Capital income tax burden: international comparison
- The taxation of capital – both at the individual and corporate level – is the subject of much debate and affects economic growth by reducing incentives to save and invest.
- It is useful to compare the taxation of capital income between countries; this is not trivial. Countries have many different tax rates, exemptions, and special rules.
- We calculate the average tax burden on capital income from aggregate statistics by dividing total capital tax revenue by total capital income. This is an approximation because taxes on household capital income are not observed directly. Our method is well used in the literature, but to our knowledge it has not been used to compare the taxation of capital between countries in recent years.
- Among the 30 OECD countries for which data are available, the average tax burden on all types of capital income is 29%, with a range of 9% in Lithuania to 50% in Canada. In general, English speaking countries tend to have high capital taxes and Eastern European countries tend to have the lowest capital taxes. Data is from 2018 in most cases.
The taxation of capital is a complex and controversial issue. Capital taxes include personal capital income taxes and corporate income taxes, as well as property and transaction taxes.
Capital is very mobile. This is especially true for the corporate income tax base, as many companies today have a global perspective when making investment decisions. High taxes on profits and capital gains also reduce the return to entrepreneurship. High capital and corporate tax rates exacerbate the well-known problems caused by the tax system; for example, the incentive for companies to borrow rather than finance investments with equity, the interest being generally deductible. In addition, the taxation of capital gains on realization leads to blocking effects on the financial and real estate markets.
The classic result of economics – thanks to the results of the models of economists Christophe Chamley (in 1986) and Kenneth Judd (in 1985) – is that capital should not be taxed at all. The taxation of capital distorts the savings decisions of individuals. By reducing the return on savings, taxes on capital penalize those who defer their consumption rather than consuming their income as they earn. Due to compound interest, the taxation of capital penalizes savings more the longer the savings horizon. For long horizons, the distortion is very important. This leads to lower savings, lower capital stock and lower GDP. Therefore, not taxing capital is in everyone’s interest, even those who spend all they earn. The Chamley-Judd result is still debated but remains the benchmark of the academic debate on the taxation of capital.
As the taxation of capital is a much debated and economically important issue, it is useful to compare tax levels between countries. Data on statutory tax rates, as well as effective corporate tax rates calculated from statutory tax rates and various factors determining the corporate tax base, are readily available from international organizations such as the OECD. However, we are not aware of recent comparisons of the average tax burden on all types of capital income calculated from national accounts aggregates.
Results: average tax burden on capital income
We report the average tax burden on capital income for 30 OECD countries in 2018, or 2017 in a few cases. The average among countries is 29 percent, with a range of 9 percent in Lithuania to 50 percent in Canada. Eastern European countries, along with Ireland, have the lowest levels of capital taxation. Many Eastern European countries have a fixed or near-fixed income tax and a low corporate tax. For example, the corporate tax rate in Hungary is 9% and Estonia and Latvia only tax distributed profits. English-speaking countries, on the other hand, tend to tax capital relatively heavily.
To calculate the average tax burden on capital income, we use the formula of the economist Enrique Mendoza and his co-authors from their study published in 1994 which is well used in the literature:
Looking first at the numerator, corporate income tax, property tax, and transaction taxes (like stamp duty) are easily identifiable. The main challenge is to identify the amount of capital income tax paid by households. Most countries apply a global income tax, in which capital and labor income are taxed together. Mendoza et al. (1994) solve this problem by assuming that the tax rate on capital income is the same as the tax rate on all income. They therefore multiply the average personal income tax rate by the capital income of households plus the mixed income of owner-run businesses. This implies that the formula is an approximation, but since we are using the same formula for all countries, the ranking should not be affected much.
Sweden is one of the countries that tax capital and labor income separately. This fact can be used to test the robustness of the formula. When using actual tax revenues rather than assessed on capital income, Sweden’s capital income tax rate drops by 8 percentage points, which means that Sweden goes from 5th to 5th. 8th place in the ranking of countries. Thus, there is some uncertainty in the method, but we still believe that it is the most robust and transparent way to compare the average tax burden on capital income between countries.
The denominator is the economy’s total operating surplus, the national accounts term for profits. At the aggregate level, this corresponds to the total income from capital, including rent charged to owners. The operating surplus can be declared gross or net, depending on whether the depreciation of the capital stock is deducted. We use net operating surplus in our calculations because it is conceptually the best definition of capital income: owners of capital care about how much they can keep after spending to replace depleted capital. The ranking of countries changes little if we use the gross operating surplus instead.
We calculate tax rates for the 30 OECD countries for which data are available. Data is for 2018 for all countries except Australia, Greece, South Korea and the United States, where data is for 2017.
 Christophe Chamley, “Optimal taxation of capital income in general equilibrium with infinite lives”, Econometric 54: 3 (1986), and Kenneth L. Judd, “Redistributive tax in a simple perfect foresight model”, Journal of Public Economics 28 (1985).
 For a recent discussion of Chamley and Judd’s results, see Ludwig Straub and Iván Werning, “Positive Long-Run Capital Taxation: Chamley-Judd Revisited,” American Economic Review 110: 1 (2020), and Varadajan V. Chari, Juan Pablo Nicolini and Pedro Teles, “Optimal Capital Tax revisited”, Journal of Monetary Economics 116 (2020).
 See David Carey and Josette Rabesona, “Tax Ratios on Labor and Capital Income and on Consumption”, in Peter Birch Sørensen (eds.), Measuring the tax burden on capital and labor (Cambridge, Mass .: MIT Press, 2004), and Cara McDaniel, “Average tax rates on consumer, investment, labor and capital in the OECD 1950-2003”, Mimeo, Arizona State University, 2007 for older calculations, and Peter Birch Sørensen, “Measuring Taxes on Capital and Labor: An Overview of Methods and Problems”, in Peter Birch Sørensen (ed.), Measuring the tax burden on capital and labor (Cambridge, Mass .: MIT Press, 2004), for a discussion of conceptual issues.
 Enreique G. Mendoza, Assaf Razin and Linda L. Tesar, “Effective Tax Rates in Macroeconomics: Cross-National Estimates of Tax Rates on Income and Factor Consumption”, Journal of Monetary Economics 34: 3 (1994).