Does Parental Choice Foster Segregated Schools?

• In Event: School Choice and School Improvement

Abstract

The Netherlands differs from most other developed countries, including the United States, with respect to its strong historical commitment to parental choice of schools. The Dutch experience can provide insights for U.S. policymakers about how a system with more parental choice might play out over time. Of central interest for this paper is the relationship between parental choice and segregation of students by educational disadvantage. Studies from both the United States and around the world have shown that parental choice often leads to segregated schools (Fiske & Ladd, 2003; Cullen et al, 2005; Bifulco & Ladd, 2007).

The Segregating Effects of Parental Choice

Greater segregation is consistent with predictions from the following model in which there are only two types of families: advantaged and disadvantaged. The sociology and economics literatures provide at least three reinforcing motivations for advantages families to choose schools that serve similarly advantaged families. The first motivation, the outgroup avoidance theory, is that some advantaged families would prefer to minimize contact with the other group. An alternative motivation, neutral ethnocentrism, posits that members of each group prefer to be with members of their own group. The third motivation relates to school quality. To the extent that the quality of schools serving advantaged students is higher than that of schools serving disadvantaged students, advantaged families who care about quality once again have an incentive to select schools serving advantaged students.

Methodology and Data

We use administrative data from 1997-2005 to examine the extent to which educationally disadvantaged immigrant pupils at the primary level are segregated from other pupils. For this purpose a disadvantaged immigrant is defined as a first- or second-generation non-Western immigrant whose parents have limited education. We use five separate measures that reflect various aspects of the extent to which disadvantaged immigrants are segregated from other students. The five measures are grouped into two categories and calculated separately for each city. Measures of isolation highlight the extent to which disadvantaged immigrant pupils are concentrated in schools with other pupils similar to themselves, and hence are isolated from more-advantaged students. Measures of imbalance highlight the extent to which pupils of the two types are unevenly distributed across schools and are invariant to a city’s overall proportion of disadvantaged immigrants.

Results

The analysis indicates that segregation levels are very high—both absolutely and relative to comparable measures for the United States—and that they have been rising. Almost 80 percent of the disadvantaged immigrant students in the four cities are in schools having more than a majority of students like themselves and that the percentage increased over the nine-year period. The proportion of such students in schools having more than 70 percent disadvantaged immigrants is correspondingly lower but still exceeds 60 percent in all years, reaching a peak in 2002. Consistent with those two measures, the isolation index indicates that the typical disadvantaged immigrant student living in one of the four cities was in a school having 70 percent or more disadvantaged immigrant students throughout the period.

Author

• Helen F. Ladd, Duke University