Aiming higher: Spillover effects of affirmative action in higher education ∗

Abstract

Institution-specific affirmative action programs aim to reduce gaps in educational attainment across social groups. However, there is little evidence on whether institutionspecific programs increase overall minority enrollment, or whether students who enroll at one sponsoring institution simply substitute away from other universities of equal quality. I exploit the roll-out of a large affirmative-action program to measure the overall effect of institution-specific affirmative action on educational outcomes for low-income students. Using administrative data on applications and enrollment, I estimate the spillover effects of a program sponsored by an elite French college. For each student that enrolled at the sponsoring institution as a result of the program, two additional students enrolled at another elite college. These spillovers are driven by increases in students’ applications to selective colleges. I rule out fixed costs in college applications as an explanation for this phenomenon. Instead, I suggest that students hold incorrect beliefs about their academic ability and chances of admission at selective universities, and that the affirmative action program corrected these beliefs upward.