Factors Associated with Voter Resistance to Public Educational Expenditure

A comprehensive regression and correlation analysis was performed using two samples of Ohio school districts. The purpose was to describe measurable socio/economic factors associated with varying levels of educational effort and to pinpoint policy implications for state or local decision makers. Evidence exists that larger districts are associated with lower cost per pupil and that wealth, measured in different ways, is associated with expenditure per pupil. State programs do somewhat equalize educational efforts, but they are far from achieving equality. OHIOJ. SCI. (6): 275, 1981 State funding programs to primary and secondary schools in Ohio have an objective of providing equal educational opportunity to the youth in the state. State policies must take into account the different levels of local effort that occur in different school districts in Ohio. The emphasis of my analysis is to identify measurable factors associated with different levels of local support for public education. The results include a statistical description of the system and policy implications for state and/or local decision makers. The measure of support selected for this analysis is an expenditure per pupil for Ohio school districts. Expenditure per pupil serves the dual role of a quality variable and an efficiency variable. High expenditure per pupil may denote low pupilteacher ratios, modern supplies and equipment, etc., or it could denote waste and inefficiency. In addition, "costs" may differ between schools due to differences in teaching training and experience, rural-urban cost differentials, or differences in pay scales. For the purpose of this paper, I concentrate on the quality measures, although I will also attempt to control for cost differences. It must also be assumed for the purpose of this demonstration that high expenManuscript received 1 May 1979 and in revised form 15 January 1981 (#79-29). diture per pupil is more likely to be associated with better school programs than is lower expenditure per pupil. It is worth noting that other researchers confirm many of the measurements indicated in this paper and offer several additional insights. Parsons (1978) found that school districts with low local per pupil operating revenues generally have larger percentages of operating funds coming from real property that includes agricultural land. These districts represent areas of Ohio that have a sparsity of population, larger agricultural areas, and that generally have property with low economic value. In addition, lower millage is found where industrial properties often provide a relatively large number of dollars per pupil. Gensemer (1978), using a Lorenz Curve Analysis, found that total revenues were distributed more in favor of high fiscal capacity districts in 1977-78 than they had been in 1974-75, indicating Ohio is moving away from fiscal neutrality over this period. METHODS The statistical methods used for my study were correlation analysis and linear and quadratic regression analysis. I chose a sample of 50 school districts, utilizing data from the 1973-74 school year. Ten independent socio/economic or demographic variables were used in the regression. Quadratic terms in this regression were also introduced in order to test for economies of scale. An additional sample of 146 districts was selected, and 15 independent variables