Racial segregation and the black–white test score gap☆
Section snippets
Peer group and segregation effects: sources and evidence
Racial or ethnic segregation can affect the relative educational achievement of black students through several mechanisms. One of the most widely discussed channels is a peer exposure effect, arising from the fact that students' outcomes depend on the expectations and achievement of their peers, and from a presumed correlation between these characteristics and the racial composition of the peer group. A second is that the black enrollment share at a school may be correlated with school quality (
Empirical framework
To illustrate our empirical approach and its relationship to a standard peer effects model, we begin with a simplified specification that focuses on school-level influences. We then extend this model to allow for neighborhood effects as well.
Assume that a student's test score depends on his or her own characteristics, the racial composition and other characteristics of his or her schoolmates, school resources and quality, and an unobserved error with a school-level component that may vary by
Data sources and sample overview
Our primary source of student achievement data is a sample of SAT records for roughly one third of test takers in the 1998–2001 high school graduation classes.17
Basic models
Table 3 presents our estimates of the model given by Eq. (7). The dependent variable is the black–white SAT score gap in the city. All the models include main effects for the overall fraction black and Hispanic in the city's schools, dummies for 5 census divisions, and the black–white gap in an inverse Mill's ratio formed from the race-specific SAT participation rates at each school.31
Confounding influences
Table 3, Table 4 indicate that there is a relatively strong relationship between segregation and the black–white achievement gap, and that this relationship cannot be attributed to selective test participation. More tentatively, the link appears to run through neighborhoods rather than schools. In this section we investigate several potential biases that might lead us either to overstate the effect of neighborhood segregation or to understate the independent effect of school segregation.
Within school segregation?
One potential explanation for our finding that school segregation has little or no effect on relative achievement is that in cities with highly segregated neighborhoods, school integration efforts are offset by programs and behaviors that lead to within school segregation (Clotfelter et al., 2003, Clotfelter, 2004, Eyler et al., 1983).40
Indirect effects of school quality and peer characteristics
As noted in the discussion of Eq. (5), our coefficient estimates capture direct minority exposure effects as well as any indirect effects associated with relative resources and peer characteristics that can be predicted by the relative exposure of black and white students to minority peers. As a final step in our analysis we explore the potential contributions of one type of indirect effects: those arising from differences in the relative incomes of schoolmates and neighbors (Wilson, 1987).42
Summary and conclusions
In this paper we present new evidence on the effects of racial segregation on the relative achievement of black students. Building from a model in which the racial composition of school and neighborhood peer groups exerts both direct and indirect causal effects on student achievement, we show that the black–white achievement gap in a city will vary with the relative segregation of schools and neighborhoods in the city, and that this aggregated design eliminates many of the biases that arise in
References (43)
- et al.
Interactions-based models
The scholastic aptitude test: participation and performance
Economics of Education Review
(1987)- et al.
An economic analysis of acting white
Quarterly Journal of Economics
(2005) - et al.
Race and school quality since Brown v. Board of Education
Brookings Papers on Economic Activity
(1992) - et al.
The Shape of the River: Long-Term Consequences of Considering Race in College and University Admissions
(1998) - et al.
School quality and black–white relative earnings: a direct assessment
Quarterly Journal of Economics
(1992) - et al.
Racial segregation and the black–white test score gap
- et al.
Selection Bias in College Admissions Test Scores
(2007) After Brown: The Rise and Retreat of School Desegregation
(2004)- et al.
Segregation and resegregation in North Carolina's public school classrooms
North Carolina Law Review
(2003)
Will affirmative action policies eliminate negative stereotypes?
American Economic Review
Equality of Educational Opportunity
Cincinnati's experiment in Negro education: a comparative study of the segregated and mixed school
Journal of Negro Education
Interpreting the evidence on life cycle skill formation
Are ghettos good or bad?
Quarterly Journal of Economics
The rise and decline of the American ghetto
Journal of Political Economy
Measuring peer group effects: a study of teenage behavior
Journal of Political Economy
Resegregation: segregation within desegregated schools
Can schools narrow the black–white test score gap?
Wage comparisons—a selectivity bias
Journal of Political Economy
Desegregation and black dropout rates
American Economic Review
Cited by (286)
Black-Friendly businesses in cities during the Civil Rights Era
2024, Journal of Urban EconomicsThe peer effect of migrant children on students’ non-cognitive outcomes: Evidence from China
2023, International Journal of Educational DevelopmentThe long run effects of de jure discrimination in the credit market: How redlining increased crime
2023, Journal of Public EconomicsMinority travel disparities and residential segregation: Evidence from the 2017 national household travel survey
2022, Transportation Research Part D: Transport and EnvironmentCitation Excerpt :The results varied slightly by gender but not by race or ethnicity. Turning to education outcomes, a number of studies have demonstrated that giving Black and Latino children who live in disadvantaged neighborhoods preferential access to better quality schools is reflected in improved test scores, high school graduation rates, and college prospects (see, for example, Card & Rothstein, 2007; and Dobbie & Fryer, 2011). Preferential access in this case means the opportunity to enroll in a better school, not necessarily a shorter or briefer trip to school.
THE IMPACT OF PUBLIC SCHOOL CHOICE: EVIDENCE FROM LOS ANGELES’S ZONES OF CHOICE
2024, Quarterly Journal of EconomicsContemporary differences in residential housing values along historic redlining boundaries
2024, Real Estate Economics
- ☆
We are grateful to the Andrew Mellon Foundation and the College Board for assistance in obtaining the SAT data used in this study, and to Jacob Vigdor, Jon Guryan, and Sarah Reber for supplying other data. We thank Florence Neymotin and Ashley Miller for outstanding research assistance, and Ken Chay, Nicole Fortin, Jeff Kling, Thomas Lemieux, Justin McCrary, anonymous referees, and seminar participants at Berkeley, Columbia, Syracuse, Wharton, Yale, NBER, and the Universities of Connecticut, Illinois, and Maryland for helpful comments and suggestions. Card's research was supported by a grant from the NICHD and Rothstein's by Princeton's Center for Economic Policy Studies.