This paper examines the link between the composition of tax revenues and the level of income per capita in the long run. I find that the “tax and growth ranking” suggested by some recent empirical studies is not robust under different assumptions about heterogeneity across countries of the long-run and short-run coefficients in the underlying econometric model. Evidence for significant tax structure effects depends on long-run parameter homogeneity restrictions, underlying pooled mean group estimation, which are found to be invalid.
