This paper examines the relationship between house price levels, school performance, and the racial and ethnic composition of Connecticut school districts between 1995 and 2000. A panel of Connecticut school districts over both time and labor market areas is used to estimate a simultaneous equations model describing the determinants of these variables. Specifically, school district changes in price level, school performance, and racial and ethnic compositions depend upon each other, labor market wide changes in these variables, and the deviation of each school district from the overall metropolitan area. The specification is based on the differencing of dependent variables, as opposed to the use of level or fixed effects models and lagging level variables beyond the period over which change is considered; as a result the model is robust to persistence in the sample. Identification of the simultaneous system arises from the presence of multiple labor market areas in the sample, and the assumption that labor market changes in a variable due not directly influence the allocation of households across towns within a labor market area. We find that towns in labor markets that experience an inflow of minority households have greater increases in percent minority if those towns already ahve a substantial minoritypopulation. We find evidence that this sorting proces is reflected in housing price changes in the low priced segment of the housing market, not in the middle and upper segments.
Journal of Economic Literature Classification:D1, D4, I2, R2, R5Schools and Housing Markets: Clapp and Ross page 2
Schools and Housing Markets: An Examination of School Segregation and Performance in ConnecticutThe U.S. educational system is characterized by tremendous diversity in terms of both school composition and performance across school districts, but often by substantial homogeneity within districts especially within smaller suburban districts. This homogeneity is the result of a system in which access to education is primarily accomplished by obtaining residential housing. Considerable evidence exists indicating that school quality as captured by test scores or other performance measures is heavily influenced by the socio-economic characteristics of the school (Hanushek, 1986), and has a large influence on property values in each school district (Ross and Yinger, 1999). Moreover, the high price of housing in high quality school districts and the land use decisions in those districts influence the composition of those schools and may create substantial barriers to access for many low income and minority households. In this context, the price of housing and the characteristics of schools are determined simultaneously by a complex process in which households sort over the housing stock and across communities, and communities may intervene in this process by regulating land use. Almost all research in the area has been dedicated to address estimation biases caused by this simultaneity as opposed to direct attempts to study ...