Alternative hypotheses exist regarding the impact of local sales and income taxes on local governments' taxing and spending decisions. One hypothesis is that local governments use sales and income taxes to pay for spending increases and leave property tax collections unchanged, while an equally plausible alternative is that local governments use sales and income taxes to reduce property taxes. Traditional models that restrict the impact of these local taxes to be the same across all local governments are not able to capture both types of behavior. The methodological difficulty lies in allowing for differences in behavior with no a priori information on which cities belong in which category. In this article, the authors use panel data to estimate a mixture model of spending and property tax response to the existence of local taxes. These empirical results provide evidence to support both hypotheses. These differences are both substantive and statistically significant.