by Anh-Huy Nguyen

Abstract
California Proposition 209 outlawed race-based affirmative action in the University of California (UC) system in 1998. However, the UC system subsequently shifted towards race-blind affirmative action by also reweighing factors other than race in the admissions process. To evaluate the hypothetical changes in the science persistence rate and graduation rate of all applicants if racial preferences had been removed entirely, I estimate baseline and counterfactual admissions models using data from between 1995-1997. Using a general equilibrium framework to fix the total number of admits and enrollees, I find that the removal of racial preferences leads to a cascade of minority enrollees into less selective campuses and a surge of non-minority enrollees into more selective campuses. The improved matching between students and campuses results in higher science persistence rates and graduation rates across the pool of all applicants. In particular, the gains are driven by minority students who were admitted under racial preferences, because the gains from better matching across UC campuses outweigh the losses from potentially being pushed outside the UC system. Non-minority students who are originally rejected under racial preferences also benefit, as some are induced into the system in the counterfactual, where they are more likely to graduate. I also investigate claims that applicants may have strategically gamed during the admissions process by misrepresenting their interest in the sciences in order to maximize their admissions probability. While there exist incentives to apply in different majors across the campuses, I find evidence that applicants often fail to game optimally, suggesting that they may not be fully informed of their relative admissions probabilities in the sciences and non-sciences.

Professor Peter Arcidiacono, Faculty Advisor

JEL Codes: I23, I28, J24, H75

View Thesis