Revisiting the effect of affirmative action on minority student outcomes

Abstract

The use of affirmative action has been at the forefront of the debate regarding the fairness of the college admissions process for years. Utilizing data from the National Longitudinal Survey of Freshman (NLSF), M. Fischer and Massey (2007) test the effects of race-based admission policies. Providing a specific focus on cumulative GPA, they conclude that the effect of affirmative action is minimal. Testing the robustnessof this claim, I replicate their results. Their regression specification suffers from issues of high correlation between their SAT and individual affirmative action variable. When adjusting for these problems, affirmative action actually has a negative effect on cumulative GPA. There is a 0.10 reduction in GPA for every 100-point increase in the difference between an individual’s SAT score and the institutional average.