The Supreme Court announced last week that it would hear a challenge to the University of Texas’ use of racial preferences in undergraduate admissions. Since the court last reviewed college affirmative action in 2003, a body of empirical research has emerged showing that racial preferences can hurt their purported beneficiaries by catapulting them into schools for which they are inadequately prepared. Placed in classrooms pitched above their current level of knowledge, they learn less than they would if they were among peers whose academic skills more closely mirror their own. This “mismatch” effect is particularly relevant to the University of Texas case, Fisher v. University of Texas, because the university claims that it needs to admit students according to race in order to achieve “classroom diversity.” Mismatch theory predicts — correctly — that using racial preferences will have the opposite effect.
Here are the details: black and Hispanic undergraduates at the University of Texas cluster in certain programs like education and are only minimally present in more challenging majors, according to the Fifth Circuit opinion upholding Texas’ policy. A study out of Duke University demonstrates how this happens. Black freshmen arrive at Duke overwhelmingly planning to major in natural sciences and economics, but over half of them drop out of those fields and switch to the humanities and soft social sciences, leaving the hard sciences largely the province of whites and Asians. Whether a Duke student will switch out of the hard sciences is wholly a function of his incoming academic qualifications, measured by SATs and high school GPA: black students whose academic qualifications match those of their white and Asian peers are no more likely to drop out of quantitative fields than other students. The average black SAT score at Duke, however, is over one standard deviation below those of whites and Asians, thanks to the university’s use of racial preferences, and it is those preference beneficiaries who are exiting the science classes in disproportionate numbers. Had they instead found themselves in a freshman chemistry class geared toward learners with similar academic backgrounds as their own, they would have been far more likely to persist in their science careers, as the success of historically black colleges in graduating science majors demonstrates.
The Duke findings accord with common sense — none of us with middling math skills is going to be able to keep up or even feel comfortable at MIT — and they have been replicated in a variety of contexts. Freshmen with credentials more than one standard deviation below their science peers at the University of California are half as likely to end up with science degrees as similar students attending schools where their credentials are closer to those of their classmates, according to a working paper by UCLA law professor Richard Sander and statistician Roger Bolus. Kenyan students in tracked classes learn more than students in academically heterogeneous classes where their academic skills are far from the mean, MIT economist Esther Duflo and colleagues at UCLA and Harvard have shown. Black law students have a much harder time passing the bar than white students with similar LSATs because the black students are much more likely to have been snatched up by elite law schools and placed in faster-paced classes than they are ready for.
Even the Fifth Circuit acknowledged the mismatch effect. The University of Texas’ practice of accepting all students from the top 10% of their high school class, no matter how academically weak those schools are, has further widened the “credentials gap” between minority and nonminority students, judge Patrick Higginbotham observed. That gap in turn “risks driving away matriculating minority students from difficult majors like business or the sciences,” he wrote, making Texas’ goal of “classroom diversity” as elusive as ever.
Several amicus briefs seeking Supreme Court review of the Fifth Circuit opinion in Fisher v. University of Texas brought mismatch theory to the court’s attention (here and here). If the court invalidates Texas’ racial preferences, expect mismatch research to play a significant role in its reasoning. The burden will then shift to advocates of affirmative action to justify their claim that racial preferences are a boon to their recipients, but so far, they have failed to make that case.