College Admission and High School Integration

Discrimination is an economic concern because it distorts not only the allocation, but also groups’ payoffs from decisions made before the market, for instance school or neighborhood choice. Policies such as affirmative action aiming at desegregation as a response to discrimination must therefore be evaluated also in terms of their effects on earlier choice. We find that a policy only operating at a later stage, conditioning on earlier individual choice, but not on exogenous markers such as race or gender, may achieve desegregation with respect to that marker in both stages. An example for this is a college admission rule based on relative performance at school. If groups that are to be integrated are disadvantaged ex ante, this policy rewards some advantaged individuals for integrating at school. We present empirical evidence for a decrease in segregation at the high school level as an unintended consequence of introducing of the Texas Top Ten percent college admission rule.