Affirmative action, meritocracy, and efficiency

This article provides a framework for comparing meritocratic and affirmative action admissions policies. The context of the analysis is admissions to public universities; admission rules are evaluated as part of the public investment problem faced by a state government. Meritocratic and affirmative admissions policies are compared in terms of their effects on the level and distribution of human capital. I argue that (a) meritocratic admissions are not necessarily efficient and (b) affirmative action policies may be efficiency enhancing relative to meritocratic ones. Both these claims, as well as their negations, depend on features of individual behavior for which there is little empirical evidence. The implications of this absence of evidence are then explored, with a focus on policy evaluation when equality and efficiency are both desiderata. I argue that standard statistical decision theoretic approaches do not apply to the affirmative action case, even if equality and efficiency are rendered commensurable based on a scalar payoff function. In this context, I suggest that a presumption for equality-enhancing policies leads to support for affirmative action, but I emphasize the contingent nature of this conclusion.