Abstract
This Selected Issues paper analyzes the major factors that may have contributed to the marked widening of the external current account imbalance in Greece since the mid-1990s, with the deficit reaching one of the highest levels (in relation to GDP) among advanced economies. The factors reviewed include developments in relative cost and other competitiveness indicators, business cycle asynchronization, and idiosyncratic supply shocks and immigration. Potential implications of euro area entry are discussed, in particular the impact of declining interest rates in the context of monetary union.
III. Tax Reform in Greece—Cleaning the Augean Stables1
1. Prior to 2002, the Greek tax system had not undergone a fundamental review and reform in many decades. Over time, it had become exceedingly complex, difficult to administer and comply with, and had developed many features that misallocated resources and hindered horizontal and vertical equity.
2. Recognizing these shortcomings, the authorities enacted measures in late 2002 aimed at improving the system’s neutrality and simplicity, reducing administrative costs, and promoting employment and entrepreneurship. This included measures to simplify tax recordkeeping, increase the use of electronic technologies in tax administration, streamline VAT administration, reform personal income, gift and inheritance taxes, and eliminate a number of relatively minor, yet costly to administer, taxes and duties.
3. The reforms adopted should help to simplify the tax system and improve equity, but a number of important issues remain to be addressed. These include limiting or eliminating many remaining exemptions and deductions in corporate taxation, reducing further labor income taxation (through reform of the pension system), introducing a modern property tax system, eliminating remaining minor “nuisance” taxes, and reforming the local tax system.
4. However, future tax reform efforts would need to be undertaken within the context of macroeconomic exigencies. Greece’s general government debt ratio, in excess of 100 percent of GDP, poses a substantial obstacle to reducing its tax burden in the medium term. In addition, pension expenditures, at l2½ percent of GDP already among the highest in the EU, are set to almost double by 2050, by far and away the largest projected increase among advanced economies.2 Thus, any efforts at reducing the overall tax burden, as has been attempted in recent years in other EU economies, would be constrained by the need to impose at least commensurate reductions in expenditures to ensure fiscal sustainability.
5. This chapter is organized as follows. Section A compares the size and structure of Greece’s prereform tax yield to those in other advanced economies, and examines its average effective tax ratios. Section B reviews some of its prereform statutory features, and compares it with the reforms. Section C examines areas in which further reforms could be considered, and Section D concludes.
A. Macroeconomic Aspects of the Prereform Greek Tax System
6. In order to fully understand the tax system’s current structure, it is useful to examine briefly Greek fiscal developments in recent decades. While small overall surpluses in the 1960s were replaced by small deficits in the 1970s, the general government debt ratio remained quite low (Figure 1). In the 1980s, however, primary expenditures rose rapidly, especially from increased social transfers and public employment, while revenues lagged, with resulting deficits in excess of 10 percent of GDP and the debt-to-GDP ratio increasing rapidly. Attempts at fiscal consolidation began in the early 1990s, but were achieved entirely through a further rise in the revenue ratio, while primary expenditures have remained broadly constant as a share of GDP since the mid-1980s. At present, the general government has a small deficit (1.2 percent of GDP in 2002), although the debt ratio, in excess of 100 percent of GDP, has yet to substantially decline.
Greece: General Government Developments, 1960-2002
(Percent of GDP)
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Sources: Ministry of Economy and Finance; OECD; and Fund staff calculations.7. As part of the recent fiscal consolidation, Greece’s tax revenue yield has increased quite markedly. Although it was previously below European and OECD averages, its recent rapid rise places it on par with these averages, and above those of other EU economies with similar or even higher levels of per capita incomes (Figure 2). This has been the result of increased personal income tax revenues, reflecting both fiscal drag, as tax schedules were not adjusted fully for inflation and wage growth, new taxes on interest income and stock exchange transactions, and improved tax administration, which also affected social security contributions and the VAT (Figure 3).
Greece: Tax Revenue -- International Comparisons
(In percent of GDP)
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Source: OECD Revenue Statistics.1/ 1998-2000Greece: Tax Revenues, 1965-2001
(Share of GDP)
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Source: OECD Revenue Statistics.8. The composition of Greek tax revenues differs in some important respects from averages in other major advanced economies (Figure 4). In general, Greek revenues are more heavily dependent on consumption-based taxes and social security contributions, and less so on direct taxation, especially personal income taxes. This reflects a combination of factors, including the relative size of corresponding tax bases, the structure of the tax code, including the interaction of tax schedules with deductions, exemptions, allowances and credits, and the level of tax compliance (an issue of particular importance in the Greek context).
Greece: Tax Composition--International Comparisons, 1965-2001
(Share of Total Tax Revenues)
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Source: OECD Revenue Statistics.9. One factor affecting the differences in the magnitude and composition of Greece’s taxes reflects relative differences in the size of its tax bases compared to those in other countries. One notable difference results from the continued high relative share of the self-employed in Greece. Dependent employment accounted for only 58.8 percent of the employed in Greece in 2002, compared to 83.6 percent in the euro area (France was the next lowest, at 70.7 percent, and Sweden had the highest rate was at 94.2 percent). This reflects the still sizable role played by small farmers, as well as by self-employed shopkeepers and small hoteliers. As a result of this economic structure, the share of wages in GDP is the smallest in Greece among OECD economies for which data are available (Figure 5).3 In contrast, Greece has the highest share of net operating surplus to GDP, about double the EU and OECD averages. This component includes the value added of the self-employed, which conceptually contains both a return to labor and capital, and thus overstates capital’s (and understates labor’s) true income share in the economy. As regards the third major tax base, Greece’s consumption share is the second highest in the OECD, although its difference from the averages is not as extreme compared to those for income shares.
Greece: Relative Tax Bases--International Comparisons
(1996-2000 Averages)
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Sources: Carey and Robesona (2002); OECD (2001); and Fund staff calculations.1/ 1996-1999.2/ 1996-1998.10. A second factor affecting Greece’s tax yield are the various rates of taxation. However, determining the “effective” tax rates goes beyond a simple comparison of statutory rates, and requires examining the influence that various deductions, allowances, exemptions and tax credits have in determining actual tax liabilities. Two broad approaches to examining this have been adopted. The first is to calculate effective marginal tax rates based on the tax code, and requires substantial detailed data. Examples of this approach include the OECD’s Taxing Wages project for labor taxation, and King and Fullerton (1984), and Devereux and Griffith (1998) for capital taxation. This requires constructing many cases under various combinations to obtain representative aggregate estimates.
11. This chapter instead uses average effective tax rates (AETRs) to assess effective tax rates in Greece. This second approach was initially proposed by Mendoza, Razin, and Tesar (1994), and has become increasingly popular as a proxy for effective marginal tax rates primarily because of its simplicity. It uses internationally comparable national aggregate revenue data and proxies tax bases with internationally comparable national accounts data. Data sources and methodology underlying the AETR calculating are discussed in the Appendix. Thus, it reflects actual revenue generated from potentially vastly different tax systems, effectively accounting for a multitude of details regarding various deductions, statutory tax rates, allowances and credits.
12. The AETR approach, nevertheless, has a number of drawbacks as a proxy for effective marginal tax rates, which are thought to have a more important influence on economic decisions.4 Among the more important implicit assumptions, which may not hold, are: (i) that the taxes are borne by those paying them (i.e., not shifted to others); (ii) that the calculated (backward-looking) average tax rates are good proxies for expected future marginal tax rates; (iii) that national accounts-based tax bases are good proxies for actual bases (e.g., capital gains are included in many tax receipts, but not in national accounts tax bases, thereby overstating tax rates); (iv) and that loss carry forwards are not significant. Nevertheless, they remain a good “first brush” measure of economies’ relative tax reliance. Subsequent work has suggested further refinements (see the references in footnote 3), but in most cases the variations in tax rates over time within a country, and relative country rankings over the most recent period using alternate tax estimates, tend to be highly correlated.
13. The AETR analysis suggests that prior to the tax reform Greece’s labor and consumption tax rates were comparable with average EU rates, but that capital taxation was much lower (Figure 6).5 However, the contribution of personal income taxation to total labor taxation was lower than in any other country except Korea, while the social security component exceeded that in any other country.6 Excises in Greece contributed slightly more than average toward total consumption taxation, with a somewhat smaller contribution from the VAT. Other, more sophisticated estimates of AETRs are found to suggest broadly comparable levels, patterns, and relative tax rates across countries.7
Greece; Average Effective Tax Ratios-International Comparisons
(1996-2000 Averages)
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Sources: Carey and Robesona (2002); and staff calculations.1/ 1996-1999.2/ 1996-1998.3/ Split between personal income tax and payroll tax/social security contributions not available.14. As with Greece’s overall tax yield, its AETRs have risen steadily over the past decade (Figure 7). In fact, the AETR on labor now exceeds that for the EU, while that for capital, despite increasing significantly, still lags the EU average by over 15 percentage points. The pattern changes somewhat if an attempt is made to split the self-employed’s incomes and tax payments between capital and labor (rather than to allocate all of them both to capital), with the labor tax ratio falling by about 9 percentage points (as the effective tax on self-employed labor is much lower than that on dependant employment) and the capital tax ratio rising by about 4 percentage points. This revised capital tax ratio remains well below international averages and well below the AETR for labor (while the two ratios are roughly comparable based on OECD averages). Greece’s AETR on consumption has been broadly comparable to the EU average since 1987 (the year it introduced the VAT).
Greece: Average Effective Tax rates, 1975-2000
Citation: IMF Staff Country Reports 2003, 157; 10.5089/9781451816150.002.A003
Source: Carey and Rabesona (2002).B. Major Features of the 2003 Tax Reform
15. This section outlines the major features of the 2003 reform, contrasting them briefly with the prereform system.8 In addition to the discussion of specific tax reforms, a few introductory comments are useful. There was widespread dissatisfaction with the Greek tax system, amid perceptions of inequities, with a resulting widespread reluctance to comply. The economy remains sharply divided between wage earners and pensioners, who have little scope to avoid tax obligations withheld at source, and the self-employed, who enjoy ample opportunities. In response, tax administration efforts had increased significantly in the last decade, including greater reliance on presumptive measures (“objective criteria” in the Greek parlance) in determining self-employed tax liabilities. While this has succeeded in improving compliance (although tax evasion is acknowledged to remain a significant problem), it also engendered increasing resentment, and given an inevitable degree of arbitrariness in applying these “objective” criteria, claims of abuse by the tax authorities and public corruption.
16. Due to the influence of various special interests, as well as the policies of successive governments, the Greek tax system became exceedingly complex, increasing administrative costs and hindering compliance. Targeted exemptions, allowances or other benefits for specific activities, occupations, or politically influential (at some time) social groups littered the legal code, which combined with the absence of single consolidated tax code, increased complexity, and complicated compliance as well as equity objectives. In addition “third-party” taxes, imposed again to benefit favored groups, accreted, with high administrative costs, and doubtful economic and efficiency and equity implications.9 Continuous revisions to the tax code increased uncertainty, and frequent tax amnesties had also eroded perceptions of fairness of the tax system and compliance.
17. In response to the perceived shortcomings, the authorities initiated a reform process. A committee was established, assigned to review the tax system, and prepare reform proposals. Upon receipt of its report in April 2002 (Tax Reform Committee, 2002), and following a dialogue with social partners, the government adopted two laws, covering simplifications of bookkeeping requirements and VAT compliance (Law 3052/September 24, 2002), and reforms to the personal income, capital income, and property transaction and inheritance and gift taxes (Law3091/December 24, 2003). The major provisions of these laws are described in Box 1. The authorities estimate that the reforms would result in a revenue loss to the budget of €900 million (equivalent to 0.6 percent of GDP) in 2003, and an additional €540 million (0.3 percent of GDP) in 2004. Although no official estimates are available for the separate reform components, the authorities estimate that personal income tax reforms are expected to cost €760 million, while measures relating to corporations and liberal professions is expected to cost €690 million. The adopted measures, while themselves significant steps toward improving the tax system’s simplicity and neutrality, and reducing administrative and compliance costs, were characterized by the authorities in the 2003 budget as “initial” reforms.
Tax Reform Measures1
Personal income taxation
Simplify the system by abolishing a large number of existing reductions, exemptions, allowances and special treatment of incomes, and substantially reducing the number of taxpayers that have to file tax returns;
Reduce the number of tax brackets from 6 to 5 (and to 4 in 2004), while reducing the top marginal tax rate to 40 percent (from 42.5 percent), and substantially increasing the size of the zero-bracket;
Increase the tax exemption for children;
Transform remaining expense reductions into tax credits, and capping them;
Abolish most “objective criteria”, except for those pertaining to high incomes, and those with inadequate bookkeeping;
Harmonize at 15 percent the interest income tax rate;
Corporate income taxation
Align the corporate tax rate on companies not listed on the Athens Stock Exchange with listed companies at 35 percent;
Increase deductions for group life insurance premia:
Allow for the option of choosing among two depreciation schedules for expensing investment;
Property taxation
Reduce the property transfer tax rate from 9-11 percent to 7-9 percent;
Impose a 3 percent tax on offshore companies’ domestic real estate holdings;
Increase the zero-bracket amount, and reduce the inheritance and gift tax rate in general (especially for transfers between parents and children and for first residences);
Reduce and transform to a single rate the inheritance and gift tax rate for listed and most nonlisted companies;
Other taxes
Abolish many stamp duties;
Tax administration
Reduce and simplify tax record keeping requirements, especially for small- and medium-sized firms;
Lengthen from two to three months the VAT tax filing period for small businesses, thereby easing administration and compliance costs;
Increase reliance on electronic submissions and control;
Codify the tax laws into one single consolidated text.
Personal income taxes
18. Personal income taxes are paid by employees, pensioners, and the self-employed, based on a progressive schedule, net of social security contributions, and of deductible expenses. The schedule in 2002 contained six brackets, ranging from 0 to 42½ percent. The statutory tax rates suggested relatively steep progressivity (OECD, 2001). In addition, the schedule had not been fully adjusted for the effect of inflation and real wage growth for revenue reasons, resulting in fiscal drag. However, a wide range of income exemptions and allowances were often effected through the deduction of certain expenses. This provided benefits at the marginal tax rate, which especially benefited those with higher incomes, and thereby strongly blunted any redistributive effects of the tax schedule, as well as imposing large revenue losses to the budget and complicating tax administration and compliance costs.10
19. Reforms to the personal income tax system should improve horizontal and vertical equity and reduce sizably administrative costs. Increasing the size of the zero-bracket amount (by 19 percent for wage earners and pensioners, and by 13.5 percent for others) and tax deductions for children is expected to reduce significantly the number of individuals required to submit income tax filings, thereby lowering administration and compliance costs (Table 1).11 The elimination of many income deductions, and the capping of those remaining and their replacement with tax credits should also simplify tax administration and improve vertical equity. Harmonizing interest income taxation (at 15 percent) should reduce tax-induced capital misallocation. Abolishing most “objective criteria” for self-employed income assessments, if accompanied by necessary improvements in bookkeeping requirements (which entails risks to fiscal revenues but is essential in shifting to a modern tax system), should improve the horizontal equity and transparency of the tax system.
Greece: Personal and Corporate Income Tax Rates, 1999-2003
Greece: Personal and Corporate Income Tax Rates, 1999-2003
| Personal Income Tax | Corporate Income Tax | |||||
|---|---|---|---|---|---|---|
| Wage Earners and Pensioners | Nonwage Earners and Professionals | Listed | Nonlisted | |||
| Total Income | Tax rate | Total Income | Tax rate | Tax rate | Tax rate | |
| In euros | In percent | In euros | In percent | In percent | ||
| 1999 | ||||||
| 5,576 | 0 | 4,696 | 0 | 35 | 40 | |
| 7,954 | 5 | 7,954 | 5 | |||
| 12,723 | 15 | 12,723 | 15 | |||
| 22,246 | 30 | 22,246 | 30 | |||
| 47,646 | 40 | 47,646 | 40 | |||
| over | 45 | over | 45 | |||
| 2000 | ||||||
| 6,749 | 0 | 5,869 | 0 | 35 | 40 | |
| 7,954 | 5 | 7,954 | 5 | |||
| 12,723 | 15 | 12,723 | 15 | |||
| 22,246 | 30 | 22,246 | 30 | |||
| 47,646 | 40 | 47,646 | 40 | |||
| over | 45 | over | 45 | |||
| 2001 | ||||||
| 7,043 | 0 | 6,163 | 0 | 35 | 40 | |
| 8,363 | 5 | 8,363 | 5 | |||
| 13,363 | 15 | 13,363 | 15 | |||
| 23,363 | 30 | 23,363 | 30 | |||
| 50,033 | 40 | 50,033 | 40 | |||
| over | 42.5 | over | 42.5 | |||
| 2002 | ||||||
| 8,400 | 0 | 7,400 | 0 | 35 | 37.5 | |
| 8,400 | 5 | |||||
| 13,400 | 15 | 13,400 | 15 | |||
| 23,400 | 30 | 23,400 | 30 | |||
| over | 40 | over | 40 | |||
| 2003 | ||||||
| 10,000 | 0 | 8,400 | 0 | 35 | 35 | |
| 13,400 | 15 | 13,400 | 15 | |||
| 23,400 | 30 | 23,400 | 30 | |||
| over | 40 | over | 40 | |||
Greece: Personal and Corporate Income Tax Rates, 1999-2003
| Personal Income Tax | Corporate Income Tax | |||||
|---|---|---|---|---|---|---|
| Wage Earners and Pensioners | Nonwage Earners and Professionals | Listed | Nonlisted | |||
| Total Income | Tax rate | Total Income | Tax rate | Tax rate | Tax rate | |
| In euros | In percent | In euros | In percent | In percent | ||
| 1999 | ||||||
| 5,576 | 0 | 4,696 | 0 | 35 | 40 | |
| 7,954 | 5 | 7,954 | 5 | |||
| 12,723 | 15 | 12,723 | 15 | |||
| 22,246 | 30 | 22,246 | 30 | |||
| 47,646 | 40 | 47,646 | 40 | |||
| over | 45 | over | 45 | |||
| 2000 | ||||||
| 6,749 | 0 | 5,869 | 0 | 35 | 40 | |
| 7,954 | 5 | 7,954 | 5 | |||
| 12,723 | 15 | 12,723 | 15 | |||
| 22,246 | 30 | 22,246 | 30 | |||
| 47,646 | 40 | 47,646 | 40 | |||
| over | 45 | over | 45 | |||
| 2001 | ||||||
| 7,043 | 0 | 6,163 | 0 | 35 | 40 | |
| 8,363 | 5 | 8,363 | 5 | |||
| 13,363 | 15 | 13,363 | 15 | |||
| 23,363 | 30 | 23,363 | 30 | |||
| 50,033 | 40 | 50,033 | 40 | |||
| over | 42.5 | over | 42.5 | |||
| 2002 | ||||||
| 8,400 | 0 | 7,400 | 0 | 35 | 37.5 | |
| 8,400 | 5 | |||||
| 13,400 | 15 | 13,400 | 15 | |||
| 23,400 | 30 | 23,400 | 30 | |||
| over | 40 | over | 40 | |||
| 2003 | ||||||
| 10,000 | 0 | 8,400 | 0 | 35 | 35 | |
| 13,400 | 15 | 13,400 | 15 | |||
| 23,400 | 30 | 23,400 | 30 | |||
| over | 40 | over | 40 | |||
Social security contributions
20. In contrast to relatively modest, by international comparisons, personal income taxes, Greek social security contributions are among the highest in the OECD. However, this varies widely between dependent employees, who can not easily evade taxation, and the self-employed, who, as with personal taxation, have widespread opportunities and incentives to do so. Contribution rates also vary widely by sector, with some completely exempt (e.g., farmers), while others are replaced or supplemented by third-party taxes. The overall effect of this structure is to provide an incentive to be self-employed and to discourage secondary earners to enter the labor market.12
21. Following the failure of recent efforts to address the unsustainability of the pension system, reductions in social security contribution rates are unlikely. Earlier “modest” reform proposals were forcefully rejected by social partners, and the government, aside from undertaking some minor adjustments (including lengthening somewhat the pension base period), and unification of replacement rates for all pensioners (itself a retrograde step, as it increased benefits for some workers and thereby worsened the financial position of an already unsustainable system), has no intention to revisit pension reforms in the near future, and therefore to allow for a reduction in contribution rates.13
Capital income taxation
22. Effective capital income tax rates vary widely, depending on the taxpayers’ legal form and way in which it is financed, invested, and distributed. Some forms of capital income are included in the personal income tax (e.g., rents) and others (capital gains, interest and dividends) are subject to separate fixed scheduler rates (varying from 0 to 15 percent, including, before the reform, differing rates for different sources of interest income). The statutory corporate income tax rate (at 35 percent for companies listed on the stock exchange and 37½ percent for nonlisted companies) was somewhat higher than statutory international averages (see Tax Reform Committee, 2002). However, the effective rate of corporate income taxation was fairly low, reflecting a wide range of tax incentives.14 The marginal effective tax wedges in manufacturing displayed relatively less distortions across physical assets, but were strongly biased in favor of debt financing. In addition, differences in the tax rate on incorporated firms and partnerships, together with lower social security contributions by the self employed, provide a bias in favor of unincorporated businesses (OECD, 2001).
23. Individual interest income tax rates were unified at 15 percent, reducing the previous preferential treatment for government debt. There were no major changes to corporate taxation aside from unifying the rate for unlisted companies at 35 percent. Companies are now allowed greater deductions for group life insurance plans for their workers, and the taxation of severance payments to redundant workers was reduced to facilitate separations. Companies can also now chose between two rates of depreciation for expensing fixed asset acquisition costs. However, the wide variations in effective capital taxation depending on legal form, source of financing, and form of distribution remains.
Indirect taxes
24. Greece levies VAT and excises on consumption goods in a system largely harmonized with practices in other EU countries. Greece’s standard rate, at 18 percent, is broadly comparable with the EU average, and its reduced rate categories (14 percent for “necessities,” such as fresh food, pharmaceuticals, energy and some services, and 4 percent for “cultural goods,” such as publications and theater tickets) are not appreciably different in scope than in other EU economies reduced rates (13, 6, and 3 percent respectively) apply in some islands. Improved tax compliance has increased the VAT’s tax yield substantially in recent years.15 Minimum excise tax rates have been harmonized with EU legislation, but petroleum excises in Greece remain well below EU averages, while levies on oil for industrial uses is double the average (which boosts users’ production costs). A number of other specific taxes (on automobiles, stamp taxes on various transactions, and turnover taxes) also exist, and are generally costly to administer, while generating relatively low revenues.
25. While no major reforms were made to the VAT and excise systems, revenue requirements restricted the desired elimination of stamp duties and other specific taxes. Only 30-40 percent of the total stamp duties were abolished, eliminating only about one quarter of their revenue yield, but the authorities plan for their further limitations in the future.
Property taxes
26. The contribution of property taxes to the total tax yield in Greece is similar to that in other countries, but its composition is much different, relying almost exclusively on transfer taxes. In contrast, property ownership taxes supply half of all property tax revenues in the EU. However, the lack of comprehensive land register precludes its imposition in Greece (as well as the extension of the VAT to new housing construction). Instead, revenues were generated by taxes of 9-11 percent on property transactions, which provided strong incentives to understate property values and inhibited labor mobility. Taxes on inheritances and gifts were steeply progressive (up to 25 percent for the value of estates in excess of about €212,000, and up to 60 percent above the same amount among nonrelated individuals) also leading to tax avoidance measures (including the creation of offshore companies).
27. Measures to reduce the taxation of property transfers and impose taxes on offshore holdings of domestic real estate should improve capital allocation and reduce tax avoidance. Initial steps in this area (reducing the transfer tax to 7-9 percent, reducing the inheritance and gifts tax on first residences, drastically cutting—to 0.6-1.2 percent—the same tax for listed and nonlisted companies, and imposing a 3 percent tax on offshore companies Greek real estate holdings) are useful to limit previous impediments to resource reallocation and improve tax compliance.
Local government taxes
28. Local government taxes are all but absent, contributing less than one percent of general government revenues. Instead, local governments rely on shared tax revenues collected by the central government and on central transfers. As a result, there is little local accountability to use centrally provided resources efficiently, or ability to tailor the provision of government services to reflect local preferences. The 2003 reforms did not include any provisions relating to local taxation.
Tax reform and AETRs
29. As a major focus on the tax reform was to simplify the system and lower administrative costs, it is not anticipated that there would be significant effects on average effective tax rates. Increasing the zero-bracket for personal income taxes, and reducing other marginal rates should reduce the labor tax ratio, but this would be partly offset by reductions in deductions and allowances, and the contribution of personal income taxes to the labor tax ratio was already very limited. Improving tax compliance among the self employed should increase somewhat capital’s tax ratio (and labor’s as well if social security contributions are also increased).
Tax administration
30. Despite a significant improvement in tax administration in recent years, tax compliance remained problematic, the costs of administering many small “nuisance” taxes was quite high, and the recent strengthening of the powers of the tax authorities had led to increased perceptions of arbitrariness. Government policies since the mid-1990s focused on improving the efficiency of administering the tax system, with beneficial results for the tax yield. A school for training tax administrators was created, as was a tax police force. Electronic systems were created that could detect stop-payments and, when fully implemented, would allow for cross-checking. Nevertheless, the large degree of self-assessment of tax liabilities by the self-employed, as well as the numerous, small-yielding “nuisance” taxes (including stamp duties and third-party taxes) burdened tax administration. As discussed above, attempts to limit tax avoidance by the self-employed through the use of “objective criteria” (based on physical characteristics of the business or the owner’s living standards) in tax auditing, while largely successful in improving tax compliance and generating revenues, engendered much dissatisfaction.
31. Adopting many of the tax reform measures discussed above should streamline tax administration. In addition, simplification of tax record requirements for small- and medium-sized businesses should improve tax compliance and reduce administrative costs. This should apply in particular to the VAT, where the tax period was raised from two to three months for some 800,000 small businesses. Similarly, implementing electronic methods for tax filings and administration should increase administrative efficiency and eventually, with the advent of cross-checking, further limit tax evasion. Finally, plans to codify the reformed tax system into a single, consolidated text should improve transparency.
C. Areas for Future Reforms
32. While the measures adopted in the recent reforms should improve resource allocation, increase horizontal and vertical equity, and reduce administrative and compliance costs, the reforms could usefully be supplemented by additional steps. The need for future reforms is well recognized by the government, and the reforms could include the following:
further reduce distortions in the effective rate of capital income taxation by limiting the wide range of corporate income tax incentives (including for employment creation and mergers);
create a comprehensive land registry (as is intended by 2005), to allow for the introduction of a modern value-based property tax, and for the extension of the VAT to newly built structures. This would also allow for a further reduction in property transfer taxes, thereby potentially improving labor mobility;
reform the pension system to allow eventually for a reduction in the comparatively high rates of social insurance contributions, which likely inhibit employment creation by far more than the personal income tax burden;
eliminate remaining “nuisance” taxes (stamp duties, third-party taxation, etc.), which often boost production costs and worsen competitiveness, and, more importantly, are costly to administer and in some cases have undesirable equity implications;
reform intergovernmental tax competencies to increase local government flexibility and accountability, including through shifting the bulk of the newly introduced property tax revenues to localities (and allowing for local rate determination).
D. Conclusion
33. The Greek authorities have initiated a substantial tax reform effort, which addresses a number of significant shortcomings in the tax system, and should improve resource allocation and equity in the Greek economy. This includes shedding light on the need to change a system that was overly complex, ever-evolving (and thus uncertain) and contained many horizontally and vertically inequitable features. Reforms focused on broadening personal income tax bases through the transformation of many deductions into limited tax credits should improve vertical equity, while the replacement of “objective criteria” with improved bookkeeping requirements should over time improve tax compliance and horizontal equity. Lower property transfer tax rates, and increased exemptions, should improve capital allocation and labor mobility, but substantial progress in this area awaits the introduction of a modern property tax. Tax administration and compliance costs should be reduced by the simplification of VAT recordkeeping, increasing reliance on electronic means, and the elimination of a number of stamp duties.
34. Nevertheless, the authorities themselves consider the efforts to date as “initial,” and intend to extend reform efforts further. The creation of a single consolidated tax code should increase transparency and reduce complexity and compliance costs. Consideration could also be given to reducing corporate tax incentives, eliminating remaining “nuisance” taxes, reforming intergovernmental taxation, and reforming the pension system to allow for a reduction in social security contribution rates.
APPENDIX Calculation Of The Average Effective Tax Ratios(AETRs)16
35. In order to calculate the AETRs on capital, τk, and on labor, τl, it is necessary to calculate the AETR on total household income, τh. This is then used to allocate total household personal income tax revenues to labor and capital, assuming the same average household tax rate applies to both. The household AETR is calculated as follows;
This states that the average effective household tax rate is the ratio of personal income taxes paid (OECD Revenue Statistics code 1100) to the sum of operating surpluses of private unincorporated enterprises (OSPUE), property income (PEI) and wages (W).
36. The effective tax rate on labor, τl, can then be calculated as follows:
where the taxes include personal income taxes paid on wages (at the average household rate), total social security contributions (2000) and payroll and workforce taxes (3000). The denominator contains employers’ gross labor costs, including wages as well as employer-paid social insurance contributions (2200).
37. The effective tax rate on capital, τk, is calculated as follows:
The first term in the numerator accounts for personal income taxes allocated to capital (note that this assigns all income from unincorporated enterprises to capital). The second, third and fourth terms include direct corporate income taxes (1200), recurrent taxes on immovable property (4100) and taxes on financial and capital transactions (4400), respectively. The denominator is the economy’s total (net of depreciation) operating surplus.
38. Finally, the effective tax rate on consumption, τc, is calculated as follows
where the numerator includes both general taxes on goods and services (5110), as well as specific excises (5121). The denominator includes the both private and government nonwage purchases of goods and services, net of taxes paid on these items, to reflect the tradition of expressing the tax rate as a share of the base price of the items excluding taxes.
Variable Names and Symbols Used
OECD Revenue Statistics
1100 = Taxes on income, profits, and capital gains of individuals
1200 = Taxes on income, profits, and capital gains of corporations
2000 = Total social security contributions
2200 = Employers’ social security contributions
3000 = Payroll and workforce taxes
4100 = Recurrent taxes on immovable property
4400 = Taxes on financial and capital transactions
5110 = General taxes on goods and services
5121 = Excises
National Accounts
CP = Private final consumption expenditures
CG = Government final consumption expenditures
CGW = Compensation of employees paid by producers of government services
OS = Total (net) operating surplus of the economy
OSPUE = (Net) Operating surplus of private unincorporated enterprises
PEI = Households’ property and entrepreneurial income
W = Wages and salaries
References
Carey, David, and Josette Rabesona, 2002, “Average Effective Tax Rates on Capital, Labor and Consumption,” paper to presented at CESifo 2002 Summer Institute (Paris).
Carey, David, 2003, “Tax Ratios on Labor and Capital Income and on Consumption, OECD Economic Studies, forthcoming.
Carey, David, and Harry Tchilinguirian, 2000, “Effective Tax Rates on Capital, Labor, and Consumption,” OECD Economics Department Working Paper No. 258 (Paris. OECD).
Devereux, M. P., and R. Griffith, 1998, “The Taxation of Discrete Investment Choices,” Institute for Fiscal Studies Working Paper 98/16.
Eurostat database, 2003, Population and Social Conditions database (Luxembourg).
International Monetary Fund, 2003, Greece—Staff Report for the 2003 Article IV Consultation.
King, Mervyn, and Don Fullerton, eds., 1984, The Taxation of Income From Capital: A Comparative Study in the United States, the United Kingdom, Sweden and West Germany (Chicago: University of Chicago Press).
Lutz, Mark, 2001, “Greece’s Labor Market: Grappling with High Unemployment,” Greece: Selected Issues and Statistical Appendix, IMF Country Report No. 01/57 (Washington: International Monetary Fund).
Lutz, Mark, 2002, Greece: Selected Issue—An Overview of Pension Reform, IMF Country Report No. 02/58 (Washington: International Monetary Fund).
Mendoza, Enrique, Assaf Razin, and Linda Tesar, 1994, “Effective Tax Rates in Macroeconomics: Cross-Country Estimates of Tax Rates on Factor Incomes and Consumption,” NBER Working Paper No. 4864 (Cambridge, MA: National Bureau of Economic Research).
Organization for Economic Cooperation and Development, 2000, “Tax Ratios: A Critical Survey,” OECD Tax Policy Studies No. 5 (Paris: OECD).
Organization for Economic Cooperation and Development, 2001, Economic Surveys: Greece (Paris: OECD).
Organization for Economic Cooperation and Development, 2002 Analytical Database, December.
Organization for Economic Cooperation and Development, Taxing Wages (various issues).
Oxford Analytica, 2002, “Greece: Tax Reform May Not Go Far Enough For Voters,” October 17, www.oxweb.com.
Republic of Greece, 2002, State Budget 2003, Volume Beta: Tax Expenditures (Athens).
Tax Reform Committee, 2002, Report for the Reform of the Greek Tax System: Proposal of Experts (in Greek; Athens).
Prepared by Mark Lutz.
See Lutz (2002).
These shortcomings, and attempts to overcome them, are dealt with more comprehensively in Carey and Tchilinguirian (2000), OECD (2000), and Carey and Rabesona (2002, 2003).
The AETR tax rates using the Mendoza et al methodology are taken from Carey and Rabesona (2002), while the separate contributions of personal income tax and social security levies to Greece’s labor AETR, and of the VAT and excises to the consumption AETR, were calculated by the author.
And the latter exceeds EU and OECD averages by more than one standard deviation.
Alternative measures of AETRs, incorporating increasingly realistic, albeit somewhat more complicated, allocations of aggregate tax revenue and national income components, have been suggested by Carey and Rabesona (2002, 2003), Carey and Tchilinguirian (2000), and OECD (2000). Carey and Rabesona (2002) report that the correlation coefficients between their baseline and the Mendoza and others’ AETR estimates are for the vast majority of countries, including Greece, 0.8 or greater. Moreover, the Spearman rank correlation coefficient estimates of the tax rates based on Mendoza 1996—2000 averages and Carey and Rabesona’s (2002) baseline rates were 0.94 for both labor and consumption, and 0.89 for capital, This further suggests that Greece’s tax ratio position relative to OECD partners is robust to alternate tax measures.
Reviews of the prereform system are relatively brief, as good discussions are contained in OECD (2001) and Tax Reform Committee (2002).
It was not clear how many third-party taxes existed, as many did not appear in the state budget, but were collected directly by the benefited party. The Tax Reform Committee (2002) estimated that there were roughly 300 of these taxes (and the revenues of those that were included in the budget were equivalent to 0.6 percent of GDP). The entire system of third-party taxation clearly reduced budgetary transparency, often had regressive consequences for income distribution (with some taxes going to well-paid professions), and distorted resource allocation.
The Tax Reform Committee (2002) reported that total deductions amounted to 10.5 percent of total declared income, costing the budget about 0.6 percent of GDP. They also provided evidence of their regressive nature from studies of the incidence of tax deductions and exemptions. It is interesting to note that this existed in the context of Greece having a measure of income inequality that is the second highest in the EU (Eurostat, 2003).
The tax credit for children was transformed into increases in the tax exempt amount in the first bracket of the tax scales. The effective tax relief for one child was increased from €90 to €150, for two children from €210 to €300, and for three children from €615 to €3,510. In total, Oxford Analytica (2002) estimates that up to 3 million of the 4.7 million individuals who paid personal income tax in 2001 would be zero-rated, and one million of them will not have to file a return but simply a declaration stating their zero-rated status.
OECD (2001). Regarding labor market rigidities, see Lutz (2001).
The unsustainability of the Greek pension system is discussed in Lutz (2002) and in the staff report.
Unlike personal income tax expenditures, those for many corporate tax activities are hard to quantify, as they may be included in enterprises’ books (e.g., donations), but are not included in their tax filings. Nevertheless, the government has estimated in the 2003 draft budget that those that they were able to quantify resulted in a revenue loss to the government equivalent to 0.8 percent of GDP in 2001 (Republic of Greece, 2002).
The effective VAT rate, calculated as the revenue divided by consumption net of VAT revenues, has risen steadily in recent years, from 10.8 percent in 1995 to 13.6 percent in 2001, above the EU average, (at 10.5 percent) while the effective rate expressed in percent of the statutory rate has risen from 60 percent to 75.6 percent over the same period, higher than the EU average (54.1 percent, OECD, 2001)
These equations are contained in Mendoza, Razin, and Tesar (1994). The terms used in them are listed in the table on the following page.
