School Desegregation, "Tipping," and Private School Enrollment

Abstract

This paper examines the role of school desegregation in the demand for private school enrollment by whites. A model is presented that emphasizes the interdependence of whites' private enrollment decisions and introduces the possibility of tipping points in from public schools. In order to assess the importance of racial composition on private school enrollment, data from two samples are analyzed. This empirical analysis supports the hypothesis that desegregation has a significant effect on private school enrollment and that this effect is greater in districts with high proportions of nonwhites. In northern as well as southern cities, there is a growing recognition that the problem of white flight from desegregating schools threatens the long-run efficacy of desegregation policies. There are fears that this flight-to predominantly private schools or suburban systems-may bring about the virtual resegregation of schools in this country. In addition, these trends may be fostering the separation of low income students from higher income students, especially where private school enrollment is involved. Whether these fears are in fact justified is open to question, however. This paper examines one aspect of the problem-the alternative to the public schools that is provided by enrollment in private schools. The central aim is to assess the impact of public school desegregation on private enrollment among households. Special attention is given to two questions of particular interest in the evaluation of school policy: (1) whether there is a tipping point in from public to private schools, and (2) how family income modifies the effect of The author is Assistant Professor of Economics, Bureau of Business and Economic Research, University of Maryland. * I would like to thank Martin Feldstein, Richard Freeman, John Kain, Thomas Schelling, members of the Harvard labor seminar, and the editor and a referee of this Journal for their comments on earlier versions of this paper. Support was provided by the Ford Foundation, the National Science Foundation, and the University of Maryland Computer Science Center. [Manuscript received September 1974; accepted April 1975.] The Journal of Human Resources XI * 1 This content downloaded from 207.46.13.115 on Sat, 08 Oct 2016 05:16:13 UTC All use subject to http://about.jstor.org/terms