TAX POLICY UNDER THE CURSE OF LOW REVENUES:

Transcription

1 Friedrich Ebert Stiftung Büro Rumänien TAX POLICY UNDER THE CURSE OF LOW REVENUES: THE CASE OF ROMANIA by Daniel Daianu, (professor of economics, SNSPA, Bucharest Ella Kállai, (chief economist, Alpha Bank Romania), Bucharest Laurian Lungu (managing partner, Macroanalitica, Bucharest 18 October

2 Friedrich-Ebert-Stiftung Office Romania 21 Emanoil Porumbaru Str., 1st Floor, Ap. 3 Sector 1, Bucharest Tel.: /-83 Fax: fes@fes.ro 1 We thank Dan Armeanu, Nicolae Chidesciuc, Giacomo Corneo, Margit Schratzenstaller-Altzinger, Dieter Vesper and Radu Vranceanu for their comments on an earlier draft. We also thank the Friedrich Ebert Stiftung for the financial support in undertaking this analysis. It goes without saying that we bear sole responsibility for its content Friedrich Ebert Stiftung Romania

3 Table of Content Pg. Introduction 5 1. The European Experience and Policy Responses During the Crisis Tax Sensitivity to the Business Cycle Tax Policy Response in the EU 8 2. The Romanian Experience Tax Revenues Dynamics ( ) The early Downsizing ( ) The Recovery ( ) Peak-to-trough ( ) Stabilisation (-2008) Stylised Facts of Current Tax Revenues Tax Revenues Size The Flat Tax and Changes in the Tax Revenues Structure Structure by Tax Categories Structure by Economic Function Legal Tax Rates and the Tax Burden Implicit Tax Rates and the Tax Collection Efficiency Taxpayers Households Companies Summing Up The Black Economy in Romania and its Impact on Tax Revenues Withstanding the Current Economic Crisis and Fiscal Consolidation Final Remarks and Policy Recommendations 38 Appendices 42 References 52

4 2 List of Tables Table 1.1 Expansionary Stimulus Measures and Consolidation Measuresin EU, % of GDP 6 Table 1.2 Tax Revenue Sensitivity in EU 8 Table 1.3 Tax Measures in the EU 9 Table Main Macro Indicators in Romania, EU-27 and NMSs 16 Table Tax revenue-to-gdp ratio in South Eastern Europe 17 Table Structure of Tax Revenues by Tax Categories 19 Table Structure of Tax Revenues by Economic Function of Taxes 20 Table Notional and Collected Tax Revenue-to-GDP ratios in Table Notional and Collected Tax Revenue pre, post and in Table Households' Budget Income Structure 30 Table Households' Budget Income Structure by Deciles 30 Table Structure of Households' Budget Expenditures 32 Table Structure of Households' Budget Expenditure by Deciles 32 Table 3.1 Fiscal Evasion and the Size of Informal Economy 34 Table 3.2 Sensitivity Analysis of Improving Tax Evasion, % of GDP 35 Table 4.1 Structure of Tax Revenues , % of GDP 37 List of Tables in Appendix 4 Table A4.1 Public Debt in EU 48 Table A4.2 Changes of the Main Taxes in Romania, Table A4.3 Main Legal Taxes in NMSs 51 List of Graphs Fig. 1 Revenues and expenses of consolidated budget 11 Fig 2 Total revenues and tax revenues of consolidated budget 11 Fig. 3 Main statutory tax rates in Fig. 4 Romanian tax revenue-to-gdp gap 17 Fig. 5 Legal Tax Rates, Fig. 6 ITR-to-LTR ratio 25 Fig. 7 LTR, ITR relationship in NMSs, Fig. 8 Tax revenue and LTR in NMSs, List of Graphs in Appendix 1 Tax Revenues Romania vs.eu-27 and NMSs Fig. A1.1 Fig. A1.2 Fig. A1.3 Fig. A1.4 Fig. A1.5 Fig. A.1.6 Fig. A1.7 Fig. A1.8 Tax revenues including social contribution Tax revenues including social contribution (cyclically adjusted) Indirect taxes Direct taxes Social contribution Consumption taxes Taxes on labour Taxes on capital

5 Tax Policy Under The Curse of Low Revenues: The Case of Romania 3 Fig. A1.9 Implicit tax rate on consumption Fig. A.1.10 Implicit tax rate on labour Fig. A1.11 Personal income tax and social contribution paid by individuals Fig. A1.12 Corporate income tax and social contribution paid by employers List of Graphs in Appendix 2 Households as taxpayers Fig. A2.1 Income tax & social security contribution share vs. social benefits share of poorest household decile, % total Fig. A2.2 Income tax & social security contribution share vs. social benefit share of richest household decile, %total Fig. A2.3 Income tax and social security contribution share vs. social benefit share of middle income household deciles, % total Fig. A2.4 Income tax and social security contribution-to-income ratio for low, high, middle income household deciles Fig. A2.5 Income tax and social security contribution share vs. social benefit share of employees household Fig. A2.6 Income tax and social security contribution share vs. social benefit share of agricultural household Fig. A2.7 Distribution of tax burden among households deciles Fig. A2.8 Distribution of social benefit among household deciles Fig. A2.9 Income tax and social security contribution share vs. social benefit share of retiree household Fig. A2.10 Income tax and social security contribution-to-income ratio by household type List of Graphs in Appendix 3 Companies' arrears to consolidated general budget Fig. A3.1 Arrears to consolidated general budget, by type Fig. A3.2 Arrears to consolidated general budget of state companies by type Fig. A3.2 Arrears to consolidated general budget of private companies by type Fig. A3.4 Arrears to consolidated general budget, by companies' ownership Fig. A3.5 Arrears to state budget, by companies' ownership Fig. A3.6 Arrears to social insurance budget, by companies' ownership Fig. A3.7 Arrears to consolidated general budget by type and sectors, June 2010

6 4 Glossary Country abbreviations Commonly used acronyms BE Belgium NMSs New Member States BG Bulgaria EU European Union CZ Czech Republic EU-25 European Union (25 Member States) DK Denmark EU-27 European Union (27 Member States) DE Germany EMU Economic and Monetary Union EE Estonia EC European Commission IE Ireland ESA95 European System of Accounts 1995 EL Greece IMF International Monetary Fund ES Spain OECD Organization for Economic Co-operation and Development FR France GDP Gross Domestic Product IT Italy PPS Purchasing Power Standard CY Cyprus PIT Personal Income Tax LV Latvia CIT Corporate Income Tax LT Lithuania SSC Social Security Contribution LU Luxemburg LTR Legal Tax Rate HU Hungary ITR Implicit Tax Rate MT Malta HH Households NL Netherlands AT Austria PL Poland PT Portugal RO Romania SI Slovenia SK Slovakia FI Finland SE Sweden UK United Kingdom

7 Tax Policy Under The Curse of Low Revenues: The Case of Romania 5 Introduction Romania has registered persistently low budget revenues over the years. The Romanian tax revenue-to-gdp ratio has been far below the average level of both the European Union (EU-27) 2 and the New Member States (NMSs ) for many years. The decade long growth cycle hid significant structural imbalances in the public budget. The global financial crisis, which erupted in 2007, has had a strong negative impact on the Romanian economy. The ensuing fall in GDP growth lowered tax revenues and forced the authorities to come up with a fiscal consolidation package. A choice made for financing the mounting budget deficit and securing financial stability was an international loan package, which was agreed upon in May The attached economic programme aimed at stabilising and consolidating Romania's fiscal position. But measures for raising tax revenues, which are particularly low in Romania, are still to work their way, or are awaited. This paper attempts to identify reasons why tax revenues in Romania are the lowest (as a share of GDP) among the EU-27 countries. It takes a broader perspective by looking at the main sources of tax revenues over the last two decades. Implications of the policy regime change following the introduction of flat tax in are considered. It also does a few comparisons with other countries from Central and Eastern Europe by looking at the main tax revenue components. The analysis looks at effects of the shadow economy on fiscal revenues by attempting to quantify the revenue loss due to economic activities, which are not taxed. After examining EU wide responses to the current crisis a basic part deals with dynamics of Romanian public budget revenues. The last section lists policy recommendations. 1. The European Experience and Policy Responses during the Crisis The current financial crisis has raised, once again, the issue of public sector fiscal sustainability. The costs of bailing out the financial sectors have placed a large burden on public sector debt for years to come in a number of countries. But, the choice between pursuing fiscal austerity or economic policies that stimulate growth continues to remain a hard one, though one could argue that short term stimulus can be reconciled with longer term fiscal consolidation. While the US, for instance, favoured by the global status of its currency could afford to use monetary and fiscal policy for spurring economic growth, for small open economies which run large budget deficits and need access to external finance fiscal consolidation can hardly be delayed in the short and medium term. Recent economic developments in Europe have shown that indebted countries face now more difficulties than they encountered in the past, when they needed to rollover their debts. Marked differences in competitiveness and the state of public finances have increased sovereign risk premia for countries both within the euro-zone and Europe in general. The crisis has led to large increases in budget deficits and government debts in many countries. In simulations performed under current scenarios, government debt in advanced economies is forecast to rise, on average, by about 35 percentage points of GDP between 2007 and Moreover, primary deficits are expected to remain high even as economic growth picks up. Large scale fiscal adjustments in most developed countries were needed before the crisis as pensions and health care costs had been on the rise. The cost of current financial crisis would merely add on to those. Fiscal consolidation is expected to be highly challenging as countries would have to maintain sustainable debt levels and structural primary balances would have to improve considerably, be it gradually though. A sharp increase in fiscal risk and rising problems with the financing of private debt, have prompted discussions of the urgency of fiscal consolidation and the need to reform the public sector and improve the system of taxation. 2 NMS includes Bulgaria, the Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Slovenia and Slovakia.

8 6 It is obvious that the scale and composition of adjustment need to be tailored to the specific conditions of individual countries. The necessary adjustments would have to come primarily from structural fiscal reforms. IMF (2010, February) list a series of actions which could be taken such as: reforms aimed at stabilising entitlement-spending-to-gdp ratio; measures to lower other primary spending in relation to GDP; or increased revenue, for instance by broadening tax bases but also tax rate hikes. The crisis has had a sizable impact on the public finances of European economies. The challenge of addressing high and rising budget deficits and public debts across the EU is a stringent one. Prior to the crisis, in 2007, the fiscal position of most EU countries seemed to be one of a relative strength, especially in most of the NMSs. According to the EC (2010a), in 2007 government deficits amounted to less than 1% of GDP on aggregate in EU27 while in 2010 they went beyond 7% of GDP. Public debt has also been going up markedly, rising from 59% in 2007 to almost 80% of GDP in 2010 as higher government borrowing was triggered by widening budget deficits. This strong deterioration in public finances has four main causes: the fall in tax receipts caused by a lower level of economic activity, the adoption of discretionary support measures introduced by governments across the EU countries in an 3 attempt to prop up economic growth, effects of automatic stabilisers and bail out 4 operations of the financial industry in the main. Table 1.1 below depicts the magnitude of stimulus and consolidation measures in EU countries during 2009 and 2010 under the European Economic Recovery Plan. Table 1.1 Expansionary stimulus measures and consolidation measures in EU, % of GDP* Discretionary stimulus 2009 Consolid 2009 Discretionary stimulus 2010 Overall Out of Overall Out of which, which, T P T P Consolid 2010 BE BG CZ DK DE EE IE EL ES FR IT CY LV LT LU In December 2010 the European Council endorsed the European Economic Recovery Plan (EERP), which allowed for the introduction of a discretionary fiscal stimulus aiming to boost demand. A large part of stimulus measures implemented in both 2009 and 2010 are temporary and would be phased out by Since other industries were also propped up by government intervention (ex: the car industry in the US, Germany, France, etc)

9 Tax Policy Under The Curse of Low Revenues: The Case of Romania 7 HU MT NL AT PL PT RO SI SK FI SE UK EU * T- temporary measures, P permanent measures. Differences are due to rounding effects. Source: Adapted from EC (2010d) Several points can be made: In general, consolidation measures were much larger than stimulus measures so that the net effect on the budget balance has been positive In New Member States (NMSs) which were deeply affected by the crisis and confronted by liquidity crises (sudden stops), such as Romania, Hungary, or the Baltics, consolidation measures prevailed. The scope of fiscal stimulus was reduced in all these countries. 5 Temporary stimulus measures were also favoured instead of permanent measures. Consolidation measures tended to be more pronounced in 2010 compared to 2009 as the magnitude of needed official adjustment became more evident was also the year that triggered a sovereign debt crisis in the EMU. Except Hungary, NMSs do not have large public debts. But budget deficits have gone up dramatically in the wake of this crisis. Moreover, not a few NMSs were running meaningful structural deficits prior to the crisis. Subsequent to the crisis, a few countries, the Baltic countries, Hungary, Bulgaria and Romania among others, were forced to implement fiscal consolidation programs due to the impairment of economic growth --against the backdrop of a highly unfriendly external environment that has been entailed by the turmoil in financial markets. But, as Becker et al (2011) note, fiscal consolidation has to take into account the risk of adding public deleveraging to the ongoing private deleveraging, a factor which could harm economic recovery and future economic growth. 1.1 Tax Sensitivity to the Business Cycle 6 Economic growth plays an important role in the evolution of tax revenue ratios. Higher growth tends to increase tax receipts while recessions decrease them. Between 1999 and 2002 tax revenues in the EU fell, as economic activity slowed down, but subsequently rose again until 2007, before the current crisis reversed, once again, this trend. However, once the effect of the business cycle is removed, the data shows that there was almost no structural increase in the overall tax ratio, this appears to have been rather flat throughout the period. Cyclical factors contribute decisively to fluctuations observed in tax ratios data. 5 Such as financial aid programs for medium-sized enterprises, financial incentives for companies which hired new employees, etc. 6 Tax revenue ratio measures the proportion of budgetary tax revenues in GDP.

10 8 In fact, cyclically adjusted revenues were even marginally higher, about 1%, in 2003 than in 2007 (EC, 2010b). This marginal fall could be attributed to corporate income tax cuts during the boom years. The deterioration of economic conditions, starting with 2008, impacted negatively tax revenues, which reached their lowest level since Thus, the apparent increase incurred during period in EU revenue tax ratios was due to the exceptionally good global economic situation. In fact, structural tax revenues have been falling markedly over that period. 7 Tax revenue sensitivity is an important parameter because it allows to establish a link between tax revenues and economic activity. Table 1.2 below presents estimates of tax revenue sensitivity in the EU countries: Table 1.2 Tax Revenue Sensitivity in EU Member State Tax Revenue Sensitivity Member State Tax Revenue Sensitivity Member State Tax Revenue Sensitivity Member State Tax Revenue Sensitivity BE 0.47 EL 0.42 LU 0.48 RO 0.28 BG 0.35 ES 0.38 HU 0.45 SI 0.42 CZ 0.36 FR 0.44 MT 0.35 SK 0.27 DK 0.50 IT 0.49 NL 0.39 FI 0.41 DE 0.40 CY 0.39 AT 0.43 SE 0.48 EE 0.29 LV 0.26 PL 0.33 UK 0.40 IE 0.36 LT 0.26 PT 0.41 EU Source: OECD, Commission services & 2010 Taxation Trends Several remarks can be drawn from the data above: Countries showing the highest sensitivity of tax revenues to economic developments are Denmark, Italy, Sweden and Luxembourg. Here economic growth has been either dismal, as in the case of Italy, preventing disparities between structural tax and unadjusted tax revenues emerging, or tax policies have been pursued wisely, paying attention to the cyclically adjusted tax values. Thus, these countries did not have, in general major problems in financing their deficits in the face of the slowdown in economic growth. The Baltic countries, Romania and Slovakia exhibit the lowest response to the economic cycle. However, even though these countries present the lowest tax sensitivity, tax revenues in most of them were impacted dramatically by the financial crisis since the cyclical component was extremely high vis-à-vis average EU-27. All of these countries experienced higher growth rates in the years prior to the crisis, thus leaving them exposed to the rapid deterioration in tax revenues when economic growth fell abruptly. 1.2 Tax Policy Response in the EU The tax policy response across the EU members during the current financial crisis has been quite complex and has been influenced by the initial budgetary conditions, the structure of the economy as well as the extent to which each individual economy was affected by the crisis. Automatic stabilisers played their role but the degree to which they did so varied in each country. In general EU governments took an activist stance and the course of action for each individual country has been influenced by a series of factors such as: 7 Tax revenue sensitivity is an indicator that measures how tax revenues are influenced by economic activity.

11 Tax Policy Under The Curse of Low Revenues: The Case of Romania 9 The extent to which the provision of social services (pensions, health care) was provided by the public sector. Some EU countries have a share of these services provided by the private sector, for these the budgetary impact was smaller. Technical factors. Some EU members provide social or economic assistance via tax reductions rather than direct government spending (EC, 2010c). Social transfers are exempted from taxes and social security contributions in some EU countries but not in others. The shares of direct and indirect taxes in the overall tax structure. A tax system where direct taxes have a large share tends to allow for higher redistributional effects. Thus, economies where redistributional effects are larger, have been inclined to use more direct taxes, which are also more 'visible' to the electorate. There is a noticeable distinction between the old EU members and the new ones when it comes to the structure of tax rates. The former tend to raise almost equal share of revenues from direct taxes, indirect taxes, and social security contributions, while the latter have a lower share of direct taxes Table 1.3 below summarises the tax measures taken by the EU members during the current financial crisis. Table 1.3 Tax Measures in the EU Statutory Rate Base or special regimes Corporate Income Tax Increase LT, HU, PT BE, BG, IE, EL (09-13), IT, LT (09-11), HU Decrease CZ, EL(10-14), HU, LU, SI, SE, LT AT, BE (10-11), ES (09-11), CY, IT, LT, NL, PL, PT, RO, SE, SK, UK (09-11) Personal Income Tax Increase EL, IE, FR, LV, PT, SI, UK DK, EE, EL, ES, IE, HU, LV, LT, PT Decrease AT, DE, DK, FR, FI, HU, LV, LT, RO AT, BE, BG, DE, DK, ES, FI, HU, MT, IE, IT, LV, LU, NL, PL, PT, RO, SK, SI, SE Social Security Contributions Increase CY, EE, HU, PT, RO, SK, FI BG, CZ, EE, LV, LT Decrease BG, CZ, HU, RO, SE FI Value Added Tax Increase CZ, EL, ES, EE, HU, LV, LT, FI EE, LV, LT Decrease IE, FI, UK (08-09) BE, DE, CY, FR, LT, MT, HU, NL, RO, SI, FI Excise Duties Increase BG, DK, EE, EL, ES, IE, HU, LV, LT, PT, DG, FI, EL, LV PL, RO, SI, FI Decrease IT, LT, PL, SK BG Source: European Commission (2010a) The EU members' response varied significantly. It can be observed that, within all main tax categories, some EU member countries introduced both tax increases and tax cuts over the past two years. This behaviour could be attributed to responses to the different phases of the crisis or the implementation of changes within the same tax component such as reducing tax breaks while introducing new incentives. Several remarks can be made on the EU tax responses to the financial crisis.

12 10 Some countries which had fiscal space implemented cuts in corporate and personal income taxation. Although this measure was not of an immediate help to companies which were already making losses, the government revenue loss was deemed to outweigh the benefits 8 of the positive signal to investors. Tax increases were prevalent in excise duties and VAT. With respect tot the latter, the tendency has been to raise the base rates. Support of households purchasing power. The reduction of personal income tax was usually implemented through an increase in household allowances rather than lowering rates. Also, in a few cases top rates have been raised, following fairness concerns. Social security contributions have generally been increased. Although the majority of the measures adopted has had an estimated budgetary impact of well below a half point of GDP, the overall impact of the adjustment turned out to be higher ex post. Adjustments in the tax rate sometimes ended up to around 1% of GDP. Reforms of the VAT, the PIT or the reforms of social security, as well as some excise rate increases, have often involved large amounts in terms of GDP. Many countries have pursued significant changes in the tax mix. Bulgaria, for instance, shifted 9 the burden of taxation from social contributions to indirect taxes. In Latvia and Slovenia direct tax increases almost compensated for decreases in social contributions and indirect taxes. In Greece, a strong decline in both direct and indirect tax revenues was partly offset by increases in social security contributions. 2. The Romanian Experience We examine the tax revenues performance in Romania during the last two decades. We present briefly tax revenue dynamics and tax policies implemented during transition and up to 2008, the end of the expansionary cycle. Then, we analyse main stylised facts of tax revenues in terms of size, statutory tax rates, efficiency of tax collection and taxpayers' response during the expansionary cycle between 2000 and We also provide a simple model for estimating the dimension of the shadow economy. Last, we present the tax policy response during the current crisis and estimated tax revenues for the following two years. 2.1 Tax revenue dynamics (the ) Over the last twenty years, after the economy started its transition from central planning to market economy tax revenues exhibited a declining trend. They declined abruptly up to 1994, fell again mildly up to 1999 being influenced by the electoral cycle thereafter. The total budget revenue, as percentage of GDP, registered a maximum at 37.5% in 1992 before reaching the precrisis maximum of 32.7% in (Fig.1). The tax revenue followed an identical path declining from 33.5% in 1992 to 27.8% in 1998 and reaching 30% in 1999 (Fig. 2). The almost a decade long decline of the consolidated budget revenue in the 90s is the result of the adjustment of the public sector to the new emerging economic environment. Several causes/explanations were advanced. 8 There have been other measures adopted by many EU countries, aimed at supporting business investment, such as favourable depreciation allowances or investment tax credits. Incentives have been usually granted for a limited period of time. 9 Indirect taxes could be a larger source of tax revenue if net migration is negative (i.e. out-migration is larger than in-migration) or when a part of working contracts in the economy are not officially registered.

13 Tax Policy Under The Curse of Low Revenues: The Case of Romania 11 The first was that a disorganisation effect à la Blanchard (1997) due to the abolishment of the 10 centralised coordination resulted in lower tax collection. Departing from an economy almost totally state owned, in just nine years around 61% of GDP was produced by privately owned firms. The second cause was the formidable shrinking of the tax base. The GDP was 83% of the 1989 level in 1998 and fully recovered only in The number of employees, the main segment of taxpayers, has declined by 35% between 1989 and The third cause was successive tax rate cuts, downward adjustments in income tax brackets, the many tax exemptions and tax holidays encouraging the perpetuation of soft budget 11 constraints. The downsizing of the state was an anticipated process for all countries from Central and Eastern Europe. von Hagen and Traistaru (2004) estimated the government expenditure-to-gdp ratio as a linear function of trade-to-gdp ratio and GDP per capita based on a data set including 22 OECD, 11 Latin American and 10 Central and Eastern European countries from They showed that given the openness and per capital income levels in Central and Eastern Europe, the governmental sectors of these countries were oversized compared to the other countries in the sample and a process of downsizing relative to GDP was to occur. % GDP Fig. 1 Revenues and expenses of consolidated budget Fig. 2 Total revenues and tax revenues of consolidated budget % GDP revenues expenses revenues tax revenues tax revenue cyclically adjusted Source: IMF (1998), IMF (2006), Ministry of Public Finance, European Commission (2011) The increase during the recession and the decline during the expansionary period of the cyclically adjusted tax revenues, indicate an enhanced pro-cyclical fiscal policy (Fig. 2). It appears that over these years, the authorities were able to reverse the declining trend of tax revenues twice: in 1997 and. On the first occasion the adopted measures led to two years of tax revenues gain equivalent to 7% of GDP which was lost between 1999 and On the second occasion, the tax revenues gain was equivalent to just 1% of GDP and lasted only one year. The description of discretionary tax changes implemented during 1990 and 2008 is the aim of the next section. The two decades long period is divided into sub periods reflecting the cyclical move of tax revenue-to-gdp. 10 See also Daianu and Vranceanu (2000) 11 Kornai used first this notion for the case of command systems which were seen as shortage economies (1980) 12 This is an analytical construct reflecting uniquely implemented discretionary tax policies and ignoring the effects of the growth cycle. To derive the cyclically adjusted tax revenues the European Commission relies on a methodology based on elasticities with respect to the output gap of different component of tax revenues (personal income tax, corporate income tax, social contribution and indirect taxes).

14 The early downsizing ( ) The breakdown of central planning in 1989 imposed fundamental changes in taxation policy. A replacement for the compulsory transfers from state enterprises to state budget, representing either part of the benefits, turnover tax or social security contribution was needed. The profit tax was introduced in 1990 to replace the communist - era confiscatory profit transfer tax. Early reforms focused on the reduction in the number and the statutory level of tax rates. The 67 breaks (corresponding to tax rates between 5% and 77%) in 1990 were successively reduced to two in 1992 (a low level of 30% and a high level of 45%) and further to a uniform 38% in In 1990 there was a personal income tax schedule with 13 brackets and marginal rates from 6% to 45%. The top marginal level was further increased to 60% in 1993, but the number of exempted taxpayers continued to grow. The social security contribution rates, after initial increases in 1990 and 1991, were kept unchanged at 35% of gross wages through These increases resulted in sharply lower compliance facilitated by increasing tolerance to nonpayment of obligations by large public enterprises. The declining compliance was stimulated further after 1995 when regular employment were increasingly governed by civil contracts of employment, designed initially for small time contractual employment and not liable to social security contribution. The standard 18% VAT was introduced in mid-1993 to replace the turnover tax. This rate together with the reduced rate of 9% introduced in 1994 for a number of basic and educational goods was unchanged until The international experience suggests that revenue collection of 9% of GDP should have been well within reach. However, actual collection declined from 5.3% of GDP in 1993 to 4.9% in This poor performance reflects weak tax administration burdened by many exemptions, multiple rates and slow reimbursements, which undermined the self-enforcing character of VAT and fuelled incentives for evasion. Between 1992 and 1996 the consolidated budget deficit reached an average of 3.1% of GDP 13 which together with a similar sized quasi-fiscal deficit helped to secure the early recovery of output compared to many other transition economies, but also created chronic inflation and balance of payment imbalances, which came to the fore at the end of The recovery ( ) In an effort to strengthen revenue collection from companies, the 1999 budget law suspended for one year all tax holidays and incentives generously and successively awarded on short-term assessments of revenue collections and investment promotions trade-offs (Table A4.2) Thus, Romania with a corporate tax level of 38% had the highest effective tax rate on investment 14 exceeding the statutory tax rate and the highest corporate tax revenues related to GDP in the region in 1999 (IMF, 2003). 13 IMF (2003) identifies several forms of quasi fiscal deficit accumulation: a) subsidies through the extension of directed low interest rate credit by the central bank to agricultural and energy-intensive sectors, b) increases in lending of state owned commercial banks to state owned sectors, c) the central bank's sales of foreign exchange at an appreciated exchange rate to the energy sector, d) a pickup in the extension of government loan guarantees, e) general increase in payment arrears by state owned companies. 14 The effective tax rate on investment is an analytical construct, which relate an investment projects' tax liability to its before-tax income in economic terms (including the effects of depreciation, operating expenses, inflation). The effective tax rate could be higher than the statutory rate if company profits were affected by inflation, but the value of investment and depreciation allowances were not benefiting from re-evaluation. Effective tax rate is below statutory rate when tax holidays reduce tax liabilities.

15 Tax Policy Under The Curse of Low Revenues: The Case of Romania 13 The personal income taxation went through several changes in 1997 and The spread between the top and bottom rates as well as the number of brackets were reduced. In parallel, 15 the statutory rate on social security contribution was increased to 43% in 1998 and 60% in 1999, a level by far the highest compared to the region and western Europe, reflecting both the establishment of the National Health Insurance House modelled on the German system, as well as the deterioration in the finances of the pension fund. For the standard VAT rate was raised to 22% from 18% and the reduced rate to 11% from 9%. The measures were effective. The tax revenues bounced back in 1999 to a level last recorded in This was a unique experience among transition economies, which have witnessed either continuous erosion or bottoming out of revenues Peak-to-trough ( ) In 2000 a global income tax law covering all sources of personal income was introduced. The top rate was reduced simultaneously with the increase in the bottom rate (Fig. 3). To mitigate the impact on low-income earners, the general tax-free allowance was more than doubled to about 40% of average wages. The social security contribution rates were reduced gradually from 60% in 2000 to 49% in In an attempt to increase the tax base, the liability of civil contracts was extended to all social security contribution not just wage tax and health insurance contribution as they have functioned since they were first introduced in The lapse of one-year moratoria on corporate tax holidays in 2000 would have put the corporate taxation to the lowest effective tax rate in the region, at less than half of the next lowest effective tax rate (IMF, 2003). To avoid that, the statutory tax rate on profits was reduced from 38% to 25% with a 5% preferential rate for profits arising from export activities, an investment tax allowance of 10% was introduced and all tax-holiday and investment incentive legislation previously suspended in the 1999 budget law were abrogated. In 2000 the standard and reduced VAT rates were unified at 19%, down from 22% and 11% respectively, and many tax exemptions were eliminated. Until 2004 Romania became one of the few countries in Europe with a uniform low VAT rate. In 2004 the reduced VAT rate for several products like medicine, books, cultural events, hotel accommodation was reintroduced. The cuts in all main taxes -social contributions, corporate income tax, VAT eroded the tax revenue to GDP ratio. 15 Social security contributions are collected by the pension and unemployment funds, health insurance, the risk, accident and handicapped special funds, as well as a special fund to top up wage payments in the education sector.

16 14 Fig. 3 Main statutory tax rates in % VAT % Tax on wage/income min LTR max LTR 2009 % Tax on profit % Social security contribution min LTR max LTR LTR Source: IMF (2003), Romanian Fiscal Council (2010) Stabilisation (-2008) In the four-bracket system of personal income tax with tax rates ranging from 18-40% was 16 replaced by a flat tax system with a tax rate of 16%. This rate is applicable to corporate income instead of the previous 25%, income from independent activities, royalties, income from movable and immovable property, but also to short term capital gains on listed shares. The social security contributions were gradually reduced from 49% in to 41.8% in The statutory VAT rate was unchanged at 19%. The tax revenue increased by an equivalent to 1% of GDP between and 2007, but next year 0.3% tax revenue-to-gdp ratio was lost since the contractionary cycle set in starting with the third quarter of The flat tax did not bring in substantially higher tax revenues and has not diminished the shadow economy significantly, as some expected. What lessons can be drawn? The main one is that the fiscal policy, which consisted of gradual tax rate cuts, did not bring more revenues compared to GDP. Tax rate cuts without accompanying measures for increasing the tax base and tax enforcement just weakens the fiscal policy stance. 16 A similar tax policy was introduced in Slovakia in There, a unique tax rate at 19% was implemented for VAT, corporate and personal income tax. In 2008 Bulgaria introduced a 10% flat tax rate for personal and corporate income, the Czech Republic introduced a 15% flat tax for personal income, Estonia, Latvia and Lithuania had flat tax rates on personal and corporate income in the first half of the 1990s already.

17 Tax Policy Under The Curse of Low Revenues: The Case of Romania 15 The fact that only the fiscal policy package which consisted of augmented statutory tax rates and measures for tax base increase had lasting positive effects on the tax revenue is telling. The second lesson is that increasing the tax base and tax enforcement have to be built-in priorities of the tax administration. Otherwise, the erosion of tax revenues occurs. The third lesson is that the changes in the tax code were approximately in line with some theoretical claims (Mankiw et al, 2009). Prior to the gradual reduction of marginal tax at high income was pursued. According to the theory the optimal marginal tax rate schedules depend on the distribution of ability. After the tax code was built around a flat income tax. A flat tax is attractive due to its simplicity and easier implementation, especially in countries, which are afflicted by precarious public administration and endemic corruption. Nonetheless, its economic effects are not uncontroversial. And its impact on trust and social capital, via income distribution, should not be underestimated. This social dimension of tax policy acquires additional salience during a period of crisis, when social cohesion and fairness in burdensharing matter tremendously. Since the introduction of VAT, the optimal principle that only the 19 final goods should be taxed uniformly was partially applied Stylised Facts of Current ( ) Tax Revenues The analysis focuses on the tax revenue dynamics during the expansionary cycle between 2000 and 2008 and tries to capture effects of the introduction of the flat income and profit tax. For this purpose the period is split into pre and post ; features of the tax revenue evolution in comparison to Eu27 and NMSs are presented (Table 2.2.1). The pre ( ) period is characterised by a standard VAT rate at 19%, a progressive personal income tax (PIT), statutory corporate income tax (CIT) at 25% and declining social security contribution (SSC) rate from 60% to 49%. During this period the GDP/capita in PPS increased by 10.3% each year, the number of employees by 0.2%. The tax revenues per GDP declined by 2.6% each year, while the cyclically adjusted tax revenue rates declined by 4.8%. The period witnessed an impressive growth, the GDP/capita in PPS increased by 14% each year, the number of employees by 1.7%. However, although the tax revenues to-gdp increased by 0.2% yearly, the cyclically adjusted tax revenue-to-gdp declined by 3.2% yearly. The tax code of the period was characterised by the same standard VAT rate from pre, a flat PIT and CIT at 16% and a declining SSC rate from 49% to 41.8%. EU-27 and NMSs followed the same trends but with slower dynamics both pre and post. 17 Depending on the assumptions made about the distribution of ability Mirrlees (1971) suggests zero top marginal tax rate, Tuomala (1990) finds a decrease in marginal tax rate as income levels increase, while Diamond (1998) and Saez (2001) find that top optimal marginal tax rate can be quite high. 18 Mirrlees (1971) finds that optimal taxation can yield linear tax schedule. Saez (2001) challenges this view and finds optimal tax rates that increase steadily for high income. In an application to Germany, Bach et al. (2001) find an optimal to marginal tax rate close to 67% when the social planner is rawlsian. Diamond (1998) find that the shape of optimal marginal tax rates has a U-form, contradicting to some extent the idea that a linear schedule is close to optimal. 19 Atkinson and Stiglitz (1976) argue that taxes should be uniform across final goods. Diamond and Mirrlees (1971) argues that taxes on intermediate goods ought to be avoided, because they distort the allocation of factor inputs. The latter view has been glaringly refuted by the current crisis: financial intermediation has expanded enormously in recent decades while extracting an undue rent from the economy. Therefore it should be taxed appropriately, to help downsize it and reduce speculative behaviour. 20 The analysis is based on data from European Commission, which uses ESA95, available for the period 1995 and The difference between the tax revenue-to-gdp ratio computed by the European Commission and the tax revenue-to-gdp ratio from the IMF country reports used in Fig. 2 from Section 2.1 ranges between 0.1% and 0.9%. 21 Weighted average, before 1998 weighted average in EU-25 according to EC(2010). 22 Simple average of the figures for Bulgaria, the Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Slovenia and Slovakia according to EC(2010).

18 16 Table Main Macro Indicators in Romania, EU-27 and NMSs GDP/capita PPS Employees, m Tax revenue/gdp Tax revenue/gdp cyclically adjusted GDP/capita PPS Employees, m Tax revenue/gdp Tax revenue/gdp cyclically adjusted GDP/capita PPS Employees, m Tax revenue/gdp Tax revenue/gdp cyclically adjusted Pre Post Annual % change Romania EU NMSs Annual % change Source: GDP/capita PPS and employees from Eurostat database, tax revenue/gdp and tax revenue/gdp cyclically adjusted from European Commission (2010c) Tax Revenue Size The overall tax revenue to GDP ratio (including social security contribution) was 27% in 2010, the lowest in the EU-27. In 2009 the Romanian tax revenue-to-gdp ratio was by 8 percentage points below EU-27 average and near 4 percentage points below NMSs average (Fig. A1.1). The gap persisted since 1995 and settled onwards 2000 in the range between 10-12% of GDP relative to EU27 average and 4-6% of GDP relative to NMSs average (Fig. 4). It is worthwhile to mention that the synchronisation of the business cycles tended to stabilise the tax-revenue-to- GDP gap. In when Romania was in recession and the other European states were growing the tax revenue to-gdp ratio gap widened. During , when all EU member states were growing the tax revenue-to-gdp ratio stabilised and in 2009 when most of EU member states contracted the tax revenue-to-gdp ratio narrowed.

19 Tax Policy Under The Curse of Low Revenues: The Case of Romania 17 Fig. 4 Romanian tax revenue-to-gdp gap % GDP vs. EU27 vs. NMS Source: based on European Commission (2010c) The persistent small tax revenues-to-gdp ratio is puzzling. An argument would be that Romania is a developing country with low GDP per capita (Gupta, 2007). Against this argument pleads the tax revenue- to- GDP ratio in countries from SE Europe with GDP per capital equal or smaller than Romania's (Table ). Bulgaria, Croatia, Montenegro and Serbia are able to collect more tax revenues than Romania by several percentage points of GDP. The small tax revenue to GDP ratio in Romania shows considerable potential for tax collection improvement, unlike in many EU member states, where the tax revenues-to-gdp ratio is much superior. Table Tax Revenue-to-GDP ratio in South Eastern Europe GDP/capita (US$) Tax revenue-to-gdp ratio Albania Bulgaria Croatia Macedonia Montenegro Romania 7500 Serbia Source: Albania-IMF 2010 Article IV Consultation Preliminary Conclusion of the Mission March 19, 2010, IMF Country Report 09/73; Bulgaria-and Romania European Commission, Croatia-IMF Country Report 10/179 and 9/185, Macedonia IMF Country Report 11/42, Montenegro-IMF Country Report 9/88 and 11/100; Serbia-IMF Country Report 9/158, 10/25 and 11/ The Flat Tax and Changes in the Tax Revenue Structure Development influences the structure of tax revenues (Gordon and Li, b). Poor countries collect less revenue as a fraction of GDP than is collected in richer countries. Gordon and Li (a) showed that while developed countries rely on broad based income and consumption tax making little use of tariffs or seignorage as sources of revenue, the poor countries make

20 18 much less use of broad-based taxes, relying instead on excise taxes, tariffs and seignorage. This section examines changes in the tax revenue structure that have been entailed by the flat tax introduction in. This is done by comparing the average tax structure pre and post with the tax revenue structure from EU-27 and NMSs. It is worthwhile to mention that all old EU Member states have progressive personal income tax while six out of the nine NMSs (excepting therefore Poland, Hungary and Slovenia) apply flat rate personal income taxes. The expectation is that the divergences in the tax revenue structure comparative to EU-27 to have strengthened after, while the divergences compared to NMSs to have got milder Structure by Tax Categories Out of the three fiscal revenue categories (direct -Fig. A1.4, indirect -Fig.A1.3 and social 25 security contributions -Fig. A1.5) the revenues collected through indirect taxes are the least pro-cyclical. This is because they tend to evolve proportionally with the output and, therefore, the ratio to GDP should be relatively stable over the cycle. Direct taxes are most procyclical because of the sensitivity of corporate taxes to the business cycle and because of the 26 progressive nature of personal income taxes. Social security contributions, which are closely related to wage bill, tend to be less responsive to the cycle. Taking into account that the analysed period was a growth period, the above characteristics of various taxes would predict an increased share of direct taxes in total tax revenues at the expense of social security taxes and to a lesser extent of indirect taxes, other conditions unchanged. Indeed, this is what happened in NMSs. In Romania, the tax system was changed in and altered the cycle driven expected structural changes (Table ). The tax revenue structure in the pre period shows that: the two main pillars of public finance in Romania are indirect taxes (the yearly generated revenue was 11.8% of GDP on average) and social security contributions (the yearly generated revenue was 10.2% of GDP on average), like in NMSs but unlike in EU-27 where the public finance relies equally on direct, indirect and social security contributions. the composition of indirect taxes is similar to that in other NMSs. The only difference is the larger share of import duties, explained by the fact that Romania joined the EU later than most NMSs, except Bulgaria, and therefore trade integration with the EU lagged behind; direct taxes (the yearly generated revenue was 6.3% of GDP) are equally generated by personal income tax and corporate income tax, unlike in EU-27 and NMSs where the largest part is generated by personal income tax. the largest source of social security contributions is employers' contribution (the yearly generated revenue was 6.8% of GDP on average) as in EU-27 and NMSs. The marked difference compared to both NMSs and EU-27 is the very small contribution of self employed, which suggests that this category remained, basically, outside official taxation. 23 Direct taxes include personal income tax, corporate income tax and other income on capital taxes corresponding to other taxes on holding gains, taxes on receivings from lottery or gambling, other taxes on income, taxes on capital defined as other current taxes and capital taxes. 24 Indirect taxes include Vat type taxes, excise duties and consumption taxes, other taxes on products, other taxes on production. 25 Social security contributions include compulsory employers' actual social contribution, compulsory employees' social contribution and compulsory social contribution by self and non-employed persons. 26 According to EC (2010a) the introduction of a flat tax combined with a tax-exempt threshold does not necessarily decrease tax progressivity. When income shocks concentrate in the income region somewhat above the threshold of the flat tax, progressivity could be greater under the flat tax. This is due to the fact that in this region the introduction of a flat tax generates, above the threshold, an increase in the marginal tax and, below the threshold, for very low income, a zero marginal tax rate.

21 Tax Policy Under The Curse of Low Revenues: The Case of Romania 19 Table Structure of Tax Revenues by Tax Categories Average % % Romania EU -27 NMSs Average Average Average Average % % % % % % % % Average % % GDP GDP GDP GDP GDP GDP Total Indirect tax VAT Excise Tax Import duties Other taxes on production Direct tax Personal income tax Corporate income tax Social security contribution employers employees Self-employed The post tax revenue structure changed in Romania and NMSs but not in the EU-27. The weight of indirect taxes in total tax revenue increased consistently (the yearly generated revenue was 12.6% of GDP on average) due to VAT, unlike in NMSs, at the expense of social security contribution. The weight of direct taxes in total tax revenue declined, unlike in other NMSs where it increased (generated by corporate income tax) The weight of social security contributions in total tax revenue declined (the yearly generated revenue was 9.6% of GDP on average) due to smaller share of both employers' and employees' contribution. Although the share of contribution of self employed in total tax revenues rose somewhat, it remained three times smaller than in EU-27 and NMSs Structure by Economic Function The structure of tax revenues according to the economic function played by taxes consumption tax (Fig. A1.6), labour tax (Fig. A1.7) and capital tax (Fig. A1.8)- reveals that the 27 Consumption tax includes value added type taxes, taxes and duties on imports, taxes on products, other taxes on production (taxes on international transactions, taxes on pollution, under-compensation of VAT, poll taxes, expenditure taxes, payments by households for licenses. 28 Tax on labour includes all taxes directly linked to wage and mostly withheld at source, paid by employees and employers including actual compulsory social contribution, all taxes and compulsory social contribution on transfer income (social transfers paid by state and benefits from old age pension schemes) of non-employed persons, where these could be identified. 29 Capital is defined broadly including physical capital, intangible and financial investment and savings. Capital tax includes taxes on business income in a broad sense not only taxes on profits but also taxes and levies that could be regarded as a prerequisite for earning, profit such as the real estate tax or the motor vehicle tax paid by enterprises and taxes on capital stocks of households or their transactions.

22 20 most prominent source of revenue had changed between pre and post from labour tax to consumption tax, unlike in EU-27 and NMSs, where labour tax remained the main source of tax revenue all over the period (Table ). The pre shows that: The tax revenue had two pillars with very similar strength labour tax (the yearly generated revenue was 12.1% of GDP on average) and consumption tax (the yearly generated revenue was 11.1% of GDP on average) unlike in EU-27 and NMSs where labour tax revenue was the leader exceeding the revenues generated by consumption tax by several percentage points. The larger share of revenues generated by consumption tax in total tax revenues both in Romania and NMSs compared to EU-27 is the consequence of a deliberate tax policy to rely on a tax which is considered growth friendly. Otherwise, the role of this tax in total tax revenue would have converged toward the EU-27 average, since the indirect taxes on which it is based is largely harmonized at EU level. More than half of the tax on employed was paid by employers unlike in EU-27 where employees are more burdened or in NMSs where the burden is equally shared by employers and employees. The non-employed are very little taxed as compared to both EU-27 and NMSs. The share of the tax on capital in total tax revenue was higher than in NMSs but lower than in EU-27. The annual tax revenue on capital was 5.2% of GDP on average. The largest component of the tax revenue on capital originated from income of corporations (55% compared to 34% in EU-27 and 49% in NMSs). Self employed contribution to tax revenue on capital was minimal compared to both EU-27 and NMSs and was offset by the unusually large (by EU27 and NMSs standards) contribution of households. The capital tax revenue-to- GDP ratio followed an ever declining path in Romania from a value equal to the EU-27 average in 1995 and converged to the NMSs average in The reason was that the preparation of the accession into EU imposed the gradual liberalisation of the capital 30 account and the free cross-border movement of capital flows. Through lower taxes on 31 capital the EU candidate countries tried to attract capital and thereby boost real convergence. The structure of tax revenue on capital indicates that Table Structure of Tax Revenues by Economic Function of Taxes Total Tax on consumption Tax on labour Tax on employed labour, of which Romania EU -27 NMSs Average Average Average Average Average Average % % % % % % % % % % % % GDP GDP GDP GDP GDP GDP Piatkowski and Jarmusek (2008) showed that changes in the average of other countries' CIT rate in Eastern Europe had strong statistical significance in explaining changes in CIT rate in individual countries. A one pp change in the average of other countries statutory CIT rate resulted in a pp change in CIT rate in a particular country. 31 The evolution of effective marginal and average tax rates on corporation between 1998 and 2009 (EC, 2009b) indicates an abrupt declining path for both Romania and NMSs. In pre compared to the effective marginal tax rate on corporate declined by 9.5pp and the effective average tax rate by 11.5pp. The corresponding declines in NMSs were much milder, by 4.7pp and 5pp respectively. Post the decline continued, compared to pre the effective marginal tax rate lost another 5.6pp, while the effective average tax rate 7.9pp in Romania. Both declines were milder than those registered by NMSs. Post in Romania the effective marginal tax rate was still above the NMSs rate, but the effective average tax rate ended below the NMS rate by 2pp.

23 Tax Policy Under The Curse of Low Revenues: The Case of Romania 21 - paid by employers - paid by employees - Tax on non employed labour Tax on capital Capital and business income income of corporations income of households income of self-employed Stocks of capital/wealth Source: European Commission (2010c) Post, the consumption tax revenue-to-gdp ratio, at 11.9% on average, exceeded the EU-27 average but lagged behind the NMSs average, the capital tax revenue-to-gdp ratio at 5% on average reached parity with NMSs average and around half of the EU-27 average, while the labour tax revenue-to-gdp ratio lagged behind both NMSs average and EU-27 averages by 8pp and 4pp respectively. The structural changes of the period point to the followings: The share of the consumption tax in total tax revenues increased at the expense of both capital and labour tax, unlike in NMSs where both the share of consumption and capital tax increased. The changes in EU-27 favoured capital taxes at the expense of both consumption and labour taxes. The contribution of self-employed to tax revenue on capital almost doubled but still remained far below the contribution of self-employed in NMSs and EU-27. The lower share of labour taxes revenues was due to the reduced share of labour taxes paid by employers. In NMSs both employers and employees' labour tax share declined but the share of non-employed labour tax increased. In EU-27 the share of the labour tax paid by employees declined. It looks like that the tax burden has been reduced for the scarcest resource in the economy, the capital in Romania, labour in EU-27 and both in NMSs. Taxes on non-employed labour are almost non-existent post as pre. These taxes brought 1.7% of GDP in EU-27 and 0.8% of GDP in NMSs (the double of the level registered pre ) Legal Tax Rates and the Tax Burden The low tax revenues might be due to low tax rates. This possibility is examined by comparing the four main tax rates, which generate 84% of the total tax revenue, with the EU-27 average and NMSs. At the first sight CIT and PIT rates are below the EU27 and NMSs average, VAT rate is about the same and SSC level is above both EU27 and NMSs averages (Fig.5, Table A4.3). Compared to Bulgaria the tax rates are substantially higher whereas the collected tax revenues are smaller by near 1 pp.

24 22 VAT VAT SSC PIT SSC PIT CIT CIT NMS RO EU-27 RO BG Fig. 5 Legal tax rates, 2009 We compute the notional tax revenue (Table ) by applying each of the above four 32 statutory tax rates to their corresponding tax base (compensation of employees for PIT and SSC, gross operating surplus for CIT and final consumption for VAT). According to the notional tax revenue-to-gdp ratio Romania has the second smallest tax burden after Bulgaria. The collected tax revenue-to GDP generated by the four main taxes is smaller than in Bulgaria by 0.7pp. Since Bulgaria's total tax revenue-to -GDP was larger by 2.6pp than Romania's in 2009, the main source of the difference and the strength of Bulgarian fiscal position reside in the tax revenue generated by excise taxes. The gap between notional tax revenue-to-gdp ratio and the collected tax revenue-to GDP ratio is middle sized, indicating that the collection/evasion efficiency is neither the largest nor the smallest among NMSs. Table Notional and Collected Tax Revenue-to-GDP ratios in 2009 RO BG CZ EE LV LT HU PL SI SK EU27 VAT rate Notional tax revenue/gdp Collected tax revenue/gdp gap PI T rate Notional tax revenue/gdp Collected tax revenue/gdp gap CIT rate Notional tax revenue/gdp PIT includes taxes on several types of income (investments, revenues from cession of goods use, prizes and gambling, pensions, revenues from independent activities, transfers of real estate properties from the personal patrimony) besides wages, therefore compensation of employees might understate the tax base. However, the tax revenue obtained from wages represented the largest and increasing weight in PIT generated revenues: 68% pre and 73% post of the total tax revenue generated by PIT.

25 Tax Policy Under The Curse of Low Revenues: The Case of Romania 23 Pre RO BG CZ EE LV LT HU PL SI SK EU27 Collected tax revenue/gdp gap SSC rate Notional tax revenue/gdp Collected tax revenue/gdp gap Total notional tax revenue/gdp Total collected tax revenue/gdp Gap Source: notional tax revenue/gdp own computation based on Eurostat data, collected tax revenue/gdp from Romanian Fiscal Council (2011) In order to separate effects of the flat income tax from effects of recession we compute the notional tax revenue and collected tax revenue generated by CIT, PIT and VAT separately for pre and post periods (Table ). We find the following characteristics for the pre interval: The notional tax revenue relative to GDP ratio was just by 1pp higher than the NMSs average due to the larger notional tax revenue for CIT. The collected tax revenue was however 3 pp lower than the NMSs average pointing to a similar difference in the gap between notional and collected tax revenue relative to GDP. The largest gap between the notional and collected tax revenue occurred for CIT, but compared to NMSs the largest difference in the gap was registered for PIT, signalling a very inefficient tax collection in both cases. Post there were significant changes: The notional tax revenue-to-gdp was below the NMSs average by 2.5pp, due especially to the notional tax revenue to-gdp ratio associated to PIT, which was by almost 4pp below the NMSs average. Indeed, the large reduction in the statutory PIT rate was not compensated by the rise of the tax base and therefore the notional tax revenue-to-gdp ratio halved. The notional tax revenue-to-gdp ratio associated to CIT declined by 64% compared to pre, while the CIT rate declined by only 36%. After the lower CIT rate the tax base shrank as well. The gap between the notional and collected taxes fell below the NMSs average for both PIT and CIT, but went beyond the NMSs average for VAT. Table Notional and collected tax revenue, pre, post and in 2009 Notional tax revenue/gdp, annual average Post 2009 Pre Collected tax revenue/gdp, annual average Post 2009 Pre Post CIT NMSs RO Gap 2009

26 24 Notional tax revenue/gdp, annual average Collected tax revenue/gdp, annual average PIT NMSs RO VAT NMSs RO CIT+PIT+VAT NMSs RO Source: notional tax revenue/gdp own computation based on Eurostat data, collected tax revenue/gdp pre and post from European Commission (2010c), collected tax revenue/gdp 2009 from Romanian Fiscal Council (2011) Implicit Tax Rates and Tax Collection Efficiency The implicit tax rate (ITR) computed for a given tax type is a backward looking effective tax rate which if applied to the chosen tax base gives the amount of collected tax revenues. The ITR for 34 consumption was stable since at around 18% but below both the EU-27 and the NMSs average (Fig A1.9). The tax rate gap was 2.8pp compared to EU27average and 3.8pp compared to NMSs average. However, comparing the consumption tax to-gdp ratio with the implicit consumption tax rate reveals that one percentage point of implicit consumption tax brought one-sixth percentage point of GDP above both EU-27 (0.55pp) and NMSs (0.658pp) averages. 35 The ITR on labour has been declining continuously from a maximum equal to EU-27 average reached in In 2008 it was below the EU-27 average by 7pp and below NMSs average by 4.4pp. One percentage point of the implicit tax rate brought one-fourth percentage point of GDP below EU-27 (0.54pp) and NMSs (0.63pp) averages (Fig.A1.10). The implicit tax-to-legal tax ratio measures tax collection efficiency and shows how much of the potential tax revenue, given the proxy for the tax base chosen, was really collected. The implicit tax-to-legal tax ratio (Fig. 6) computed (see Box 2) for four principal taxes VAT, households' income tax, corporate income tax and social security contribution- situates Romania below NMSs for VAT and social security contribution. Box 2 Methodological notes ITR for VAT, social security contribution, PIT and CIT were computed using Eurostat database on national accounts and main tax aggregates. ITR for VAT is computed as the ratio between value added type taxes and household final consumption expenditure. Gap 33 Implicit taxes can be computed for each tax type as the ratio of total tax revenue of that tax type to a proxy of the potential tax base defined using the production and income accounts of the national accounts. 34 Computed in EC (2010c) using as proxy for the potential tax base the final consumption expenditure of households on the economic territory of the country. 35 Computed in EC (2010c) using as proxy for the potential tax base the compensation of employees working in the economic territory defined as total remuneration, in kind or cash, payable by an employer to an employee in return for work done (includes gross wages and employers' contribution to social security as well as to private pensions and related schemes.

27 Tax Policy Under The Curse of Low Revenues: The Case of Romania 25 Box 2 Methodological notes ITR for social contribution is computed as the ratio between actual social contribution and compensation of employees; ITR for PIT is the ratio between taxes on individual or household income plus taxes on other income and compensation of employees. Taxes on individual or household income do not include taxes on profit and holding. ITR for CIT is the ratio between taxes on the income or profits of corporations including holding gains and gross operating surplus and gross mixed income. Household final consumption expenditure, according to the methodology of national accounts, includes all purchases of goods and services made by households and covers goods and services purchased from the market and own-supplied, the later representing the formal economy. The least efficient tax collection is the corporate income tax. Households' income tax collection registered a significant improvement since 2006 catching up with the efficiency of VAT and social security contribution collection: little above half of the potential. Corporate income tax is collected in a proportion of 33% of potential. The switch to the flat tax pushed the dependency of public finances on taxes with the highest compliance but still below the corresponding compliance rates in NMSs. The low implicit tax rate might have several explanations: first, the legal tax rate in Romania is smaller than the legal tax rate in NMSs; second, reduced tax rates (VAT and PIT in non-flat-tax-countries); third, there are more tax exemptions and loopholes and more tax avoidance possibilities in Romania with negative impact on the tax base; fourth, the efficiency of tax collection is much lower in Romania than in other NMSs. Some of these possibilities are examined next. % Romania % NMSs VAT households' income tax corporate income tax social security contribution Fig. 6 ITR-to-LTR ratio VAT households' income tax corporate income tax social security contribution Source: own computation based on Eurostat definitions and data The estimation of the particular shape of the relationship between legal tax rate and ITR (see Box 3) in the case of VAT, households' income tax, corporate income tax and social security contributions in NMSs including Romania in 2009 is presented in Fig. 7. A polynomial regression of second degree best fits the relationship, confirming the fact that in the context of the year 2009 the relationship between ITR and legal tax rate is non-linear. The underlying assumption is that in all NMSs countries the tax base and tax collection respond in the same way to the changes in the legal tax rates. If that is credible, taking into account the common legacy from the

28 26 past influencing behaviour and the shared trajectory of EU integration influencing policies, the relationship might indicate the changes needed (either in tax code and rates or tax collection) for a better ITR. Box 3 Implicit and legal tax rate relationship The relationship between implicit tax rate (ITR) and the legal tax rate (LTR) is indirect. ITR is the ratio between the amount of tax collected (TC) and the tax base (TB), each of which depends on LTR and other variables which are ignored in order to simplify the discussion. TC ( LTR ) ITR ( LTR ) = TB ( LTR ) How ITR does vary when LTR changes, ceteris paribus, is given by the derivative of (1) (1) ditr ( LTR ) dltr dtc ( LTR ) dtb ( LTR ) TC ( LTR ) = dltr - dltr (2) TC TB TB ( LTR ) It is the differences between the elasticity of taxes collected and tax base with respect to legal tax rate augmented by the ratio between taxes collected and tax base. The relationship between ITR and LTR might be of either sign, an upward or downward sloping curve. It is reasonable to assume that the tax collected and the tax base respond in the same way, either upwards or downwards, to an increase in LTR. When both move upwards the relationship between ITR and LTR is positive (an upward sloping function) when the elasticity of tax collected is higher than the elasticity of tax base and the other way round (a downward sloping function) when the relationship between the elasticities is reversed. The situation is opposite when the tax base and tax collected move downward with an increase of LTR. Of course, it can happen that tax collected and the tax base do not respond in the same way to an increase of LTR. In that case the relationship between ITR and LTR is non-linear. The particular shape of the complex relationship between ITR and LTR for the case of NMS countries for the year 2009 is depicted in Fig. 6. a/ Romania has lower ITRs for all four taxes than the experience of NMSs indicates it should have given the level of LTR. b/ the weakest link between legal and implicit tax rate is in the case of corporate income tax. This hints at the fact that companies have several ways to avoid income tax payment all over NMSs c/ higher legal corporate income tax leads to lower implicit tax. The fact that Romania belongs to the group of countries with the second lowest legal corporate income tax might be an advantage for attracting capital and job creation. Compared to the group the implicit tax in Romania is the lowest by half percentage point suggesting a potential for tax collection improvement. d/ the strongest link between legal and implicit tax is in the case of households' income tax, indicating a high degree of tax compliance by individuals. The higher the LTR for households' income the higher the ITR. As in the case of corporate tax, Romania belongs to the group of countries with the second lowest legal personal income tax and compared to the countries from this group Romania's implicit tax is the lowest by one percentage point gap distance from the second lowest implicit tax in the group.

29 Tax Policy Under The Curse of Low Revenues: The Case of Romania 27 e/ the relationship between the implicit tax and legal tax in case of VAT is U shaped. It might be the consequence of moving between corners. NMSs introduced VAT in the 90s and later as EU candidates took the measures necessary to comply with the European Law. When VAT was introduced most NMSs choose rates close to lower corner (18%) and when forced to increase tax revenues they resorted to VAT increases ending in the higher corner (23-24%). In either corner the implicit tax is the same, the efficiency of collection is similar. Since July 2010, Romania jumped from the low corner to the higher corner, as other NMSs did previously. f/ Romania's social security contributions rate would require an implicit tax rate by 5pp higher than the actual one. The efficiency of social security contributions collection seems to be far behind the level the NMSs experience would indicate. implicit tax rate, % VAT RO y = x x R 2 = legal tax rate,% implicit tax rate, % Holuseholds' income tax 6 4 RO 2 y = x x R 2 = legal tax rate, % Corporate income tax Social security contribution implicit tax rate, % y = Ln(x) R 2 = RO implicit tax rate, % y = x x R 2 = RO legal tax rate, % Fig. 7 LTR, ITR relationship in NMSs, 2009 legal tax rate, % The relationship between legal tax rates and the generated tax revenues as share in GDP (Fig. 8) is estimated for year 2009 and NMSs including Romania. The changes in legal tax rates explain half or more of the changes in tax revenue-to-gdp in case of households' income tax and social security contributions. The relationship is very weak in case of CIT, indicating that other 36 characteristics of the tax design (than the legal tax rate) are crucial for the tax revenue. VAT stands in between CIT, on the one hand, and PIT and social security contribution, on the other hand, in terms of the strength between legal tax rate and tax revenue-to-gdp ratio. Romania collects the amount of revenue for CIT and PIT as indicated by the NMSs experience, but lags behind in the case of VAT and SSCs. The integration of social security contributions collection and tax administration decided in 2003 appeared not to bear fruits. The reason for integration at 36 The level of CIT revenues/gdp crucially depends on the share of corporations liable for CIT and their profits in overall firm profits.

30 28 that time was twofold. First, to improve administrative efficiency and provide better services by centralizing the fragmented collection services and enforcement activities undertaken in separate agencies within a unified revenue administration. Second, to get better tax and social contribution collection compliance. The fact that the results are rather disappointing after five years indicates that besides tax collection problems there are structural weaknesses in the tax design impeding significant tax revenue. tax revenues /G DP, % VAT, NMS in 2009 RO y = x x R 2 = legal tax rate,% Tax revenue/g DP, % Holuseholds' income tax, NMS in 2009 RO y = x x R 2 = legal tax rate, % tax revenue/gdp, % Corporate income tax, NMS in 2009 y = x x R 2 = legal tax rate, % RO tax revenue/g DP, % Social security contribution, NMS in 2009 y = x x R 2 = legal tax rate, % RO 2.3 Taxpayers Fig. 8 Tax revenue and LTR in NMSs, 2009 According to the National Authority for Fiscal Administration 7.7m taxpayers were registered, 37 out of which 1.3m companies, in The voluntary payment compliance improved from % in 2009 to 78.9% in 2010, while the voluntary registration compliance improved from 83.9% to 84.6%. Around one quarter of those who are liable to pay taxes do not pay. The burden of direct taxes and social security contributions falls almost equally on individuals and companies. The share of taxes paid by individuals in direct taxes and social security contributions taken together was 48.2% (EU % and NMSs 47.5%) in 2009 EC, 2011). The tax paid by individuals-to-gdp ratio fluctuated between 5% and 7% from 1997 to 2008 and climbed to 8% in 2009 being by 5 percentage points below the EU-27 average and 1 percentage point below NMSs average (Fig A1.11). The tax paid by companies-to-gdp ratio after o period of high volatility between 1995 and 2000 embarked on an abrupt declining path up to 2002 and stabilised thereafter at 9%. In 2008, another declining phase followed (Fig A1.12) leading the 37 The share of tax revenue in total tax revenue originating from taxpayers paying all their tax contribution in time. 38 The share of taxpayers voluntarily submitting their income statement in the total estimated number of taxpayers.

31 Tax Policy Under The Curse of Low Revenues: The Case of Romania 29 corporate income tax-to-gdp ratio to 8.5% in 2009 compared to 10% in both NMSs and EU-27. Studies show (EC, 2010c) that income taxes are pro-cyclical, while social security contributions are rather countercyclical. The cyclical nature of income tax and social contributions taken together depends on the compounded sensitivity to the cycle of both. In case of employees the counter-cyclicality of social security contribution more than offset the pro-cyclicality of income tax, while in the case of companies the effects work in the other way round Households Around 7.4m households were registered in Romania in % of total households were with an employed household head (employee households), 44% with a retired household head 40 (retiree households) and 8% with an agricultural worker as household head (agricultural 41 households). 77% of taxes paid by individuals were the contribution of employee households (Fig A2.5), while the retiree and agricultural households contributed by 17% (Fig A2.9) and 1.5% (Fig A2.6) respectively in What did determine this uneven distribution of tax burden among households? The explanation is revealed by the structure of households' budget. Households' income has two main components: money income and in kind income. The money income consists of gross salaries, income from independent activities, income from social 42 provision and income from proprietorship. The income in kind consists of equivalent value of free or lower price provisions from economic units and the equivalent value of consumption of agricultural products from own resources. The structure of personal income tax revenue shows that the main contributors are the employees, wage tax representing 68.3% of total personal income tax revenue pre (on average over the 4-year-period before ) and 73.5% post (on average over the 4-year-period after 2004). Pensioners and agricultural workers are little taxed. Pension tax revenue represented 0.6% pre and 1.3% post in total personal income tax revenue. The tax revenue from agricultural income is like non-existent (the annual revenues in the last decade being below 0.01% of GDP). This figure should be judged against the share of agriculture in GDP, which fluctuates in a range of 8-12%. This situation contributes to the low level of tax revenues in Romania. The average annual share of gross salaries in total income of the average household was 43% pre and 50% post (Table ). Although, the share of gross salaries in total households' income increased post for all households, it still varies hugely among different households from 80% in case of employees households to 7.7% and 21% in case of agricultural households and retiree households 43 respectively. The uneven distribution of wage income among households generates the uneven distribution of tax burden among households. 39 The analysis is based on aggregated data from Households Budget Surveys conducted yearly by the National Statistical Institute since The number of retirees increased by 70% between 2000 and 1991 due to the massive early retirement used as a substitute for unemployment. Between 2010 and 2000 the number of retirees has declined by 10%, reaching 5.5m. The retiree-to-employee ratio has declined from 1.4 pre to 1.2 post. 41 The other categories were households with an unemployed head (3.9% of the total), households with an own account worker as head (4.9% of the total) and households with an employer as head (1.8% of the total). We decided not to report on these types of households first, due to their size in total households and second due to the fact that the data was not provided in each year's survey. 42 Include unemployment benefits, pensions, children allowances, scholarships and other allowances of social protection. 43 The wage income in retiree and agricultural households is obtained by the other households/ members than the head.

32 30 Table Households' Budget Income Structure Gross salaries Independent activities Social provision Money income Income in kind From own agricultural products Other income Income-to- GDP/capita Total HH Employee HH Agricultural HH Retiree HH Average Average Average Average Average Average Average Average Total income Source: Households' budget survey, Households' budget income structure by deciles shows how skewed towards the richest 30% of households the tax burden is (Fig. A2.7). Only for these households the share of gross salaries in total income exceeds half, the country average (Table ). For all deciles, except for the richest, the share of total income in GDP/capita declined post. For all deciles, except the poorest, the share of gross salaries in total income increased post. The implementation of flat tax and the reduction of social security contribution lowered the tax wedge for the 44 employed person with low income from 44% pre to 41.8% post stimulating work and pulling out some informal activity into formal and taxable activity and thereby raising the tax base. Table Households' Budget Income Structure by deciles D1 D2 D3 D4 D5 D6 D7 D8 D9 D Money income Gross salaries Independent activities Social provision Income in kind From own agricultural products Other income Income-to- GDP/capita Total income Source

33 Tax Policy Under The Curse of Low Revenues: The Case of Romania Money income Gross salaries Independent activities Social provision From own agricultural products Other income Income-to- GDP/capita Income in kind Total income Source: Households' budget survey, The tax revenue contribution of the richest household decile in total taxes paid by individuals had a hump shape evolution between 1998 and 2009 with a maximum reached at 41.5% in 2004 (Fig. A2.2). After the introduction of the flat income tax rate in, the tax contribution continued to fall up to 33% in The poorest household decile's tax contribution was 0.5% of the total (Fig.A2.1). The middle-income households (5 and 6 deciles) tax contribution share in total was U-shaped with a minimum in These households contributed to the total individuals' tax by 15% in 2009 (Fig.A2.3). Half of individuals' tax burden fell on the richest 30% of households in 1998, then on the richest 20% of households in 2004 before shifting back on the richest 30% in 2009 (Fig.A2.7). Both pre and post the social benefit distribution among households was hump shaped (Fig. A2.8). The largest social benefit share from total was directed to the seventh household deciles. The share from the total social benefit received by the richest 3 household deciles was higher than the share from total social benefit received by the poorest household decile and rose post compared to pre. Taking into account that social benefits include pensions, unemployment benefit, children allowances and other social protection benefits the above distribution might indicate that social benefits are improperly targeted by not favouring the poorest individuals. Nevertheless, the income inequality between the richest and poorest households deciles moderated after the provision of social benefits. Pre the richest households decile income was higher than the poorest household decile income 4.9 times when the social benefit provision was ignored and just 4.1 times when the provided social benefits were included in the income. Post the corresponding ratios were 6 and 4.9. The compound effect of flat tax and the provision of social benefit among households raised the inequality. Households' budget income is spent for consumption representing on average more than half 47 of the budget, taxes and social contributions weighting 14.2% and expenses for agricultural products from own production, mostly untaxed, representing around 17% of the budget post (Table ). 45 Income taxes and social contributions. 46 Based on data from households' budget survey organized yearly since Taxes and social contributions include: taxes on wages, taxes on pensions, taxes on independent non-agricultural activities, contributions for pensions, contributions for unemployment benefit, contributions for health insurance.

34 32 This untaxed part of the budget hugely differs among different types of households being the smallest in the case of employee households and the largest for agricultural households. Although the share of this do it by yourself activity type diminished post it remained sizable and creates inefficiency for VAT tax collection. A fiscal policy in which the whole public finance relies mainly on indirect taxes will not create enough revenues unless it creates incentives for the large hidden part of the economy to come into the open. The process started 48 but at a very low speed. A simple computation shows that total households' consumption from own production would represent 6.5% of GDP. Out of this 1% of GDP would represent 49 agricultural households' consumption from own production. Table Structure of Households' Budget Expenditures Consumption expendit Taxes and social contribution Expenditures for agricultural products from own production Other expenditures Total HH Employee HH Agricultural HH Retiree HH Average Average Average Average Average Average Average Average Source: Households' budget survey, The untaxed part of consumption is increasing with poverty (Table ). The share of expenses for agricultural products from own production was more than half pre and became 45% post in the poorest household decile. Only in the richest 40% of households this kind of consumption represents less than in the average household. Table Structure of Households' Budget Expenditures by deciles D1 D2 D3 D4 D5 D6 D7 D8 D9 D Consumption expenditures Taxes and social contribution Expenditures for agricultural products from own production Other expenditures Taking into account that the total income of the average households is 0.38 of GDP/capita and the consumption from own production represents 17% of total income 49 Taking into account that 8% of total households are agricultural households, the average agricultural households include persons, the total income of average agricultural household represent 0.25 of GDP/capita and the share of consumption from own production in total income represented 46.7%.

35 Tax Policy Under The Curse of Low Revenues: The Case of Romania 33 Consumption expenditures Taxes and social contribution Expenditures for agricultural products from own production Other expenditures Source: Households' budget survey, The tax contribution in total budget expenses increased for all households during the last decade from 13% pre to 14.6% post. For employee households the increase was from 21.6% to 23.1%, from 1.8% to 2.4% for agricultural households and from 6.3% to 6.6% for retiree households. The reason was the increase of tax base namely the increase of remunerated work. The tax contribution in total budget expenses for the middle-income households (5 and 6 deciles) increased from % pre to 10-12% post, for the richest household decile the increase was from 22.4% to 25.5%, while for poorest households decile there was almost no change (Fig. A2.4) Companies The corporate income tax was 2.5% of GDP and the social security contributions paid by 50 employers amounted to 6.1% of GDP in 2009 (Fig. A1.12). In the same year companies accumulated arrears to consolidated general budget representing 3.36% of GDP up by 0.85pp compared to the previous year (Fig.A3.1). In the year when all these arrears would have been paid the taxes paid by companies to-gdp ratio would have been similar to EU-27 average. Compared to the arrears to the consolidated budget registered during the previous recession ( ), current arrears are half. 45% of the arrears to the consolidated budget represent arrears to state budget (Fig. A3.3) in terms of unpaid VAT and income tax and 49% are arrears to the social insurance budget (Fig. A3.2). 60% of total arrears are generated by public companies (Fig.A3.4 and Fig. A3.5). The sectoral breakdown of arrears points out mining, manufacturing and services as major arrears generators: they generate 30.5%, 21% and 21% respectively of all arrears to consolidated general budget (Fig.A3.7). 2.4 Summing Up The tax revenue-to-gdp ratio is the lowest in EU-27. The tax revenue-to-gdp gap settled after 2000 between 10-12% of GDP relative to EU-27 and 4-6% of GDP relative to NMSs. The main pillars of the tax revenues are indirect taxes and social security contributions (like in other NMSs); according to their economic function they are the consumption tax and labour tax (like in other NMSs). The labour tax burden fell entirely on employees and employers; self employed and non-employed are almost tax exempted. The tax burden measured by the notional tax revenues is among the lowest in NMSs; the gap between the notional tax revenue and collected tax revenue is middle sized among NMSs 50 The burden of social security contribution paid by employers is partially carried by employees in form of lower gross market wages.

36 34 The efficiency of tax collection, although it has improved over time, especially in the case of CIT and PIT, is still far below the level of other NMSs Although agriculture contributes to GDP in the range of 8-12% (depending on weather conditions) its share in tax revenues is pathetically small; this reflects the size of for ownconsumption agriculture, the few modern forms of agricultural activity and, not least, tax evasion. The declining trend of the tax revenue-to-gdp ratio was twice reversed over the last two decades. First, in 1997, when a gain of tax revenue was obtained through the increase of VAT and of social security contributions. The second time was in, but less efficient and not lasting, when the flat PIT and CIT tax was introduced. For the rise in the VAT rate in 2010 is still to early to make a judgement The introduction of the flat tax corroborated with successive reductions in SSC had several consequences. First, the weight of indirect and consumption tax revenue in total tax revenue increased. The efficiency of collecting VAT, indicated by the evolution of ITR, exceeded the efficiency of PIT and CIT, but remained far behind the VAT efficiency reached in NMSs. Second, judging upon the evolution of the share of expenditures on agricultural products from own production in households' budget, a proxy of informal activity, some informal and do by yourself activity was pulled into the open between pre- and post. The process was slow and was related to still large tax wedge on low-income employment. The increased social benefits to agricultural households reduced the consumption from own production and presumably increased the tax base for VAT. Third, the income inequality between the richest and poorest households has increased. Companies' tax arrears are large. If added to the amount of taxes paid the collected tax revenue from companies would become comparable to EU27 average. 3. The Black Economy in Romania and its Impact on Tax Revenues Estimates of the Romanian shadow economy vary by quite a margin. There are studies which attempt to measure it using various methodologies. For instance, AT Kearney estimates the size of the Romanian shadow economy in 2009 at 29% of GDP, Andrei et al (2010) at 36%, the Romanian Fiscal Council at 20%, Alexandru et al (2009) at 27% (in 2008), OECD at 34% (in 2000). Overall, the estimates of the Romanian shadow economy vary across time roughly between 19% and 38% of GDP. Table 3.1 below considers three possible values for the shadow economy, a minimum of 19% of GDP, a close to a presumed maximum estimate of 35% of GDP and an in-between value of 27% of GDP. It then presents estimates of total fiscal evasion for each of these three values. The scope of this exercise is to form an idea about a possible range for total fiscal evasion attributed to the shadow economy as well as identifying the corresponding shares of revenue lost in SSC, VAT and PIT. Table 3.1 Fiscal Evasion and the Size of Informal Economy (estimates for the year 2010) GDP shadow economy (bn RON) GDP shadow economy as % of 2010 GDP Total fiscal evasion (as % of GDP), of which: SSC VAT PIT Source: Authors calculations

37 Tax Policy Under The Curse of Low Revenues: The Case of Romania 35 Several points can be noted: First, if the shadow economy amounted to 35% of 2010 GDP, its size would be RON 180 Bn. In this case, full compliance of paying taxes would bring to the budget revenues equivalent to 16% of GDP (assuming a VAT rate of 24%). VAT budgetary revenue alone would rise by 8.4% of GDP. If the size of the shadow economy were 27% of GDP, or RON 139 Bn, fiscal evasion would amount to some 12.3% of GDP. This estimate of the shadow economy is close to the figure of 29.4%, reported by A.T. Kearney (2010). Again, VAT revenues would bring to the budget more than half of the foregone revenues. If the informal economy were 19% of GDP, or of RON 98 Bn. the corresponding value for total fiscal evasion would be 8.7% of GDP. This is slightly lower than the shadow economy figure 51 estimated by the Romanian Fiscal Council (2011) of However, given the estimates of 52 informal economy in other developed countries in the EU, this figure looks suspiciously low. From the data in Table 3.1 it can be inferred that reducing the size of the informal economy could be an important mechanism through which budgetary revenues would be increased. Obviously, assuming that the shadow economy would be brought to light 100% would be unrealistic. Table 3.2 below estimates the impact on budgetary revenues if 30%, 50% and 70% respectively of the shadow economy were to be brought to light under various assumptions for the size of the shadow economy. What is important to note however, is that the figures for SSC (social security contributions), VAT (value added tax) and PIT (personal income tax) in Table above were multiplied by their corresponding index efficiency indicators for year So, no improvements in revenues collection have been assumed, although these are likely to be made in the years to come. Table 3.2 Sensitivity Analysis of Improving Tax Evasion, % of GDP Share of informal economy brought to light The size of the informal economy, % of total output SSC VAT PIT Increase in budgetary revenues SSC VAT PIT Increase in budgetary revenues SSC This figure applies to the year 2009 but between 2006 and 2009 the Romanian Fiscal Council estimates of the share of the shadow economy stayed within 19-20% of GDP. 52 See for instance AT Kearney (2010) where the size of the informal economy is estimated at 22% in Italy, 17.8% in Belgium, 14.6% in Germany, 25.9% in Poland, 23.5% in Hungary, etc. Or OECD (2009) which estimates the size of the Romanian informal sector at 34.4% in These are 0.81 for PIT, 0.58 for both VAT and SSC (see for instance Romanian Fiscal Council 2011).

38 VAT PIT Increase in budgetary revenues Source: Authors Calculations The results show that gains to public sector revenues would be enormous even if only half of the informal economy were to be brought to light. In this case the equivalent of 4% of GDP in extra revenues could be raised if the size of the shadow economy is assumed to lie somewhere around 30% of GDP - a value which seems to be close to reality. 4 Withstanding the Current Economic Crisis and Fiscal Consolidation In autumn 2008 the Romanian authorities implemented several tax changes in order to mitigate the impact of the global financial crisis. These changes included temporary tax exemptions on capital gains from trading securities on the Romanian stock market (in force from 1 January 2009 and applicable for a period of one year). In December 2008, the government introduced a 5% VAT rate for the construction of social dwellings and subject to conditions, private dwellings 54 not exceeding 120sqm and a value of RON 380,000 (about 90,000). But fiscal consolidation has got the upper hand following the sharp fall in budget revenues while the economy plunged in Without corrective measures the budget deficit would have entered a double digit territory easily. A legitimate question is why Romania adopted corrective measures (fiscal consolidation) in spite of having a small debt ratio about 18% at the end of The stark reality is that the large external deficits and massive short-term external borrowing by the non-government sector created prerequisites for a boom bust cycle prior to the eruption of the financial crisis. Once the latter hit a sudden stop in external finance put large portions of the economy to a standstill, hitting investment and consumption and lowering budged revenues dramatically. The outcome was a huge surge in the budget deficit. Its funding became a policy conundrum and fiscal consolidation was a must under the circumstances. Likewise, financial stability was at stake. Urgent fiscal measures were adopted and an international loan package was convened upon.. After a series of yearly reductions (the latest with effect as of 1 January 2009) the budget law of 2009 led to an increase in social security contribution rates by 3.3 pp with effect on 1 February An increase of 2.3pp falls on the employers. Employees with normal working conditions must contribute for social security at 10.5%, while employers contribute at a rate of 20.8%. Furthermore, the government increased the excise duties on alcohol, beverages, cigarettes and fuel as from April Under the spectre of the inability to stop the rise in the budget deficit in 2010, as against 2009, and in the context of a worsening international crisis (the sovereign debt crisis has also come up in that year) a drastic measure was introduced: in mid 2010 the VAT was increased to 24% from 19%. 54 The government launched in 2009 a social program for new homebuyers, fully guaranteeing their mortgage loans. Up to end 2010 there were issued state guarantees amounting to 1.5bn. Most houses bought through the program were old houses not subject to VAT. The tax reduction had rather limited budgetary impact.

39 Tax Policy Under The Curse of Low Revenues: The Case of Romania 37 The measures adopted in 2009 and 2010 stopped the declining trend of tax revenues relative to GDP, which started in The changes in taxation of recent years have enhanced the role of indirect taxes, as some argue should be the case since they are, purportedly, more growth friendly (Johansson et al, 2008) and least pro-cyclical. But in order to strengthen further their role and maximise the revenue collected other measures to bring the hidden economy into the open are badly needed. Otherwise, the current tax structure will hardly be able to bring additional revenues to the public budget. This inability seems to be acknowledged, since projections regarding tax revenues in 2012 do not expect fiscal revenues to rise significantly (Table 4.1). The expansion of budget revenues is projected to come from the growth of other revenue sources (EU funds absorption). Fiscal revenues (including social security contributions) are foreseen to increase by only 0.6% of GDP in 2011 and by only 0.4% in 2012 as compared to As a matter of fact, the recently announced fiscal strategy for does not envisage budget revenues piercing a ceiling of 34% of GDP. This fact begs a critical question again: why tax revenues cannot be increased in Romania? It is true that the poor development level of agriculture does play a role in it, but it cannot explain this situation convincingly. Table 4.1 Structure of Tax Revenues in , %GDP Direct taxes (1) Indirect taxes (2) Social contribution (3) Total Fiscal revenue (4)=(1)+(2)+(3) Non - fiscal revenue (5) Other revenues The challenge of fiscal consolidation stays enormous. For 2011 the pace of budget deficit reduction is about 2.5% of GDP (from 6.9 to 4.4%) and this endeavour is supposed to internalise a part of the inefficiency of public sector companies about 0.5% of GDP. Things are even more complicated in 2012 when the targeted budget deficit is supposed to come down to cca 3% of GDP, as in most EU countries. The EU economic governance reform will make fiscal/budget consolidation a must for Romania too and the European semester will operate fully as a 55 disciplining fiscal policy tool. The challenge of fiscal consolidation is to be judged in view of mounting contingent liabilities on the public budget; these liabilities include debt repayments, arrears and the entitlements to public sector employees. It is true that the inflation tax may alleviate this challenge (bearing in mind that Romania continues to score a pretty high inflation rate in the EU). The bottom line is that raising tax revenues, via combating fiscal evasion (principally) has acquired a strategic policy dimension in Romania. It should, arguably, be the main avenue for undertaking fiscal/budget consolidation. The importance of raising tax/budget revenues is to be judged in conjunction with the need to figure out a more optimal tax structure. Although the EU imposes rules of the game (including the competition law and the prohibition of aid schemes) the competitiveness challenge, at both a national and EU level (the Asian competition) demand a more flexible fiscal policy. (6) Total revenue (7)=(4)+(5)+(6) e e Romania signed the Europact plus in 2011

40 38 By the latter is meant a policy that should be supportive of activities that can enhance the implementation of Europe 2020 strategy and specific national aims. This strategy is a de facto industrial policy at national and EU levels. Combating fiscal evasion and bringing parts of the back economy into the open would repair a fractured sentiment of fairness in the Romanian society. The pains of fiscal correction have amplified perceptions, entertained by not a few, that the flat tax introduction has favoured the better off people, in the main, and has increased income inequality, with implications for social cohesion and trust as a public good in society. How much of this sentiment is justified is not the topic of this investigation. But, it is indisputable that higher fiscal/budget revenues would enable an enhanced provision of public goods. If the reduction of tax evasion (the black economy) would be complemented by a big jump in the EU funds absorption the overall effect for the public budget and for the economy as a whole would be pretty sizeable. Fiscal revenues could also be raised by adjusting royalties and making land/property taxes more appropriate. One should not forget that public goods have the virtue of crowding in private economy output. Last, but not least, raising tax revenues has to be combined with a much higher efficiency of public expenditure. There is a big waste of public resources in Romania owing to resource misallocation, rent-seeking, incompetence, etc. This challenge implies a firm action against corruption, politicization of public administration and political cronyism. 5. Final Remarks and Policy Recommendations It is an acknowledged fact that the tax revenue to GDP ratio in Romania is extremely low, in fact it is the lowest among the EU-27. Given that the overall level of taxation is relatively high, especially when compared to other NMSs, there is much leeway for raising this ratio significantly. Even if agriculture related tax revenues would rise sharply (these could amount to 3% of GDP) the room for increasing overall tax revenues is more than significant. This can be achieved by increasing tax compliance rates and by bringing parts of the shadow economy into the open. Fiscal revenues could also be raised by increasing royalties and property taxes. While direct taxes are not high, taxes on labour handicap job creation. Addressing this imbalance, by reducing, for instance, SSCs may look attractive, but it could affect negatively the budget balance in the short term. And Romania can hardly allow itself slippages in this regard in view of a menacing conditions prevailing in the international environment. It would be easier to reduce SSCs if tax revenues would increase beforehand, possibly following a strong and effective fight against fiscal evasion and a broadening of the tax base. Likewise, tax allowances as a means for stimulating domestic savings and supporting a reorientation of resource allocation toward tradable sectors could also be taken into account. This said, however, efforts should not be made in isolation; they should be pursued bearing in mind the overall objective of fiscal consolidation. Raising tax revenues and improving the efficiency of public expenditure should be pursued simultaneously so that fiscal sustainability is achieved. The public debt sovereign crisis in Europe underscores the importance of achieving fiscal/budget consolidation. Although Romania's public debt is still relatively low compared to 56 most EU countries, its growth rate over the last two years is clearly unsustainable. 56 See Lungu (2011) and Dumitru and Stanca (2011).

41 Tax Policy Under The Curse of Low Revenues: The Case of Romania 39 Public debt grew from 18% of GDP at the end of 2008 to above 36% in It is therefore important to keep control over the size of public debt even though its current level is perceived to be low. The efforts to raise tax revenues should be viewed as a medium term objective. In this respect several aspects should be borne in mind. 1. There is a paramount need to continue fiscal consolidation In March 2011 Romania signed the Euro Pact Plus, which places ceilings on EU members' budget deficits and public debt of 3% of GDP and 60% of GDP respectively. These constraints would be legally binding at the national level. However, the effort to achieve the 3% budget deficit target in 2012, and lower in subsequent years, would be quite sizable. It would entail raising the tax revenues to GDP ratio and further reducing government spending as a share to GDP. In the first step the budget deficit would need to be reduced from 6.5% of GDP in 2010 to 4.4% in To this however, the cost of arrears (as contingent liabilities) would need to be added, amounting to 0.5% in 2011 and some 4.8% of GDP over several years. Notably, a success in raising tax revenues by combating fiscal evasion would help immensely the objective of fiscal consolidation. A large increase in EU funds absorption would diminish the pressure to cut overall expenditure. Fiscal revenues could be increased also by raising royalties. Enhancing the efficiency of budget expenditure would be a tremendous plus. A better targeted provision of social benefits toward the most in need would diminish the funding needs and would help combat income inequality and, therefore, protect social cohesion. 2. Define a clear tax policy concept Usually, national tax policies favour either labour income taxation or consumption taxation. In Romania both these tax levels are high relative to taxes in other EU countries. Labour income taxation, which includes income tax, social security contributions, health and unemployment contributions are among the highest (see Table A4.3) At 24%, the current consumption tax is only one point below the EU ceiling of 25%. This underscores the importance of reducing the labour tax, and, in time, VAT. The reasons for doing this are twofold. First, this is likely to improve work incentives. Second, such actions would leave room for tax policy to react in case a tightening of fiscal policy would be needed in the future. Such an approach involves a forward-looking attitude to tax policy. Optimal taxation is an aspect that has received little attention in Romania. The results in this study tend to suggest that taxation is far from optimal. There is a large shadow economy and the potential for increasing tax revenues is pretty high. There is also scope for using tax policy as a means for improving resource allocation, be it under the constraints of EU rules; fiscal policy could stimulate domestic savings via tax allowances and could be supportive of activities that can improve Romania's economic competitiveness. One should mention here the issue of fairness, which has not been addressed properly following the introduction of the flat tax and which has been given a higher profile by burden-sharing considerations. These arguments suggest that tax policy is far from being optimal in Romania. An analysis of tax optimality should be a separate study, which should be assessing effects of possible changes in tax policy on tax revenues and macro- and microeconomic performance.

42 40 3. Reducing the size of the shadow economy Estimates of the shadow economy show that Romania ranks together with Bulgaria at the top of the EU. Tax evasion is pretty high as it is reflected in data, in the level of tax revenues. Simplified compliance procedures, increased efficiency in detecting non-compliance, severe penalties for those who practice and perpetrate tax evasion and, not least, drastic punishment (including lay-offs) of public sector employees should be considered by the authorities. A similar approach improved incentives for business and labor formalization in Bulgaria, for instance. Penalties for tax evasion should be very high such as to act as a deterrent. 4. Increase administrative effectiveness and efficiency 57 Consolidating local tax offices. This should lead to significant economy of scale, which would raise revenues collected per staff member. In parallel efforts should be focused on further developing the e-tax system, as more taxpayers would need to be convinced to file their tax returns electronically. Improving information systems and information technology management as well as the operational capacity. Improve human resource management function and strategy 5. Adopt a 'Transparency and Credibility' package Historically and after the big fall in the early 90s, the tax revenues to GDP ratio has been relatively stable in Romania, irrespective of the changes in the levels of taxation. Several things could explain this outcome. The first is the weaknesses of institutions responsible for collecting tax revenues. Here political will plays an important role. The second is the population's perception that the value of public services offered in exchange for the taxes it pays is extremely low. Limited progress in building the country's physical infrastructure, the quality of the services in public sector health and education are just a few examples. Such a perception could be very damaging, especially when it becomes entrenched, because it could lead to increased efforts from both employers and employees to avoid paying taxes. The third explanation relates to an issue already touched upon in this paper: to fairness. Government spending on various projects is often perceived to be made to firms belonging to a political clientele. The lack of transparency pertaining to the use of funds and their selective allocation erodes population's trust in public institutions. Moreover, efforts of tax inspectors often concentrate excessively on companies, which regularly pay taxes. In addition, the big waste of resources compounds a huge problem. The fourth explanation is that individual ethical conduct, an important disciplining device based on the principle of self regulation, is permanently dissuaded. The contagion effect is large when a sizable part of market participants evade taxes without any consequences. Given these aspects, one possible way for the authorities to increase tax revenues would be to address the above causes simultaneously, by adopting a comprehensive package of fiscal and budget expenditure reform. Details would need to be sorted out but the general ideas would be the following: 57 In Bulgaria during the revenue administration reform the number of field offices were reduced from 340 to 29, the staff by 25% and the cost of revenue collection declined from 1.4% of revenues in 2002 to 0.8% in 2008.

43 Tax Policy Under The Curse of Low Revenues: The Case of Romania 41 Announce full transparency of government spending programs. Spending on goods and services should be assigned following an open bidding process. Benchmarks for public spending on these goods and services should be clearly defined and any extra spending incurred over the benchmark value should entail consequences. The names of those who decided on major public projects be made public via the internet. Announce clear strategies for investments in infrastructure, health care and education by nominating specific projects together with their costs and completion time. Local public utilities: have tariffs and costs be published on the internet so as to foster benchmarking and competition. Cost-benefit analysis should be done effectively (no public investment above a certain threshold be done without such a thorough analysis). Budgeting and human resource management in the public sector should shift from inputs to objectives and performance. The culture of follow up has to change dramatically in Romania, for the better. Set up an independent Audit Office to monitor public sector spending. Currently there is an institution which audits public sector accounts, Curtea de Conturi, but there is more to be done in this regard. Make it easier for companies and individuals to pay their taxes. Simplify tax forms and improve mechanisms for online payment. Announce a firm plan for tax reductions. VAT should be kept unchanged in the first phase. Labour taxes should be reduced by cutting SSC substantially provided tax revenues permit it (fiscal consolidation proceeds accordingly). The plan should have a 2-3 years horizon and its continuation should be conditional on the increased compliance in paying taxes. Phase 2 would involve reductions in VAT, if the budgetary surplus is achieved and could be 58 sustained. Raise penalties for tax evasion so that these would act as a deterrent. 58 Between 2002 and 2008 the public sector administration reform in Bulgaria saw CIT reduced from 23.5% to 10%, PIT (max rate) from 29% to 10%, SSC from 35.7% to 26.5%, dividend income tax from 15% to 7%, while compliance rates for all these taxes rose strongly.

44 42 Appendix 1. Tax revenues Romania vs. EU-27 and NMSs % GDP Fig.A1.1 Tax revenues including social contribution ro EU27 NMS % GDP Fig.A1.2 Tax revenues including social contribution (cyclically adjusted) ro EU27 NMS 2007 % GDP Fig. A1.3 Indirect taxes % GDP Fig. A1.4 Direct taxes ro EU27 NMS ro EU27 NMS % GDP Fig. A1.5 Social contribution % GDP Fig. A1.6 Consumption taxes ro EU27 NMS ro EU27 NMS % GDP Fig. A1.7 Taxes on labor % GDP Fig. A1.8 Taxes on capital ro EU27 NMS ro EU27 NMS

45 Tax Policy Under The Curse of Low Revenues: The Case of Romania 43 % Fig. A1.9 Implicit tax rate on consumption % Fig. A1.10 Implicit tax rate on labour % GDP ro EU27 NMS Fig. A1.11 Personal income tax and social contribution paid by individuals % GDP ro EU27 NMS Fig. A1.12 Corporate income tax and social contribution paid by employers ro EU27 NMS ro EU27 NMS Source: European Commission (2010)

46 44 Appendix 2. Households as taxpayers social benefit Fig. A2.1 Income tax & social security contribution share vs. social benefits share of poorest household decile,% total social benefit (left scale) tax contribution tax contribution (right scale) social benefit Fig. A2.2 Income tax & social security contribution share vs. social benefit share of richest household decile, % total social benefit (left scale) tax contribution tax contribution (right scale) 14.0 Fig. A2.3 Income tax and social security contribution share vs. social benefit share of middle income household deciles, % total Fig. A2.4 Income tax and social security contribution-to--income ratio for low, high and middle income household deciles social benefit decile 5-social benefit decile 5-tax contribution decile 6-social benefit decile 6-tax contribution tax contribution decile 1 decile 5 decile 6 decile 10 tax contribution Fig. A2.5 Income tax & social security contribution share vs. social benefits share of employees households, % total tax contribution (left scale) social benefit social benefit (right scale) tax contribution Fig. A2.6 Income tax &social security contribution share vs. social benefits share of agricultural households, % total tax contribution (left scale) social benefit (right scale) social benefit 40 Fig. A2.7 Distribution of tax burden among households deciles 15 Fig. A2.8 Distribution of social benefits among households deciles 35 %of total %of total d1 d2 d3 d4 d5 d6 d7 d8 d9 d10 0 d1 d2 d3 d4 d5 d6 d7 d8 d9 d10 pre post pre post

47 Tax Policy Under The Curse of Low Revenues: The Case of Romania Fig. A2.9 Income tax & social security contribution share vs. social benefits share of retiree households, % total Fig. A2.10 Income tax and social security contribution -to-income ratio, by households type tax contribution social benefit tax contribution (left scale) social benefit (right scale) employees households retirees households agricultural households Source: INSSE, Households'Budget Survey

48 46 Appendix 3. Companies' arrears to consolidated general budget % GDP Fig. A3.1 Arrears to consolidated general budget, by type % GDP Fig. A3.2 Arrears to consolidated general budget of state companies, by type dec. dec.2006 dec.2007 dec.2008 dec.2009 iun.2010 dec. dec.2006 dec.2007 dec.2008 dec.2009 iun.2010 social insurance special funds state budget local budgets social insurance special funds state budget local budgets % GDP Fig. A3.3 Arrears to consolidated general budget of private companies, by type % GDP Fig. A3.4 Arrears to consolidated general budget, by companies ownership dec. dec.2006 dec.2007 dec.2008 dec.2009 iun dec. dec.2006 dec.2007 dec.2008 dec.2009 iun.2010 social insurance special funds state budget local budgets public companies state companies % GDP Fig. A3.5 Arrears to state budget, by companies ownership dec. dec.2006 dec.2007 dec.2008 dec.2009 iun.2010 % GDP Fig. A3.6 Arrears to social insurance budget, by companies ownership dec. dec.2006 dec.2007 dec.2008 dec.2009 iun.2010 public companies state companies public companies state companies

49 Tax Policy Under The Curse of Low Revenues: The Case of Romania % Fig. A3.7 Arrears to consolidated general budget, by type and sectors 80% 60% 40% 20% 0% total total insurance special funds state budget local budget Agriculture Mining Manufacturing Utilities Construction Trade Services Real Estate Source:NBR

50 Appendix 4. Memento Items Table A4.1 Public Debt in EU EU-27 BE BG CZ DK DE EE IE EL ES FR IT CY LV LT LU HU MT NL AT PL PT RO SI SK FI SE UK Source: European Commission (2010d) 48

51 Table A4.2 Changes of the main taxes in Romania Energy suppliers and mining exempted Period Tax rates Deduction/exemptions Profit tax 1990 Progressive tax rates differentiated upon enterprises profitabili ty Enterprises with foreign ownership exempted for 2 years after taxable incomes obtained, and for the following three years it was possible a reduction by 50% of the tax. For reinvested profits in Law 71 awarded long term tax holidays selectively to foreign investors Beginning 1991 Progressive taxes ranging between 5% (annual profit of ROL ) and 77% (annual profit >ROL 955m) with 67 brackets End 1991 Two brackets 30% min rate and 45% max rate % % Emergency Ordinance 92 introduces new tax holidays and tax cuts (including on customs duties and VAT) % Law 241 extended the tax holidays and tax cuts to a larger class of beneficiaries % 5% preferential rate for profits arising from export activities 10% investment tax allowance All expenses are deductible except certain categories such as penalties, fines, protocol expenditures exceeding a certain level, provisions, etc. 16% Wage tax April 1991 Progressive rates accord ing to Tax Policy Under The Curse of Low Revenues: The Case of Romania 49 income brackets ranging from 5% and 60% 1998 Progressive rates according to income brackets, ranging from 21% to 45% 2000 Yearly taxation standard comprising progressive rates according to brackets of taxable yearly income ranging from 18% to 40%

52 50 16% VAT % Reduced rate 9% meat, fish, milk, edible fats, medications for human and veterinary use, live animals, agricultural and land -improvement works, chemical and mineral fertilizers, fresh vegetables and fruits (until April 1997) % Reduced rate 9% for the above items plus advertising and publicity activities, eggs of domestic species of fowl, flour and flour pastes, sugar, rice, uniforms for preschool and elementary school pupils, articles of clothing and footwear for babies, urban transport for travellers, prostheses and orthopaedic products % Reduced rate 11% for the above items % % Reduced rate: 9% medicine, books, magazines, admission to cultural events, hotel accommodation % Reduced rate; 9% for the above items and 5% for construction of social and, subject to conditions, private dwellings not exceeding 120sqm and a value of RON ( 90000) % Reduced rate; 9% for the above items and 5% for construction of social and, subject t o conditions, private dwellings not exceeding 120sqm and a value of RON ( 90000) Social security contributions % 3% employees contribution % 5% employees contribution % 5% employees contribution % Decrease in employers and employees contribution % Increase in employers contribution, reduction in employees contribution % Reduction in employers contribution % Reduction in employers contribution % Reduction in employers con tribution % Reduction in employers and employees contributions % Increase in employers and employees contribution

53 Tax Policy Under The Curse of Low Revenues: The Case of Romania 51 Table A4.3 Main legal tax rates in NMSs Pre RO VAT PIT CIT SSC Bg VAT PIT CIT SSC CZ VAT PIT CIT SSC EE VAT PIT CIT SSC LV VAT PIT CIT SSC Lt VAT PIT CIT SSC Hu VAT PIT CIT SSC PL VAT PIT CIT SSC SI VAT PIT CIT SSC SK VAT PIT CIT SSC Source : European Commission, IMF

54 52 References Alexandru Adriana AnaMaria, Dobre Ion and Ghinararu Catalin 2009, Estimating the size of the Romanian Shadow Economy Using the Currency Demand Approach Tudorel Andrei, Andreea Iluzia Iacob, Stelian Stancu, Bogdan Oancea 2010, Quantitative Techniques used for the Informal Economy Analysis at National and Regional Level, Informatica Economică vol. 14, no. 3/2010. Atkinson A.B. and Stiglitz J. E. (1976), The design of Tax Structure, Journal of Public Economics 6, pp Bach S., Corneo G. and Steiner V. (2011), Optimal top marginal tax rates under income splitting for couples, CEPR Discussion Paper 8435 Becker Torbjorn and Daniel Daianu, Zsolt Darvas, Vladimir Gligorov, Michael Landesmann, Pavle Petrovic, Jean Pisani-Ferry, Dariusz Rosati, Andre Sapir, Beatrice Weder Di Mauro, (2010), Whither Growth in Central and Eastern Europe? Policy Lessons for an Integrated World, Bruegel Blueprint Series; Blanchard O. (1997), The Economics of Transition in Eastern Europe, Clarendon Lectures, Oxford University Press Daianu, Daniel and Radu Vranceanu (2000/2001), Pitfalls of Taxation Policy in Transition Economies, Acta Oeconomica, Vol.51 (1), pp.3-15 Diamond P. A. and Mirrlees J. A. (1971), Optimal Taxation and Public Production, Review of Economic Studies, 39, pp Diamond P. A. (1998), Optimal income taxation: An example with a U-shaped pattern of optimal marginal rates, American Economic Review, 88, pp83-95 Dumitru, Ionut and Stanca, Razvan (2011), Fiscal Discipline and Economic Growth The Case of Romania in Selected Papers, Scientific Romanian Diaspora Conference. European Commission (2009a), Taxation Trends in EU European Commission (2009b), Effective Tax Levels using The Devereux/Griffith Methodology European Commission (2010a), Taxation paper No 24: 'Monitoring tax revenues and tax reforms in EU Member States - Tax policy after the crisis'. European Commission (2010b), Taxation Paper No 20: 'The 2008 financial crisis and taxation policy'. Written by Thomas Hemmelgarn and Gaëtan Nicodème European Commission (2010c), Taxation trends in the European Union 2010 European Commission (2010d), Public Finances in the EMU 2010, European Economy 4 European Commission (2011), Taxation trends in the European Union 2011 Gordon R. and Li W. (a), Puzzling tax structures in developing countries: a comparison of two alternative explanations, NBER Working Paper Gordon R. and Li W. (b), Tax structure in developing countries: Many puzzles and a possible explanation, NBER Working Paper Gupta A.S. (2007), Determinants of Tax Revenue Efforts in Developing Countries, IMF Working Paper 184 IMF(1998), Country Report 123 IMF(2001), Country Report 16 IMF(2003), Country Report 12 IMF(2004), Country Report 220 IMF(2006), Country Report 169 IMF (2010, February), Strategies for Fiscal Consolidation in the Post-Crisis World. IMF(2010, March), Albania-2010 Article IV Consultation, Preliminary Conclusion of the Mission Johansson A., C. Heady, J. Arnold, B. Brys and L. Vartia (2008), Taxation and Economic Growth, OECD Economics Department Working Paper 620 Kornai, Ianos (1980), 'The Economics of Shortage, Amsterdam, North-Holland A.T. Kearney (2010) The Shadow Economy in Europe Laurian, Lungu (2011), Fiscal Sustainability in Romania in Selected Papers, Scientific Romanian Diaspora Conference Mankiw N.G., Weinzierl M. and Yagan D. (2009), Optimal taxation in theory and practice, NBER Working Paper Mirrlees J. A. (1971) An Exploration in the Theory of Optimum Income Taxation, Review of Economic Studies, vol 38, no 2, pp OECD (2009) Competition Policy and the Informal Economy, Policy Roundtables Piatkowski M. and Jarmusek M. (2008), Zero Corporate Income tax in Moldova, Tax comnpetition and its implication for eastern Europe, IMF Working Paper 203 Rodrik D. (1998), Why do More Open Economies have Bigger Governments?, Journal of Political Economy, Vol 106, pp Romanian Fiscal Council (2010), Pozitia finantelor publice in Romania Comparatii internationale Romanian Fiscal Council (2011) Annual Report Saez E. (2001), Using Elasticities to derive Optimal Income Tax Rates, Review of Economic Studies 68, pp Schneider F. (2010), Size and Development of the Shadow Economy of 31 European Countries from 2003 to 2010 Tuomala M. (1990), Optimal Income Tax and Redistribution, Clarendon Press, Oxford Von Hagen J. and Traistaru I. (2004), Macroeconomic adjustment in the new EU Member States, SUERF Studies (European Money and Finance Forum 2006/4) edited by Morten Balling

55

56 TAX POLICY UNDER THE CURSE OF LOW REVENUES: THE CASE OF ROMANIA Bucharest Black Sea 21 Emanoil Porumbaru Str., 1st Floor, Ap. 3 Sector 1, Bucharest Tel.: /-83 Fax: fes@fes.ro