Globalisation, capital mobility and convergence of effective tax rates

This paper explicitly addresses the issue of international capital mobility and convergence of effective tax rates. After reviewing the existing econometric literature on the relation between tax burden and economic integration, the paper updates the approach of effective tax rates (ETRs) introduced by Mendoza et al. (1994) introducing a distinction between ETRs on mobile capital and ETRs on immobile capital. This fills a gap of the empirical literature, usually compounding taxes on corporations and taxes on immovable property under the same heading of ‘capital tax rates’, even though the expected reactions of these two forms of ‘capital’ to economic integration might be significantly different. An econometric analysis relating ‘tax burden’ and some measures of trade openness and capital mobility is performed. Evidence is provided that capital mobility affects the convergence of tax rates on mobile capital, making more difficult for countries to differentiate taxes on mobile tax bases. Far from being support for the race-to-the-bottom hypothesis, the paper argues that there is a significant push towards the homogeneity of tax burdens on mobile capital, which is – to some extent – support for a milder version of the ‘efficiency hypothesis’. The econometric analysis is carried out including the main European countries, Japan, the United States, Australia and Canada. These latter countries are particularly important for the perspective adopted in this paper, as they experienced a liberalisation of capital flows before its introduction in Europe at the beginning of the Nineties. Rather than to a race-to-the-bottom, taxes on mobile tax bases may race-to-some-average, indicating that the main effect of capital mobility could be that of preventing significant differentiation of effective tax rates rather than driving them to zero. Instead of using the levels of effective tax rates, this study uses their coefficient of variation across countries for each given year (CVAR). In alternative, the absolute value of the difference between each country’s effective tax rate and the corresponding average (DIFF) will also be used. Both measures pick the main feature of tax competition, which, if any, is that of making differentiation costly, as large tax differentials may give rise to move capital across borders. The robustness of our regression to alternative methods and, in particular, the strong evidence that the most recent period is particularly valuable to test the effects of tax competition adds to this literature in the expected direction, i.e. that economic integration makes more difficult for countries involved to differentiate the effective tax burden on mobile capital. While the convergence of effective tax rates on mobile capital is partly driven by economic integration, relatively more immobile tax bases should not be affected by openness. In particular, there is no particular reason to expect that taxes on immobile capital should converge across countries as, by definition, immobile capital cannot easily move from one country to another. The same line of reasoning may apply, to some extent, to labour (at least unskilled) and consumption. Conversely, if any, intense tax competition on one tax base might induce more dispersion of other tax bases, if countries act under a tax revenue constraint. This would lead to a either a positive or no relation of the coefficient of variation of immobile capital, labour and consumption with capital mobility. Results suggests that countries that cannot differentiate the tax burden on capital (because of its mobility) may more easily succeed to differentiate the tax burden on labour (which is a relatively immobile factor). Indirect support to this conclusion also comes from the irrelevance of outward FDI in driving the convergence of taxes on both immobile capital and consumption. The more plausible explanation is that both taxes on immobile capital and consumption may constitute a sort of backstop to the convergence of tax rates on capital. In the case of consumption, it must be also considered that effective tax rates on consumption are already more homogenous across countries, compared with capital and labour, and this leaves much less space to converge.