| Recent discussions of "globalization" have highlighted the prospect that a country's ability to tax and regulate---in ways that differ appreciably from its neighbors---is diminishing over time. This paper sets out and empirically tests a model of interjurisdictional tax competition among national governments, where the basis for competition is the securities transaction tax (STT), an ad valorem tax levied on the value of equity trades.;Using a newly assembled panel data set for 16 countries and 19 years, the econometric work indicates that a 1% increase in the rival's tax is associated with a 0.3% rise in own tax in the short-run, and a 1% rise in the long-run. In other words, countries match each other's tax changes in the long run. In addition, the impact of a change in rivals, taxes on own tax is greater, the lower are barriers in international capital markets. Finally, the tax is shown to depend negatively on the openness of the economy, on the size of the financial sector, and on the share of government expenditures in the economy.;The elasticity of trading volume with respect to the tax is estimated in two ways. First, as in previous studies of the STT, the tax is assumed to be exogenous; this yields an estimate of around -0.3. Exogeneity of the tax is rejected, however. Accounting for the endogeneity results in an elasticity that is three times larger, a more reasonable estimate given the government's revenue maximization problem.;Finally, firms' cross-listing decisions are considered. Cross-listing abroad is negatively associated with rivals, taxes; conversely, cross-listing from abroad is negatively correlated with own tax. Cross-listing is also more likely to be undertaken by larger firms. Further, there is a positive relationship between cross-listings from abroad and both market size and turnover rate, suggesting that enhanced liquidity is an objective in cross-listing. Further, it is shown that the adverse relationship between the tax and trading volume is made worse, the higher is the proportion of cross-listed stocks. Finally, the slope of the best-response function increases in the proportion of cross-listed stocks on a given market. |
