Debates about affirmative action in higher education typically revolve around abstract ideas of justice: Opponents believe racial preferences are an immoral form of discrimination, while supporters contend that preferences provide essential redress to equalize competition and correct past injustices. Pragmatic questions about the size, effectiveness and costs of preferences tend to be shunted aside, partly because good data has been hard to come by.
All that is starting to change.
Over the past few years, researchers have begun to produce large datasets that make it possible to compare the fortunes of minority students who attend universities that use varying levels of admissions preferences. In many contexts, scholars find that students perform better, both in the shortterm and the longterm, when students’ credentials are closer to those of their classmates. When students are surrounded by peers who have much higher credentials, they often have more trouble persisting in a difficult major, graduating from college or getting a good job.
This phenomenon is known as the “mismatch effect,” and last month I published a study in the Stanford Law Review, trying to determine whether the mismatch effect operates in law schools. Legal education is an ideal place to study these effects because nearly all law schools rely on similar, largely numerical admissions criteria and have standardized grading schemes. And nearly all states require law graduates to take bar exams before they can practice law. My study focused on black law students and compared black and white outcomes.
I found that law schools almost universally use very large preferences for blacks to achieve something very close to racial proportionality. The credentials gap between white and black students is about 30 times larger than it would be in a race-blind regime.
Starting a highly competitive curriculum with a large academic disadvantage, blacks wind up clustered in the bottom tenth of the class at nearly all law schools. I estimate the mismatch effect increases the number of black dropouts from law school by 40%, and increases the number of blacks failing their first bar exam by 80%.
The mismatch effect appears to operate in the job market as well. Law firms — once thought to be single-minded in their determination to recruit lawyers from the most elite schools possible — turn out to weigh law school grades more heavily than school prestige. The typical black law graduate, I estimate, loses about $10,000 in annual earnings because large preferences induce her to make a bad trade-off between law school prestige and law school grades.
These findings are controversial, of course, and have already sparked other interpretations of the data. But that is the larger point. The affirmative action debate is entering a new era, away from an exclusive focus on ideology and toward a more robust discussion of its costs, its benefits and how effective it really is as a social policy. A long-standing code of silence about how preferences operate is coming to an end — and as we learn and understand more about them, everyone should benefit.