Abstract:
This paper uses meta-analysis to evaluate the results of 42 studies and 641 individual estimates of the effect of taxes on economic growth in OECD countries. Our analysis addresses a number of difficult coding issues such as: implications of the government budget constraint for interpretations of tax effects; units of measurement for economic growth rates and tax rates; implications of equation specifications that measure short-run, medium-run, and long-run effects; length of time period (annual data versus multi-year periods); and other factors. Our main findings are: Estimates in the literature are characterized by significant (negative) publication bias. Controlling for publication bias, we find that increases in unproductive expenditures funded by distortionary taxes and/or deficits have a significant, negative effect on growth; while increases in non-distortionary taxes to fund productive expenditures and/or government surpluses have a significant, positive effect. The estimated differences in these policies indicate that there is scope for tax policy to have a meaningful impact on economic growth. Finally, we find weak evidence that taxes on labour are more growth retarding than other types of taxes, while the evidence regarding other types of taxes is mixed.