Pre-College Human Capital Investments and Affirmative Action: A Structural Policy Analysis of US College Admissions

I study a structural model of college admissions framed as a multi-object, asymmetric, all-pay auction with bid preferences and single-unit demands. Auction theory provides a parsimonious representation of endogenous investment incentives in matching markets. I establish semiparametric identification and show that the identifying assumptions employed are minimal, given the available data. I then propose an estimator based on a two-stage procedure pioneered by Guerre, Perrigne, and Vuong [19] for estimation of first-price bidding models with private information. Structural estimates produce counterfactual experiments to compare color-blind admissions and the current US form of affirmative action (AA) on investment and welfare. An AA ban would result in a large migration of minority students out of the best schools and into the lowest quality schools, as well as a significant drop in minority investment and college graduation rates. A more extreme form of AA, a representative race quota, induces more HC investment by minorities, but involves a larger welfare loss. However, if the social planner trading off racial equity and efficiency is indifferent between the actual US AA and a color-blind alternative, then counterfactual estimates imply that a quota is strictly preferred to both. Date: Original version: October 2009; This version: June 2013.