There has been surprisingly little empirical work explaining why countries choose different bundles of taxes. Early research by Musgrave and Hinrichs focused on the amorphous distinction between direct and indirect taxes. More recent research has examined the use of trade taxes and inflation, with little appreciation for the fact that countries are choosing a complete revenue mix. We examine the determinants of tax composition in 100 countries using data averaged over the periods 1975-80, 1981-85, and 1986-92. The dependent variables are tax revenues from the different sources as fractions of total revenue (or of GDP), including personal income taxes, corporate taxes, social security & payroll taxes, domestic goods & services taxes, trade taxes and tax rates, property taxes, and nontax revenues from seigniorage and public enterprises. All tax sources have increasing marginal political costs attached to their use. Consequently, as the scale or size of total revenues increases, reliance on all revenue sources can be expected to increase, though some tax bases may be used more heavily than others. We find strong evidence supporting this scale effect. Our empirical analysis also provides strong support for the prediction that countries tend to rely more heavily on tax sources for which the base is relatively large, and there is evidence that the use of taxes that depend on widespread literacy increases as educational attainment rises. While an explanation of the tax mix that relies on economic variables measuring scale, tax bases, and administration and enforcement costs works reasonably well even in a diverse sample of countries, we find that tax composition also varies with the nature of the political regime. Socialist countries tend to make more use of corporate, sales, and excise tax sources than other regimes, perhaps due to the greater ease with which the activity of businesses can be monitored, a stronger ideological interest in taxing business, or a reduced need to use taxation of individuals to accomplish social goals. We also find that more repressive governments rely less on personal income taxation, possibly because this tax source requires a higher degree of voluntary compliance than other forms of taxation. At a general level, the results as a whole are of interest for at least two reasons. First, they add to the set of stylized facts that may serve as a basis for further theoretical work that can unify the experience of different countries around the world. Secondly, the ability to model the tax mix of a diverse sample of countries raises interesting questions about the possibilities for tax reform. To the extent that international differences in the mix of taxes are predictable, proposals for reform that do not take the underlying forces into account are likely to be unsuccessful, or may be very costly if they do succeed in altering the existing equilibrium.

Additional Metadata
Keywords political economy of taxation, tax mix, tax bases, administration costs, political regime
JEL Economic Models of Political Processes: Rent-Seeking, Elections, Legislatures, and Voting Behavior (jel D72), Structure, Scope, and Performance of Government (jel H11), Taxation, Subsidies, and Revenue: General (jel H20), Personal Income and Other Nonbusiness Taxes and Subsidies (jel H24), Business Taxes and Subsidies (jel H25), Other Sources of Revenue (jel H27)
Publisher Department of Economics
Series Carleton Economic Papers (CEP)
Kenny, Lawrence W., & Winer, S. (2001). Tax Systems in the World: An Empirical Investigation into the Importance of Tax Bases, Collection Costs, and Political Regime (No. CEP 01-03). Carleton Economic Papers (CEP). Department of Economics.