Journal Issue
Share
Chapter

CHAPTER 3 Tax Reform: Principles, Context, and Administration

Author(s):
Benedict Clements, Juan Toro R., and Victoria Perry
Published Date:
October 2010
Share
  • ShareShare
Show Summary Details

Increasing the Tax27 Ratio: Principles and Experience

For countries looking to substantially increase tax revenue, standard principles (equity, efficiency, ease of implementation) apply—but their application faces emerging challenges:

  • Equity. Substantially increased inequality in many countries over recent years28 heightens equity concerns as reflected, for instance, in the increased attention paid to high net wealth individuals (OECD, 2009c). Heavy age-related government spending results in large lifetime transfers towards the baby-boomers, so inter-generational equity suggests they might reasonably bear a substantial part of any increased tax burden (through consumption taxation, for example, which reaches them when they spend accumulated savings);
  • Efficiency. Uncoordinated tax-setting, given the increased international mobility of capital, goods, and people, can lead to collectively inefficient outcomes. This heightens the case for international cooperation and, in its absence, strengthens the efficiency case for taxing relatively immobile bases (notably real estate and natural resources). Changed understanding of efficient policy—notably in relation to the climate, and perhaps taxation of the financial sector—and the prospect of sluggish growth may also impact the preferred tax mix; and
  • Implementation. New approaches are required to collect taxes more effectively, including stronger international collaboration, enhanced legal frameworks, strengthened compliance strategies and collection systems, and intensified use of new technologies (to support real-time information management, increased use of pre-populated returns, electronic tax invoices, and, with potential implications for policy design, more extensive personalized pricing).29

Theory gives little practical guidance on how best to increase the tax ratio—beyond the unspectacular prescription that if policy is initially optimal, all marginal tax rates be increased equi-proportionally. At an optimum, the welfare cost of changing some tax instrument to raise an additional dollar of revenue—its marginal cost of public funds (MCPF)—must be the same for all instruments: otherwise, welfare could be increased without loss of revenue by shifting from the instrument with a higher MCPF to one with a lower one. Starting from such an optimum, the best way to raise additional revenue is by increasing all marginal tax rates in the same proportion.30 More generally (and plausibly), the first place to look for more revenue is the tax instrument with the lowest MCPF.

There is no consensus on the precise MCPFs of alternative tax instruments, but there is increasing evidence on their relative efficiency. Calculating MCPFs requires taking views on both efficiency (estimates vary widely) and equity (values differ), and so cannot yet firmly guide policy. Empirical work has, though, led to some broad consensus that:

  • The corporate income tax can be particularly distortionary. Tax effects on investment,31 and hence long-run growth, can be powerful;32 and
  • Broad-based consumption taxes and property taxes are less harmful to growth than income taxes.33 Taxing consumption is equivalent to taxing accumulated assets and labor income: so it falls partly on a completely inelastic base—previously existing assets—and partly on a base less internationally mobile than capital income.

What contribution could relatively efficient tax policy measures make toward fiscal adjustment in advanced countries with large fiscal gaps? As will be apparent, it is not possible in such an exercise to go beyond an illustrative approach. That said, however, in the United States, the United Kingdom, France, Germany, Japan, and Italy, for example (Table 3.1), reasonably efficient possible measures for excises, real property taxes, and VAT policy improvements, and the introduction of efficient carbon prices in the United States and Europe (with the revenues captured by government), could raise perhaps a weighted average of 2.8 percent of GDP. If Japan were to increase the rate of its already efficient VAT to 10 percent, and the United States to introduce a broad based VAT at the same rate, an additional 2.6 and 4.5 percent of GDP, respectively, could be raised. And these approximations do not include estimates of any increases in overall income tax revenues.34

Table 3.1.Estimated Potential Revenue Increases in Advanced G-20 Countries with Large Adjustment Needs1(In percent of GDP)
CountryReduce VAT policy GAP by halfTobacco and alcohol excises 2Fuel excises 34Property taxes 5TotalVAT at 10 percent rate 6Full auctioning / Taxation of carbon emissions 7Total
France3.80.10.315.1n/a0.25.3
Germany2.40.20.313.8n/a0.64.5
Italy3.10.30.314.6n/a0.55.1
Japan0.30.90.312.42.605.0
United Kingdom3.300.203.5n/a0.55.0
United States00.30.600.94.50.86.1
Unweighted avg.2.10.30.30.73.40.4
Weighted avg. (GDP) 81.10.30.40.42.20.5
Sources: Staff estimates and other estimates as discussed in the Tax Policy Options section below.

Figures do not include any increases from base broadening or rate increases in income taxes.

Based upon raising rates for alcohol and tobacco to the 2006 average level of each tax across the six countries shown, where existing rates are below the mean.

Based on raising gasoline and diesel rates by 10 cents per liter in each case other than the United States.

Raising the U.S. tax to 30 cents per liter would raise an additional 0.6 percentage points of GDP existing in the United States.

Increase revenue from property taxes to yield average ratio to GDP in the United States, Canada, and the United Kingdom.

For Japan, estimate of increased revenue from doubling VAT rate to 10 percent; for the United States, approximation of receipts from introduction of broad based federal VAT at 10 percent.

Estimates for European countries derived by weighting allocation of emission rights based upon per country levels of emissions in 2007; a small proportion of these revenues would represent double counting of the carbon emission externality correcting portion of fuel excises.

Sources: Staff estimates and other estimates as discussed in the Tax Policy Options section below.

Figures do not include any increases from base broadening or rate increases in income taxes.

Based upon raising rates for alcohol and tobacco to the 2006 average level of each tax across the six countries shown, where existing rates are below the mean.

Based on raising gasoline and diesel rates by 10 cents per liter in each case other than the United States.

Raising the U.S. tax to 30 cents per liter would raise an additional 0.6 percentage points of GDP existing in the United States.

Increase revenue from property taxes to yield average ratio to GDP in the United States, Canada, and the United Kingdom.

For Japan, estimate of increased revenue from doubling VAT rate to 10 percent; for the United States, approximation of receipts from introduction of broad based federal VAT at 10 percent.

Estimates for European countries derived by weighting allocation of emission rights based upon per country levels of emissions in 2007; a small proportion of these revenues would represent double counting of the carbon emission externality correcting portion of fuel excises.

Current Tax Structures35

Initial positions—tax levels and the mix of taxes—vary greatly:

  • Tax revenue in percent of GDP (Table 3.2) varies from under 10 percent to over 40 percent. Tax ratios tend to increase with per capita income, but this is far from a complete explanation of the differences: they vary widely even at similar levels of income (Figure 3.1). Nontax revenues, of course, can make an important contribution to overall revenue effort, especially in resource-rich countries making heavy use of royalties or state enterprises: Saudi Arabia presents an extreme case (see Figure 3.2); and
  • Relative reliance on different revenue sources also varies greatly (Table 3.3). The empirical literature finds in particular that reliance on income taxes increases with national income (Martinez-Vasquez, Vulovic, and Liu, 2009), and somewhat tentatively, that reliance on labor taxation is lower the higher is the dependency ratio, suggesting an unwillingness of workers to finance the elderly.36
Table 3.2.OECD and Other G-20 Countries: Tax Revenue Structure1(In percent of GDP)
Last Available YearTotal Revenue and Social Security Contributions 3Tax RevenueTaxes on Income, Profits, and Capital GainsSocial Security TaxesTaxes on Payroll and WorkforceProperty TaxesDomestic Taxes on Goods and ServicesInternational Trade Taxes
Totalof which:Totalof which:
IndividualCorporations and Other EnterprisesUnallocableGeneral sales, turnover or VATExcises 2OtherTotalof which:
Import dutiesExport dutiesOther
Argentina200833.231.15.31.73.30.35.12.33.29.57.81.70.04.40.93.50.0
Australia200737.130.818.411.37.10.00.01.42.77.74.03.00.00.50.50.00.0
Austria200748.042.312.77.25.00.514.22.70.611.98.03.60.30.00.00.00.0
Belgium200748.143.916.56.210.20.013.60.02.311.07.53.50.00.00.00.00.0
Brazil200836.635.86.72.52.81.413.70.00.610.59.00.20.00.60.60.00.0
Canada200740.533.316.612.43.70.54.80.73.37.64.51.70.00.20.20.00.0
China200820.918.05.01.23.70.00.311.88.50.92.5-1.40.6-2.00.0
Czech Republic200742.037.49.49.10.30.016.20.00.410.46.63.80.00.00.00.00.0
Denmark200755.648.729.014.814.00.31.00.21.916.410.46.00.00.00.00.00.0
Finland200747.443.016.95.211.70.011.90.01.112.98.74.20.00.00.00.00.0
France200749.643.710.47.53.00.016.21.23.510.87.43.00.00.00.00.00.0
Germany200743.936.211.39.12.20.013.20.00.910.67.03.20.00.00.00.00.0
Greece200740.032.07.54.72.60.211.70.01.411.38.32.80.20.00.00.00.0
Hungary200744.839.510.07.12.80.012.90.60.814.57.94.22.40.00.00.00.0
Iceland200747.940.918.56.310.02.23.10.02.515.711.64.20.00.40.40.00.0
India 3200722.818.66.32.24.10.02.62.62.22.2
Indonesia 4200820.413.36.60.55.34.21.00.70.50.3
Ireland200735.830.812.18.73.40.04.70.22.510.97.43.50.00.00.00.00.0
Italy200746.943.514.711.13.8-0.313.00.02.111.06.23.00.40.00.00.00.0
Japan200731.028.310.35.54.80.010.30.02.54.92.51.80.00.20.20.00.0
Korea200725.026.58.44.44.00.05.50.13.47.54.23.10.00.80.80.00.0
Luxembourg200741.036.512.87.45.40.010.20.03.69.95.94.00.00.00.00.00.0
Mexico200721.418.05.05.02.80.30.39.23.70.40.00.30.30.00.0
Netherlands200745.837.510.97.73.30.013.60.01.211.07.53.50.00.00.00.00.0
New Zealand200735.722.515.05.12.40.00.00.110.58.42.10.01.01.00.00.0
Norway200756.143.621.04.916.10.09.10.00.611.78.33.30.00.10.10.00.0
Poland200740.034.98.03.15.00.012.00.01.213.28.34.90.00.00.00.00.0
Portugal200743.136.49.45.63.90.011.70.01.413.68.15.40.00.00.00.00.0
Russia200740.035.910.04.06.00.05.16.05.10.80.08.61.57.10.0
Saudi Arabia200867.36.70.50.50.00.80.8
Slovak Republic200734.729.45.80.25.30.311.70.00.411.37.53.80.00.00.00.00.0
South Africa200730.916.68.58.20.00.60.31.710.47.52.50.01.31.30.00.0
Spain200741.037.212.34.67.40.312.10.03.09.56.33.20.00.00.00.00.0
Sweden200753.648.318.714.93.80.012.62.71.212.99.23.60.10.00.00.00.0
Switzerland200736.828.913.22.311.00.06.70.02.46.43.92.50.00.20.20.00.0
Turkey200731.723.75.64.01.60.05.10.00.911.05.15.50.00.30.30.00.0
United Kingdom200737.836.114.310.93.40.06.60.04.510.56.63.40.00.00.00.00.0
United States200729.928.313.910.83.10.06.60.03.14.52.21.10.00.20.20.00.0
Unweighted average533.311.96.95.30.48.80.41.810.26.83.00.20.60.30.20.0
Sources: IMF, Government Finance Statistics; International Financial Statistics; World Economic Outlook; and OECD.

General government.

Including taxes on specific services.

Gross tax revenue of the central government plus state tax revenue.

Central government.

For each revenue item, only countries for which data are available are included in the calculation.

Sources: IMF, Government Finance Statistics; International Financial Statistics; World Economic Outlook; and OECD.

General government.

Including taxes on specific services.

Gross tax revenue of the central government plus state tax revenue.

Central government.

For each revenue item, only countries for which data are available are included in the calculation.

Figure 3.1.OECD and Other G-20 Countries: Tax Revenue and GDP Per Capita

Sources: IMF; and OECD.

Figure 3.2.OECD and Other G-20 Countries: Tax Revenue Structure

(In percent of total tax revenue)

Sources: IMF; and OECD.

Table 3.3.Tax Revenue Structure1(In percent of tax revenues)
Last Available YearTax RevenueTaxes on Income, Profits, and Capital GainsSocial Security TaxesTaxes on Payroll and WorkforceProperty TaxesDomestic Taxes on Goods and ServicesInternational Trade Taxes
Totalof which:Totalof which:
IndividualCorporations and Other EnterprisesUnallocableGeneral sales, turnover or VATExcises 2OtherTotalof which:
Import dutiesExport dutiesOther
Argentina200810017.15.510.61.016.47.310.430.425.05.40.114.12.811.30.1
Australia200710059.836.723.10.00.04.78.924.913.09.60.01.71.70.00.0
Austria200710030.022.55.81.733.76.31.426.018.37.60.00.00.00.00.0
Belgium200710037.529.38.20.131.00.05.123.516.37.20.00.00.00.00.0
Brazil200810018.87.17.93.838.40.01.829.425.00.60.01.61.60.00.0
Canada200710049.837.411.01.414.42.09.922.913.65.20.00.70.70.00.0
China200810027.56.920.60.01.565.647.34.713.6-7.63.3-10.80.0
Czech Republic200710025.111.613.40.043.50.01.227.717.610.10.00.00.00.00.0
Denmark200710059.651.77.40.52.00.53.831.821.410.40.00.00.00.00.0
Finland200710039.330.39.00.027.70.02.629.319.59.80.00.00.00.00.0
France200710023.917.16.80.037.12.88.024.617.07.00.00.00.00.00.0
Germany200710031.225.16.10.036.60.02.529.219.48.70.00.10.10.00.0
Greece200710023.414.78.00.636.40.04.332.522.89.20.60.10.10.00.0
Hungary200710025.218.27.00.032.71.52.036.920.010.96.00.10.10.00.0
Iceland200710045.433.96.15.47.70.16.136.125.910.20.01.01.00.00.0
India 3200710033.711.722.00.014.114.111.911.9
Indonesia 4200810049.73.839.631.87.85.53.52.1
Ireland200710039.328.410.90.015.40.78.234.424.110.30.00.00.00.00.0
Italy200710033.725.68.8-0.730.00.04.925.214.27.00.90.00.00.00.0
Japan200710036.419.616.80.036.40.09.017.38.86.50.00.60.60.00.0
Korea200710031.816.715.10.020.80.212.828.315.811.50.03.03.00.00.0
Luxembourg200710035.020.114.90.027.80.09.826.315.310.90.00.00.00.00.0
Mexico200710027.727.715.31.41.751.420.42.50.01.71.70.00.0
Netherlands200710029.120.48.70.036.20.03.328.519.88.70.00.00.00.00.0
New Zealand200710062.942.114.26.70.00.05.329.423.55.90.02.92.90.00.0
Norway200710048.122.126.00.020.80.02.826.719.17.70.00.20.20.00.0
Poland200710023.115.27.90.034.30.03.436.223.512.70.00.10.10.00.0
Portugal200710025.915.810.10.032.10.03.836.824.112.70.00.00.00.00.0
Russia200710027.911.116.80.014.116.614.22.30.023.94.219.80.0
Saudi Arabia20081007.67.60.012.512.5
Slovak Republic200710019.98.610.21.039.80.01.435.822.912.90.00.00.00.00.0
South Africa200710053.727.326.30.01.91.05.433.624.38.20.14.34.20.00.1
Spain200710033.119.812.40.932.60.08.023.716.27.50.00.00.00.00.0
Sweden200710038.730.97.90.026.15.72.425.719.16.40.20.00.00.00.0
Switzerland200710045.935.310.60.023.30.08.220.413.17.30.00.70.70.00.0
Turkey200710023.717.06.80.021.70.03.846.421.323.00.01.21.20.00.0
United Kingdom200710039.530.19.40.018.40.012.629.118.29.30.00.10.10.00.0
United States200710049.038.111.00.023.40.011.015.87.73.90.00.70.70.00.0
Unweighted average 510035.023.011.81.424.31.05.530.120.08.50.62.11.60.60.0
Sources: IMF, Government Finance Statistics; International Finance Statistics; and World Economic Outlook.

General government.

Including taxes on specific services.

Gross tax revenue of the central government plus state tax revenue.

Central government.

For each revenue item, only countries for which data are available are included in the calculation.

Sources: IMF, Government Finance Statistics; International Finance Statistics; and World Economic Outlook.

General government.

Including taxes on specific services.

Gross tax revenue of the central government plus state tax revenue.

Central government.

For each revenue item, only countries for which data are available are included in the calculation.

These deep differences point to the need for country-specificity in designing revenue adjustment programs. There are nevertheless common themes, from both design and administrative perspectives.

Tax Policy Options

Tax reform must be considered as a package, but in light of common lessons and challenges on key instruments. What matters for the fairness of a tax system, for instance, is not the distributional impact of any tax considered in isolation, but that of all taxes (and indeed spending) combined. While ‘tax-by-tax’ policy design is thus to be avoided, effective reform does require recognizing the limits and potential of each instrument.37

Consumption taxes

Value-added tax (VAT)

The VAT is a mainstay of the tax systems of almost all G-20 and emerging countries. Saudi Arabia and the United States are the only G-20 members without one; India is currently introducing a federal level VAT to be coordinated with its relatively new state-level VATs. Elsewhere in the G-20, the VAT raises, on average, over 5 percent of GDP and about 20 percent of revenue (Table 3.4): it has proved a relatively efficient source of revenue38—one, that is, with a relatively low MCPF.

Table 3.4.Current VAT Rates and Efficiency in G-20 Countries
C-efficiencyVAT revenues as percent of GDPCurrent Standard RateCurrent Other Positive Rates
Canada503.15.0
Japan692.65.0
Australia513.810.0
Indonesia523.710.05; 10; 15.0
Korea614.210.0
South Africa657.414.0
Mexico333.715.010.0
United Kingdom436.517.55.0
China, P.R.: Mainland686.017.013.0
Russia485.618.010.0
Turkey375.518.01.0; 8.0; 26; 40
Germany506.219.07.0
France457.119.62.1; 5.5
Italy396.120.04.0;10.0
Brazil517.320.5Multiple (25 rates)
Argentina466.921.010.5; 27.0
Sources: IMF staff calculations; International Bureau of Fiscal Documentation (IBFD); and PricewaterhouseCoopers.
Sources: IMF staff calculations; International Bureau of Fiscal Documentation (IBFD); and PricewaterhouseCoopers.

However, exemptions and excessive rate differentiation compromise the effectiveness and implementation of the VAT. Exemption—charging no VAT on sales but denying refund of tax paid on inputs—undermines the logic of the VAT by taxing intermediate transactions. Multiple rates are less damaging in policy terms, but the most common rationale—improving equity—is generally unpersuasive for G-20 countries: the rich generally spend absolutely more on items which are taxed at low rates to assist the poor; and most G-20 countries have, or could develop, instruments that are better targeted to equity objectives. In the United Kingdom, for example, eliminating zero-and reduced-rating, while increasing income-related benefits to protect the poor, would raise net revenue of around 0.75 percent of GDP (Crawford, Keen and Smith, 2008).39 Tax administration—and the compliance burden—is also adversely affected by multiple rates and exemptions.

There is substantial scope for improving the revenue performance of the VAT in almost all countries. The effectiveness of a VAT is conveniently assessed by its ‘C-efficiency,’ defined as VAT revenue divided by the product of the standard rate and aggregate private consumption: for a VAT with no exemptions, a single rate, and full compliance, C-efficiency would be 100 percent.40 In practice, many VATs are far from this: many countries could raise significant revenue by modestly increasing C-efficiency, with no need to increase the standard rate: Italy, for example, would gain around 2.5 percent of GDP by raising C-efficiency to the G-20 average (Appendix 9).

Broadly speaking, the scope for administrative improvement is especially large in emerging countries, and that for policy improvement, especially large in advanced countries. While informative, C-efficiency measures in themselves give little clue as to precisely where improvements in the VAT might be found. It can, however, be decomposed into components relating to the VAT “compliance gap” and the “policy gap.”41Table 3.5 illustrates this for selected countries. What is striking is that (though there are, of course, marked exceptions) while C-efficiencies are much the same for both groups, this reflects the offsetting effects of a significantly higher compliance rate in advanced countries combined with policy design that is, if anything, poorer. For example, the proportional revenue gain from moving to the high level of compliance in France is nearly three times as large for emerging economies as for advanced economies; while that from moving closer to Latvia’s efficient policy design is slightly larger in advanced countries than in emerging countries.

Table 3.5.Additional VAT Revenue from Policy and Administrative Improvements, 2006 Figures
VAT Revenue in percent of:VAT RateC-efficiencyVAT Compliance GapVAT Policy GapPotential Extra Revenue (in percent of GDP) from: 1
Tax RevenuesGDP
Improved policyImproved compliance 2
Max. improvement Reducing gap by halfMax. compliance Reducing gap to 15%
Emerging Economies
Argentina29.96.921.04621414.92.31.90.5
Mexico20.43.715.03318605.62.80.80.1
Hungary30.57.420.04923374.32.22.20.8
Latvia39.18.321.04922385.12.52.30.7
Lithuania36.17.518.05022364.32.12.10.7
Brazil30.77.317.552n/a3.81.92.00.6
Indonesia30.13.710.052n/a1.91.01.00.3
India17.13.012.535n/a3.61.80.80.2
China36.76.017.068n/a1.00.51.60.5
S. Africa28.27.414.065n/a1.60.82.00.6
Bulgaria39.511.820.068n/a1.91.03.20.9
Romania28.68.119.050n/a4.82.42.20.6
Russia15.05.618.048n/a3.71.81.50.4
Turkey29.35.518.037n/a6.33.21.50.4
Average29.17.118.65021433.81.91.80.5
Advanced EconomiesMax. Improvement Reducing gap by halfMax. compliance Reducing gap to 7%
France42.27.119.6457527.53.80.50.0
Germany27.16.216.05010444.92.40.70.2
Italy21.06.120.03922506.23.11.71.2
UK21.76.517.54313506.53.31.00.5
Australia12.93.810.051n/a2.61.30.60.1
Japan14.22.65.069n/a0.70.30.40.1
Korea20.94.210.061n/a1.80.90.60.1
Canada9.23.15.050n/a1.40.70.30.1
Average21.14.912.95113493.92.00.70.3
Sources: WEO; GFS; Reckon LLP (2009); and IMF staff estimates.

For countries where no VAT gap estimate is available, the average (21 percent for emerging and 13 percent advanced economies) of those available has been used.

Improving VAT compliance is likely to have an indirect positive effect on income tax compliance which is not reflected in these figures.

Sources: WEO; GFS; Reckon LLP (2009); and IMF staff estimates.

For countries where no VAT gap estimate is available, the average (21 percent for emerging and 13 percent advanced economies) of those available has been used.

Improving VAT compliance is likely to have an indirect positive effect on income tax compliance which is not reflected in these figures.

Guiding principles for VAT reform include:

  • Reducing exemptions and eliminating reduced rates is generally the best way to increase VAT revenue, unless low efficiency is caused by weak administration. Much could be done without increasing the standard rate in many countries. In Mexico, for instance, the reduced border rate of 10 percent serves little useful purpose; and the reduced rate in Germany costs 0.8 percent of GDP; on average, even reducing this exemption/rate “policy gap” by half could raise nearly 2 percent of GDP for both emerging and advanced economies.
  • There can be substantial revenue gain from cutting large VAT compliance gaps. Latvia, for instance, could raise 1.6 percent of GDP by reducing its VAT compliance gap to that of France; reducing the compliance gap to 15 percent in emerging, and 7 percent in advanced, economies could raise an estimated 0.5 percent and 0.3 percent of GDP, respectively.
  • Where neither structure nor administration is problematic, rates could be raised with minimal distortion. In Japan, for example, C-efficiency is high but the (single) rate is low: substantially increasing the rate in such cases is a reasonably sure way to raise more revenue at minimal welfare cost.

For countries without a VAT, introduction is the leading option for substantially enhancing revenues. In the United States, for example, a VAT at 13 percent might raise 6 percent of GDP (Graetz, 2005; other recent estimates give comparable revenue per percentage point of the VAT rate for a broad-based VAT with few exemptions). Late adopters would benefit from avoiding the errors of ‘old’ VATs, such as the overly-broad exemptions to which the EU is locked in (Cnossen, 2003).

Excises

Many countries have scope to increase significantly revenues from tobacco and alcohol excises. Receipts are noticeably lower in the emerging G-20 (Table 3.6), where the arguments for cigarette taxation, in particular, may be especially strong. In the advanced economies, their yield (especially for alcohol) is in trend decline (falling by about 0.5 percentage point of GDP in the United Kingdom since 1995, for instance) reflecting not just changing consumption patterns but also falling real tax rates. Policymakers have moderated rate increases for fear of excessive cross-border shopping and smuggling:42 enhanced cross-border cooperation, in both design and implementation, may be required to realize the potential gains. The minimum excise rates within the EU illustrate the possibilities, but also the difficulty: politics has meant that many rates are low (zero, for some alcoholic drinks).

Table 3.6.Excise Revenue from Tobacco and Alcohol Consumption in Selected G-20 Countries(In percent of GDP)
TobaccoAlcoholTotal1
19952200719953200719952007
Australia0.260.520.140.170.400.69
Brazil0.350.110.200.100.550.21
Canada0.470.460.130.090.600.55
China 4n/a0.02n/an/an/an/a
France0.530.520.230.050.760.57
Germany0.570.580.210.140.780.72
India0.310.23n/an/an/an/a
Italy0.530.660.060.070.590.73
Japan0.420.440.420.290.840.73
Korea R.0.520.280.540.291.060.58
Mexico0.140.180.120.190.260.37
Russia0.100.150.400.200.500.35
UK1.000.580.750.581.751.15
USA0.180.170.150.100.330.27
Sources: IMF; OECD; and national authorities.

Data for Turkey are for combined tobacco and alcohol only, and for 2006 to 2009.

1999 for Australia, 1997 for India, 2000 for Mexico, and 2001 for Russia.

1999 for Australia, 2000 for Mexico, and 1998 for Russia.

Does not include profits from tobacco monopoly.

Sources: IMF; OECD; and national authorities.

Data for Turkey are for combined tobacco and alcohol only, and for 2006 to 2009.

1999 for Australia, 1997 for India, 2000 for Mexico, and 2001 for Russia.

1999 for Australia, 2000 for Mexico, and 1998 for Russia.

Does not include profits from tobacco monopoly.

The low level of fuel taxation in many advanced countries means that the potential revenue gains from more efficient tax levels are substantial. Among G-20 countries, fuel tax revenues in Japan, Mexico, and the United States are especially low (Figure 3.3a, 3.3b). Coady, and others (2010) project that the forgone revenues in G-20 countries from taxing below $0.30 cents per liter (the lower end of their benchmarks for efficient fuel tax levels) could reach $490 billion by the end of 2010. Possible efficiency gains from the taxation of diesel may be especially marked, given the preferential tax treatment it receives in many G-20 countries (Figure 3.3b). In addition, the fact that fuel taxes are often used as a second-best alternative to more efficient tax instruments (e.g., congestion charges) suggests that the net revenue effect of replacing these components of the fuel tax with their more efficient alternative may be positive.43

Figure 3.3a.Motor Fuel Tax Revenues

(In percent of GDP)

Sources: OECD Revenue Statistics 2009; OECD database of environmental taxes; http://www.taxpolicycenter.org/taxfacts/listdocs.cfm?topic2id=80 Figures for 2007, except Australia (2008); and France, Mexico, Turkey (2006).

Figure 3.3b.Motor Fuel Taxes for Selected G-20 Countries

Sources: IEA Prices and Energy Taxes 2009 Third Quarter; Energy Information Administration; Parry and Strand (2010). National taxes and average local taxes combined where appropriate. Figures for 2009 except India (2008); and standard unleaded gasoline, except France, South Africa, Turkey, and the United Kingdom (premium).

Car taxes in some cases have unexploited potential. They vary greatly—one-off registration fees, annual ownership fees, taxes on new sales—and some of the concerns to which they are tailored (road use and emissions) are better targeted by other instruments. Nevertheless, this is another convenient tax handle that some could exploit further: Mexico, France, and the United States, for example, raise less than half of the 0.4 percent or so of GDP collected in Germany, Italy, and the United Kingdom. Further, they can be an instrument of progressivity, especially for developing/emerging countries, if rates are varied according to size or type of vehicle.

Scope for new types of excises is limited. The empirical evidence required to warrant rate differentiation across countries is rarely firm enough to outweigh implementation costs, and taxes addressed to environmental harm (beyond fuel excises/carbon taxation) have little revenue potential—that not being their main purpose. Taxing telecom services is sometimes suggested, partly to tap rents that cannot be reached directly. But the drawbacks are substantial: network externalities are important in early stages of the product cycle; distinguishing personal from business use is hard; and auctions can be a more effective way of extracting rents.

Income tax

Corporate income tax (CIT)

The increased international tax competition over the past two decades is likely to continue. There is substantial evidence that the significant decrease in statutory rates of CIT since the mid-1980s (Figure 3.4a)—by an average of about 15 percentage points across the OECD—reflects strategic competition in tax-setting, not simply some common trend (Devereux, Lockwood, and Redoana, 2008). One instance of this is that the highest corporate tax rates in the G-20 (and hence perhaps the greatest pressures for reduction) are found in large economies: notably the United States.44 Movements towards territorial rather than worldwide taxation in the United Kingdom—that is, taxing corporations only on their income derived within the country, rather than on all of their income no matter where derived if they are headquartered or otherwise deemed to be domestic companies—as is often also discussed for the United States,45 are a further symptom of this competitive trend, and would also be a possible source of its intensification.46

Figure 3.4a.Corporate Income Tax Revenue and CIT Statutory Rate in OECD Countries

Source: OECD.

CIT revenue had, until the crisis, remained strong47—but this cannot be relied on looking forward (Figure 3.4b). To the extent that it reflected increased incorporation as CIT rates fell relative to personal income tax rates (PIT) (de Mooij and Nicodeme, 2008), resilience could continue if CIT rates keep falling (though with some offsetting reduction in PIT receipts). Some argue that the strength of CIT revenue reflected rates being above revenue-maximizing levels, but this remains contentious (Brill and Hassett, 2007; and Clausing, 2007) and any such effect must ultimately vanish. The strength of CIT revenue also reflected a large contribution from the financial sector48 that has now fallen substantially, and may be permanently reduced by regulatory reform. While there remains scope for base-broadening in many countries, potential revenue gains from this in the G-20 seem fairly modest: there have already been significant base-broadening measures, notably in relation to depreciation (Devereux, Griffiths, and Klemm, 2002) and, in some cases—as with China, and for example, in the EU state aid rules—a scaling-back of incentives.

Figure 3.4b.Corporate Income Tax Rate and CIT Revenue in Selected G-20 Countries, 1995–2008

Sources: Government Finance Statistics; International Financial Statistics; World Economic Outlook; OECD; and IMF staff estimates.

Unprecedented international coordination would be required to limit/reverse pressures on CIT rates and revenues. Those who see the CIT as particularly damaging to growth would of course welcome its demise. It does though serve as a backstop to the PIT and, potentially, a relatively efficient tax on rents (that is, earnings in excess of a “normal” return to capital). However, given the ease with which profits can be shifted to low-tax jurisdictions, it can play this role fully only if policies are coordinated across countries: the MCPF of the CIT may be much lower when policy is coordinated than from a unilateral perspective. Coordination might take a variety of forms—agreement on minimum tax rates, on scaling back incentives, some form of formulary apportionment, or more limited agreements (to deal with hybrid entities, for instance; see Thuronyi, 2010). The recent progress on combating the use of tax havens, discussed below, is limited to information exchange, and does not address these more sensitive topics of tax rates and design.

Prospects are brighter in resource-rich economies. Though not immune to pressures of international tax competition, the element of location-specific rent in resource returns provides a potentially robust source of relatively non-distorting revenues.49 This is indeed an important source of revenue in many advanced and emerging countries (Figure 3.5): eleven of the G-20 are major oil and gas producers;50 others are major ore and metal exporters. Most are sufficiently able to diversify the risks of natural resource exploitation to make profit/cash-flow based instruments more efficient than fixed fees and royalties, yet some—including the United States and Russia—still place heavy reliance on the latter.51 Movement towards explicit rent taxation, including through auction, could produce a marked revenue enhancement.52 This is not to argue that average effective tax rates are necessarily low (in any case, these will vary with price and project), but that tax structures could be modified both to promote investment and to secure for governments higher shares of resource rent in profitable projects.

Figure 3.5.Minerals Contribution to Total Government Revenues

(In percent of GDP1)

Sources: National authorities; World Economic Outlook; and IMF staff estimates.

Notes: *U.S. mineral revenue data excludes corporate incomes taxes; data for Brazil, Indonesia, and Saudi Arabia reflects 2008 levels; all the rest reflects 2007.

12008 GDP or most recent available year.

Personal income tax (PIT)

The personal income tax is generally considered key to the pursuit of equity in the tax system, though the effectiveness of this is tempered by the incentive effects (on both real activity and compliance) of increasing effective marginal rates of PIT. Incentive effects on the labor supply of primary workers are generally modest (Blundell and Macurdy, 1999) including for high earners: the substantial reduction in top marginal tax rates in Russia on movement to a flat tax of 13 percent, for instance, has been found to have small effects (Ivanova, Keen, and Klemm, 2005; Gorodnichenko, Martinez-Vasquez, and Peter, 2009). Tax effects on the participation decisions of secondary workers can be substantial, however, even at currently historically low levels of progressivity and top marginal tax rates (Table 3.7). Account needs to be taken also of high effective marginal rates implied by the withdrawal of benefits, including earned income tax credits; and better targeting of these, as discussed in Chapter 2, will amplify these effects. There is significant evidence that higher rates of PIT risk encourage tax avoidance (through the use of deductions, for example) and evasion, particularly for higher net income individuals (Saez, Slemrod, and Giertz, 2009). They are increasingly important as a source of revenue, reflecting increased inequality in recent years, making it natural to look to them for an increased contribution; but they also have a greater facility for avoidance (Box 3.1) making this difficult to do. Increased rates of social contributions, discussed earlier, can cause compliance difficulties at the lower end of the income distribution, but this can be addressed, in part, by integrating tax and social contribution administrations.

Table 3.7.Total Tax Revenue and PIT Revenue as Percent of GDP, and PIT Top Marginal Rates for G-20 Countries1
CountryLast Available YearTotal Tax RevenuePIT RevenueRatio PIT to Total RevenuePIT Top Marginal Rate
Argentina200831.11.75.535
Australia200730.811.336.745
Brazil200835.83.228.928
Canada200733.312.437.439 / 48.33
China, P.R.: Mainland200818.01.26.945
France200743.77.517.140
Germany200736.29.125.145
India200718.62.211.730
Indonesia200813.3435
Italy200743.511.125.643
Japan 5200728.35.519.640
Korea, Republic of200726.54.416.7356
Mexico20071828
Russia200735.94.011.113
Saudi Arabia 720086.7
South Africa200730.98.527.340
Turkey200723.74.017.035
United Kingdom200736.110.930.140
United States200728.310.838.1357
Sources: www.bus.umich.edu/OTPR/otpr/OTPRdataV3.asp (The World Tax Database of the University of Michigan); KPMG (2008) database; PriceWaterhouseCoopers (2008); and IBFD (2008).

General government.

Includes withholding tax on wages and half the revenue of tax withheld on capital.

Sum of the federal and provincial top marginal rates. Lowest rate corresponds to Alberta (flat 10 percent rate) and highest to Nova Scotia.

Central government only.

The rate is for PIT but revenue includes also the inhabitant tax composed of a 10 percent tax on income earned in the previous year and a poll tax.

The rate will be reduced to 33 percent from 2012.

Rate is for the federal PIT but revenue includes that of state and local PITs.

Sources: www.bus.umich.edu/OTPR/otpr/OTPRdataV3.asp (The World Tax Database of the University of Michigan); KPMG (2008) database; PriceWaterhouseCoopers (2008); and IBFD (2008).

General government.

Includes withholding tax on wages and half the revenue of tax withheld on capital.

Sum of the federal and provincial top marginal rates. Lowest rate corresponds to Alberta (flat 10 percent rate) and highest to Nova Scotia.

Central government only.

The rate is for PIT but revenue includes also the inhabitant tax composed of a 10 percent tax on income earned in the previous year and a poll tax.

The rate will be reduced to 33 percent from 2012.

Rate is for the federal PIT but revenue includes that of state and local PITs.

There is a significant scope in some countries, however, for base-broadening and simplification within the PIT, which could raise substantial revenue. For example, Japan and Korea have relatively high top marginal PIT rates (respectively, 40 and 35 percent), but have relatively low PIT ratios (5.5 and 4.4 percent of GDP) compared to other advanced G-20 countries. Such reforms would likely improve equity, given the nature of many of the base narrowing provisions presently existing. And in some countries that are heavily reliant on the PIT and in need of large fiscal adjustment, there may be little choice but to raise intermediate marginal rates in the PIT schedule.

Other

Carbon pricing

Pricing greenhouse gas emissions—by taxing carbon or auctioning emissions permits—could raise large sums. Globally efficient pricing could raise US$50–660 billion annually,53 increasing over the next decades as the efficient price rises faster than emissions falls. The completeness of coverage (by country and emission source) this presumes is unachievable in the near term, but realistic short-term sums are still substantial. Current legislative proposals for emissions trading in the United States have revenue potential of about US$870 billion over 2011–19: roughly US$100 billion annually, or ½ percent of GDP—15 percent of the cumulative forecast fiscal deficit for that period (Congressional Budget Office, 2009a, b). Revenues from such schemes might appropriately be reduced by compensating poor consumers and some offsetting of fuel and other taxes; and their cross-country allocation will depend on arrangements for trading emissions rights (IMF, 2008). Nevertheless, carbon pricing provides a clear opportunity for substantially increasing revenue while enhancing efficiency and sustainable growth.

Box 3.1.Taxing High Net Income Individuals (HNIIs)

This is an area of growing importance and difficulty. Those with the highest incomes pay a substantial share of all PIT: the top 0.1 percent of taxpayers in Germany, for instance, pay 8 percent of PIT; and in the United States, the top 0.7 percent pays 37 percent. HNIIs account for an average of 20–25 percent of total PIT revenue among G-20 countries. At the same time, however, they pose significant risk of non-compliance: HNIIs draw a significant fraction of their income from sources offering great opportunities for avoidance and evasion, including non-cash compensation (bonuses, stock options, and fringe benefits), entrepreneurial income, and investment: in the United States, capital gains alone have accounted for about one third of total income for taxpayers at the top of the income spectrum. HNIIs often have access to off-shore investment vehicles, which can facilitate non-compliance. An estimated 7–16 percent of assets of those with high wealth are held offshore, 1 though this varies greatly by region (with roughly 30 percent of Latin American and Middle Eastern assets held offshore, but less than 5 percent in North America and Japan). Little is known about the revenue cost of the evasion and avoidance associated with HNIIs, but many tax administrations believe it to be substantial (and see signs of this in the encouraging results of voluntary disclosure initiatives discussed below). They are conscious too of the danger that perceptions of the richest not paying their “fair share” will erode compliance more widely.

Combating avoidance and evasion among HNIIs requires not only increased enforcement, but also anti-abuse legislation and addressing fundamental tax distortions. For example, a common way to shelter income is by using tax (but not economic) loss-generating schemes to offset other income: countries have responded to this by, for instance, disallowing use of passive losses to offset income and ignoring transactions without economic substance, but scope for game-playing remains. Tax planning by transforming one type of income into another—often recharacterizing ordinary income as (preferentially-treated) capital gains—is invited by applying sharply different tax rates to different types of receipts. A lower tax rate on all forms of capital income—as under a dual income tax—would mean both fewer resources wasted on tax planning and reduced incentives for cross-border evasion.

1OECD (2009c) defining “high net wealth individuals,” (HNWIs) as individuals with at least US$1million in net investable, non-residential assets.

Realizing these gains requires limiting the free allocation of permits and extending the scope of carbon pricing. Around two-thirds of the potential revenue (from 2013–20) from schemes proposed in the EU, the United States, and Australia, is forgone under current plans to award almost all permits free of charge, conferring large windfall profits.54 Swift transition to full auctioning55 could raise hundreds of billions of dollars. Revenue (and efficiency) would also be enhanced by broadening the base of carbon pricing. There is little economic rationale, in particular, for the current exclusion of international transportation fuels not merely from carbon pricing but from any fuel excise:56 taxing them could generate US$150–200 billion over the coming decade in the G-20, though substantial international coordination would again be required.

Property taxes

Property taxes57 are a promising source of increased revenue for some countries, but there are practical obstacles. They currently yield around 3 percent of GDP in Canada, the United Kingdom, and the United States, but well below 1 percent in other G-20 countries. Efficiency and fairness argue strongly for firm use of property taxes: they are relatively benign for growth; raise few issues of international coordination; and, while their incidence is still not fully understood (Sennoga, Sjoquist, and Wallace, 2008), they seem to be borne mainly by the well-off. Obstacles to their wider use include administrative complexities and costs (including the development of efficient cadastre and valuation mechanisms), and the unpopularity that their transparency can bring. The (appropriate) assignment of property taxes predominantly to lower levels of government may pose challenges for increased revenue raising. This, though, is another area with clear potential for significant and relatively efficient medium-term revenue enhancement in several countries.

Improving Tax Compliance58

Significant tax gaps are widespread in the G-20. VAT compliance gaps (the difference between actual and potential VAT revenues) are 20 percent in some (Mexico, Italy, and some other EU countries), but nearer to 10 percent in others (France and Germany). Compliance is generally very high for income taxes withheld or subject to third-party reporting, but for other sources of income is commonly very low: for small traders, for instance, the gap is over 50 percent in the United States. Improving revenue administration and combating tax abuse could yield considerable revenue—the discussion on value added taxes above estimates that extra revenue equivalent to 0.8 percent of GDP could be collected by reducing the VAT gap in G-20 countries in the coming years.

Pervasive tax abuse significantly erodes revenue through:

  • Informality—estimates of the size of the informal economy in high-income countries range from 8–30 percent of GDP (Schneider, 2009);
  • Aggressive tax planning—contrived schemes pushing the boundaries of legal interpretation;
  • Offshore tax abuse—evasion and avoidance through tax havens and bank secrecy jurisdictions;59
  • Tax fraud—mostly through false tax refund and credit claims, including by organized crime: EU VAT fraud losses, for instance, were estimated to be $80–$140 billion in 2006 (International VAT Association, 2007); and
  • Unpaid tax debts—weak payment compliance and enforcement, resulting in large stocks of tax debt.

The crisis has aggravated compliance problems;60 restoring tax discipline is an immediate priority. For example, in the United Kingdom, the VAT gap increased by 3 percentage points between 2007–08 and 2008–09, and Lithuania’s VAT debt more than doubled in the first half of 2009. Taxpayers who have drifted towards informality need to be brought back into the system, and there may be a resurgence of contrived tax schemes as the appetite for risk increases and corporations seek to restore their financial positions. Revenue agencies must be alert to new schemes, such as abuse of the very substantial tax losses emerging from the crisis—$1.1 trillion of bank losses and write-downs have been reported (OECD, 2009c).

Pressure to reduce tax gaps presents an opportunity to improve revenue administration and tax compliance through medium-term systemic solutions. Improving the medium-term fiscal position requires reshaping revenue administration. There are four priorities: intensifying international collaboration, especially in exchanging tax information; developing sound risk-based compliance strategies; strengthening legal frameworks, including the powers of revenue agencies (e.g., in accessing information and conducting audits); and exploiting new information technology to better align tax compliance management with businesses’ lifecycles.

Recent advances in international collaboration in tax information exchange and transparency are an important step forward—but implementation is critical and further opportunities remain for stronger cooperation. Recognizing the need for more global responses, the G-20 has enhanced its support of OECD efforts to establish international standards of tax information exchange and transparency. This has resulted in a large increase in the number of bilateral tax information exchange agreements with bank secrecy and tax haven jurisdictions. Continued international resolve and cooperation will be necessary to ensure that commitments under the agreements are met; technical assistance may need to be provided to tax havens to improve their administrative capacity and legal frameworks to facilitate timely information exchange. The need for stronger cross-country collaboration is evident in other areas too, for example, to be more effective in responding to criminal fraud. The EU has recognized that lack of collaboration between Member States has contributed to the vulnerabilities exploited by a raft of multi-billion dollar intra-community VAT frauds and, more recently, frauds associated with the trading of carbon credit permits. Improved systems of information exchange between EU revenue agencies would enhance early warning of emerging revenue risks; joint investigations should be expanded. Cross-country alignment of more effective domestic responses to cross-border evasion (such as through voluntary disclosure compliance programs—discussed below) should be also pursued.

Fundamental strengthening of compliance improvement strategies is crucial. Driven by risk management approaches, this entails:

  • Efficient gathering and administration of taxpayer and third-party information utilizing modern technology and streamlined processes to reduce compliance costs and facilitate modeling of revenue risks during all stages of the taxpayer’s business lifecycle;
  • Robust revenue analysis and identification of emerging compliance risks; and61
  • Development of appropriate responses to mitigate identified risks—mitigation strategies will vary depending on the underlying reasons for non-compliance. For example, audits and penalties are a fitting response to deliberate evasion, while education and assistance are appropriate to situations where taxpayers do not understand the law. Importantly, mitigation strategies should seek to achieve wide impact and enduring compliance within the broader taxpaying community.

In emerging economies—where, as discussed earlier, revenue possibilities from sustainable compliance improvement are greater—tackling endemic tax abuses to enhance the taxpaying culture requires significant capacity building in core systems of revenue administration (including in compliance-related areas of risk management, audit, collection enforcement, taxpayer services, and dispute resolution).62 Through comprehensive reform efforts, revenue agencies in emerging economies can play an important role in fostering formalization, by helping new entrepreneurs and taking visible enforcement action against the shadow economy to establish tax discipline. In advanced economies, where systems of administration are more robust, the central compliance challenge is more about combating aggressive tax planning, offshore evasion, and tax fraud. These compliance risks require domestic and global responses and often novel approaches, like the recent voluntary disclosure programs aimed at bringing taxpayers involved in offshore tax abuse into compliance. These programs are an integral part of wider strategies to achieve enduring tax compliance; their success rests on large scale financial information gathering by revenue agencies, enhanced detection capabilities, and a commitment by the authorities to follow through with strong enforcement action, including prosecution, against those who choose to continue cheating the tax system.63 Unlike voluntary disclosure programs, traditional tax amnesties in some countries have focused on short-term repatriation of revenues, without enhancing the compliance management capabilities of their revenue agencies and promoting sustainable compliance improvement. Tax amnesties have sometimes been implemented through anonymous one-off payments (normally a fixed fee) via the banking system, without provision of information to tax administrations, and no-questions-asked policies that preclude future audits of tax years covered under these amnesties.

Legal frameworks need to be enhanced to address compliance risks and pervasive tax abuse:

  • Countries can do more within existing tax and financial regulatory structures—making the most of existing legal powers, data and intelligence in relation to financial flows requires the highest levels of cooperation and information exchange between revenue agencies and corporate regulators, banking supervisors, anti-money laundering regulators, financial intelligence units, border management and other law enforcement agencies.
  • Further legislative solutions need to be adopted to combat offshore tax abuse—it may be appropriate to impose stronger domestic sanctions and other disincentives (e.g., stiff fines and criminal prosecution of evasion and its facilitation; and policy measures to discourage transactions with uncooperative low tax jurisdictions, such as introducing withholding taxes on funds sent offshore and denying certain expense deductions).
  • Aggressive tax planning needs to be tackled with firm countermeasures, including development of common good practice in anti-avoidance rules—an effective set of general anti-avoidance rules should be available to revenue agencies as part of the tax litigation armory. Development of a model set of principles—based on best practice64—to guide all G-20 countries in the drafting of effective general anti-avoidance rules would be a major step forward. Procedural rules should also be developed to assure that taxpayers cannot avoid scrutiny of their questionable transactions by playing the audit lottery. These might include mandatory disclosure of specifically identified transactions or, more generally, uncertain tax positions that put significant amounts of revenue at risk. Domestic and international codes of tax practice for banks, other large corporations and tax intermediaries—with appropriate incentives to comply—should also be pursued. Strong penalties for promoters, facilitators, and users of contrived and opaque tax schemes should be adopted.
  • Tax litigation needs to be streamlined—good practices in negotiated settlements can also minimize instances of costly and lengthy litigation.

Intensifying the use of modern information technology in delivering revenue administration will significantly improve compliance management and reduce compliance costs. Besides basic internet-based services (e.g., tax information and return filing) widely adopted in several countries, revenue agencies should intensify the adoption of electronic solutions to automate and align economic agents’ tax compliance and business cycles. Good examples of this direction include on-line taxpayers’ registration and termination of business, automatic gathering of third-party information, business-to-government standard financial reporting as a by-product of natural business processes, and use of electronic invoices with the potential of real-time transaction monitoring and verification of VAT compliance. There are also other successful technology-based innovations that could be adopted more widely including, for example, automated risk-based selection systems, on-line auction of seized assets, pre-populated tax returns, on-line compliance reporting services, and accounting systems for promoting formalization of small taxpayers. The opportunities presented by these and other technological solutions, in a context of the key directions discussed earlier to enhance international transparency and strengthen compliance management, pave the way to reshaping revenue administration to meet the compliance challenges of the digital age and globalized economy.

    Other Resources Citing This Publication