THE DETERMINANTS OF SCHOOL BUDGET ELECTION OUTCOMES IN NEW YORK STATE: A FORECASTING MODEL (EDUCATIONAL FINANCE, PROPERTY TAXATION)

This research examines the effect sociodemographic, school district, and tax related factors have on the outcome of school budget elections in New York State. The basic hypothesis is that tax related factors associated with the election are the most important determinants of school budget voting. The dissertation shows how the distribution of costs as reflected in tax burdens and increases in taxation at the time of the referenda effect voting behavior.;The relative importance of variables associated with each of the three models are assessed using multiple-partial coefficients of determination. Two regression models are developed for the 294 suburban school districts in our sample. They are evaluated in terms of their ability to forecast voting behavior subsequent to that used in the statistical analysis. An ex post forecast is performed utilizing measures of the forecast error variance, standard error of forecast, 1-U square statistic, and root mean square error. Path analytic techniques summarize the structural relationships among the variables in the final model and assess their relative importance. The six variables in the final model explain about half of the variation.;The findings of this research affirm the importance of economic and fiscal factors in school budget elections. The tax related factors made by school officials at the time of the referenda are the most important.;The research design tests hypotheses proposed by three models, and develops a forecasting model that is capable of predicting school budget election outcomes. A statistical analysis of the data is conducted in three stages: (1) a comparison of the differences of means of success and defeat districts using a t-test; (2) hypothesis testing of individual variables using bivariate regression analysis; (3) the development of predictive models using multivariate regression analysis.