Empirical research on the effects of differential business taxation across jurisdictions relies on the appropriate measurement of the burden of tax in each location. While numerous summary measures have been proposed and used in various contexts to make such comparisons, most fail to account for the full effects of each state's tax system and the interactions of state tax systems with both local and federal taxes. This paper addresses these issues and employs an approach used in recent state tax reform studies to measure tax burdens. The advantages of this "representative firm" approach over traditional measures are discussed, and its empirical significance is tested. ; The empirical properties of tax burdens measured via a representative firm are very different from simpler, conventional measures. Taxes are estimated to play a negligible role in decisions concerning where to locate capital investment. This result, coupled with recent results showing the responsiveness of sales and employment decisions to apportionment changes, is consistent with the hypothesis of the existence of a hierarchy of behavioral responses to tax changes.