Tax Burden, defined as the ratio of total tax revenues over personal income, is frequently used to measure state tax policy. The authors analyze the empirical relationship between changes in Tax Burden and changes in tax policies from 1987 to 2000 using states’ forecasts of revenue impacts of new tax legislation. Their two major findings have important implications. First, they demonstrate that income-induced, nontax policy changes are a significant determinant of changes in Tax Burden. These income effects are likely to cause misinterpretation when Tax Burden is used as a variable in economic growth regressions. Second, they estimate that approximately half of the total variation in Tax Burden is due to changes in nontax policy factors. This finding quantifies the extent of the “mismeasurement” problem that has been discussed, but not analyzed, in previous literature. In concluding, the authors promote the use of alternative approaches for estimating the economic effects of taxes.