Association for Public Policy Analysis & Management

Works like Alesina et al. (2004) and Hanushek and Yilmaz (2006) have suggested that there may be a relationship between the arrangement of school districts (e.g. governmental density, the number of districts per 100,000 residents) within a county or region and the efficacy of the educational systems operated therein. This work will develop an interactive model conceptually similar to Ulfarsson and Carruthers (2006) that will simultaneously estimate educational, governance, and housing outcomes in a general equilibrium system to test this hypothesis. In short, are schools more effective when, for instance, the entire county shares a school district, or when each community has their own?

District-level data will be collected for 609 school districts from the Ohio Department of Education. Separate specifications of this model will consider both raw achievement on standardized tests and value added ratings. Other variables in this model will include measures of housing cost, which are broken out by district in the American Community Survey, and data on educational taxes, collected by the Ohio Department of Taxation. Statistics describing the geography of school districts will be drawn from the Census of Governments and Metropolitan Power Diffusion Index data from the University of Pittsburgh (Briem 2011, unpublished).

By simultaneously estimating both fragmentation and educational spending and achievement alongside housing market outcomes, the model is able to do more than simply evaluate whether or not local governance structure has an impact on the efficacy of schooling; it will also test the legitimacy of the Tiebout (1956) hypothesis by determining if the value of educational excellence with respect to expenditures is capitalized in the cost of housing, as would be predicted by Oates (1969) and Wheaton (1993). If markets are efficient, it should be impossible to relocate to a higher-performing and/or lower-cost school district without paying more for housing, all else being equal. If not, this suggests opportunities for families of limited means to secure housing elsewhere and improve their personal finances and/or children’s educational performance.