A Reply to Finis Welch

Abstract

Welch's reanalysis of the results reported in my earlier article provides a different, although not contradictory, picture of the long-term impact of school desegregation programs on the enrollment of white students in public schools. His results, derived from an assessment of the dynamic implications of the lagged regression results I reported, clearly indicate that affected school districts do not recover the enrollments they lose because of the introduction of a school desegregation action. Since my results addressed only the question of whether the rate of the response of the enrollment of white students to increased contact with black students changes as a result of the implementation of a desegregation action, Welch's results are an important extension. Welch also summarizes the results of an original analysis of which he was one of the co-authors. However, he fails to note that my analysis differs from his and previous works with respect to the large number of districts that are included, encompassing the enrollment experiences of districts of different sizes and locations. In this regard, it is not clear how one can compare Welch's results from a descriptive analysis of 125 districts to those of my study, since the two samples are not representative of a larger universe. This is not to suggest that Welch's results are not plausible; rather, it suggests that their significance cannot be assessed within the same context as those I presented. Welch makes several inferences from the results reported in my earlier article to which it is appropriate that I should respond within the context of the conclusions reported in the summary section of my article (Wilson 1985, pp. 151-152). My results indicated that patterns of withdrawal and nonentrance can be induced by the broader ecological processes of residential differentiation and selection, as indicated by the fact that increased interracial contact led to similar declines in the enrollment of white students in districts that did not desegregate and in districts that eventually desegregated, when the latter districts were somewhat removed from the desegregation controversy. This finding should not be surprising, since the character of the school environment is an important factor that families consider in selecting a residential location. Although increased interracial contact is the mechanism that seems to trigger declines in enrollment, racial prejudice and parents' perceptions of lowered educational standards, the breakdown of discipline, and the physical and verbal abuse of their children are probably the underlying causes. The extent of acceleration in the rate of decline in school enrollment occasioned by the implementation of a desegregation action is probably linked intricately to the pervasiveness of the changes mandated by the action itself. This would explain why the phase variable in my analysis had no effect that was independent of a change in interracial contacts. Welch interpreted the phase-dependent responses to change in exposure among implementing districts-as evidence of the influence of the programmatic component of desegregation programs. He suggests that the lower rate of response that occurred during the preimplementation phase is similar to what would be observed under a voluntary plan, while the higher rate of response that occurred during the implementation phase is similar to what would be observed under a plan calling for the mandatory reassignment of students. In addition, Welch questions whether the results reported in Table 5 of my article would be the same if the characteristics of plans were permitted to vary by phase. This is an empirical question, a definitive answer to which must await further analysis. Moreover, if Welch's hypothesis is plausible, one wonders why districts that were under court order, which usually involved the mandatory assignment of pupils, did not experience greater declines in enrollment (see Tables 4 and 5 and footnote 7 in my 1985 article). Courts, via the institution of mandatory plans, have been most active and more effective in districts with higher levels of school segregation and high concentrations of black students. The implementation of voluntary plans, usually by the district's own initiative, however, occurred mainly in small-to-moderate-sized districts outside the South with low concentrations of Address all correspondence to Professor Wilson at the Center for Demography and Ecology, Social Science Building, University of Wisconsin, Madison, WI 53706-1393.