Association for Public Policy Analysis & Management

In recent years, several states have decided to discontinue affirmative action in admissions to public universities statewide.  There is a growing body of research on the effects of these affirmative action bans on enrollment, and the usual finding is that they reduce minority enrollment at selective colleges.  However, the effects of affirmative action bans on racial segregation have not yet been directly studied.

This paper uses a difference-in-differences estimator on data from a panel of U.S. states for 1995-2003 to estimate the effects of having a statewide affirmative action ban on various segregation indexes, including exposure indexes and dissimilarity indexes.  The preliminary results suggest that statewide affirmative action bans may actually be reducing racial segregation in higher education.  One possible explanation for this seemingly perverse finding is that there is a U-shaped relationship across colleges between the average SAT score and the percentage of the student body that is black.  If affirmative action bans cause minority students to cascade down from very selective institutions to slightly less selective institutions (while having little effect on those attending non-selective institutions), it is possible that an affirmative action ban could lower segregation as measured by standard segregation indexes.  However, it is worth keeping in mind the caveat that the effect of an affirmative action ban on segregation can vary depending on the exact nature of the reshuffling of students to colleges, as well as the additional caveat that the effect of affirmative action on segregation is just one of a number of considerations that policymakers may want to take into account when deciding on affirmative action policy.