Voting in a Local School Election: A Micro Analysis
IN recent years empirical studies of local school finance have relied to a large extent on the median voter and related models, tested with data aggregated to the precinct, school district, or local level.1 While there are advantages to using aggregated data, the limited availability of data on the distribution of income, property tax payments, and other variables, as well as the possibility of bias associated with the grouping of households into aggregated units, suggest some important disadvantages.2 This paper attempts to analyze the demand for local public education using individual household data obtained through a survey of voters in two local school elections in a Detroit suburb.3 In section I a model of voting in a school election is presented. The model assumes that individual voters determine their desired level of educational expenditures per pupil by maximizing a utility function subject to a budget constraint. Individuals decide whether to vote for or against a given millage request by comparing their desired expenditure level with the actual and proposed levels. On the basis of some assumptions concerning the stochastic nature of the individual utility functions, our analysis suggests that the probability of a yes or no vote can be estimated using a binary logit form.4 In section II the model variables and estimates of the model parameters are presented and discussed. The estimation results are interpreted in the context of the voting model presented in section I and are compared to the results of several educational expenditure studies. In section III the outcome of the two local elections is analyzed, with an attempt made to explain the passage of the second election, in light of the failure of the first. In particular, the model is used to test the reaction of voters to the state circuit-breaker legislation which was enacted after the first election. Some concluding remarks are presented in the final section. Finally, some tests for the presence of bias in the survey responses are described in the appendix.
- DOI
- 10.2307/1924901
- Volume
- 59 (1)
- Pages
- 30
- Export
- BibTeX
- Sources
- openalex crossref