Report No: ACS20151 . Republic of Moldova Moldova Public Finance Review Towards More Efficient and More Sustainable Public Finances . August 2016 . GMF03 EUROPE AND CENTRAL ASIA . . . Standard Disclaimer: . This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: . The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750- 4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Report No. ACS20151 Moldova Public Finance Review Towards More Efficient and More Sustainable Public Finances August 2016 Macroeconomics and Fiscal Management Global Practice Europe and Central Asia Region Document of the World Bank CURRENCY AND EQUIVALENT UNITS (as of August 30, 2016) Currency Unit = Moldovan Lei (MDL) US$1=19.77 MDL GOVERNMENT FISCAL YEAR January 1 – December 31 ACRONYMS AND ABBREVIATIONS ASYCUDA Automated System for Customs' NHIC National Health Insurance Company Data BEEPs Business Environment and OECD Organization for Economic Cooperation Enterprise Performance Survey and Development CIT Corporate Income Tax OLS Ordinary Least Squares DSA Debt Sustainability Analysis PAYG Pay-As-You-Go ECA Europe and Central Asia PISA Program for International Student Assessment EU European Union PIT Personal Income Tax FDI Foreign Direct Investments PPP Purchasing Power Parity FMIS Financial Management Integrated PSP Public Sector Performance System GDP Gross Domestic Product SSC Social Security Contributions HCC Health Care Contributions STS State Tax Service IMF International Monetary Fund VAT Value Added Tax GFS Government Finance Statistics VET Vocational Education and Training WEO World Economic Outlook WDI World Development Indicators MIC Middle-Income Countries Vice President: Cyril Muller Country Director: Qimiao Fan Practice Director: Satu Kahkonen Practice Manager: Ivailo Izvorski Task Team Leader: Ruslan Piontkivsky Contents Acknowledgements .................................................................................................................................................................. i Executive Summary .................................................................................................................................................................. i Chapter I. Fiscal Sustainability and Fiscal Stance....................................................................................................... 1 Macro Developments and Fiscal Pressures ......................................................................................................... 1 Fiscal Sustainability ...................................................................................................................................................... 2 Fiscal Policy ...................................................................................................................................................................... 5 Moldova’s Fiscal Rule ................................................................................................................................................... 6 Fiscal Policy and Private Savings ............................................................................................................................. 7 Recommendations ......................................................................................................................................................... 9 Spotlight: Good Practices in Fiscal Consolidation .................................................................................................... 10 Lessons from Latvia, Estonia, and Lithuania .................................................................................................... 10 Chapter II. Spending ............................................................................................................................................................. 13 Government Spending over Time ......................................................................................................................... 13 Moldova’s Spending in Comparative Perspective .......................................................................................... 14 Economic Classification ............................................................................................................................................. 18 Functional Classification ........................................................................................................................................... 24 Efficiency of Public Spending .................................................................................................................................. 29 Recommendations ....................................................................................................................................................... 31 Chapter III. Revenues ........................................................................................................................................................... 33 Trends and Composition ........................................................................................................................................... 33 Policy and Administration ........................................................................................................................................ 37 Recommendations ....................................................................................................................................................... 40 Annexes ..................................................................................................................................................................................... 42 References ................................................................................................................................................................................ 46 List of Tables Table 1: Summary of Main Recommendations.................................................................................................................... ii Table 2: Econometric Estimation of Private Savings Determinants in Moldova: Results ................................. 8 Table 3: Spending Measures, Baltic Countries, 2009 –10..................................................................................... 11 Table 4: Changes in Annual Wages, Baltic Countries, 2009 ................................................................................ 11 Table 5: Pension Reforms, Baltic Countries, 2008–10 ........................................................................................... 11 Table 6: Revenue Measures, Baltic Countries, 2008–10 ....................................................................................... 12 Table 7: PSP Indicators, Moldova and Peer Countries ................................................................................................... 29 Table 8: Tax Preferences and Main Tax Exemptions ...................................................................................................... 37 List of Figures Figure 1: Real GDP Growth in 2000–15 ................................................................................................................................. 1 Figure 2: General Government Spending in 2014–15 ..................................................................................................... 1 Figure 3: General Government Balance, 2000–14 Average .......................................................................................... 2 Figure 4: General Government Debt, 2000–14 ................................................................................................................... 2 Figure 5: Moldova’s Public Debt and Guarantees, 2015 .................................................................................................. 3 Figure 6: Sources of Moldova’s Deficit Financing .............................................................................................................. 3 Figure 7: Headline and Cyclically Adjusted Fiscal Balances, 2000–15 ...................................................................... 5 Figure 8: Output Gap and Fiscal Impulse, 2000–15 .......................................................................................................... 5 Figure 9: Actual and Simulated Fiscal Balance, 2000–15 ............................................................................................... 6 Figure 10: Long Term Change in Spending, 2000–09 to 2010–15, Economic Classification ......................... 13 Figure 11: Long Term Change in Outlays, 2000–09 to 2010–15, Functional Classification........................... 13 Figure 12: Short Term Change in Spending from 2010 to 2015, Economic Classification ............................. 14 Figure 13: Short Term Change in Outlays from 2010 to 2015, Functional Classification ............................... 14 Figure 14: General Government Spending, Moldova and Comparators ................................................................. 15 Figure 15: General Government Spending, Moldova, Peers, and Selected ECA Countries .............................. 15 Figure 16: General Government Spending and GDP per capita, Moldova and Other Countries ................... 15 Figure 17: General Government Spending, and HDI, Moldova and Other Countries ....................................... 15 Figure 18: Income and Government Effectiveness, Moldova and Other Countries ........................................... 16 Figure 19: General Government Spending and Government Effectiveness, Moldova and Other Countries ............................................................................................................................................................................................................... 16 Figure 20: Average General Government Spending, Moldova and Peers, 2000–14 ......................................... 16 Figure 21: General Government Spending, Moldova and Selected Countries ...................................................... 16 Figure 22: Average General Government Spending: Economic Calcification, Moldova and Peers, 2000– 14.......................................................................................................................................................................................................... 17 Figure 23: Spending by Function, Moldova and Selected Countries ........................................................................ 17 Figure 24: Number of Public Employees and the Wage Bill ........................................................................................ 18 Figure 25: Structure of the Public Wage Bill ...................................................................................................................... 18 Figure 26: Budget Spending, Wages, and GDP, 2000–15 .............................................................................................. 18 Figure 27: Staff Positions in Public Employment, 2009 and 2015 ............................................................................ 18 Figure 28: Where the Public Wage Bill Has Grown the Most ...................................................................................... 19 Figure 29: Structure of Spending on Goods and Services ............................................................................................. 21 Figure 30: Total and Goods and Services Spending, 2000–15 .................................................................................... 21 Figure 31: Positive Contributions to Growth of Goods and Services Spending ................................................... 21 Figure 32: Negative Contributions to Growth of Goods and Services Spending ................................................. 21 Figure 33: Evolution of Spending on Transfers to the Population, 2001–15 ....................................................... 22 Figure 34: Structure of Spending on Transfers to the Population ............................................................................ 22 Figure 35: Evolution of Capital Spending, 2009–14........................................................................................................ 22 Figure 36: Structure of Capital Spending............................................................................................................................. 22 Figure 37: Contributions to Average Growth of Capital Spending, 2009–14 ....................................................... 23 Figure 38: Pension System Dependency Ratio, 2014–70 .............................................................................................. 25 Figure 39: Structure of Social Care and Social Insurance Spending ......................................................................... 25 Figure 40: Enrollment in Education Institutions.............................................................................................................. 26 Figure 41: Educational Spending by Function................................................................................................................... 26 Figure 42: Pupil/Teacher Ratios ............................................................................................................................................. 27 Figure 43: Educational Spending by Economic Classification .................................................................................... 27 Figure 44: Health Spending by Economic Classification, 2009–14........................................................................... 28 Figure 45: NHIC Spending by Economic Classification .................................................................................................. 28 Figure 46: Theoretical Production Possibility Frontier................................................................................................. 30 Figure 47: Data Envelope Analysis, VRS input-oriented ............................................................................................... 30 Figure 48: Long Term Changes in Public Revenues, 2000–09 to 2010–15 ........................................................... 33 Figure 49: Short Term Changes in Public Revenues, 2010 to 2015 ......................................................................... 33 Figure 50: Average General Government Revenues including Grants, 2010–14................................................ 34 Figure 51: General Government Revenue and GDP per Capita, Moldova and Other Countries.................... 34 Figure 52: General Government Revenue Structure, Moldova and Peers, 2013 ................................................. 35 Figure 53: Average Revenues from Goods and Services Taxes and CIT, Moldova and Peers, 2010–14 ... 36 Figure 54: CIT Rate, Moldova and Regional Peers, 2013 .............................................................................................. 36 Figure 55: Average Revenue Collected by Type of Payer, 2008–14, ........................................................................ 36 Figure 56: Average Revenue Collected by Sector of the Economy, 2008–14, ...................................................... 36 Figure 57: Administrative Burden of Tax Compliance, Moldova and ECA, 2006–15 ........................................ 39 Figure 58: Companies Paying Bribes to the Tax Administration, 2013 .................................................................. 39 List of Boxes Box 1: Lessons from Ukraine’s Debt Crises .......................................................................................................................... 4 Box 2: Types of Fiscal Rules ........................................................................................................................................................ 6 Box 3: Why the Fiscal Policy-Savings Relationship Matters: Testing Ricardian Equivalence ......................... 7 Box 4: Saving-Investment Balance in Moldova ................................................................................................................... 8 Box 5: The Public Sector Performance (PSP) Indicator ................................................................................................. 29 Box 6: Tax Reform in the Slovak Republic ......................................................................................................................... 40 _______________________ Acknowledgements This report was prepared by a World Bank team led by Ruslan Piontkivsky and comprising Marcel Chistruga, Anastasia Golovach, Kamer Karakurum-Ozdemir, Maryna Sidarenka, and Ayberk Yilmaz. The report also benefited from comments and suggestions received from Oleksii Balabushko, Lucia Casap, Barbara Cunha, Irina Guban, Andrea Guedes, Ewa Joanna Korczyc, and Iryna Shcherbyna. Viorica Strah provided invaluable administrative assistance. The team has benefited from the guidance of Qimiao Fan (Country Director), Ivailo Izvorski (Practice Manager), Alexander Kremer (Country Manager), and Faruk Khan (Program Leader). The team is grateful to Government officials from the Ministry of Finance, National Bank of Moldova, National Statistics Bureau, and other branches of Government and development partners of Moldova for their cooperation in conducting the analysis. Executive Summary This report1 argues that Moldova’s government could reduce fiscal risks to the economy by reducing the size of spending and improving its efficiency, making the tax system simpler, with fewer tax preferences, and strengthening the tax administration. After years of robust performance, in 2015–16 Moldova’s public finances came under pressure. Fraud in the banking sector and economic recession have pushed up public debt, reducing fiscal space and undermining investor confidence. Grants from donors have fallen. Less revenue has forced the government into an ad hoc spending adjustment, with an abrupt reversal of the recent welcome trend of higher capital spending. The government could address these challenges along three dimensions: fiscal stance and sustainability, spending, and revenues. First and foremost, nonetheless, it should concentrate on gradually reducing current spending. As experience in other countries has demonstrated, fiscal consolidation based on spending cuts in a context and circumstances similar to these in Moldova, may yield better results than one based on tax increases. The first dimension to consider is safeguarding fiscal sustainability. Moldova managed to reduce its debt from over 80 percent of GDP in 2000 to just over 20 percent in 2008. After the global crisis, debt hovered above 30 percent of GDP before called public guarantees for defrauded bank depositors escalated it to more than 45 percent in 2015. While this level of public debt is broadly in line with the average for low middle-income countries, fiscal sustainability may be challenged by the lack of market access, limited fiscal deposits, and the risk that further contingent liabilities in the financial sector may emerge and need to be covered. It is recommended that the authorities promote development of the domestic government securities market and work hard to limit the scope of contingent liabilities. In recent years, Moldova’s fiscal policy has been a-cyclical, that is, neutral relative to the economic cycle. However, the fiscal rule the government introduced in 2014 -- stipulating that the fiscal deficit, excluding grants, should not exceed 2.5 percent of GDP, although the deficit ceiling could be exceeded by the amounts of externally financed investment projects -- would neither improve fiscal sustainability nor promote counter-cyclicality. A retroactive analysis applying Moldova’s rule to 2000–15 finds that the fiscal balance and debt would have been worse than what actually happened in terms of both sustainability and counter-cyclicality. It is therefore recommended that the authorities improve the fiscal rule to make it a more effective anchor for sustainability and output stabilization. The second dimension is reducing the size and improving the efficiency of spending. At 37.8 percent of GDP on average for 2000–14, Moldova’s government spent more than countries with similar per capita income, degree of human development, and government efficiency. Improving governance is an important prerequisite for possible expansion of expenditures in the future. Today Moldova spends more on public consumption than comparators and is a clear international outlier on education outlays. Starting in 2010, following a decade of increasing expenditures, the government began to reduce spending, most notably on education and social protection, as it created more fiscal space for capital spending. However, it needs to do more to contain current expenditures and to secure critical public investments. Today, Moldova’s public spending is far from efficient: according to the Public Sector Performance indicator2 , more efficient regional comparators achieve similar outcomes using 40 percent lower revenues. Spending on education and health is particularly high, but even beyond these 1 This report builds on and logically continues a series of analytical reports under the Moldova Programmatic Public Expenditure Reviews (PERs). Recent PERs have focused on public investment management, intergovernmental fiscal relations, and public support to agriculture. 2 Public Sector Performance indicator compares countries on a variety of socioeconomic indicators and relates public spending to the outcomes of public policies, see Box 5 for details. i sectors there is room for efficiency gains. Pension system is becoming socially unsustainable. The efficiency of spending could be improved by reforming the pension PAYG system, (including raising the retirement age), better targeting social assistance, optimizing the hospital, primary and secondary school networks, reducing public employment, strengthening procurement practices, and revising how the government invests its capital. The third dimension is making the tax system simpler. Moldova collects more revenues than peer countries but also depends on external grants. As in most other Eastern European countries, the revenue structure is skewed toward taxes on goods and services (indirect taxes). Moldova’s tax revenues have been declining as tax exemptions proliferated. While the tax administration has improved, continuing deficiencies in capacity and governance cannot deal with the problem of high informality. Moldova could make the tax system simpler, more efficient, and revenue-enhancing by reducing tax preferences, increasing the nontaxable amount of the personal income tax, improving property valuation, increasing excises, improving tax administration, reducing compliance costs, and simplifying the tax structure. Most importantly, the government needs to deal with tax expenditures, since tax initiatives over the last 15 years resulted in the adoption of a wide range of reduced tax rates and tax exemptions, with significant costs for the budget. In the short term, additional revenues might supplement expenditure cuts to safeguard fiscal sustainability. In the longer term, though, Moldova would need to find substitutes for external grants, so that they gradually become a relatively smaller source of revenues. A brief summary of the recommendations along the three dimensions is presented in Table 1. The details are presented in the main body of the report. Table 1: Summary of Main Recommendations Area Recommendation Safeguarding fiscal  Carefully monitor contingent fiscal liabilities and come up with clear, more sustainability stringent rules for providing government guarantees. (see Chapter 1)  Make the fiscal rule more effective.  Develop the government securities market. Reducing the size and  Reform the pension PAYG system to preserve fiscal and social sustainability. improving the efficiency  Better target social assistance. of spending  Continue optimizing the primary and secondary education network and (see Chapter 2) expand formula financing to other education sub-sectors.  Optimize the hospital network.  Reduce the number of public employees.  Reinforce procurement practices.  Better regulate public capital investment. Making the tax system  Reduce tax expenditures. simpler  Make the PIT more progressive by increasing the nontaxable amount to the (see Chapter 3) minimum living standard.  Improve property valuation to collect more from real estate taxes.  Raise excise tax rates over the medium term.  Improve tax administration, reduce compliance costs, and simplify the tax structure. ii Chapter I. Fiscal Sustainability and Fiscal Stance Moldova is currently experiencing fiscal pressures despite low deficits over the last 15 years that resulted in a much reduced level of public debt. Policy slippages led to a recent decline in official financing, while guarantees issued in 2014–15 to cover depositors in three failed banks pushed debt past 45 percent of GDP and undermined investor confidence. Fiscal sustainability may be further challenged by the lack of market access, limited fiscal deposits, and the risks that further contingent liabilities from the financial sector may need to be covered. Meanwhile, Moldova’s fiscal policy has been mostly neutral to the economic cycle for the last 15 years. However, the fiscal rule introduced in 2014 would neither improve fiscal sustainability nor promote counter-cyclicality. The authorities should carefully watch potential contingent fiscal liabilities and ensure that the fiscal rule becomes an anchor for sustainability and output stabilization. Macro Developments and Fiscal Pressures 1. After more than a decade of mostly high growth, Moldova’s economy fell into recession in 2009 and again in 2015. Since the early 2000s, the economy grew by 5 percent a year on average; the poverty rate as nationally defined dropped from 26 percent in 2007 to 11 percent in 2014; and growth in the consumption of the bottom 40 percent of the welfare distribution was consistently higher than the average. However, in 2015 an adverse external environment, summer drought, and a banking crisis put the brakes on economic growth (Figure 1). The deep recession of Moldova’s eastern partners and the trade restrictions Russia imposed on imports from Moldova slashed remittances from Russia and halved exports to that country, and exchange rate depreciation pushed up inflation. Hot weather and low precipitation caused a double-digit decline in agricultural GDP. And a banking crisis brought on by massive fraud in three banks lowered confidence in the banking sector, leading to significant interest rate increases and reduced credit to the private sector. Among the monetary and fiscal costs of resolution of the insolvent banks (equivalent to 12 percent of GDP) were higher public debt, lower foreign exchange reserves, damage to business confidence, and fewer macroeconomic buffers to sustain economic shocks. The economy contracted by 0.5 percent in 2015 (see Annex 1 for the main macro indicators). Figure 1: Real GDP Growth in 2000–15 Figure 2: General Government Spending in (Percent) 2014–15 (Percent Real Change) Source: World Development Indicators and Moldovan Source: Moldovan authorities, World Bank calculations. authorities. Note: G&S stands for Goods and Services. 1 2. Since 2015 the contraction in the economy and a shortfall in external financing has been putting heavier fiscal pressure on the government. Moldova also received only a fraction of planned external grants and financing due to political instability and delays in economic reforms. Moreover, the cost of debt service shot up as currency depreciation and monetary tightening pushed up real interest rates. To adapt to the changed circumstances, since mid-2015 the government has been rationing expenditures. Some categories, notably capital outlays, have been cut dramatically as the government protected the public wage bill and social protection spending (Figure 2). In the medium term, these pressures are likely to be aggravated by adverse demographics, social spending commitments, and recognition of contingent liabilities in the banking sector that are estimated at 12 percent of GDP. The challenge for the government now is to adjust the fiscal stance without unduly slowing growth; maintain fiscal sustainability; and ameliorate the negative impact on the living standards of the poor and vulnerable. Fiscal Sustainability 3. For the past 15 years, low deficits helped Moldova keep its public debt sustainable. For 2000–143 the deficit averaged 1.3 percent of GDP, lower than the average for Europe and Central Asia (ECA) and significantly lower than for other middle-income countries with per-capita income similar to Moldova’s (Figure 3). Likewise, Moldova’s public debt was on average lower than in MICs and close to the ECA average (Figure 4). Moldova managed to reduce its debt from over 80 percent of GDP in 2000 to just over 20 percent before the 2008–09 financial crisis. After the crisis, debt rose to 30 percent of GDP before the payment on public guarantees to bank depositors elevated it to above 45 percent in 2015. Figure 3: General Government Balance, Figure 4: General Government Debt, 2000–14 2000–14 Average (Percent of GDP) (Percent of GDP) Source: Find My Friends, using IMF WEO. Source: Find My Friends, using IMF WEO. 4. According to the Joint IMF-World Bank Debt Sustainability Analysis (DSA) Moldova’s debt remains sustainable, but it is sensitive to an increase in contingent liabilities. While sensitive to exchange rate depreciation and a sudden increase in other debt-creating flows, Moldova’s debt dynamics are in general projected to remain sustainable in both the baseline and alternative scenarios. The DSA concludes that pursuing prudent fiscal policy and advancing structural reforms continue to be necessary to ensure debt sustainability (IMF 2015). 3All cross country comparisons in this report are based on data up to 2014 due to data availability for a broad range of peer countries. 2 5. The government’s heavy reliance on foreign borrowing may also be risky. Foreign financing was very important for deficit financing in recent years. In 2010–15, on average 64 percent of the deficit was financed by net external borrowing (Figure 5). While official long-term financing dominates the composition of public external debt (Figure 6) given the lack of access to private international capital markets and the thin domestic securities markets,4 a large part of official financing is in the form of budget support that depends on sustained progress in advancing reform. Political instability and delays in reform implementation could substantially reduce the available financing. This could further reduce the government’s ability to finance planned spending, as happened in 2015 (World Bank 2016). Figure 5: Moldova’s Public Debt and Figure 6: Sources of Moldova’s Deficit Guarantees, 2015 Financing (Percent of Total) (Percent of GDP) Source: Moldova authorities. Source: Moldovan authorities, World Bank calculations. 6. Fiscal sustainability is being challenged by risks of additional contingent liabilities and low growth, low labor market participation, and negative demographic changes. Given the very low formal employment rate, driven by an unsustainably low retirement age and a persistent fall in the labor force participation of both men and women, by 2015 there were 1.4 inactive adults per active one (aged 15+). Combined with an increasing share of elderly, most of whom are inactive, Moldova is looking at an exceptionally high economic dependency ratio, which persists through different scenarios for labor force participation. Current demographic projections estimate that the pension system dependency ratio (number of pensioners per contributor) will rise from 77 percent in 2014 to a peak of 108 percent in 2055 because the future workforce (the contributors) will be smaller than the current workforce. Finally, the risks of additional contingent liabilities, especially in the financial sector, are not negligible because the root causes of the previous banking crisis have not yet been fully dealt with. Following the public disclosure of the banking fraud, the authorities began to take the first steps to address the governance problems throughout the financial sector. However, more needs to be done to strengthen the independence, powers, and supervision capacity of the regulators (NBM and NCFM), and to ensure shareholder transparency and good corporate governance in financial institutions. The recent example of Ukraine illustrates how an accumulation of quasifiscal liabilities can make public debt unsustainable (Box 1). 4In 2015, the stock of government securities denominated in local currency was only 4.2 percent of GDP. Development of the government securities market is one of the priorities in the National Development Strategy, “Moldova 2020”, http://particip.gov.md/public/files/strategia/Moldova_2020_proiect.pdf 3 Box 1: Lessons from Ukraine’s Debt Crises Figure B1.1: Public and Publicly Guaranteed Debt in Moldova and Ukraine, 2014–15 Percent Ukraine’s example show how quickly public debt may become unsustainable. In 2007, (Percent of GDP) Ukraine’s fiscal deficit was 2 percent of GDP and its public debt amounted to 13 percent. In 2014, the overall deficit (including the deficit of state-owned gas supply company Naftogaz) reached 10 percent of GDP, public and guaranteed debt exceeded 70 percent, and in 2015 payments of debt interest reached 4.5 percent of GDP. In November 2015 Ukraine had to restructure external public debt with private creditors. Public and publicly- guaranteed debt stabilized at about 80 percent of GDP in 2015 (Figure B1.1). It continues to be a major source of vulnerability for Ukraine. Source: Moldova and Ukraine authorities. A combination of procyclical fiscal policies, economic crises with currency depreciation, and a proliferation of quasi-fiscal operations led to unsustainability:  Pre-2008 Ukrainian fiscal policy was procyclical and budget spending grew rapidly due to high increases in wages, subsidies, and transfers. While the headline fiscal deficit stayed relatively low at 2 percent, in 2007 the cyclically adjusted deficit was 3.5 percent.  Ukraine then experienced two deep recessions, one in 2009 and one in 2014–15, both accompanied by steep currency depreciations. Real GDP declined 15 percent in 2009, and 17 percent in 2014–15 (figure B1.2). In 2015, Ukraine’s economy was 17 percent smaller than in 2007. Likewise, as a result of several waves of currency depreciation, the Ukrainian hryvnya lost 80 percent of its value between 2007 and 2015. Since a large share of Ukraine’s public debt was denominated in foreign currency, the combination of recession with depreciation forced a rapid rise in the debt-to- GDP ratio.  Coverage of the deficits of Naftogaz and recapitalization of the state-owned banks aggravated fiscal unsustainability (Figure B1.3B1.3). The Naftogaz structural deficit grew rapidly because the authorities were reluctant to fully align gas tariffs with the surge in imported gas prices. Currency depreciations increased the required tariff adjustments denominated in local currency. In 2014 the Naftogaz deficit reached 5.6 percent of GDP.  Meanwhile, during both recessions the financial sector and the government stepped in to recapitalize state-owned and systemically important banks. Public spending on bank recapitalization and the Deposit Guarantee Fund (DGF) peaked at 2.4 percent of GDP in 2015. These additional needs were financed through new public borrowings. Figure B1.2: GDP and Exchange Rate Dynamics Figure B1.3: Ukraine’s General Fiscal Balance (2007=100) (Percent of GDP) (Percent) 4 Source: Ukrainian Statistics Service, National Bank of Source: Ministry of Finance, IMF. Ukraine. Fiscal Policy 7. Moldova’s fiscal policy has been acyclical for the last 15 years. During the boom years of the early and mid-2000s, Moldova kept public finances almost balanced or ran a fiscal surplus. When the economy dived into a 6 percent recession during the 2008–09 global crisis, the authorities swiftly responded by raising the headline fiscal deficit. As a result, the underlying structural fiscal deficits have been much less volatile, moving in the 1–3 percent range throughout the period (Figure 7). Therefore, fiscal impulses have been very small, with a close to zero correlation between them and the output gap (Figure 8). In 2014–15, fiscal impulses were almost zero. Figure 7: Headline and Cyclically Adjusted Figure 8: Output Gap and Fiscal Impulse, Fiscal Balances, 2000–15 2000–15 (Percent of GDP) (Percent of Potential GDP) Source: Moldovan authorities, World Bank calculations. Source: Moldovan authorities, World Bank calculations. 5 Moldova’s Fiscal Rule 8. The fiscal rule introduced Figure 9: Actual and Simulated Fiscal Balance, 2000–15 in Moldova in 2014 could neither (Percent of GDP) improve fiscal sustainability nor promote counter-cyclical fiscal policy. Fiscal rules are typically adopted to ensure debt sustainability, stabilize output, and limit the size of government. Moldova introduced a budget balance rule (see Box 2), stipulating that, from 2018 and on, the fiscal deficit, excluding grants, should not exceed 2.5 percent of GDP, although the deficit ceiling could be exceeded by the amounts of externally financed investment projects. A rule Source: Moldovan authorities, World Bank calculations. that does not adjust for the stage of the business cycle is unlikely to help stabilize output and introduce counter-cyclical fiscal policy. In addition, the rule is not explicitly linked to debt sustainability. A backward recalculation of Moldova’s rule applied to 2000–15 shows that this fiscal rule would have performed worse than what actually happened in terms of both sustainability and counter-cyclicality. If the fiscal deficit had been at the ceiling following the rule, the deficit would have averaged 2.2 percent of GDP for the 15 years—0.9 percentage point higher than the actual (Figure 9). For instance, the fiscal deficit would have been 4.2 percent of GDP in 2015 compared to the actual 2.3 percent. While its sustainability impact would have been relatively small, it would have performed much worse in stabilizing output. For instance, it would not have required the government to run fiscal surpluses during the period of high economic growth in the early to mid-2000s. Box 2: Types of Fiscal Rules Governments adopt fiscal rules to ensure debt sustainability, to stabilize output, and to limit the size of government. To achieve these objectives, a fiscal rule imposes long-lasting constraints on fiscal policy through numerical limits on budgetary aggregates. There are 4 main types of fiscal rule:  Budget balance rules require that the budget be balanced; the target can be specified as the overall, primary, cyclically-adjusted, or over-the-cycle balance. Examples include Mexico and Paraguay (overall balance), Colombia, Chile, Norway, and Romania (structural balance).  Debt rules set an explicit target limit for the public debt-to-GDP ratio. This type of rule is used by Bulgaria, Poland, Lithuania, Peru, New Zealand, and Switzerland, among others.  Expenditure rules put permanent limits on total, primary, or current spending in absolute terms, growth rates, or percent of GDP. Poland, Spain, Finland, Japan, and France use expenditure rules.  Revenue rule: set a ceiling or floor on revenues with the goal of boosting revenue collection, preventing an excessive tax burden, or both. Examples include Panama, Norway, Canada (Alberta), and Kazakhstan. Due to trade-offs between the properties of different fiscal rules, many countries have combined elements of several. Most common are combinations of a budget balance rule with a debt rule or with an expenditure rule. The combination of budget balance with expenditure rules is often used in high income countries, while the combination of a budget balance rule with a debt rule is more common in emerging markets. Source: MFM Forum—Learning Module on Fiscal Rules, May 2, 2016 6 Fiscal Policy and Private Savings 9. It is crucial to understand how fiscal policy and private savings are linked in order to gauge the impact of a potential fiscal stimulus on the economy. While macroeconomic stability is the main objective of fiscal consolidation, and fiscal policy therefore aims primarily to reduce the deficit, the discussion in this section goes a step further—to a situation where fiscal policy is used to stimulate the economy. Moldova improved its public finances in 2010–14, as indicated by the declining fiscal deficit. However, growth prospects for the medium term are not so encouraging. Meanwhile, fiscal performance has recently deteriorated, mainly due to the liabilities that arose from the banking crisis and a notable decline in grant revenue. While fiscal dynamics are currently under pressure, an equally relevant issue is how a potential fiscal stimulus would impact future policymaking in an environment where a fiscal anchor is operating and fiscal adjustment is back on track. It may therefore be a useful exercise to look into the trade-offs between revenue and spending policies and private savings so as to inform fiscal policy and understand how it may constrain the economy. Box 3 summarizes the conceptual framework for such an analysis; Box 4 presents the savings and investment profile of Moldova. Box 3: Why the Fiscal Policy-Savings Relationship Matters: Testing Ricardian Equivalence How effective fiscal policy is depends on, among other factors, the response of private savings to a change in fiscal stance. Changes in the budget deficit that affect public savings may be partly counteracted by changes in private savings. In fact, in many OECD economies fiscal adjustments with the objective of stabilization have been associated with inverse movements in private savings (OECD 2004). There may be many reasons for this, because fiscal measures can be offset by private saving responses through a variety of channels. The primary one is through disposable income; the marginal propensity to consume out of an additional unit of income is generally less than one in the long term (additional income is partly saved and partly consumed). Another offset can occur when private savings go up indirectly because of higher interest rates caused by higher budget deficits. Finally, these could be complemented by the negative effects on asset prices of increased government debt (OECD 2004). As a result, a tax cut by the government to stimulate the economy, for example, may be less effective if the offset coefficient between public and private saving is high. The Ricardian equivalence tells that since consumers are forward-looking, how government spending is financed does not affect consumption decisions, and therefore it does not affect aggregate demand. The hypothesis to be tested is whether a 1 percentage point increase in public savings leads to a less than 1 percent increase in total savings. Full Ricardian equivalence implies that a 1 percentage point decline in public saving (e.g., through a tax cut or a rise in spending) leads to an equivalent increase in private saving—when consumption is kept constant, higher disposable income implies higher private saving. In other words, there is a one-to-one negative relationship between public and private savings. In practice, full Ricardian equivalence hardly holds, as has been demonstrated in empirical studies in both the developed and the developing world. The analysis for this section tests the impact of the various components of public balances (different policy tools) on private savings. Source: OECD 2004 10. The results of the econometric estimation in Moldova suggest that, controlling for other factors, there is a powerful negative association between public and private savings. The baseline analysis tested the relationship between public and private savings while controlling for real GDP per capita, the interest rate on deposits, and the real exchange rate (see Annex 2 for data and methodology). All coefficients are statistically significant at a 95 percent confidence interval except the coefficient for the constant term. As shown in Table 2, a 1 percentage point increase in public savings (driven by an increase in revenues, a cut in expenditures, or both) leads to a 0.7 percentage point 7 decline in private savings. As expected, the coefficient for real GDP per capita is positive, which suggests that private savings tend to increase as countries climb the income ladder. Similarly, the positive coefficient for the deposit rate indicates that households and firms react rationally to higher interest rates by saving more. Finally, the coefficient for the real exchange rate has a negative relationship with private savings. This is not surprising: Real depreciation makes foreign goods more expensive for local consumers, leading to a decline in consumption and an increase in private savings. Coefficients for female labor force participation, the inflation rate, and government spending were also added to the regressions but turned out to be insignificant when considered together with the baseline variables. Table 2: Econometric Estimation of Private Savings Determinants in Moldova: Results (OLS with Hac VCov Matrix, 1998–2014) Dependent Variable: Private Sector Savings Variable Coefficient t-Statistic Prob. Constant 0.0026 0.3527 0.7304 Public savings -0.7277 -3.9863 0.0018 Real GDP per capita 0.3383 3.1263 0.0088 Deposit rate 0.7652 4.3190 0.0010 Real exchange rate -0.2081 -2.8268 0.0153 R-squared 0.7450 Adjusted R-squared 0.6601 Source: World Bank calculations 11. The estimations of the impact of revenue policy on savings were inconclusive. The analysis also considered alternative specifications by including income and consumption taxes in addition to the control variables. Both estimations yielded statistically significant results for all variables except the constants, at a 95 percent confidence interval. However, though the coefficient for tax revenues from income was positive, the coefficient for tax revenues from consumption was negative. These variables capture the actual fiscal realizations in nominal terms rather than the tax rate. As a result, higher revenues from consumption taxes are a reflection of consumption, but higher revenues from income taxes reflect better incomes. Thus, the results are intuitive in the sense that heightened consumption is negatively correlated with private savings but higher income is positively correlated. Box 4: Saving-Investment Balance in Moldova Both savings and investment as a share of GDP have been volatile for the last 20 years. Although low in the late 1990s and early 2000s, both savings and investment rates shot up in the run-up to the global economic crisis, although the pace of increase for the latter was faster, reaching 34 percent in 2008 and widening the saving-investment gap (Figure B4.1). Starting in 2009, the economy seems to have entered a new phase of lower but stable levels of investment at a narrow 23-24 percent interval. Concomitantly, savings declined sharply, reaching the low of 11 percent in 2011 before beginning to increase again slowly. As a reflection of this volatility, the large current account deficit in 2007-08 had narrowed down to 6 percent by 2014. Public and private savings moved in opposite directions until 2009. Private savings, as expected, represent a higher share of total savings. Public and private savings moved in opposite directions in 1997 – 2010 before they began to co-move in 2011 (Figure B4.2). This general negative correlation between the public and private components of savings suggests a negative offset coefficient (less than 1), which is tested in this section. Figure B4.1: Moldova: Savings Investment Figure B4.2: Moldova: Gross Public and Private Balance and CAD, 1996– 2014 Savings, 1997–2013 (Percent of GDP) (Percent of GDP) 8 Source: Moldovan authorities, WDI online database, Source: Moldovan Statistics Bureau. http://data.worldbank.org/data-catalog/world- development-indicators Recommendations 12. Moldova must move to preserve its fiscal sustainability and avoid procyclical policies. Recent developments have increased the risks to Moldova’s fiscal sustainability because the government protected all depositors in three failed banks. Meanwhile, the risks of additional contingent liabilities, especially in the financial sector, are not negligible, though the authorities started to address the causes of the previous banking crisis and to improve the transparency of shareholding. Meanwhile, the current rule could make fiscal policy procyclical. Therefore, it is recommended that the authorities:  Carefully monitor contingent fiscal liabilities and come up with clear, more stringent rules for providing government guarantees. The government should carefully monitor contingent fiscal liabilities primarily in the financial sector but also in the energy and (other) state-owned enterprises.  Make the fiscal rule more effective. The government could improve the use of its fiscal rule as an anchor for sustainability and output stabilization by reconsidering the exclusion of investment projects with external financing from the deficit ceiling, and adjusting the rule to account for the economic cycle. The deficit rule could be supplemented by an operational expenditure rule.  Develop the government securities market. The authorities should also promote development of the domestic government securities market so that it can finance an increasing share of public financing requirements. 9 Spotlight: Good Practices in Fiscal Consolidation 13. While fiscal consolidations depends on specific country circumstances, the economics literature suggests that fiscal consolidations based on spending cuts are often preferable to tax increases. Based on the experience of the OECD countries in 1970-2007, Alesina and Ardagna (2010) conclude that spending-based adjustments are more likely to reduce the deficit and public debt than an adjustment based only on tax increases. Based on a historical analysis of fiscal consolidation in advanced economies, IMF (2010) suggests that adjusting spending is less likely to lead to a recession. However, when consolidation must be large, adopting multiple instruments, both spending and revenue items, increases the probability that debt stabilization will succeed (OECD 2012). Annex 3 provides data on how spending and revenue measures might contribute to fiscal consolidation in OECD countries. 14. “Smart” expenditure reform may be thought of having three pillars: achieving efficiency gains in service delivery, ensuring the sustainability of spending, and establishing fiscal institutions. Efficiency gains could be achieved, for example, through improvements in public administration – both on wages and employment, public health care, public education, and other public services. Social spending reforms could focus on improving the targeting and managing the scope of social transfers and public pensions. At the same time, more vulnerable households need to be protected. All these could be complemented by institutional reforms to promote spending control and ensure sustainability. 15. When large fiscal adjustment is needed, spending cuts are typically accompanied by revenue increases. It is good practice for governments to first explore the scope for broadening tax bases to keep tax rates low and direct their attention to mobilizing revenues from the least distorting taxes. The consolidation package should focus on measures that reduce the distortions that are most harmful to growth, and tilt revenue structures toward taxes and charges that are the least inefficient or that correct externalities. Lessons from Latvia, Estonia, and Lithuania 16. Estonia, Latvia, and Lithuania had very high growth rates up to 2008, when they were hit hard by the global crisis. The cumulative output loss associated with the recession in the Baltics was almost twice the size of the losses suffered by the hardest-hit countries in the 1997-98 Asian crisis (Groenendijk and Jaansoo 2015). Latvia suffered the largest loss of output, and recovered the latest; the crisis was the shortest in Lithuania. Estonia`s economy contracted by 4.1 percent in 2008 and 14.1 percent in 2009, Latvia`s by 2.8 percent in 2008, 17.7 percent in 2009, and 1.3 percent in 2010, and Lithuania`s GDP fell by 14.8 percent in 2009. In parallel, general government balances deteriorated significantly. Between 2008 and 2011, the general government budget deficit averaged 0.9 percent in Estonia, 6.1 percent in Latvia, and 7.1percent in Lithuania. 17. All three Baltic countries undertook sizable fiscal consolidations in 2008–10. All three states front-loaded their fiscal consolidation, put more emphasis on cutting spending, and complemented that with revenue measures. The total consolidation packages over 2008-10 amounted to 13.8 percent of GDP in Estonia, 14 percent in Latvia, and 13.4 percent in Lithuania. Most of the spending cuts took place in 2009, when they accounted for 70 percent of the total consolidation packages in Estonia and Latvia and 78 percent in Lithuania (Table 3; Groenendijk and Jaansoo 2015). In Latvia, the relative importance shifted more to the revenue side but remained on the spending side in Estonia and Lithuania. The Baltic countries do not have an independent monetary policy: during 2009- 10, all operated currency boards and subsequently all three joined the Eurozone. 10 Table 3: Spending Measures, Baltic Countries, 2009 –10 (Measures taken in each country) Estonia Latvia Lithuania Hiring / Pay freeze X X - Wage reduction X X X Staff reductions X X X Reorganizations X X X Efficiency cuts X - X Health X X X Education - X X Pensions X X X Unemployment X X X Other social security/welfare X X X Infrastructure X - - Source: Groenendijk and Jaansoo 2015. 18. The Baltic countries moved to decrease their public wage bills. Estonia and Latvia implemented hiring and pay freezes. More importantly, all three countries cut public wages at the outset of the crisis—Latvia by 18 percent, Lithuania by 10 percent, and Estonia by 8 percent (Table 4). Civil servants also saw benefits cuts as additional pay funds, training funds, and one-time support schemes were abolished. All three countries complemented these wage measures with staff reductions and public sector reorganizations (Groenendijk and Jaansoo 2015). Table 4: Changes in Annual Wages, Baltic Countries, 2009 (Percent) Estonia Latvia, Percent Lithuania Public services -4.5 -9.7 -11.0 Public administration -7.6 -18.0 -10.0 Education -2.5 -9.9 8.0 Source: Groenendijk and Jaansoo 2015. 19. Spending cuts in health, education, and pensions complemented the reduction in public wages. Spending on health fell in all three countries; for example, in Latvia it declined cumulatıvely by 0.8 percent of GDP in 2009–10 (Groenendijk and Jaansoo 2015). Latvia and Lithuania implemented reforms to encourage saving and efficiency in education by introducing a funding system based on fixed amounts per student. In Lithuania, the educational reforms yielded savings of 0.3 percent of GDP in 2009 and 0.5 percent in 2010. All three countries also raised the retirement age and temporarily froze pensions. Estonia restructured its pension schemes, and Estonia and Lithuania reduced pension payments (Table 5). Table 5: Pension Reforms, Baltic Countries, 2008–10 (Measure taken in each country) Estonia Latvia Lithuania Raise in retirement age (including early retirement) X X X Restructuring the pension schemes X - - Reduction of pension payments X - X Temporary pension freeze X X X Extension of contribution periods - X - Source: Groenendijk and Jaansoo 2015. 11 20. Revenue measures accompanied spending cuts in the Baltic countries. All three increased excise duties on tobacco, alcohol, and fuel (Table 6). Moreover, the standard VAT rate rose in all three countries, from 18 percent to 21 percent in Latvia and Lithuania and to 20 percent in Estonia. Reduced VAT rates were also increased (Groenendijk and Jaansoo 2015). Estonia and Lithuania abolished several VAT exemptions. In addition, Lithuania introduced a real-estate tax in 2011, and all three Baltic countries moved to broaden the tax base in order to boost revenues (Groenendijk and Jaansoo 2015). Table 6: Revenue Measures, Baltic Countries, 2008–10 (Measures taken in each country) Estonia Latvia Lithuania Raising the VAT rate X X X Raising excise rates X X X Tobacco X X X Alcohol X X X Fuel X X X Raising the tax rate for unemployment insurance X Broadening the base for personal income tax by reducing the number of X X X allowances Postponement of planned income tax reduction X Drop in the income tax rate from 25 percent to 23 percent in 2009, then a X rise to 26 percent in 2010, followed by a reduction to 25 percent in 2011 Increase in corporate income tax rate in 2009, from 15 percent to 20 X percent (reversed in 2010) Introduction of a progressive real-estate tax, with higher rates for buildings - X - with higher value in 2009; tax rates doubled in 2011 VAT exemptions abolished X X Real-estate tax introduced in 2011 - - X Personal income tax rates reduced - - X Exemptions made to excise duties X Before 2010 the personal income tax exemption could be applied to the X - - first child; in 2010 and after, the exemption applied only to the second and subsequent children. Tax bases broadened X X X Source: Groenendijk and Jaansoo 2015. 12 Chapter II. Spending Moldova’s government spends more than countries with similar per-capita income, degree of human development, and government efficiency. It spends more on public consumption and is a clear international outlier in education outlays. Starting in 2010, the government began to reduce spending, most notably on education and social protection and it created more space for capital spending. However, Moldova’s government is still far from efficient: comparators produce similar outcomes using 40 percent less revenue. The efficiency of its spending could be improved by reforming the pension PAYG system, better targeting social assistance, optimizing hospital and as primary and secondary school networks, reducing public employment, strengthening procurement practices, and revising how it invests its capital. Government Spending over Time 21. Compared to the previous decade, in 2010–15 higher transfers to citizens elevated budget spending by 1.7 percentage points, to an average of 39.3 percent GDP. Spending had been growing gradually through the 2000s and peaked at 45.3 percent of GDP in the crisis year of 2009. As Figure 10 shows, in 2010–15, compared to 2000–09, transfers to the population increased the most (1.8 p.p.), driven by higher pensions, followed by capital investments (0.8 p.p.), particularly in road infrastructure, and spending on goods and services (0.8 p.p.), driven by the health sector. Meanwhile, a reduction in the public debt-to-GDP ratio helped lower interest outlays. Transfers for production, mainly agricultural subsidies, decreased 0.7 p.p. In the functional classification, outlays on social protection went up by 2.1 p.p., mostly because of higher spending on pensions (Figure 11). Figure 10: Long Term Change in Spending, Figure 11: Long Term Change in Outlays, 2000– 2000–09 to 2010–15, Economic Classification 09 to 2010–15, Functional Classification (Percentage Points of GDP) (Percentage Points of GDP) Source: Ministry of Finance, World Bank calculations. Source: Ministry of Finance, World Bank calculations. 22. Over 2010–15, the government reduced current spending and created space for capital spending. Due to education reforms and reductions in spending on public order and national security, the total public wage bill decreased by 1.8 p.p. to 8.5 percent of GDP in 2015, though that is still high by international standards (Figure 12). Spending on goods and services also went down (by 1.1 p.p.), and so did transfers to the population (1p.p.), which is the biggest spending item, mainly comprised of pensions and social payments. Higher support from foreign grants and loans helped to push up capital spending on infrastructure, with public investment in roads have gone up almost 2 p.p. since 2010. In 13 2014 capital spending set a 15-year record of 8.5 percent of GDP, before reversing in 2015 as external assistance defined. As Figure 13 shows, by functional classification, the adjustment came from efficiency gains, especially due to school network and staff rationalization in education (-2.2 p.p.) and to the new targeted program and reduction of categorical compensation schemes in social assistance (-1.4 p.p.). Figure 12: Short Term Change in Spending Figure 13: Short Term Change in Outlays from from 2010 to 2015, Economic Classification 2010 to 2015, Functional Classification (Percentage Points of GDP) (Percentage Points of GDP) Source: Ministry of Finance, World Bank calculations. Source: Ministry of Finance, World Bank calculations. Moldova’s Spending in Comparative Perspective 23. Moldova’s government spends more than other countries with similar per-capita income. At 37.8 percent of GDP on average for 2000–14, Moldova’s general government expenditures were 8.5 percentage points higher than the average for lower middle-income countries (MICs) and 3.7 p.p. higher than the ECA average (Figure 14). Compared to regional peers in the same income category (Armenia, Georgia, Kyrgyz Republic), Moldova also stands out for high spending for the last 15 years, especially in the 2006–08 pre-crisis period. Moldova’s general government spending is in fact in line with ECA countries of similar size but with much higher GDP per capita, such as the Slovak and Czech Republics. In addition, many countries reduced government spending as a share in GDP significantly over the last 15 years, while Moldova has a higher share than in the early 2000s (Figure 15). 14 Figure 14: General Government Spending, Figure 15: General Government Spending, Moldova and Comparators Moldova, Peers, and Selected ECA Countries (Percent of GDP) (Percent of GDP) Source: Find My Friends, using IMF WEO. Source: Find My Friends, using IMF WEO. 24. Moldova spends more than countries with similar levels of human development and government efficiency. Countries whose spending is similar to that of Moldova tend to have higher human development and more efficient governments (Figure 16-Figure 19). In other words, higher level of government spending is usually associated with a high level of income – a proxy for social agreement on government output -- and high government efficiency. Moldova needs to spend less given its current level of government effectiveness. This analysis highlights the importance of good governance as a prerequisite for possible expansion of expenditures, provided they provide public goods or tackle externalities. Figure 16: General Government Spending and Figure 17: General Government Spending, GDP per capita, Moldova and Other Countries and HDI, Moldova and Other Countries (Percent of GDP) (Percent of GDP) Source: Find My Friends, using IMF WEO. Source: Find My Friends, using IMF WEO and Worldwide Governance Indicators. Note: General government total expenditure (share of nominal GDP), Note: General government total expenditure (share of latest available outcome. nominal GDP), latest available outcome. Note: HDI - Human Development Index. 15 Figure 18: Income and Government Effectiveness, Figure 19: General Government Spending Moldova and Other Countries and Government Effectiveness, Moldova (US dollars by PPP, per capita) and Other Countries (Percent of GDP) Source: Find My Friends, using IMF WEO and Worldwide Governance Indicators. Figure 20: Average General Government Figure 21: General Government Spending, Spending, Moldova and Peers, 2000–14 Moldova and Selected Countries (Percent of GDP) (Percent of GDP) Source: Find My Friends, using IMF WEO Source: Find My Friends, using IMF WEO 25. Current spending dominated Moldova’s public expenditures. Moldova spent most on social security insurance (11 percent of GDP), wages (9 percent), and goods and services (8 percent). Social security insurance in Moldova is significantly higher than in the lower MICs, but on par with that ECA average, which could be explained by demographic trends and history that are similar to those of the rest of ECA. On average Moldova spent more on wages and purchases of goods and services lower MICs and ECA countries (Figure 20). Purchases of goods and services were one of the largest among all peer countries except for the Kyrgyz Republic (Figure 21). 16 26. Within current spending, Figure 22: Average General Government Spending: outlays are the highest on education, Economic Calcification, Moldova and Peers, 2000–14 health, and social protection. Its (Percent of GDP) functional structure is in line with peers except for education and health (Figure 22 and Figure 23). At 8.1 percent of GDP on average for 2000–14, Moldova’s spending on education was the highest in ECA, which averaged 4.1 percent, and one of the highest among lower MICs, which averaged 4.5 percent. At 4.7 percent of GDP on average for 2000–14 its spending on health was higher than the ECA average of 3.5 percent and the lower MIC average of 2.9 percent. Source: Find My Friends, using Education and Health databases. Figure 23: Spending by Function, Moldova and Selected Countries (Percentage of GDP) Source: IMF GFS Data base for 2013. 17 Economic Classification Wage Bill 27. The government began to cut the wage bill after the 2009 crisis, mostly by reducing public employment. Between 2010 and 2015 public employment went down 16 percent and the wage bill fell by nearly 2 percent of GDP, to 8.5 percent (Figure 24). Most of the job cuts were in education (down 25 percent), which accounts for half of public employment, a staggering number in international comparison); science (down 22 percent); and health (down 12 percent) (Figure 27). Figure 24: Number of Public Employees and the Figure 25: Structure of the Public Wage Bill Wage Bill (Percent of GDP) (Percent of GDP) Source: National authorities, World Bank calculations. Source: BOOST data base, World Bank calculations. Figure 26: Budget Spending, Wages, and GDP, Figure 27: Staff Positions in Public Employment, 2000–15 2009 and 2015 (Index) (Number of Positions) Source: National authorities, World Bank calculations. Source: Ministry of Finance. 28. Meanwhile, the fiscal pressure stemmed from smaller expenditure items. The wage bill rose the fastest for social care facilities, justice, defense, and public order (Figure 28). Wages also rose substantially for the National Patrol Inspectorate Police and the Inspectorate-General. High wage outlays in social care facilities was associated with a more dynamic increase in wages at home care 18 services and shelters for the elderly and Figure 28: Where the Public Wage Bill Has Grown disabled. The number of staff positions in the Most social assistance sector went up 33 percent. (Percentage Points) 29. The fastest wage growth was in the judiciary, although justice reform is slow. The wage bill in the justice sector registered in 2009–14 the highest cumulative contributions to the overall growth of the wage bill envelope (force5) (defined according to World Bank 2015b). The wages paid to judges went up 80 percent and almost doubled for those on courts of appeal.6 The fiscal pressures caused by higher wages for the justice sector did not improve performance; most surveys show even more popular and corporate distrust in Moldova’s justice system. According to the 2013 International Survey of Enterprises, only 31.9 percent of economic agents in Moldova consider courts to be reasonable, impartial, Source: BOOST data base, World Bank calculations. and incorrupt, compared to 38.9 percent for all of Eastern Europe. Additionally, while total cases per 100,000 inhabitants in Moldova are about half the EU average, 70 percent of economic agents do not believe that courts will protect their property rights, which discourages their intentions to invest. 30. The total wage bill is probably understated. Some wages for road construction projects are registered as capital spending (an estimated 0.2 percent of GDP in 2015) and wages in the health sector, particularly in hospitals, are in part registered as procurement of goods and services through the National Health Insurance Company (about 2 percent of GDP in 2014). In addition, the steep drop in the education wage bill in 2013 was due to the higher education system, where wages fell from 6.7 percent of total wage bill to nil in 2014, an unrealistic decline (for more detail, see the education sector section). 31. Ineffective and fragmented local government makes it difficult to reduce the public wage bill. On average municipalities in Moldova are considerably smaller than in most peer countries (World Bank 2014a). Almost 90 percent of the spending on municipal wages goes to the local executives, whose remuneration has gone up 53 percent since 2009. Education is still the largest public employer and there is room at local levels for optimization. The adjustment of 2010–12 decreased the number of education employees financed through the central budget by 60 percent, but at the local level opposition from local authorities kept the decrease to just 14 percent. About 10 percent of all local jurisdictions do not cover their spending from their own revenues. Reducing the number of local authorities and centralization and streamlining of public services, together with the introduction of effective internal controls, may relieve the pressure of wage bill spending. 5 Detailed analysis of fiscal data in this report is based on Moldova BOOST data, available for 2009-14. 6 Since 2012, the wages at Chisinau court of appeals have gone up 40 percent and doubled at the Balti court of appeals. 19 32. There are considerable differences in earnings across various public institutions. Moldova’s average wage is the lowest in the region and is 10 times smaller than in the EU15 average (Moldova’s GDP per capita is 9 times smaller than the EU15 average). However, in Moldova earnings in autonomous agencies and for managerial positions (Courts of Accounts 2015) are significantly higher than average. In fact, state agencies and national regulators usually have higher wages than the rest of the public sector. For instance, in the last three years alone, the wage bill at National Competition Council increased 160 percent and it almost doubled at the Central Election Commission and the Broadcast Coordination Council. At the same time, public entry-level positions are paid lower than average salaries, making difficult retain young and qualified personnel. Bonuses given at the discretion of managers often constitute a larger proportion of total earnings for higher-graded posts, especially at the very top of the scale. This system of remuneration is opaque. Goods and Services 33. The fiscal pressure on spending on goods and services comes from purchases of smaller, often one-off items. At 8.1 percent of GDP (Figure 30), spending on goods and services accounted for 21 percent of public spending. The biggest item, half of the total, was for mandatory health insurance services,7 followed by 22 percent for education8 (Figure 29). However, the fiscal pressure is coming not from large but from smaller items. Heightened fiscal pressure came from procurement of fuel and power service by the Agency for Energy Efficiency, which has increased 7.6 times since 2011, and from procurement by the Internal Affairs authorities.9 34. Poor practices undermine the quality and efficiency of public procurement. The quality of the supplied services and goods is not improving (see the functional analysis) and the procurement process is inefficient and slow. The majority of public acquisitions, on average almost 8 percent of GDP10, are managed by the Agency of Public Acquisition. Even after installation of a new information system, numerous legislative drawbacks, red tape, and the lack of transparency undermine efficiency. The number of contracting authorities is also high, which prevents efficiencies of scale. While opportunities for small and medium-sized enterprises are limited, Courts of Accounts and civil society periodically highlight the unusually high number of cases when the prices of procurement contracts are changed after companies win the bidding. There is also potential for cutting spending on low-priority goods by making public entities more accountable through, e.g., publication of budget planning and provisions. 7 In part, this category represents wages to the health sector. 8The dynamics of procurement in higher education institutions is distorted by a change in public financing of the sector. Funds allocated are categorized as state order and not wages. 9The four institutions that explain this are the general inspectorate of police (60 percent growth); commissioner general of policy Chisinau (up almost 500 percent growth); information and technology service at MIA (more than 300 percent growth) and Ungheni police department (1255 percent growth). 10About 22.5 percent is for construction, 6 percent for education and training services, and 14 percent for pharmaceutical and food products and fuel and cars. 20 Figure 29: Structure of Spending on Goods and Figure 30: Total and Goods and Services Services Spending, 2000–15 (Percent of Total Spending) (Percent) Source: BOOST data base. Source: Ministry of Finance and World Bank calculations. Figure 31: Positive Contributions to Growth of Figure 32: Negative Contributions to Growth Goods and Services Spending of Goods and Services Spending (Percentage Points) (Percentage Points) Source: BOOST data base, World Bank calculations. Source: BOOST data base, World Bank calculations. Transfers to the Population 35. Population aging is driving spending on pensions and social benefits. Social benefits and pensions paid through the state social insurance budget accounted for 88.3 percent of transfers in 2014 (Figure 33 and Figure 34). Since 2013, the main fiscal pressure has been coming from spending by the National Insurance House and particularly pensions (more detail below). 21 Figure 33: Evolution of Spending on Transfers Figure 34: Structure of Spending on to the Population, 2001–15 Transfers to the Population (Percent) (Percent) Source: Ministry of Finance and World Bank calculations. Source: BOOST data base, World Bank calculations. Capital 36. The government has boosted capital spending, relying increasingly on external loans and grants. Growth in capital spending has averaged almost 6 percent of GDP for the last 15 years (Figure 35) and its share in total spending rose from 11 percent in 2009 to 22 percent in 2014 (Figure 36), as the share of externally funded spending grew from 26 percent to 39 percent. Capital spending contributed on average 3.8 p.p. to general spending growth, with a clear upward trend to 2014. Figure 35: Evolution of Capital Spending, Figure 36: Structure of Capital Spending 2009–14 (Percent) (Percent) Source: BOOST data base. Source: BOOST data base, World Bank calculations. 37. Capital spending is concentrated in a few areas; road services absorb the most. The roads services, utilities and agriculture account for half of public capital expenditure (Figure 37). Road modernization accounts for one-third of capital spending and is mainly financed by external donors. 22 The share of capital spending on agriculture increased due to investment subsidies through the Agriculture Subsidy Fund and other agricultural programs started in 201411 12(World Bank 2015a). Spending on utilities went up to 8.2 percent in 2009–14 mainly because of new donor-supported water supply services. 38. Most of the fiscal pressure in 2009–14 Figure 37: Contributions to Average Growth of came from recent health, energy projects and Capital Spending, 2009–14 spending on utilities. National health (Percentage Points) investment projects begun in 2014 accelerated the most. Capital spending under the national health projects reached 0.5 percent of GDP in 2014, from 0.02 percent in 2009. The healthcare service improvement project alone accounts for 43 percent of total capital spending on national health projects. Capital spending on utilities increased to 0.7 percent of GDP in 2014, from 0.3 percent in 200913. Investments in the gas connection between Moldova and Romania have also accelerated, as well as investments in environmental protection, which have more than tripled since 2011.14 Source: BOOST data base, World Bank calculations. 39. Trends in budgeted capital spending have tended to be consistent with Moldova’s strategic objectives, except for energy, utilities, and the railway network. Increased domestic funding for road maintenance in recent years has made it possible to mobilize external funds for investments in road rehabilitation and capital repairs. However, with its extensive railway system, Moldova needs to invest in maintaining and developing it. In agriculture, capital spending mainly consists of transfers to farmers (investment subsidies). Trends in education are consistent with the rehabilitation needs of an oversized general primary and secondary school network: investment has declined but capital repairs increased. In health, with domestic resources scarce, external funding has dominated capital spending. Energy investments are held back by unresolved institutional bottlenecks. To ensure sustainable operation of energy and heating companies, and to attract much-needed private investment, the sector urgently requires governance and institutional reforms, debt restructuring, and immediate investments. The repair needs in municipal water and sanitation systems are considerable. Setting tariffs at cost-recovery level is a prerequisite for concentrated efforts to increase absorption of donor funds and better manage investment. 11 The following agricultural programs begun in 2014: National Wine and Vine Office, Rural Finance and Agricultural Business Development Program, Rural Financial Services and Marketing Program, Competitive Agriculture Project, World Bank Emergency Moldovan Agriculture Support Project, IFAD VI Project and the Viticulture and Wine Sector Restructuring Program, 0.2 percent of GDP. 12 Over 2011-2014, one of the biggest fiscal pressures stemmed from capital spending to a joint stock company Farmaco. This culminated in 2014 with a spending of 0.5 percent of GDP. 13 Main fiscal pressure came from the Water supply service development under the Ministry of Environment (490 percent increase), spending on street lights (2870 percent increase) and district planning (224 percent increase). 14 The pressure to increase this spending envelope came from the Environmental Fund in Rezina district (1.4 times increase in funds); Ministry of Environment (1.7 times increase); Criuleni district (4.7 times increase) and Drochia district (1400 times increase). 23 40. Given severe infrastructure needs and a constrained fiscal envelope, capital spending must be made more efficient. The scope for raising revenues is limited and declining external assistance could restrict funding for capital investments. Capital needs should be carefully estimated based on analysis of the existing portfolio, especially unfinished projects. Spending needs also to be adjusted to make more space for public investment in the medium term as management and absorption capacity improve. Since there are no unified databases of existing, unfinished, and frozen projects, efficiency gains should start with portfolio inventory and optimization. In general, efficiency gains might come from improvements in planning, monitoring, and evaluation of public investments. Public investments should follow the government regulation on public investment project cycle.15 Functional Classification Social Protection 41. A change in the pay-as-you-go (PAYG) pension formula in Moldova has reduced pensions relative to wages to one of the lowest levels in the region: In 2013, the average old-age pension was about 28 percent of the gross wage, one of the lowest gross replacement rates in the ECA region. In 1995 the pension formula has yielded a gross replacement rate of 45 percent, but changes introduced in 1998 pushed the replacement rate down. After 1999, the salaries on which the pension benefit formula was calculated were not revalued (valorized) to account for the cumulative growth of the average wage between when the salary was earned and the date of retirement. That reduced initial replacement rates and created differences between new and old pensions. Moreover, the policy for minimum pensions heavily redistributes toward agricultural workers at the expense of insured individuals who pay average and above-average contributions. Moreover, frequent discretionary increases in pensions undermine the fiscal sustainability of the pension system. These policies compress the benefits structure, weaken the linkage between contributions and pension benefits, and discourage contributions. 42. Low pensions create an incentive to evade contribution evasion, which causes the pension system deficit. In 1999, pension outlays amounted to 4.4 percent of GDP; in 2014 they reached 7.5 percent. Over the same period, the number of contributors shrank by more than 40 percent. Contribution revenues fell short of covering growing pension costs, which resulted in a gap of 10 percent in national social insurance expenditures. Currently, only about 30 percent of the working- age population contribute to the system but almost all elderly receive pensions. The shrinking contributor base and the high number of pensioners together drive up the system-dependency ratio. By 2020 the ratio is projected to rise from the current 0.76 pensioners: 1 contributor to 1 pensioner: 1 contributor because of the aging work force (Figure 38). The shrinking coverage has implications for poverty among the elderly and for the future cost of social protection. 15Public Investment Management Resolution (No. 1029, dated December 19, 2013) describes the public investment project life cycle and establishes criteria for screening projects. 24 43. The retirement age of 57 for women Figure 38: Pension System Dependency Ratio, and 62 for men is low by international 2014–70 standards. In 2003 relatively low life expectancy (Ratio) in Moldova was an argument to suspend raising the retirement age. Though Moldovan life expectancy at birth is indeed among the lowest in the ECA, since 2006 the life expectancy at retirement age has improved—in 2013, it was 15.7 years for men and 19.6 for women, which is similar to other transition economies in the region. Assuming life expectancy at retirement to be about 15 years, the women’s retirement age is currently too low and even men’s needs to start rising gradually. Source: Moldova Economic Security of the Elderly report (2016, forthcoming); PROST (Pension Reform Options 44. Special pensions for some groups Simulations Toolkit of World Bank) and NSIH data. make the system inequitable. Special pension Note: Dependency ratio = number of pensions per beneficiaries such as civil servants, judges, contributor. prosecutors, members of the government and Parliament, and others account for 11 percent of old-age pensioners; their pensions cost about 13 percent of total pension outlays. 45. Unless there is parametric reform and better compliance, the pension system will converge to a quasi-flat pension system paying low benefits. The gap between earnings and pensions will grow and the system will not be socially sustainable. The average replacement rate is projected to decrease steadily from 28 percent to almost 13 percent in 2047. 46. Many social assistance benefits are Figure 39: Structure of Social Care and Social poorly targeted. Outlays on social assistance are 2 Insurance Spending percent of GDP. After reforms in 2009–10, the (Percent of Total) share of GDP allocated to targeted social assistance began to go up, but after 2012 the number of beneficiaries of targeted cash transfers began to shrink. Social assistance programs are directed to the 14 percent or so of the total population who are not eligible for the social insurance pension. However, about 20 percent of total social assistance spending leaks to people in the top 60 percent, with some programs channeling well above 30 percent of their budget to people who are better-off people (World Bank, 2016 forthcoming). There are five groups of payments, but all are fragmented and too small to have much impact. The targeted program Ajutor Social and the heating allowance are the most progressive (the bottom 20 Source: BOOST data base and World Bank calculations. percent receive over 70 percent of benefits but they each cover only about 4–6 percent of the population. Improving the efficiency and equity of the 25 social safety net requires consolidation of numerous categorical programs and reallocating those funds to expanding the coverage of poverty-targeted cash transfers. 47. Social care facilities, which provide services to poor and vulnerable population groups, together cost about 1 percent of GDP, but programs, providers, and levels of government are highly fragmented. Although Moldova has made some progress in de-institutionalizing care and providing community-based services, some are still residential. Social care is decentralized to local governments, the poorest of which have few resources to fund it Moreover, local authorities often lack capacity to assess the need and plan service delivery. Moldova has made progress in engaging non- governmental organizations to provide services, but this is mostly done by relying on external funding rather than local budgets. Spending on social care facilities has doubled since 2009. The biggest items are for shelters and rehabilitation, though there are also other kinds of centers and social assistance and care services. All spending items have gone up considerably, however, with personal assistance services in particular having increased 85-fold. Education 48. Trends in education are consistent with the need to rehabilitate an oversized general primary and secondary school network. After a peak of 9.4 percent of GDP in 2009, education spending fell to 7 percent in 2014–15. Due to overstaffing, particularly at local levels and in non-teacher staff, on average 51 percent of all sector spending goes to secondary education, with lyceums and gymnasiums being the biggest recipients (30 percent). The share of preschool education was 22 percent with an upward trend due to high demand for kindergartens (Figure 40). On average 12 percent was spent on higher education and over 6 percent on extracurricular activities (Figure 401). Due to low pupil- to teacher ratios (Figure 42), 60 percent of public resources were being spent on wages, but that share has been declining along with staff positions (Figure 41). Figure 40: Enrollment in Education Figure 41: Educational Spending by Function. Institutions (Percent of Total Education Spending ) (Ratio) Source: Moldovan National Bureau of Statistics. Source: BOOST database, World Bank calculations. 26 Figure 42: Pupil/Teacher Ratios Figure 43: Educational Spending by Economic (Number of Pupils per Teacher) Classification (Percent of Total Education Spending) Source: National statistics. Source: BOOST database, World Bank calculations. 49. With the majority of resources spent on wages, nutrition, and utilities, there is little left for quality improvements. Half of non-wage current spending is nutrition (22 percent), mainly for kindergartens and secondary schools, and utilities16 (27 percent). Staff training and retraining represent 0.2 percent of all non-wage current spending, while almost the same amount goes to buying books and manuals. The teaching force needs to be strengthened. About one-third of the 32,000 primary and secondary teachers lack pedagogical certification and almost 15 percent are past retirement age. The teaching load, based on hours, is low by OECD standards, and there is a mismatch between teacher skills and school needs. With a salary 23 percent below the country average and a lack of career advancement opportunities, it is difficult to attract and retain the most talented. Moreover, school managers usually lack the basic management skills the new law requires. 50. Capital spending on education is low and mainly channeled to secondary education. With the school network optimization project, a larger share of funds is being spent to improve water, sewage, heating systems, and buildings. On average 60 percent of capital expenditures were for repairs, of which half went to secondary schools. Only a small share of capital spending went to new facilities. 51. Despite its high cost, the quality of Moldovan education is poor. The country’s 15-year-olds performed at the bottom of Europe in the OECD Programme for International Student Assessment (PISA) 2010. In fact, 57.3 percent of Moldova’s students are illiterate: they fall short of the basic reading proficiency they need to participate effectively and productively in society, and the situation is worse in rural areas. The education system is unable to produce the skills employers are demanding. A 2013 survey by the National Confederation of Employers cited a low-skilled labor force and the current education and training system as key constraints; the Global Competitiveness Index ranks Moldova 103 out of 144 countries for quality of training. Over 40 percent of firms report that “skills” are a severe constraint to their growth. 16 Including electricity (6.4 percent), natural gas (8.7 percent), heating (9.2 percent), and water supply (3 percent). 27 Health 52. Averaging 4.8 percent for the last 15 years, health is the third largest spending category. Mainly due to a large and costly network of hospitals, by the end of 2009 spending on health had reached a record high of 6.4 percent of GDP. However, it has since then health expenditure has been contained at some 5.5 percent (Figure 44). Spending by the National Health Insurance Company (NHIC) represented 80 percent of all expenditures on health. For 2009–14 health-related capital spending was very low at 0.4 percent of GDP and went mainly on refurbishment of existing physical and technical infrastructure. The capital spending increase in 2014 was due to the launch of the Healthcare Services Improvement Project (0.2 percent of GDP). Figure 44: Health Spending by Economic Figure 45: NHIC Spending by Economic Classification, 2009–14 Classification (Percent of GDP) (Percent of GDP) Source: BOOST database, World Bank calculations. Source: BOOST database, World Bank calculations. 53. Inefficiencies and imbalances in service provision between Chisinau and the rest of the country impede health service delivery; 60 percent of health spending goes to national, district, departmental, and Chisinau municipal hospitals.17 Moldova’s rate of hospital beds per 100,000 people is 32 percent above the EU average, and is concentrated in Chisinau municipality, with many overlaps between municipal, national, and departmental hospitals. Meanwhile, in rural areas, thinly distributed public resources and dilapidated infrastructure are a threat to the safety of patients. Of 71 public hospitals in Moldova, 24 conduct fewer than 1,000 surgeries a year.18 54. Although the bulk of financing is directed to hospitals, they fail to provide quality services. There is a persistently decreasing trend in hospitalization rates based on referral from family doctors (from 22 percent in 2013 to 20 percent in 2015). As a result, primary and outpatient health care is not fulfilling its prevention function, which is less costly than hospitalization. Life expectancy in Moldova is 10 years lower than in peer countries; 87 percent of all deaths are caused by noncommunicable diseases; and infant mortality is among the highest in the region. 17National Clinical Hospital, the Oncological Institute and the Mother and Child Institute are among the most ‘expensive’ hospitals. 18 CNAM administrative data suggests that 22 raion hospitals and 2 departmental hospitals (the State Chancellery and Ministry of Health hospitals) conduct fewer than 1,000 surgeries a year. 28 Efficiency of Public Spending 55. Moldova spends more than peer countries, but its spending performs less well, according to the Public Sector Performance (PSP) indicator (see Table 7 and Box 5). Moldova had one of the lowest scores on administrative performance in the PSP sample; the government scored low on securing property rights, administering justice, enforcing contracts, and providing a business-friendly regulatory environment. The score of the education sector was in fact the lowest in the sample. The health sector is also at the bottom, reflecting a very high incidence of tuberculosis and infant mortality, and low life expectancy, which along with that of Russia and the Kyrgyz Republic is very low. Mainly due to strong economic growth, Moldova’s total PSP score was only 5 percent below the average, but it was still among the lowest performers. Moldova had the highest score on the average GDP growth and above-average score on unemployment rate. Without these two indicators, Moldova’s PSP would have been 20 percent below the average. Table 7: PSP Indicators, Moldova and Peer Countries (Score) Administrative Education Health Distribution Stability Economic PSP Performance Albania 0.92 1.04 1.24 1.02 0.58 1.03 0.97 Armenia 0.95 0.96 0.64 0.71 0.97 1.60 0.97 Bosnia and Herzegovina 0.96 1.01 0.82 0.93 1.07 0.60 0.90 Bulgaria 0.89 0.95 0.88 0.89 1.14 0.68 0.91 Croatia 0.95 0.98 1.37 0.86 0.95 -0.02 0.85 Estonia 1.48 1.22 1.37 0.92 0.72 0.63 1.05 Georgia 1.12 0.97 0.55 1.06 1.12 1.78 1.09 Kazakhstan 1.00 0.97 0.49 1.59 0.88 2.11 1.17 Kyrgyz Republic 0.73 0.85 0.47 0.67 0.92 1.18 0.80 Latvia 1.10 1.05 0.90 0.91 1.14 0.40 0.92 Lithuania 1.11 1.06 0.97 0.92 0.82 1.46 1.06 Macedonia, FYR 1.07 0.98 1.07 0.96 1.27 1.06 1.07 Moldova 0.79 0.91 0.55 0.86 1.15 1.42 0.95 Montenegro 1.21 1.11 1.20 1.43 0.98 0.76 1.11 Poland 1.19 1.00 1.30 0.95 1.26 1.36 1.17 Romania 0.94 0.97 0.66 0.77 1.55 0.77 0.95 Russian Federation 0.85 0.95 0.67 1.25 0.72 1.22 0.95 Slovak Republic 0.86 0.92 1.82 1.02 1.10 1.22 1.16 Slovenia 1.14 1.10 2.40 0.96 0.87 0.34 1.13 Ukraine 0.75 1.01 0.63 1.35 0.80 0.27 0.81 Source: World Bank calculations Box 5: The Public Sector Performance (PSP) Indicator The PSP approach compares countries on a variety of socioeconomic indicators. It then relates public spending to the outcomes of public policies. The greater the positive impact of higher spending on the indicators, the more efficient a country is assumed to be. There are two sets of performance indicators: process or opportunity indicators, and indicators traditionally used for measuring macroeconomic outcomes. The first looks at administrative, education, and health outcomes. The second tries to measure the outcomes of government interaction with markets and look into income distribution and stability. Table B6.1: Composition of the Total PSP Indicator Administrative Property rights GINI index (World Bank Distribution Public Sector Performance Irregular payments and bribes estimate) Ethics and corruption Poverty headcount ratio at Judicial independence national poverty lines Favoritism in decisions of Average inflation, 1994-2003 Stability government officials Average unemployment, 1994- Undue influence 2003 29 Burden of government regulation Average GDP real growth rate, Economic 2009-2014 Stability Education Quality of primary education Coefficient of variation of Secondary education enrollment average real GDP growth Quality of the education system Quality of math and science education Health Tuberculosis cases Infant mortality Life expectancy Source: Afonso, Schuknecht, and Tanzi 2006. 56. Moldova is far from the efficiency frontier of the sample countries because it “wastes” a substantial amount of public spending. Efficiency is ranked by comparing each country’s position relative to the production possibilities frontier. The more distant a country is from the frontier, the less efficient its public sector is. The most efficient countries in the sample are Poland, Lithuania, and Kazakhstan.19 Russia, the Kyrgyz Republic, and Ukraine spend more but are less productive than Moldova (Figure 46). Figure 46: Theoretical Production Possibility Figure 47: Data Envelope Analysis, VRS Frontier input-oriented (Score) (Score) Source: World Bank calculations. Source: World Bank calculations Note: Because Poland, Kazakhstan, and Lithuania are on the frontier, their efficiency scores are equal to 1. Note: VRS- variable returns to scale. 57. Moldova could perform as well as it does even if it spent less. Its input efficiency score, using Data Envelop Analysis20, is 0.6, meaning that it should be able to attain the same level of output 19Kazakhstan’s high score is explained by high average real GDP growth, low volatility of GDP growth, low poverty rates, and relative good scores on administrative performance. In combination with a low spending-to-GDP ratio, this places Kazakhstan on the production possibility frontier. 20 Data envelop analysis (DEA) is a linear programming technique for measuring relative performance of decision making units and construction of production possibility frontiers. The method assumes that if a unit can achieve a certain level of efficiency using a given number of inputs, another unit can do the same. In this case, the calculated input efficiency determines how much less a country could spend in order to attain the same outcomes—in other words, what a country’s public 30 (public performance) using only 60 percent of the funds it currently spends (Figure 47). In this sense, Moldova is wasting 40 percent of its public resources, so there is a lot of space for making public spending more efficient. In other words, Moldova’s spending-to-GDP ratio could have been much lower and still would have got the same results. Recommendations Moldova can make its spending more efficient, with improvements especially needed in education and health; there is ample room beyond these sectors as well. The analysis in this chapter suggests the following recommendations:  Continue optimizing the primary and secondary education network and expand formula financing to other education sub-sectors. Currently general primary and secondary education budget allocations are based on a formula that takes into consideration the number of students in an institution. This has led to optimization of that network and to more efficient allocation of resources. A similar practice should be followed in pre-school, vocational education and training (VET), and higher education. The system of VET and higher education should be better aligned to the needs of the economy. Government should also allocate more resources to teacher training and educational materials.  Optimize the hospital network. Old infrastructure and the costs of Moldova’s hospitals make it difficult to reallocate resources to preventive health care. Government should rationalize the costly hospital networks while building up the quality and coverage of less costly primary and preventive care. This could be achieved by drawing up a master plan for hospital network rationalization, increasing the use of ambulatory care, building the managerial capacities pf hospitals administrators, and better monitoring health outcomes.  Reform the pension PAYG system to preserve fiscal and social sustainability. The immediate measure is to re-establish valorization of the past earnings in the pension formula along with a lower accrual rate and adequate benefit indexation. Wage valorization of the salary base and price indexation of pension benefits would improve pension adequacy for new beneficiaries and maintain the real value of future pensions. This approach corresponds to good international practice because it preserves the purchasing power of the elderly, furthers fiscal sustainability, and strengthens the link between contributions and benefits. Benefit adequacy will only be sustainable in the longer run if the retirement age is increased. The retirement age in Moldova should gradually rise to equalize the age for women at 62 and then for both genders up to 65. Increasing the retirement age, in addition to previous measures, would improve replacement rates. It would strengthen the sustainability of the pension system and allow for future reforms, such as introduction of fully-funded pensions. Supplemental pension benefits should be phased out.  Improve the targeting of social assistance. Government should increase the scope for well-targeted programs, Ajutor Social and the heating allowance, and reduce fragmentation of spending on social care facilities and categorical benefits. expenditures as a percent of GDP would be if it was on the production possibilities frontier compared to its actual public expenditure as a percent of GDP. 31  Reduce the wage bill by reducing the number of public employees. Strategic staffing and merit-based remuneration should start with the inventory of public employees and the use of a centralized fiscal oversight and reporting mechanism (FMIS). Reducing public employment at the local level should be given special attention. The fragmentation of local government authorities can be addressed by reducing the number of local authorities and by centralizing and streamlining public services. To attract and retain the necessary skilled staff, the authorities should introduce merit-based remuneration.  Reinforce procurement practices. Reducing the number of contracting authorities by unifying and matching procurement needs could be useful. Government could also reduce spending on low-priority goods by making public entitles more accountable, perhaps by publishing budget planning and provisioning and strengthening anticorruption efforts).  Better regulate public capital investment. Government should realize the potential of the new regulations on public capital investments to set up transparent and efficient mechanisms for project preparation, appraisal, and selection. It could create and regularly update a comprehensive database of investment projects, whether active and frozen. The database should be consistent with the public asset management register and other sources of information. Spending identified as inefficient, such as maintenance of unfinished construction projects, should be stopped. Asset management decisions should be guided by portfolio optimization. The government should institutionalize its monitoring and evaluation system and make the results an integral part of budget decision making. 32 Chapter III. Revenues Moldova collects more revenues than peer countries but also depends on external grants. As in most countries in Eastern Europe, the revenue structure is skewed toward taxes on goods and services. Moldova’s revenues have recently declined due to a reduction in grants from foreign donors because of policy slippages that delayed disbursements and because of the impact from the proliferation of tax exemptions. While tax administration has improved, deficiencies in capacity and governance cannot deal with the problem of high informality. Moldova could make the tax system simpler, more efficient, and boost revenues at the same time by reducing tax preferences, simplifying labor taxation, increasing the nontaxable amount of the personal income tax (PIT), improving property valuation, increasing excises, and improving tax administration. Trends and Composition 58. Revenues increased by 0.5 percent of GDP on average in 2010–15 compared to the previous decade and amounted to 37.2 percent of GDP. This moderate increase masks large shifts in composition (Figure 48). The increase in the health insurance revenues was due to introduction of the health insurance scheme in 2003. The main growth drivers of budget revenues were VAT, because of high imports and consumption (+0.7 p.p.) and external grants, because of more foreign assistance to pro-European governments since 2010 (+1.2 p.p.). Meanwhile, after the 2009 crisis, revenues from special funds and means were down 1.4 p. p. and non-tax revenues down by 0.7 p.p. Corporate income tax (CIT) collections were on average lower by 0.2 p.p., since they were zero-rated during 2008–11, while a narrower tax base because of inadequate valuation of assets explains the decrease in revenues from real estate taxes. Figure 48: Long Term Changes in Public Figure 49: Short Term Changes in Public Revenues, 2000–09 to 2010–15 Revenues, 2010 to 2015 (Percentage Points of GDP) (Percentage Points of GDP) Source: National authorities and World Bank calculations. Source: National authorities and World Bank calculations. 59. While the trend relative to the previous decade is up, more recently revenues declined by 2.5 percent of GDP from 2010 to 2015. Grants peaked at 3.7 percent of GDP in 2014, before plunging to 1.7 percent GDP in 2015, which constituted almost half of the total decline. Social security contributions also declined 0.7 p.p. because inadequate incentives to contribute (see discussion on pensions in Chapter 2) reduced the number of contributors. VAT collections declined by 1.5 p.p. because new VAT exemptions were introduced, rates were reduced, and the share of consumption in 33 GDP gradually declined. Meanwhile, CIT revenues went up 1.5 p.p. after the 12 percent rate was reintroduced in 2012 (Figure 49). 60. Moldova’s revenues are larger than those of countries with similar per capita income. In 2010–14 general government revenues, including grants, were 8 p.p. higher than the average for countries with similar incomes and 3.3 percentage points higher than the ECA average (Figure 50Figure 51). External grants have averaged 3 percent of GDP for the last 15 years—higher than Moldova’s regional peers but in line with countries with similar incomes. However, Moldova’s dependence on external grants has been trending up gradually and reached 3.7 percent of GDP in 2014. Meanwhile, other countries in the region with a relatively high share of grants managed to reduce their dependence on this source. While some large countries in the region, such as Poland and Turkey, had a higher share of fiscal collections than Moldova, countries of similar size (Lithuania, Georgia, Albania) and Romania had a lower fiscal load (Figure 52). Figure 50: Average General Government Figure 51: General Government Revenue and Revenues including Grants, 2010–14 GDP per Capita, Moldova and Other Countries (Percent of GDP) (Percent of GDP) Source: Find My Friends, using IMF WEO. Source: Find My Friends, using IMF WEO. 34 Figure 52: General Government Revenue Structure, Moldova and Peers, 2013 (Percent of GDP) Source: IMF GFS Database for 2013. 61. Similarly to most countries in Central and Eastern Europe, Moldova’s revenue structure is skewed toward taxes on goods and services while income taxes are low. For 2000–14 the former averaged 14.7 percent of GDP, a very high share among both lower MICs and globally (Figure 53). This is mostly explained by high final consumption and imports, which led to high VAT collections, and by the difficulty of verifying incomes and profits and collecting taxes on them. The CIT rate of 12 percent is among the lowest in the region, and the CIT share relative to GDP is one of the lowest in the world (Figure 54). In Moldova’s case, however, a low CIT rate has not helped to attract higher FDI inflows than peers.21 Not surprisingly, foreign investors are interested in more than just low tax rates: political stability, a predictable business environment, ease of doing business, good infrastructure, a skilled labor force, a sizable market or access to a sizable market are some of the factors helping attract foreign investors. 21 At 4.4 percent, Moldova’s share of FDI in GDP in 2014 is slightly higher than the average o f the region and of lower MICs. However, Moldova’s ability to attract FDI is poor when benchmarked against comparator economies like Albania, Croatia, and Georgia. In addition, a substantial part of the FDI stock may not be genuine foreign investment but rather a recycled funds of local origin seeking anonymity or legal protection of foreign jurisdictions 35 Figure 53: Average Revenues from Goods and Figure 54: CIT Rate, Moldova and Regional Services Taxes and CIT, Moldova and Peers, Peers, 2013 2010–14 (Percent) (Percent of GDP ) Source: Find My Friends, using IMF WEO. Source: Find My Friends, using IMF WEO and Worldwide Governance Indicators. 62. The revenue structure by sectors of the economy reflects the large presence of the state and the service-oriented nature of the economy. As Figure 55 highlights, public entities and administration paid over 27 percent of all taxes—public administration pays 7 percent of all domestic collections administered by the State Tax Inspectorate. Meanwhile, wholesale and retail trade is a clear sectoral leader, which accounted for almost a quarter of all collections, followed by other service- oriented activities: transport, communication, and real estate (Figure 56). Figure 55: Average Revenue Collected by Figure 56: Average Revenue Collected by Sector Type of Payer, 2008–14, of the Economy, 2008–14, (Percent of Total) (Percent of Total) Source: National authorities. Source: National authorities. 36 Policy and Administration 63. On the policy side, Moldovan tax initiatives resulted in adoption of a wide range of reduced tax rates and tax exemptions, with significant costs for the budget. Over the last 15 years, Moldova has introduced reduced and zero rates for VAT, reduced the CIT rate from 32 percent to 12, and to 6 percent for capital gains. PIT rates were lowered in 2004 and the number of taxable income brackets was reduced to two. However, enlargement of the tax base by abolishing tax preferences and exemptions has not accompanied the decline in rates. There is a wide spectrum of tax expenditures for small businesses, agricultural enterprises, and software development companies. Also, numerous tax expenditures in the form of lower CIT and lower VAT rates were granted to agricultural entities, residents of scientific and innovation parks, companies making investments, residents of free economic zones and recently software companies (see Table 8). Tax expenditures in agriculture alone were estimated at 0.6 percent of GDP. Tax expenditures amounted to about 60 percent of tax collections in agriculture and about 40 percent of budgeted direct spending on agriculture (World Bank 2015a). Table 8: Tax Preferences and Main Tax Exemptions Tax Description VAT Standard VAT – 20 percent Certain types of supplies – 8 percent Certain supplies, international transportation, supplies to and from the territory of Free Economic Zones, import and/or supply of goods and services destined for technical assistance projects under specific conditions, etc. – zero rate Financial services, passenger transportation services, etc. – exempt from VAT CIT Individual entrepreneurs – 7 and 18 percent Farming entities – 7 percent Legal entities – 12 percent Small and medium enterprises – 3 percent on income Capital gains tax – 6 percent PIT Standard rate – 7 percent and 18 percent Final withholding tax on certain non-wage income – 15 percent Rental income – 5 percent Dividends – 6 and 15 percent SSC and HFC Employer – 23 percent (SSC) and 3.5 percent (HFC) Employee – 6 percent (SSC) and 3.5 percent (HFC) Exemptions from Compensation of Parliament, Government, and Presidency employees related to taxable income service obligations Donations and inheritance Increase in capital shares Remuneration and wages of sportsmen Revenues of patent holders Revenues of natural persons from selling certain agricultural products Revenues of natural persons from selling certain secondary raw materials Sale of real estate Revenues from interests for individual enterprises and farmers Source: Fiscal Code of the Republic of Moldova, Title 2. 64. VAT collections, especially at customs, suffer from high noncompliance. Since 2009, VAT revenues have lagged the dynamics in import and sales due to high share of reduced and zero-rated supplies. The VAT tax gap was estimated at 4 percent of GDP in 2010–15, about 70 percent of which is attributed to VAT at customs and the rest to the Tax Service (Expert-grup 2016). 37 65. Labor taxation is complex. It ranges from relatively unprogressive taxation of full-time employment contracts to regressive taxation of self-employed workers due to lump sum payments for mandatory social and health contributions. The system also offers many opportunities for tax evasion through patent payments, self-employment contracts, and capital gains. As a result, the system does not encourage formal employment and compliance. 66. Property tax collections are among the lowest in the region and decreasing. Because real estate is not properly valued, taxes on property in Moldova are about 1 percent of total tax revenues, compared to about 5 percent in OECD countries, and it is going down. There is room for higher revenue from the property tax, if valuation is improved (Overview Note 2016). 67. Excises are low. Excise tax revenue, 3.2 percent of GDP in 2015, was significantly lower than the averages in the EU and Central and Eastern Europe (CEE) for the main products usually subject to excise duties. This is particularly true for cigarettes, petroleum fuel products, and ethyl alcohol and spirits. In 2010, total excise on cigarettes in Moldova was 5 percent of the average total excise in CEE and only 3 percent of the average in the EU. Similarly for petrol, diesel, and ethyl alcohol, and spirits for which total excise in Moldova was between 15 and 31 percent of total excise in the EU and CEE (IMF 2011). 68. High estimated tax buoyancy22 may indicate that Moldova relies a lot on discretionary revenue measures. The estimated average annual buoyancy of tax revenues for 1999–2013 was 1.5 (Papaphilippou 2014). Such a relatively high coefficient could indicate heavier reliance on discretionary measures, compared to an automatic revenue response to changes in nominal GDP. In Moldova, annual laws on the budget and on fiscal policy incorporate policy measures that could be codified and not change every year. The discretionary nature of the Moldova tax system is amplified by the large number of taxes, which complicates administration and tax policy making. 69. Tax administration has recently improved. Moldova has made substantial progress with respect to the time it takes to comply with tax requirements (Figure 57). This declining trend in compliance transaction costs was brought about by the introduction of the e-declaration system, online payment for some taxes, and the reduction in the number of declarations. 22 Tax buoyancy is an indicator to measure the responsiveness of revenue mobilization to changes in GDP. 38 Figure 57: Administrative Burden of Tax Compliance, Moldova and ECA, 2006–15 (Number per year and hours per year) Source: Doing Business. 70. However, improving services is hampered by perceptions of corruption and weakness in the tax administration. According to BEEPs (2013), about 14 percent of companies had to pay bribes in their interaction with tax administration, and small businesses were disproportionally affected (Figure 58). At the same time, weaknesses in tax administration, including inadequate attention to large taxpayers, a focus on minor compliance matters, a lack of integrity in taxpayer registration, and underdeveloped risk management throughout the tax process, all encourage the informal economy. This leads to unfair competition: according to the 2013 BEEPs, 32.3 percent of companies reported that they must compete against unregistered or informal firms. The State Tax Service (STS) recently improved risk profiling for selecting large taxpayers for comprehensive audits. As a result, the number of audits went down from 971 in 2012 to 543 in 2014, while collections from audits went up from 39 to 46 MDL million. However, information-matching (other than VAT purchase to sales) continues to be manual because information systems are fragmented. Reinforcing and expanding STS risk-based compliance management by targeting registration, filing, and payment will enhance the efficiency and increase revenue collection. Figure 58: Companies Paying Bribes to the Tax 71. Inefficient STS organization and Administration, 2013 administrative support systems make (Percent of Companies Surveyed) it difficult to improve revenue collection. STS currently has 35 local offices and the Main State Tax Inspectorate, each of which has separate legal status. The he STS Chairman appoints the heads of local offices but all other staff recruitment and human resources (HR) issues are managed locally. This is an unusual structure for Moldova, where most other public organizations, such as the Customs Service, have centralized management of human resources. Decentralized HR management severely limits selection Source: BEEPS. from the talent available and efficient allocation of staff in response to business needs. A gap analysis of management systems conducted by the Internal Control Directorate identified a lack of process 39 descriptions, criteria for performance assessments to measure the efficiency of various tax functions, inadequate human and financial resources, and risks in key business processes. The lack of mobility of staff, nontransparent procedures, and poor oversight of local operations leads to a poor governance environment and undermines STS effectiveness in managing taxpayer compliance (see Box 6 for how the Slovak Republic dealt with similar problems). Box 6: Tax Reform in the Slovak Republic In 2004 Slovakia introduced a reform to make the tax system simple, easy to administer, and neutral to different types of income. It decided to shift more tax burden from direct to indirect taxation, since direct taxes represented more than 50 percent of tax revenues. The authorities decided to take care of social and distributional policy not though taxation but through spending policy, such as means-testing benefits to the poor. The tax reform had the following features:  Broader base and lower income tax rates. Almost all exemptions and specific lower tax rates from CIT and PIT were cancelled, notably for new firms, substantially broadening the tax base and aligning investment incentives with EU state aid rules. The broader base made possible the CIT rate reduction from 25 to 19 percent without negative fiscal impact. Three PIT brackets were replaced by a single rate of 19 percent and a substantial non-taxable amount to ensure progressivity.  Special regime for the self-employed. Self-employed persons have to pay only half the social security contribution rate (35.2 percent) and taxes on 60 percent of their income. This policy is currently being reconsidered because it has created legal opportunities for evasion; the number of self-employed has increased to 20 percent of labor force. Slovak authorities are also thinking about abolishing mandatory healthcare contributions and including them in general taxation.  Cancelation of “unproductive” taxes. Slovakia completely abolished several taxes that had low returns: inheritance, property transfer, dividend tax, several local taxes. The social impacts were calculated and a tax bonus (negative income tax) on families with children was introduced.  Single VAT rate. The differentiated VAT rates of 10 and 23 percent were unified at 19 percent.  Higher excises. The authorities steeply increased excises on petrol, alcohol, and tobacco. The tax reform turned out to be almost revenue neutral and contributed to higher growth. A comparison of the estimated tax revenues that would have been received in 2004 without reform and the revenues actually received suggests that the decline in PIT revenues (estimated at 0.8 percent of GDP) and CIT revenues (0.8 percent of GDP) was almost entirely offset by the increase in VAT revenues (0.8 percent of GDP) and excise revenues (0.6 percent of GDP). In the 10 years after 2004, Slovakia grew 3.9 percent on average, one of the top three fastest-growing economies among new EU member states. Source: Based on Brook and Leibfritz 2005 Recommendations Moldova’s tax system can be made simpler, more efficient, and more revenue-enhancing. A simple tax system is easier to administer and has lower transaction costs for businesses. One step toward a simpler system would be to reduce or remove various tax privileges and special treatments, though additional steps may be needed to reap more revenue without raising tax rates. In the short term, additional revenues might supplement expenditure measures to ensure fiscal sustainability, since the economy is confronting a number of external and climate-related shocks. In the longer term, Moldova would need to find substitutes for external grants, which would gradually become a relatively smaller source of revenue. The authorities are invited to consider the following recommendations:  Reduce tax preferences. Rather than raising nominal tax rates, the government should concentrate the efforts on increasing tax productivity by reducing the scope for exemptions 40 and preferential rates. Closure of VAT loopholes has the highest revenue potential, given the number of exempted, reduced, and zero VAT rates and the high VAT administration costs. Of particular concern is the preferential treatment of agricultural products, passenger transport, and financial services. Fiscal preferences, such as zero-rated VAT supplies in the Free Economic Zones could be replaced by provision of targeted services (World Bank 2015b). The authorities could also consider introducing VAT taxation of cars. Streamlining the tax deduction and the VAT refund process and discouraging or better analyzing credit VAT will reduce the tax gap and fraudulent activities. The preferential CIT rates for farmers, SMEs, and individual entrepreneurs should be standardized and y a simple and targeted system of tax deductions or tax bonuses introduced. The authorities could consider cancelling tax preferences for some types of taxable revenues and revising PIT deductions and exemptions, especially in those in agriculture and individual entrepreneurs.  Improve tax administration, reduce compliance costs, and simplify the tax structure. A modern tax service organization could deliver better services to taxpayers. To encourage compliance, reduce burdensome bureaucratic requirements to ensure easy registration, filing, accurate declarations, and risk-based analysis. Here indirect methods applied on taxation of high-wealth individuals and making the large taxpayer units more efficient is vital. In drafting the new Tax Code, consider a simpler tax structure. At the moment, over 90 percent of budget revenues are collected from nine taxes, but there are over 100 different taxes, fees, and other sources of revenue.  Over the medium term raise excises in line with the EU integration agenda. The excise rates need to increase gradually in line with the EU requirements. The authorities could also abolish excises on certain products (gold, caviar, electro, coffee, crystal glass, etc.) where tax administration costs more than the revenues from taxation of these items brings in.  Make the PIT more progressive by increasing the nontaxable amount to the minimum living standard. In addition, the one-off patent payments could be replaced by an easy-to- administer tax deduction for self-employed professionals in order to assure that labor taxation is more progressive for everyone.  Improve property valuation to collect more from real estate taxes. Collections could go up significantly if the value of assets was adjusted to market value. 41 Annexes Annex 1. Key Macroeconomic Indicators (Percent of GDP unless otherwise indicated) 2009 2010 2011 2012 2013 2014 2015 Nominal GDP, MDL billion 60.4 71.9 82.3 88.2 100.5 112.1 121.9 Nominal GDP, U$D billion 5.4 5.8 7.0 7.3 8.0 8.0 6.5 Nominal GDP, PPP, constant US$ billion1 12.8 13.7 14.6 14.5 15.9 16.7 16.6 GDP per capita, US$ 1,525 1,632 1,971 2,046 2,243 2,244 1,822 GDP per capita, PPP, constant US$1 3,592 3,846 4,111 4,085 4,473 4,693 4,676 Population, million 3.6 3.6 3.6 3.6 3.6 3.5 3.5 GDP, % real change -6.0 7.1 6.8 -0.7 9.4 4.8 -0.5 Consumption, % real change -0.9 9.2 9.4 0.9 5.2 2.7 -1.9 Gross fixed investment, % real change -30.9 17.2 13.0 0.4 3.8 10.0 -1.2 Exports, % real change -12.1 13.7 27.4 2.3 9.6 1.0 2.3 Imports, % real change -23.6 14.3 19.7 2.5 4.4 0.4 -4.3 GDP deflator, % change 2.2 11.3 7.2 7.9 4.1 6.4 9.6 CPI, % change, average 0.0 7.4 7.6 4.6 4.6 5.1 9.7 Current account balance -8.2 -8.3 -12.1 -8.7 -6.4 -7.1 -7.2 Trade balance -38.8 -40.3 -40.6 -37.5 -36.7 -30.0 -38.8 Remittances2 20.7 21.9 22.1 25.2 25.3 24.1 21.8 Remittances, % change, US$ -36.2 13.2 21.7 8.8 10.1 -4.9 -26.7 Direct investment 4.8 4.8 4.5 3.5 3.1 3.9 4.1 Terms of trade, % change 0.1 0.0 -1.4 0.8 -0.5 -1.0 5.5 External debt 79.6 81.0 76.4 82.6 83.6 82.5 97.9 Budget revenues 38.9 38.3 36.6 38.0 36.7 37.8 35.8 Grants 2.1 2.8 2.1 1.9 2.1 3.7 1.7 Budget expenditures 45.3 40.8 39.0 40.1 38.5 39.6 38.1 Wages 11.7 10.3 9.5 9.8 8.4 8.2 8.5 Transfers to population 14.3 13.5 13.1 12.6 12.0 12.0 12.5 Capital expenditure 5.1 4.8 5.2 6.3 7.1 8.5 6.5 Fiscal balance -6.3 -2.5 -2.4 -2.1 -1.8 -1.7 -2.2 Public debt and guarantees 29.0 31.9 30.3 33.2 31.7 32.5 45.5 Source: Moldovan authorities. Notes: 1 Data are in constant 2011 international dollars. 2 Starting in 2012 BMP6 is used. 42 Annex 2: Data and Methodology for Econometric Estimation of Determinants of Private Savings The data sources used in this empirical exercise are World Development Indicators (WDI) and the publications of the National Bureau of Statistics of the Republic of Moldova. The analysis is based on the entire coverage period of 1997–2014 less one year lost due to differencing. The WDI cover a large number of variables, such as gross savings as a percent of GDP, GDP per capita (PPP, current international dollars), deposit interest rates, and the real effective exchange rate index. Unfortunately, it does not decompose public and private gross savings. However, the National Bureau of Statistics provides a breakdown of savings by nonfinancial organizations, financial organizations, public administration, services rendered to the population, and households. Thus, it was possible to construct private gross savings and public gross savings series by applying the share of public savings in total savings calculated by the National Bureau of Statistics to the gross savings obtained from WDI. The National Bureau of Statistics also provided fiscal variables, since data for recent years are not available in WDI. Fiscal variables are available for 2000–2014, and these were extrapolated back to 1997 by using the rate of changes obtained from the WDI data. The Augmented Dickey-Fuller unit root test was applied to all variables to check whether all series are stationary. However, the results suggest that most of the variables have trends and are non- stationary in levels. Thus, the analysis took first differences of all variables except real GDP per capital and the real exchange rate in order to make the series stationary. Real GDP per capita and the real exchange rate were made stationary by taking log differences to make them stationary. The estimation uses OLS with HAC VCov Matrix. 43 Annex 3: Data on Potential Contributions of Spending and Revenue Measures to Fiscal Consolidation Table A3: Potential Contributions of Spending and Revenue Measures to Fiscal Consolidation Percent of GDP AUS AUT BEL CAN CHE CZE DEU DNK ESP FRA FIN GBR GRC HUN ISL 1. Social transfers A. Family benefits 0.5 0.7 0.6 - - 0.1 - 1.4 - 1.1 1.0 1.3 - 1.4 1.0 B. Disability benefits - 0.3 0.2 - 0.5 0.5 - 1.3 0.6 - 1.0 0.3 - 0.6 - 2. Pensions A. Elimination of tax breaks 2.7 0.1 0.1 2.0 0.1 0.8 0.2 0 0 1.2 1.0 3. Health care A. Increase efficiency 0.5 1.8 2.1 2.5 0.5 1.3 1.3 2.8 1.6 1.3 3.0 3.7 3.9 1.7 2.0 4. Education A. Increase efficiency in primary and secondary 0.4 0.4 0.5 0.2 0.2 0.2 0.4 0.6 0.2 0 0.2 0.3 1.0 education B. Introduce or raise tuition - 0.4 0.4 - 0.4 0.3 0.4 0.4 0.2 0.3 0 - 0.4 0.4 0 fees for tertiary education 5. Government wage bill A. Restore public-private - 0.3 0.6 - - 0.4 0.2 2.0 1.0 - 1.0 1.8 - - - sector pay relativities 6. Reduce subsidies as share of - 2.3 0.8 - 2.4 0.7 - 1.2 - 0.2 - - - - 0 GDP to OECD average 7. Broaden VAT base 0.6 - 1.4 - - - 0.4 - 1.4 1.4 0 1.8 2.0 0.1 1.0 8. Introduce or increase taxes - 0.8 0.6 - 0.9 0.8 0.6 - 0.3 - 1.0 - 0.8 0.7 - on immovable property 9. Environmental taxes A. Cut GHG emissions via an emission trading system with 4.2 1.8 1.8 2.5 1.8 1.8 1.8 1.8 1.8 1.8 2.0 1.8 1.8 1.8 full permit auctioning Table A3: Potential Contributions of Spending and Revenue Measures to Fiscal Consolidation (continued) IRL ITA JPN KOR LUX MEX NLD NZL NOR POL PRT SVK SWE TUR USA 1. Social transfers A. Family benefits 0.7 - - - 1.2 - 0.1 1.1 0.9 - - - 1.4 - - B. Disability benefits - - - - 0.1 - 0.8 0.7 1.8 0.6 0.3 - 1.3 - - 2. Pensions A. Elimination of tax breaks 1.2 0 0.7 0.5 0.2 0.6 0.2 0.1 0.2 0.8 3. Health care A. Increase efficiency 4.8 1.1 0.8 0.6 2.0 0.7 2.7 2.6 1.5 1.5 1.0 2.7 2.7 1.5 2.7 4. Education A. Increase efficiency in primary and secondary 0.3 0.4 0.2 - 0.5 - 0.3 0.3 0.8 0.2 0.1 0.2 0.5 - 0.8 education B. Introduce or raise tuition 0.3 0.2 - - 0.4 0.1 0.2 - 0.4 0.1 0.1 - 0.4 0.4 - fees for tertiary education 5. Government wage bill A. Restore public-private 0.9 1.1 0.6 - 0.8 - 0.3 0.9 - 2.2 - 0.8 0.7 - 0.5 sector pay relativities 6. Reduce subsidies as share of - - - - 0.2 - 0.1 - 0.7 - - 0.2 0.1 - - GDP to OECD average 7. Broaden VAT base 0.4 2.6 - - - 2.5 - - 0.2 1.4 1.2 0.6 - 3.3 8. Introduce or increase taxes 0.2 0.4 - 0 0.9 0.8 0.4 - 0.7 - 0.3 0.6 0.2 0.9 - on immovable property 9. Environmental taxes A. Cut GHG emissions via an 1.8 1.8 1.2 1.8 1.8 4.2 1.8 1.8 1.8 1.8 2.2 emission trading system with 44 full permit auctioning Source: OECD 2012. aAn empty cell indicates that no information was available. Cells with a dash indicate that no savings are available from this source. AUS Australia IRL Ireland AUT Austria ITA Italy BEL Belgium JPN Japan CAN Canada KOR Korea CHE Chili LUX Luxemburg CZE Check Republic MEX Mexico DEU Germany NLD Netherlands DNK Denmark NZL New Zealand ESP Spain NOR Norway FRA France POL Poland FIN Finland PRT Portugal GBR Great Britain SVK Slovakia GRC Greece SWE Sweden ISL Island TUR Turkey USA United States 45 References Alesina, A., & Ardagna, S. (2010). Large Changes in Fiscal Policy: Taxes versus Spending. Tax Policy and the Economy, Volume 24. Antolin, P., de Serres, A., & Maisonneuve, C. (2004). Long-Term Budgetary Implications of Tax-Favoured Retirement Plans. OECD Economics Department Working Papers, No. 393. OECD. Blöndal, S., Field, S., & Girouard, N. (2002). Investment in Human Capital through Post-compulsory Education and Training: Selected Efficiency Aspects. OECD Economics Department Working Papers, No. 333. OECD Publishing. Groenendijk, N., & Jaansoo, A. (2015). Public finance Systems for Coping With the Crises: Lessons from the Three Baltic States. Public Finance and Management Volume 15, Number 3, pp. 257-289. IMF. (2010, October). Will It Hurt? Macroeconomic Effects of Fiscal Consolidation. World Economic Outlook Chapter 3. International Monetary Fund. IMF. (2014, April). Public Expenditure Reform: Making Difficult Choices. Fiscal Monitor . IMF. Joumard, I., André, C., & Nicq, C. (2010). Health Care System: Efficiency and Institutions. Economics Department Working Papers No: 769. OECD. Molnar, M. (2012). Fiscal Consolidation: Part 5. What Factors Determine the Success of Consolidation Efforts? OECD Economics Department Working Papers 936. OECD. OECD. (2012, April). Fiscal Consolidation: How Much, How Fast and by What Means? OECD Economic Policy Papers No. 01. OECD. OECD. (2012, April). What are the Best Policy Instruments for Fiscal Consolidation? OECD Economics Department Policy Notes, No. 12. Serres, A., Murtin, F., & Nicoletti, G. (2010). A Framework for Assessing Green Growth Policies. OECD Economics Department Working Papers, No. 774. OECD Publishing. Sutherland, D., Price, R., Joumard, I., & Nicq, C. (2007). Performance and Indicators for Public Spending Efficiency in Primary and Secondary Education. OECD Economics Department Working Papers, No. 546. OECD Publishing. 46