3678 7472 9200 13460 34710 53272 83241 118174 154401 October 341 971 4293 7739 9507 13884 36807 56006 86235 120930 158415 November 120 376 1091 4778 7943 9905 14823 38316 57052 90834 125149 166019 December 149 419 1212 5194 8057 10190 16331 39619 58141 93131 128092 168343 Services January 146 447 1148 4249 7051 9874 17205 51003 81446 154597 219490 February 157 482 1205 4504 7162 10084 20079 54187 83636 157754 224758 March 163 538 1309 4864 7419 10413 27795 56257 87862 160433 227905 April 193 556 1411 5160 7488 10601 30366 59446 93748 174709 231323 May 208 586 1746 5453 7680 11016 31384 60771 108909 177253 234562 June 215 608 2007 5812 7986 11181 33459 61725 121803 178704 237845 July 229 649 2446 5929 8201 12118 33997 64503 129073 184211 244505 August 250 681 2630 6155 8434 12939 35612 65705 131140 189209 250618 September 271 754 2811 6380 8688 13377 38375 69076 134920 193100 256883 October 299 817 3146 6690 9069 13831 41990 73599 144164 198861 266388 November 138 356 922 3651 6841 9512 14244 44144 75276 147768 201824 271449 December 140 398 1046 4005 6973 9792 15035 45685 76843 149643 205131 279321 Source: 1990 to 2000 :National Institute for Statistics; 2001 data: estimated using the 2000 serie and the m/n CPI published by NIS  S22: Consumer Price Index per Month, previous Month = 100 (previous month= 100) Total 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 January 114.8 119.5 111.5 104.9 102 101.2 113.7 104.9 103.0 104.3 103.7 February 107 112.5 108.2 105.9 101.4 101.9 118.8 107.2 102.9 102.2 102.3 March 106.6 110 109.2 108.3 100.9 101.7 130.7 103.8 106.4 101.8 102 April 126.5 104.7 110 106.1 101.6 101.9 106.9 102.7 104.8 104.8 102.7 May 105.1 112.1 130.4 105 101.1 105.3 104.3 102.3 105.3 101.8 101.7 June 102 104.3 105.5 102.6 101.3 101 102.3 101.3 105.1 102.8 101.6 July 109.5 103.2 113.2 101.6 102.6 107.5 100.7 101.3 101.7 104.3 101.3 August 111.2 103.4 110.8 101.8 101 103.8 103.5 100.6 101.2 101.8 102.2 September 107.3 110.1 110.9 103.9. 101.6 102.4 103.3 102.7 103.2 102.8 101.9 October 110.4 109.6 116.3 104.4 103.5 103.4 106.5 103.9 104.2 102.8 102.4 November 123.4 110.9 113.5 114.2 102.8 104.1 105.8 104.3 101.9 104.0 102.8 102.7 December 111.6 113.7 113.2 107.4 102.1 103.7 110.3 104.5 102.2 102.9 102.5 102.2 Food 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 January 108.5 121.7 107 103.9 103 101.1 110.6 104.5 102.5 106.8 103.8 February 104.1 112.4 105.7 106.9 101.5 101.9 125.2 107.7 102.5 103.1 103.1 March 112.9 107.9 115.2 111.1 100.3 101.5 131.4 102.5 104.7 102.4 102.5 April 158.6 106.2 111.8 105.9 101.4 102.4 105.5 102.1 105.2 102.3 103.3 May 101 116.3 130.3 105.5 100.7 106.9 102.8 100.7 103.1 101.9 101.9 June 98.3 104.5 100.4 101.1 100.1 100.6 101 100 100.5 103.7 102 July 110.8 101.1 110.1 100.6 103.1 105.4 99.5 99 98.8 105.2 100.1 August . 117.2 101.6 109.7 100.4 100 101.9 103.9 99 100.0 101.2 100.7 September 106.9 112.1 112 105 101.2 102.3 102.1 102.7 103.6 103 101.4 October 109.6 110.8 117.1 104.9 103.5 103.6 106 101.6 103.4 103.1 101.8 November 120.4 108.7 114.6 116.5 103.2 103.8 105.8 104.1 101.8 103.4 102.9 101.2 December 105.1 116.7 115 105.5 102.6 104.5 111.9 105.8 102.5 104.1 103 102.6 Non Food 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 January 122 119.6 117.6 105.7 101.1 101.5 117.1 102.8 102.5 102.4 102.2 February 108.8 114.1 111.9 104.7 101.3 101.8 112.3 107 103.4 101.3 101.3 March 103.4 111.8 103.1 104.9 101 101.5 127.5 105.2 108.8 101.2 101.8 April 105.1 103.5 108.5 106.3 102.1 101.3 107.9 102.3 103.7 105.3 102.4 May 108.9 109.6 132.5 104.1 101 103.9 106.4 104.1 103.4 101.9 101.8 June 105.1 104.2 108.9 103.2 101.9 101.4 102.4 102.6 107.3 103 101.2 July 109.3 104.6 114 102.7 101.9 110 101.7 102.6 102.6 103.9 102 August 106 104.9 113 102.8 101.5 105 102.7 101.9 102.2 102.1 103.9 September 107.3 107.9 111.1 102.7 101.6 102.1 103.2 101.8 102.9 103.1 102.3 October 111.4 108.7 116.7 103.6 103.3 103.2 106 105.1 103.6 102.3 102.6 November 120.2 110.5 112.3 111.3 102.6 104.2 106.8 104.1 101.9 105.3 103.5 104.8 December 124 111.2 111.1 108.7 101.4 102.9 110.2 103.4 101.9 102.5 102.4 101.4 Services 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 January 104.4 112.3 109.8 106.1 101.1 100.8 114.4 111.6 106.0 103.3 107 February 108 107.7 105 106 101.6 102.1 116.7 106.2 102.7 102 102.4 March 103.5 111.7 108.6 108 103.6 103.3 138.4 103.8 105.0 101.7 101.4 April 118.5 103.3 107.8 106.1 100.9 101.8 109.3 105.7 106.7 108.9 101.5 May 107.6 105.4 123.8 105.7 102.6 103.9 103.3 102.2 116.2 101.5 101.4 June 103.7 103.8 114.9 106.6 104 101.5 106.6 101.6 111.8 100.8 101.4 July 106.2 106.8 121.9 102 102.7 108.4 101.6 104.5 106.0 103.1 102.8 August 109.2 104.9 107.5 103.8 102.8 106.8 104.8 101.9 101.6 102.7 102.5 September 108.7 110.7 106.9 103.7 103 103.4 107.8 105.1 102.9 102.1 102.5 October 110.2 108.3 111.9 104.9 104.4 103.4 109.4 106.5 106.8 103 103.7 November 137.7 119 112.9 116.1 102.3 104.9 103 105.1 102.3 102.5 101.5 101.9 December 101.3 112 113.5 109.7 101.9 103 105.6 103.5 102.1 101.3 101.6 102.9 Source The National Institulefar Statistics and Economic Studies  S23: Consummer Price Index (yearly evolution) 1990=100 provisional 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total 270 839 2987 7072 9353 12983 33077 52624 76728 111793 150361 Food goods 286 963 3361 7940 10469 14277 35886 53264 68098 97856 132791 Non-food goods 268 788 2907 6770 8776 12206 30823 49392 75233 108336 144195 Services 236 661 2250 5642 8051 11831 32709 62819 115609 177922 240906 previous year =100 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total 270 310 356 237 132 139 255 159 146 146 135 Food goods 286 337 349 236 132 136 251 148 128 144 136 Non-food goods 268 294 369 233 130 139 253 160 152 144 133 Services 236 280 340 251 143 147 277 192 184 154 135 Source: National Institute for Statistics and staff calculations (2001)  S24: Industrial Production Price Index, per Month 1996= 100 1998=100 Total 1997 1998 1999 2000 2001 2000 2001 Industry-Total January 138.0 309.2 368.5 598.6 895.2 183.3 275.3 February 158.6 317.7 383.2 615.6 927.7 188.8 285.3 March 239.8 323.0 408.8 632.5 947.5 193.6 291.4 April 251.9 330.1 430.9 651.3 962.2 199.3 295.9 May 262.3 334.7 445.9 663.9 984.0 203.8 302.6 June 268.7 337.3 477.8 698.8 999.6 213.6 307.4 July 273.7 338.0 492.1 740.5 1029.8 225.8 316.7 August 276.5 340.3 510.1 765.9 1051.6 232.4 323.4 September 281.7 344.7 526.8 796.4 1072.4 241.8 329.8 October 286.1 349.0 548.4 826.3 1095.2 252.0 336.8 November 294.5 354.8 563.4 847.3 1110.1 260.1 341.4 December 300.2 359.5 585.4 865.6 1126.1 266.2 346.3 Source: The National Institute for Statistics and staff calculations  S25: Indices of Investment by Activities 1990 = 100 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total 74.2 73.4 79.6 100.6 111.4 114.9 108.7 104.2 98.1 101.5 Agriculture 39.2 31.9 20.3 86.5 52.2 55.5 29.0 28.7 23.6 25.9 Sylviculture, forestry and hunting 38.9 45.6 31.2 35.8 21.4 40.0 29.3 26.0 31.3 36.6 Industry I 91 96.8 91.6 89.6 112.4 121.5 114.6 112.1 102.9 95.0 Construction 43.9 55.4 61.7 136.7 146.4 180.2 230.7 177.8 168.4 206.5 Trade, hotels and restaurants 122.9 193.9 295.5 268.9 307.1 356.5 301.2 388.2 288.0 389.7 Transport and storage 52 27.4 56.4 40.3 26.0 30.2 25.2 26.2 27.1 37.8 Post and telecommunications 110.2 259.1 622.9 718.8 616.7 820.8 16 times 18.4 times 25.5 times 25.1 times Financial, banking and insurance activities 612.7 17 times 26 times 44 times 47 times 53 times 61 times 60.3 times 54.4 times 55.4 times Public administration 87.4 82.5 202 386.2 522.1 475.1 720.7 417.4 582.3 488.5 Education 124.1 55 90 160.6 179.6 317.0 350.9 258.8 135.8 12.8 Health and social assistance 127.2 89.2 78.1 71.5 120.1 126.5 120.8 141.3 75.2 19.2 Other activities of the national economy 69.7 57 52.4 62.5 133.6 77.9 54.8 37.1 31.9 39.1 1) Including electric and thermal energy, gas and water S26: Investment by Sector, 1993-2000 (In billions of ROL at current prices) 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Industry 171 503 1,411 2,952 5,402 9,187 19,772 27,568 37,172 49,394 Construction and geological exploration 6 23 78 431 679 1,310 3,676 4,227 5,767 10,673 Agriculture and forestry 32 96 196 1,529 1,420 2,427 2,890 4,088 5,933 9,881 Transport and storage 33 57 317 750 711 1,287 2,333 3,473 5,325 10,651 Post and Telecommunication 3 20 134 300 368 761 3,016 4,589 9,404 12,416 Trade ' 14 59 273 679 1,144 2,071 3,868 7,220 7,689 14,971 Education 3 4 17 67 109 295 710 768 588 83 Health and Social Assistance 4 8 21 43 105 166 356 581 457 175 Public Administration and Defence 3 8 56 292 578 809 2,703 2,287 4,423 5,356 Financial, banking and insurance sector 2 19 81 311 472 831 2,071 3,002 3,870 5,582 Other 42 93 239 651 2,007 1,802 2,741 2,712 3,319 5,805 Total Investment 314 889 2,822 8,005 12,996 20,946 44,135 60,515 83,947 124,987 in national economy, of which: state sector 283 717 1,959 4,693 6,899 10,705 20,084 na na more than 50% owned by state 29,001 41,583 52,122 investment in housing na na 165 443 558 975 2,303 3,204 4,754 (as % of total investment) 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Industry 55% 57% 50% 37% 42% 44% 45% 46% 44% 40% Construction and geological exploration 2% 3% 3% 5% 5% 6% 8% 7% 7% 9% Agriculture and forestry 10% 11% 7% 190/ 11% 12% 7% 7% 7% 8% Transport and storage 10% 6% 11% 90/ 5% 6% 5% 6% 6% 9% Post and Telecommunication 1% 2% 5% 4% 3% 4% 7% 8% 11% 10% Trade ' 4% 7% 10% 8% 9% 10% 9/0 12% 90% 12% Education 1% 0% 1% 1% 1% 1% 2% 1% 1% 0.1% Health and Social Assistance 1% 1% 1% 1% 1% 1% 1% 1% 1% 0.1% Public Administration and Defence 1% 1% 2% 4% 4% 4% 6% 4% 5% 4% Financial , banking and insurance sector 1% 2% 3% 4% 4% 4% 5% 5% 5% 4% Other 14% 100/ 8% 8% 15% 9% 6% 4% 4% 5% Total Investment 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% of which: state sector 90% 81% 69% 59% 53% 51% 46% more than 50% owned by state 48% 50% 42% investment in housing na na 6% 6% 4% 5% 5% 5% 6% na 1) for 1996 and 1997 Trade includes hotels and restaurants  S27: Employment, Unemployment, Average Monthly Net Nominal Salary Earnings, by Region, in 2000 Total North-East South-East South South-West West North-West Center Bucharest Population, as of July Ist (mn persons) 22.4 3.8 2.9 3.5 2.4 2.0 2.8 2.6 2.3 of which: male (mn persons) 11.0 1.9 1.4 1.7 1.2 1.0 1.4 1.3 1.1 female (mn persons) 11.5 1.9 1.5 1.8 1.2 1.1 1.4 1.3 1.2 Urban , % of total 54.6 43.5 56.8 41.6 45.3 62.2 52.6 60.3 88.8 Rural, % of total 45.4 56.5 43.2 58.4 54.7 37.8 47.4 39.7 11.2 lnhabitans/Km 94.1 103.8 82.1 100.6 82.2 63.7 83.3 77.5 1254.6 Employment (average number) thousand employees 4623 633 565 617 429 472 576 627 704 Registered Unemployment Rate % of civil active population 10.5 13.2 11.4 10.4 11.6 10.4 8.5 10.3 5.8 - women 10.1 12.3 11.3 9.7 10.8 10.4 7.9 10.7 6.7 Average Monthly Net Nominal Wage ROUemployee 2139138 1850572 2168914 2076885 2264133 2042069 1911542 1972759 2751848 USDollar equivalent per employee 98.6 85.3 100.0 95.7 104.4 94.1 88.1 90.9 126.9 Exchange rate, ROUUS$, ave 2000 21692.74 Contibution to Gross Value Added % of national total 13.5 13.1 13.2 9.7 9.2 12.1 12.5 16.5 Note: North East Region covers counties Bacau, Botosani, lasi, Neamt, Suceava, Vasin South East Region covers counties: Braila, Buzau, Constanta, Galati, Tulcea, Vrancea South Region covers counties: Arges, Calarasi, Dambovita, Giurgiu, lalomita, Prahova, Teleorman South West Region covers counties: Dolj, Gor, Mehedinti, Oh, Valces West Region covers counties: Arad, Caras-Severin, Hunedoara, Timis North West Region covers counties: Bihor, Bistrita-Nasaud, Cluj, Maramures, Satu Mare, Salaj Center Region covers counties: Alba, Brasov, Covasna, Harghita, Mures, Sibiu Bucharest area covers the Bucharest municipality  S28: Energy Balance 1996 1997 1998 1999 2000 Thousand tons of oil equivalent (10,000 Kcal/kg) Energy Sources - Total 53 941 5126.0 464.0 Production, o.w: 35,135 31,401 28,796 27,765 28,106 Coal ' 8,065 6,600 5,149 4,644 5,601 Natutal gas 13,764 11,908 11,195 11,192 10,968 Crude oil 6,700 6,604 6,415 6,244 6,157 Hydroelectric power 1,718 2,916 3,009 2,848 2,610 Other 4,888 3,373 3,028 2,837 2,770 Import, o.w: 18,806 19,163 15,148 10,186 10,925 Coal 2,773 3,429 2,495 1,584 1,706 Natural gas 5,654 4,030 3,773 2,538 2,712 Crude Oil 7,153 6,243 6,000 4,293 4,759 Oil products 2,981 4,018 2,712 1,530 1,470 Heavy fuel oil Electric power 193 89 102 95 67 Consumption 50,365 45,505 40,983 36,567 36,374 Population 10,618 9,673 9,412 8,757 8,433 Data Source IMF (NISES data) 1) without coking coal S29: Production, Domestic Consumption, Export and Import of Oil and Oil Products, 1980 - 1999 (thousand tons) Crude oil: Total Refined Product Domestic Total Total Domestic Production Import Supply Production Export Consumption 1985 11,092 14,626 25,718 24,987 9,689 15,298 1986 10,520 17,047 27,567 27,081 10,374 16,707 1987 9,846 21,366 31,212 30,250 11,829 18,421 1988 9,713 20,957 30,670 30,253 13,248 17,005 1989 9,573 21,809 31,382 29,821 13,375 16,446 1990 8,135 16,058 24,193 22,790 5,120 17,670 1991 6,941 8,634 15,575 15,293 2,496 12,797 1992 6,770 6,572 13,342 13,073 2,560 10,513 1993 6,830 7,581 13,771 13,111 2,676 10,453 1994 6,860 8,122 14,982 14,390 4,069 10,321 1995 6,951 8,657 15,608 13,796 4,690 9,106 1996 6,852 7,156 14,008 13,602 3,730 9,872 1997 6,750 6,245 12,995 13,166 2,882 10,284 1998 6,553 5,974 12,527 13,233 3,169 10,064 1999 6,154 4,294 10,448 10,303 1,957 8,346 Data Source IMF (Data provided by the Romanian authorities)  S30: Evolution of the Privatization Process number of privatized commercial companies (cumulated, by year end) 1993 % oftotal 1994 % of total 1995 % of total 1996 % of total 1997 %oftotal 1998 % of total 1999 % of total 2000 % of total 2001' % of total Total number of commercial companies 265 1000/0 860 100% 1479 100% 2719 100% 3832 100% 5073 100% 6420 100% 1341 100% 127 100% Small commercial companies 239 90% 711 83% 1031 . 70% 2010 74% 2930 76% 3823 75% 4576 71% 1070 80% na separately Medium commercial companies 24 9% 134 16% 404 27% 640 24% 797 21% 1066 21% 1546 24% 247 18% 108 85% Largecommercialcompanies 2 1% 15 2% 44 3% 69 3% 105 3% 184 4% 298 5% 24 2% 19 15% 1) In 2001 the number under the "medium commercial companies" heading includes the small companies equity of privatized commercial companies (bn lei and USS equivalent', cumulated, by year end) 1993 1994 1995 1996 1997 1998 1999 2000 2001 bn ROL mil. US$ bn ROL mil. USS bn ROL mil. USS bn ROL mil. USS bn ROL mil. USS bn ROL mil USS bn ROL mil. UISS bn ROL tul. US$ bn ROL mil. US$ Total number of commercial companies 64 50.1 574.65 325.2 3746.86 1453.4 6854.4 1698.7 10760 1341.1 19180 1751.4 - 29597 1621.3 12773 492.7 13089 414.2 Small commercial companies 32 24.7 120.91 68.4 354.42 137.5 778.52 192.9 1151 143.4 1705 155.7 2279 124.8 2477 95.6 na na Medium commercial companies 22 17.1 320.55 181.4 2176.35 844.2 3797.32 941.1 4835 602.7 6639 606.3 9930 544.0 5204 200.7 na na Large commercial companies ll 82 133.19 75.4 1216.1 471.7 2278.61 564.7 4774 595.0 10835 989.4 16388 897.7 5092 196.4 na na Exchange rate, leilUS$ 1276 1767 2578 4035 8023 10951 18255 25926 31597 1) Amounts in million US$ calculated using the exchange rate by end of each year (NBR data) Source. Romnaman priwatiition Team (data from Romanian authorities) and staffcalculations  S31: Private Sector Share in Total GDP and in Value Added by Branches 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 GDP, total 16.4 23.6 26.4 34.8 38.9 45.3 54.9 60.6 62.0 63.3 65.5 67.1 VA in Industry 5.7 9.2 11.8 17.4 23.3 29.9 38.5 42.1 45.9 52.8 55.7 VA in Agriculture and forestry 61.3 73.9 81.7 83.5 89.3 89.0 90.1 96.8 96.3 97.0 97.0 VA in Construction 1.9 16.1 21.0 26.8 51.6 57.8 69.3 76.6 78.6 80.6 80.9 VA in Services 2.0 16.8 18.8 29.3 39.1 58.1 66.7 71.5 75.5 75.9 76.6 Source: National Institute for Statistics: Evolution of the private Sector in Romania, 1990-2000; 2001:NIS Monthly Statistical Bulletin 1/2002 Private Sector Share in Investment, Construction, Export, Import, Retail Trade, Services delivered to Households, Employment 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Investment 4.3 8.1 15.6 26.0 36.8 39.3 39.7 35.4 40.5 50.5 58.3 63.8 Construction 3.1 4.5 13 30.2 55.5 70.8 75.5 70.5 74.7 85.6 89.3 88.4 Export (FOB) 0.2 15.9 27.5 27.9 40.3 41.2 51.4 54.8 48.9 65.7 65.7 66.7 Import (FOB) 0.4 16.1 32.8 27.2 39.2 45.4 48.3 52.4 48.3 72.1 70.1 Retail Trade .. .. .. 64.4 70.1 71.9 74.3 79.9 86.7 90.8 92.0 Services delivered to households .. .. 43.9 55.9 58.9 60.3 62.1 74.5 83.4 85.7 85.4 Average number of employees 0.7 2.4 5.3 10.9 17.5 22.1 23.0 27.8 34.6 39.4 44.5 Source: National Institute for Statistics: Evolution ofthe private Sector in Romania, 1990-2000; 2001:NIS Monthly Statistical Bulletin 1/2002  S32: General Government Consolidated Revenues and Expenditure billions of lel 1992 1993 1994 1995 1996 1997 1998 1999 2001 Total Revenue 2252 6700 15537 22642 31597 72386 111000 173634 351456 Current Revenue 2208 6652 15476 22580 31443 71802 110867 173337 351069 Tax Revenue 2019 6269 14042 20804 29257 67000 103992 164026 327124 Directtax,o.w: .. 4332 10028 14117 19523 44248 63467 100813 198043 Wage and Salaries 1 458 1325 3221 4583 6656 13946 15093 17834 37197 Profit Tax 319 754 1911 2811 3548 10780 11067 17037 22206 Social Secunty 691 2137 4602 5885 8186 17671 33425 57667 125615 Indirect Tax, o.w: .. 1937 4014 6687 9734 22752 40525 63213 129081 VAT .. 726 2268 3779 5359 11681 22493 32471 73604 Custom T-xes 87 269 562 1043 1674 3353 5741 7847 9038 Excises 744 775 1054 1485 4289 8431 16958 27293 Non-tax Revenue 189 383 1434 1776 2186 4802 6875 9311 23945 Capital Revenue 44 48 61 62 154 584 133 297 386 Total Expenditures 2530 6771 16643 25061 36810 85639 131123 193567 390317 Curent/3 2216 5786 13757 20840 30783 71859 115394 177835 352301 Goods and Services 852 2335 5924 9078 12873 26774 42738 68800 139311 Wages 453 1333 3236 4694 6568 12344 18671 26259 58189 Other 399 1002 2688 4384 6305 14430 24067 42542 81122 Interest Payments 15 188 672 989 1840 9659 17450 28796 45375 Subsidies and Transfers 1349 3262 7163 10773 16070 35426 55206 80293 170761 Subsidies and bonuses 785 1366 1913 2972 4729 6364 6211 9303 23434 Transfers 564 1895 5248 7801 11341 29062 48995 70936 147327 Capital 248 844 2729 3802 5682 12106 13530 15015 37015 Lending minus Repayments 66 141 157 419 345 1674 2198 717 1002 Deficit -278 -71 -1106 -2419 -5213 -13253 -20123 -19933 -38861 1) 1998 and 1999. actual values adjusted for collection ofheath contibutions. previously classified as wage taxes, health conributious now included with Social Security Some: IMF;2001:pelimina1y actual General Government Consolidated Revenues and Expenditure' percent of GDP Year 1992 1993 1994 1995 1996 1997 1998 1999 1999 2000 2001 Memo Item: GDP 6029 20035 49773.2 72135.5 108919.6 252925.44 368260.7 521735.6 545730.2 800308.1 1154126 Revised GDP Series Total Revenue 37.4 33.4 31.2 31.4 29.0 * 28.6 30.1 33.3 31.8 31.3 30.5 Current Revenue 36.6 33.2 31.1 31.3 28.9 28.4 30.1 332 31.8 31.2 30.4 Tax Revenue 33.5 31.3 28.2 28.8 26.9 26.5 28.2 31.4 30.1 29.4 28.3 Direct tax, o.w: .. 21.6 20.1 19.6 17.9 17.5 17.2 19.3 185 180 17.2 Wage and Salaries 7.6 6.6 6 5 6.4 6.1 5.5 4.1 3.4 3 3 3 4 3.2 Profit Tax 5.3 3.8 3.8 3.9 3.3 4.3 3.0 3.3 3.1 2.5 1.9 SocialSecurity2 11.5 10.7 9.2 8.2 7.5 7.0 9.1 11.1 10.6 10.8 10.9 Indirect Tax, o.w: .. 9.7 8.1 9.3 8.9 9.0 11.0 12.1 11.6 11 4 11.2 VAT .. 3.6 4.6 5.2 4 9 4.6 6.1 6.2 6.0 6.3 6.4 Custom Taxes 1.4 1.3 1.1 1.4 1.5 1.3 1.6 1.5 1.4 1.1 0.8 Excises .. 3.7 1.6 1.5 1.4 1.7 2.3 3.3 3.1 2.3 2 4 Non-tax Revenue 3.1 1.9 2.9 2.5 2.0 1.9 1.9 1.8 1.7 1.9 2.1 Capital Revenue 0.7 0.2 0.1 0.1 0.1 0.2 0.0 0.1 0.1 0.1 0 0 Total Expenditures 42.0 33.8 33.4 34.7 33 8 33.9 35.6 37.1 35.5 35.4 33 8 Current/3 36.8 28.9 27.6 28.9 28.3 28.4 31.3 34.1 32.6 31.9 30.5 Goods and Services 14.1 11.7 11.9 12.6 11.8 10.6 11.6 13.2 12.6 12.5 12.1 Wages 7.5 6.7 6.5 6.5 6.0 4.9 5.1 5.0 4.8 5.5 5.0 Other 6.6 5.0 5.4 6.1 5.8 5.7 6.5 8.2 7.8 7.0 7.0 Interest Payments2 0.2 0.9 1.4 1.4 1.7 3.8 4.7 5.5 5.3 4.9 3.9 Subsidies and Transfers' 22.4 16.3 14.4 14.9 14.8 14.0 15.0 15.4 14.7 14.5 14.8 Subsidies and bonuses 13.0 6.8 3.8 4.1 4.3 2.5 1.7 1.8 1.7 2.2 2.0 Transfers 9.4 9.5 10.5 10.8 10.4 11.5 13.3 13.6 13.0 12.3 12.8 Capital 4.1 4.2 5.5 5.3 5.2 4.8 3.7 2.9 2.8 3.1 3.2 Lending minus Repayments 1.1 0.7 0.3 0.6 0.3 0.7 0.6 0.1 0.1 0.4 0.1 Deficit -4.6 -04 -2.2 -3.4 -4.8 -5.2 -5.5 -3.8 -3.7 -4.1 -3.4 Souce: staffcalculanons, sing IMF data and NIS GDP data  S33: Monetary Developments 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Monetary aggregates Reserve Money (eop) 602.2 1292.1 3245.4 4690.8 7877.3 10618.0 19081.3 36153.9 50966.2 66223.5 (real annual percentage change) .. (45.8) 55.3 13.1 7.0 (46.4) 27.8 22.4 0.2 (7.7) Currency outside NBR ' 460.1 1124.0 2397.8 3951.4 5,902.4 9,657.9 12,287.9 18,817.5 27,545.8 38,387.8 Banks deposits with NBR 142.2 168.0 847.7 739.5 1974.9 960.1 6793.4 17336.4 23420 27835.7 Broad Money (e.o.p) 1856.1 4472.4 10648.7 18278.1 30334.6 62150.4 92529.9 134122.5 185060.0 270,512 (real annual percentage change) .. (39.1) 47.2 34.3 5.8 (18.4) 5.6 (6.4) (1.9) 3.9 Currency held by non-bank public 411.5 1048.7 2200.6 3760.5 5382.7 9200.1 11525 17371.6 25742 35,635 Leu denominated bank deposits 1112.5 2127.2 6090.0 10385.6 17866.3 35264.8 50803.4 66269.4 84462 119093 Foreign currency deposits 332.2 1296.4 2358.1 4132.0 7085.6 17680.4 30201.5 50481.5 74856 115784.1 MI ' (e.o.p) 1028.2 2231.3 4534.2 7083.2 11173.4 18731.8 22109.7 29668.9 46331 64308.6 Narrow Money' (e.o.p) 1523.9 3175.9 8290.6 14146.1 23249.0 44470.0 62328.4 83641.0 110204 154,728 Money multipliers(eop) Reserve money multiplier (M2) 3.1 3.5 3.3 3.9 3.9 5.9 4.8 3.7 3.6 4.1 Reserve money multiplier (M I) 1.7 1.7 1.4 1.5 1.4 1.8 1.2 0.8 0.9 1.0 Foreign currency deposits ratio to Broad money 17.9 29.0 22.1 22.6 23.4 28.4 32.6 37.6 40.4 42.8 Currency held by non-bank public as: % of broad money 22.2 23.4 20.7 20.6 17.7 14.8 12.5 13.0 13.9 13.2 % ofGDP 6.8 5.2 4.4 5.2 4.9 3.6 3.1 3.2 3.2 3.1 Memo: GDP , billion lei 6029.2 20035.8 49773.2 72135.5 108919.5 252925.4 373798.2 545730.2 800308.1 1154126.4 Inflation (Dec-on-Dec) 299.2 395.5 161.7 127.8 156.9 251.4 140.6 154.8 140.7 130.3 1) Vault cash + currency outside banks 2) MI = currency outside banks and demand deposits 3) Narrow Money = as M2 - Forex deposits Sorce: National Bank ofRomania  S34: Interest Rates on ROL Government Securities (new and rolled over issues) Interest bearing T-Bills Discount T-Bills Interest bearing T-Bills Discount T-Bills Interest bearing T-Bills Discount T-Bills Average Average n Average . Average n Average n Average nominal interest rate, nOmmal yield, p.a. % nomnal nterestrate, nOmml yield, p.a.% nomnal interest rate, nOmmal yield,p.a.% ROLbn p.a. ROLbn ROLbn p.a. ROLbn ROL bn% Jan-1996 Jan-1999 4458.7 703 Jan-2001 7629.0 50.1 Feb. 44.0 51.4 197.2 45.0 Feb 1075.0 88.6 5960.1 89.5 Feb 7248.4 51.3 Mar. 650.4 53.5 332.5 45.0 Mar 5291.5 78.3 Mar 10109.1 49.9 Apr. .339.0 54.3 Apr 8571.0 103.5 3557.8 77.0 Apr 7194.5 48.5 May 298.9 53.3 May 76573 106.9 May 6962.0 47.2 Jun. 52.3 61.1 Jun 1212.9 94.0 5736.6 100.3 Jun 3926.1 41.9 Jul. 748.7 59.8 Jul 490.4 75.0 1856.8 75.8 Jul 37.7 37.0 4693.0 35.9 Aug. 538.0 50.3 501.0 57.9 Aug 5820.4 68.8 Aug 3292.8 36.4 Sep. 650.5 50.2 Sep 18.2 59.0 3456.3 55.8 Sep 6189.5 37.8 Oct. 2629.0 50.5 Oct 3840.9 52.1 Oct 6500.0 36.4 Nov. 245.8 49.4 Nov 5063.5 58.9 Nov 4679.1 35.1 Dec. 899.4 63.3 Dec 2029.3 66.6 5127.2 76.0 Dec 5617.1 35.7 Jan-1997 3065.6 174.3 Jan-2000 713.5 66.2 6690.0 74.4 Feb. 250.3 186.0 Feb 9098.3 72.0 Mar. 899.4 144.0 Mar 3819.1 55.7 5511.3 58.0 Apr. 3626.8 172.1 Apr 965.0 49.0 7624.0 48.5 May 1254.1 92.5 May 10259.2 46.4 Jun. 899.4 48.0 Jun 2280.0 39.4 7693.4 45.6 Jul. 3626.8 44.0 Jul 2568.8 40.4 7552.7 41.8 Aug. 1132.0 60.0 1979.7 50.9 Aug 1214.8 43.2 4598.2 43.6 Sep. 3374.5 34.0 988.1 39.0 Sep 2687.0 43.3 5286.7 46.6 Oct. 898.4 45.8 Oct 1391.0 48.2 1824.1 50.2 Nov. 4500.0 80.0 763.3 87.0 Nov 1477.9 51.0 614.1 51.0 Dec. 2448.2 98.6 Dec 3871.3 49.7 Jan-1998 2849.0 104.4 Jan-2000 Feb. 2126.4 90.5 Feb Mar. 693.0 75.0 Mar Apr. 4484.6 69.5 Apr May 2911.0 47.9 May Jun 1781.1 41.3 Jun Jul. 6647.0 42.5 Jul Aug. 5014.2 47.9 Aug Sep 4360.7 50.3 Sep Oct 4999.8 57.3 Oct Nov 687.7 65.6 Nov Dec 4210.0 76.0 707.7 72.7 Dec  S35: Interest Rrates to the Non-bank Sector, 1991 - 2000 percent per anuum Commercial Banks ( average, relation Commercial Banks (average, with non-bank clients) Inflation relation with non-bank clients) Inflation Deposit Rate Lending Spread Monthly Deposit Rate Lending Spread Monthly Inflation Inflation, Inflation Inflation, (nominal) (nominal) (nominal) (%) annualized (nominal) (nominal) (nominal) (%) annualized 1991 1997 96.6 108.0 11.4 30.7 2384.9 Mar-91 8 11.2 3.2 6.6 115.3 Mar-97 73.8 91.4 17.6 23 31.4 Jun-91 8 13.2 5.2 2.0 26.8 Jun-97 33.6 49.8 16.2 3.3 47.6 Sep-91 8 13.2 5.2 7.3 132.9 Sep-97 34.1 55.6 21.5 4.5 696 Dec-91 11 29 18 13.7 366.8 Dec-97 1992 1998 Mar-92 23.5 28.5 5 10 213.8 Mar-98 40.6 62.1 21.5 3.8 56.4 Jun-92 42.6 55.6 13 4.3 65.7 Jun-98 34.3 53.2 18.9 1.3 16.8 Sep-92 39.9 49.9 10 10.1 217.3 Sep-98 33.5 47.9 14.4 2.7 37.7 Dec-92 32.9 43.6 10.7 13.2 342.7 Dec-98 42.3 58.9 16.6 2.2 29.8 1993 1999 Mar-93 31.71 46.5 14.8 9.2 187.5 Mar-99 46.1 61.9 15.8 6.4 110.5 Jun-93 33.5 48.0 14.5 5.5 90.1 Jun-99 49.1 72.6 23.5 5.1 81.6 Sep-93 33.81 53.8 20.0 10.9 246.1 Sep-99 44.0 , 68.6 24.6 3.2 45.9 Dec-93 42.5 86.4 43.9 7.4 135.5 Dec-99 41.3 62.0 20.7 2.9 40.9 1994 2000 Mar-94 60.23 105.7 45.5 8.3 160.3 Mar-00 41.2 63.7 22.5 1.8 23.9 Jun-94 59.3 105.9 46.6 2.6 36.1 Jun-00 29.9 50.5 20.6 2.8 39.3 Sep-94 55.75 81.0 25.3 3.9 58.3 Sep-00 26.4 45.4 19 28 39 3 Dec-94 49.49 61.8 12.3 2.1 28.3 Dec-00 28.6 49.2 20.6 2.5 34.5 1995 2001 Mar-95 39.42 52.1 12.7 0.9 11.4 Mar-00 29.5 50.3 20.8 2.0 26.8 Jun-95 37.55 48.1 10.5 1.3 16.8 Jun-00 27.7 46.8 19.08 1.6 21.0 Sep-95 34.09 46.1 12.0 1.6 21.0 Sep-00 24.2 42.8 18.65 1.9 25.3 Dec-95 32.4 47.5 15.1 3.7 54.6 Dec-00 23.4 40.6 17.17 2.2 29.8 1996 Mar-96 38 57.9 19.9 1.7 22.4 Jun-96 38.7 57.4 18.7 1.0 12.7 Sep-96 38.9 55.0 16.1 2.4 32.9 Dec-96 38.9 53.6 14.7 10.3 224.3 Source: National Bank ofRomania and staffcalculations  S36: Loan Classification Credits and Risk Provisions ROL billions 1. Exposure from loans granted to bank and non bank clients (and related interest) Dec.1994 Dec.1995 Dec.1996 Dec.1997 Dec.1998 Dec.1999 Dec.2000 Dec.2001 Credits granted, total (unadjusted) 9,385.3 18,453.4 26,743.7 40,084.8 66,913.7 59,446.5 71,538.1 109,142.9 Standard 3,349.0 3,096.4 2,969.2 5,464.4 8,366.3 14,611.5 66,984.5 104,736.9 Watch 2,794.2 4,973.1 7,397.0 8,546.6 10,601.9 13,543.1 833.8 697.0 Substandard 1,101.6 2,591.1 4,878.9 4,998.8 8,797.1 10,253.9 980.4 943.0 Doubtful 1,114.3 2,528.8 2,690.4 3,982.8 5,155.0 3,978.7 522.4 322.0 Loss 1,026.2 5,264.0 8,808.2 17,092.3 33,993.4 17,059.3 2,217.0 2,444.0 2. Provisioning ROL billions Total 131.0 1,785.0 2,513.6 7,312.8 16,207.6 10,056.0 2,642.2 2,813.8 Standard 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Watch 5.1 38.0 114.1 96.5 101.3 106.1 23.9 34.0 Substandard 13.6 72.6 152.7 272.3 218.3 348.8 193.5 1778 Doubtful 17.6 252.7 389.3 541.8 467.1 386.7 207.8 158.0 Loss 94.7 1,421.7 1,857.5 6,402.2 15,420.9 9,214.4 2,217.0 2,444.0 2. Romania: Solvency Ratio percent Dec.1995 13.8 Dec.1996 13.3 Dec.1997 13.6 Dec.1998 10.3 Dec.1999 17.9 Dec.2000 23.8 Dec.2001 26.9 Solvency Ratio is the ratio betwen banks' oun funds (on capital plus additional capital) and the risk iighted assets plus other risk weighted off balance sheet items Source: National Bank ofRomania  S37: Interbank Foreign Exchange Market Volume of Volue ofVolume of transactions Month transactions Exchange rate Month (mill.USD) Exchange rate (mill.USD) total change end of period average total change end of period average (leilUSD) leilUSD (lei/USD) lei/USD Jan-92 62.4 -104.2 198 194.91 Jan-96 349.3 -103.4 2,640 2,599.24 Feb. 87.9 25.5 198 197.60 Feb. 302.9 -46.4 2,859 2,773.71 Mar. 90.9 3.1 198 198.00 Mar. 314.0 11.1 2,937 2,872.60 Apr. 118.2 27.3 206 198.38 Apr. 341.7 27.7 2,913 2,911.14 May 78.3 -39.9 226 223.60 May 360.8 19.0 2,951 2,930.41 Jun. 2.4 -75.9 304 261.32 Jun. 354.6 -6.2 3,028 2,988.00 Jul. 54.3 51.8 365 349.39 Jul. 404.8 50.2 3,135 3,063.22 Aug. 25.2 -29.1 384 375.24 Aug. 371.0 -33.8 3,162 3,143.91 Sep. 55.1 29.9 426 404.14 Sep. 386.9 15.9 3,261 3,201.19 Oct. 67.7 12.6 430 430.00 Oct. 320.0 -66.9 3,375 3,295.74 Nov. 65.9 -1.8 430 430.00 Nov. 293.8 -26.2 3,591 3,478.19 Dec. 79.1 13.2 460 432.86 Dec. 466.4 172.6 4,035 3,733.89 Jan-93 42 -37.2 480 470.10 Jan-97 272.4 -194.0 5,932 4,963.40 Feb. 49 7.4 560 510.50 Feb. 342.9 70.5 7,744 6,895.40 Mar. 60 10.2 598 586.04 Mar. 488.0 145.1 6,996 7,235.90 Apr. 49 -10.0 615 603.71 Apr. 1,042.9 554.9 7,095 7,048.52 May 47 -2.1 650 621.43 May 858.8 -184.1 7,110 7,090.70 Jun. 61 13.6 735 688.50 Jun. 690.7 -168.1 7,032 7,172.29 Jul. 72 11.2 803 768.50 Jul. 881.2 190.5 7,354 7,164.26 Aug. 82 10.0 836 808.59 Aug. 759.5 -121.7 7,471 7,445.24 Sep. 83 1.0 910 870.00 Sep. 698.6 -60.9 7,613 7,528.82 Oct. 68 -15.2 1,036 984.60 Oct 889.1 190.5 7,741 7,701.91 Nov. 74 5.7 1,074 1,067.86 Nov. 789.8 -99.3 7,860 7,808.15 Dec. 112 38.1 1,276 1,140.80 Dec. 962.1 172.3 8,023 7,960.25 Jan-94 90 -22.1 1,450 1,387.16 Jan-98 947.8 -14.3 8,248 8,293.40 Feb. 87 -2.9 1,570 1,493.50 Feb. 849.9 -97.9 8,105 8,230.90 Mar. 102 15.3 1,650 1,601.30 Mar. 1,172.4 322.5 8,490 8,207.09 Apr. 144 42.1 1,659 1,670.71 Apr. 1,117.1 -55.4 8,345 8,379.60 May 142 -1.9 1,659 1,657.24 May 980.8 -136.2 8,511 8,477.25 Jun. 172 29.3 1,677 1,667.09 Jun. 933.3 -47.5 8,670 8,569.40 Jul. 193 21.3 1,690 1,685.71 Jul. 1,177.3 244.0 8,744 8,699.40 Aug. 428 235.0 1,701 1,687.83 Aug. 1,228.3 51.0 8,924 8,781.24 Sep. 330 -98.1 1,756 1,727.09 Sep 1513.6 285.3 9,162 9,041.19 Oct. 334 4.6 1,752 1,752.95 Oct 1768.4 254.8 9,592 9,380.68 Nov. 347 12.4 1,770 1,756.55 Nov 1719.6 -48.8 10,082 9,908.86 Dec. 494 147.0 1,767 1,773.90 Dec 2,220.2 500.6 10,951 10,528.59 Jan-95 302 -192.0 1780.0 1,776.00 Jan-99 1,644.0 -576.2 11,614 11,353.60 Feb. 366 64.3 1810.0 1,798.85 Feb 2302.7 658.7 12,774 12,271.00 Mar. 354 -12.1 1847.0 1,832.57 Mar 1,838.9 -463.8 14,925 14,053.52 Apr. 265 -89.0 1879.0 1,864.95 Apr 1,287.3 -551.6 14,992 14,792.62 May 342 77.0 1937.0 1,911.20 May 1,854.9 567.6 15,622 15,237.81 Jun. 343 0.7 1975.0 1,955.82 Jun 1,455.7 -399.2 15,840 15,756.50 Jul. 433 90.2 2018.0 1,994.29 Jul 1,692.6 236.9 16,037 15,920.95 Aug. 424 -9.0 2069.0 2,045.91 Aug 1,557.8 -134.8 16,220 16,100.95 Sep. 421 -2.4 2128.0 2,100.00 Sep 1,824.9 267.0 16,488 16,359.45 Oct. 423 1.7 2225.0 2,166.20 Oct 1,953.0 128.2 16,870 16,705.57 Nov. 384 -39.2 2547.0 2,395.27 Nov 2,501.9 548.9 17893 17,446.73 Dec. 453 68.8 2578.0 2,558.00 Dec 1,853.1 -648.8 18,255 17996.43 Source: National Bank of Romania  Interbank Foreign Exchange Market Volume of Months transactions Exchange rate (million USD) end of average total change period Jan-00 1753 -100 18465 18353 Feb 1669 -84 18892 18702 Mar 2091 423 19480 19207 Apr 1901 -190 20076 19759 May 1902 1 20697 20393 Jun 1638 -264 21358 21031 Jul 1732 94 21890 21601 Aug 1975 244 22973 22422 Sep 2473 498 24169 23602 Oct 2056 -417 24850 24538 Nov 1879 -177 25364 25103 Dec 1733 -146 25926 25604 Jan-01 1606 -127 26513 26243 Feb 1466 -140 27059 26815 Mar 1922 456 27566 27299 Apr 1894 -28 28214 27878 May 2021 127 28754 28493 Jun 1818 -204 29160 28952 Jul 2504 686 29623 29364 Aug 2279 -225 30044 29809 Sep 1868 -411 30465 30236 Oct 2807 939 31015 30786 Nov 2061 -746 31532 31299 Dec 2543 482 31597 31556 Source: National Bank ofRomania  S38: Balance of Payments, in million US$ NBR format (includes non-convertible currency) 1998 1999 2000 2001 Credit Debit Balance Credit Debit Balance Credit Debit Balance Credit Debit Balance 1. CURRENT ACCOUNT (A+B+C) 10,668 13,636 -2,968 10810 12,279 -1,469 13537 14,900 -1,363 15151 17,500 -2,349 A. Goods and Services 9,519 12,798 -3,279 9854 11,538 -1,684 12133 14,071 -1,938 13333 16,522 -3,189 Goods 8,302 10,927 -2,625 8487 9,744 -1,257 10366 12,050 -1,684 11385 14,354 -2,969 Services 1,217 1,871 -654 1367 1,794 -427 1767 2,021 -254 1948 2,168 -220 B. Incomes 263 705 -442 152 563 -411 325 610 -285 369 704 -335 C. Current transfers 886 133 753 804 178 626 1079 219 860 1449 274 1,175 2. CAPITAL AND FINANCIAL ACCOUNT (A+B) 7,686 4,963 2,723 5634 5132 502 5413 4180 1,233 5763 4459 1,304 A. Capital account 39 0 39 46 1 45 37 1 36 105 2 103 a. Capital transfers 39 0 39 46 1 45 37 1 36 105 2 103 39 0 39 44 0 44 35 0 35 95 0 95 0 0 0 2 1 1 2 1 1 10 2 8 B. Financial account 7,647 4,963 2,684 5588 5,131 457 5376 4,179 1,197 5658 4,457 1,201 a. Direct investment 2,063 23 2,040 1,102 77 1,025 1,122 71 1,051 1,281 127 1,154 11 2 9 8 24 -16 45 34 11 48 31 17 2,052 21 2,031 1,094 53 1,041 1,077 37 1,040 1,233 96 1,137 b. Portfolio investment 510 380 130 213 928 -715 472 371 101 1,149 566 583 32 31 1 48 39 9 39 11 28 17 25 -8 478 349 129 165 889 -724 433 360 73 1,132 541 591 c. Other capital investment 4,047 4,345 -298 3,523 3,182 341 3,693 2,717 976 3,213 2,279 934 1,165 954 211 916 672 244 672 1,078 -406 415 412 3 2,882 3,391 -509 2,607 2,510 97 3,021 1,639 1,382 2,798 1,867 931 d. In transit accounts 11 39 -28 4 21 -17 50 43 7 8 0 8 e. Barter and clearing accounts 5 8 -3 8 12 -4 25 35 -10 7 1 6 f. Reserve assets (NBR) 1,011 168 843 738 911 -173 14 942 -928 0 1,484 -1,484 0 0 0 0 5 -5 0 2 -2 0 5 -5 102 0 102 1 10 -9 14 5 9 0 6 -6 0 0 0 0 0 0 0 0 0 0 0 0 909 168 741 737 896 -159 0 935 -935 0 1,473 -1,473 0 0 0 0 0 0 0 0 0 0 0 0 3. NET ERRORS AND OMISSIONS 245 0 245 967 0 967 130 0 130 1,045 0 1,045 Source: National Bank ofRomania  S39: External Debt by Creditor millions of USD; end of period Period Medium and long term external debt By creditor Total of which: Multilateral Bilateral Private banks Debt securities ING Bank Total of which: Total Total Total of which: Credit Schroeder IMF EU IBRD EBRD CS First CS First Nomura Merrill Deutsche Salomon Boston Boston Securities Lynch Bank AG Mith Germany Switzerland Japan GB Barney 1990 230.0 ..... ................ 1991 1143.0 836 809 .. 3 .. 215 92 .. .. .. .. 1992 2479.0 1733 1032 454 215 425 321 .. .. .. .. 1993 3357.0 2059.0 1041.0 515.0 403.0 63.0 1298.0 .. .. .. .. .. 1994 4596.8 2715.4 1312.8 549.6 667.5 136.7 881.3 853.6 .. .. .. .. 1995 5482.1 2786.6 1039.2 652.6 794.2 203.1 1204.7 1175.8 64.7 .. .. .. 1996 7208.9 2720.5 650.9 632.6 1022.0 278.3 1301.3 1550.0 1224.8 .. .. 719.1 275.0 1997 85843 3391.9 641.8 641.8 1391.7 464.9 1164.9 1694.0 1755.9 335.3 .. 631.4 495.4 1998 8770.7 3688.5 538.6 461.3 1469.4 674.4 1142.5 1673.6 1880.1 357.8 .. 720.8 497.7 1999 8435.5 3874.8 458.2 206.7 1706.3 752.9 1010.8 1488.3 1158.0 309.3 .. 293.2 215.1 2000 10240.5 4340.6 452.8 209.3 1997.7 788.2 928.9 1780.7 1154.3 285.4 .. 261.2 58.0 139.5 139.5 2001 11411.5 4538.8 388.8 198.5 2025.9 803.6 870.1 1871.4 1702.7 270.7 529.4 .. 28.9 132.3 264.7 1) The tables does not cover the 'Supplier credits and "Otherpriate* Source: National Bank ofRomania: Annual Report 1999 and Bulletin 1/2002  S40: Reserve Assets and External Debt by Debtor millions of USD, end of period Total Debt Medium and long term external debt Short term Reserve of which: Liquid incl.ST By debtor External Foreign convertible monetary foreign assets Total Public Publicly P & PG Private Debt Assets currency gold in months of debt guaranteed as % of commercial debt (BK System) Imports debt total (non-guaranteed) 1990 1376.0 527.9 850.1 1140.0 230.0 .. .. .. 910 1991 1503.7 656.6 795.0 1.3 2131.0 1143.0 1027.0 .. 89.9 116 988 1992 1594.1 826.2 757.0 1.5 3240.0 2479.0 2010.0 309.0 93.5 160 761 1993 1909.5 993.7 913.9 2.0 4249.0 3357.0 2370.0 741.0 92.7 246.0 892 1994 3092.2 2030.5 1006.2 3.1 5562.7 4596.7 2982.7 1264.7 92.4 349.4 966 1995 2625.1 1522.9 1046.1 1.9 6482.1 5482.1 3388.4 1560.1 90.3 533.6 1000 1996 3144.3 2098.7 1041.5 2.1 8344.9 7208.9 4326.0 1848.4 85.6 1034.5 1136.0 1997 4670.9 3699.5 867.5 3.6 9502.3 8584.3 4815.9 2037.8 79.8 1730.6 918.0 1998 3791.7 2866.2 924.3 2.7 9899.5 9322.5 4814.4 2152.4 74.7 2355.7 577.0 1999 3653.6 2676.9 966.6 2.8 9155.7 8770.7 3948.4 2270.9 70.9 2551.4 385.0 2000 4842.2 3921.2 920.0 3.3 10616.1 10240.5 4653.6 2299.7 67.9 3287.2 375.6 2001 6380.6 5435.1 938.7 3.9 11815.1 11411.5 4982.3 2690.3 67.2 3738.9 403.6 Source: The 1999 Annual Report of the National Bank ofRomania; NBR Monthly Bulletin 1/2002   IVING Report-No.: 24756 RO Type: SR