Working Paper: NBER ID: w20804
Authors: Sandra E. Black; Kalena E. Cortes; Jane Arnold Lincove
Abstract: College admissions officers face a rapidly changing policy environment where court decisions have limited the use of affirmative action. At the same time, there is mounting evidence that commonly used signals of college readiness, such as the SAT/ACTs, are subject to race and socioeconomic bias. Our study investigates the efficacy and equity of college admissions criteria by estimating the effect of multiple measures of college readiness on freshman college grade point average and four-year graduation. Importantly, we take advantage of a unique institutional feature of the Texas higher education system to control for selection into admissions and enrollment. We find that SAT/ACT scores, high school exit exams, and advanced coursework are predictors of student success in college. However, when we simulate changes in college enrollment and college outcomes with additional admissions criteria, we find that adding SAT/ACT or high school exit exam criteria to a rank-based admissions policy significantly decreases enrollment among minorities and other groups, with the most negative effects generated by the SAT/ACT, while inducing only minimal gains in college GPA and four-year graduation rates.
Keywords: college admissions; race-blind policies; equity; efficacy
JEL Codes: I21; I23; I24; J15; J18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| SAT/ACT scores (C00) | GPA (C00) |
| SAT/ACT scores (C00) | four-year graduation rate (Y40) |
| high school exit exams (I21) | GPA (C00) |
| advanced coursework (A23) | GPA (C00) |
| SAT/ACT scores (C00) | minority enrollment (J15) |
| average GPA (C00) | diversity among admitted students (I24) |
| SAT/ACT scores (C00) | average GPA (C00) |
| SAT/ACT scores (C00) | graduation rates (I23) |