Why do Tax and Expenditure Limitations Pass in State Elections?

Abstract

This article uses information on the actual conditions in all states over the extended period 1978 to 1990 to estimate the probability of tax and expenditure limitation (TEL) passage in state elections. This probability is a function of economic, fiscal, demographic, and political factors in the state, as well as specific features of the TEL; these variables are chosen to measure demand-side and supply-side features of the political process in the state. The estimation procedures use maximum likelihood techniques; a special feature of the estimation is correction for the sample selection bias that may occur with standard probit estimation because a threshold set of conditions must exist before any limitation is even placed on a statewide ballot and a voting outcome is actually observed. The results show that increases in both property taxation and local revenues relative to state revenues increase the probability of TEL passage; however, increases in total state tax revenues are associated with a lower likelihood of TEL passage. Furthermore, changes in income and in the tax price of state and local public services are important determinants of TEL passage over time. Specific features of a limitation do not generally affect passage probabilities, although a state that has already imposed a limitation is less likely to do so again. Political and demographic features of the state also have little impact on TEL outcomes.