The literature testing for aggregate impacts of taxes on long-run growth rates in the OECD has generally used tax rate measures constructed from macroeconomic aggregates such as tax revenues. These have a number of advantages but two major disadvantages: they are typically average, rather than marginal, rates, and are constructed from endogenous tax revenues. Theory predicts a number of responses to both average and marginal tax rates, but empirical analogues of the latter tend to be at the micro level. In addition though most OECD economies are best regarded as small open economies, previous macroeconomic tests of OECD tax-growth relationships have implicitly been based on closed-economy models, focusing on domestic tax rates. This paper explores the relevance of these two aspects – "macro average‟ versus "micro marginal‟ tax rates, and open economy dimensions – for test of tax-growth effects in OECD countries. We use annual panel data on a number of average and marginal tax rate measures and find: (i) statistically small and/or non-robust effects of macro-based average tax rates on capital income and consumption but more evidence for average labor income tax effects; (ii) statistically robust GDP growth effects of modest size from changes in marginal income tax rates at both the personal and corporate levels; (iii) international tax competition, in which both domestic and foreign corporate tax rates play a role, is consistent with the data; (iv) tax effects on GDP growth appear to operate largely via impacts on factor productivity rather than factor accumulation.
