How Colleges Make Racial Disparities Worse Affirmative action sets up unprepared students for failure. Yet schools ignore this ‘mismatch’ evidence. By Richard Sander
http://www.wsj.com/articles/how-colleges-make-racial-disparities-worse-1450396886
Supreme Court Justice Antonin Scalia ignited a firestorm last week at oral arguments for Fisher v. University of Texas, a case concerning that school’s affirmative-action policies. The media pounced after Justice Scalia suggested that it might be not be a bad thing if fewer African-Americans were admitted to the University of Texas. Many rushed to call the comments racist.
Subsequent reports clarified that Mr. Scalia had been invoking the “mismatch” hypothesis, which posits that students who receive large admissions preferences—and who therefore attend a school that they wouldn’t have gotten into otherwise—often end up hurt by the academic gap between them and their college peers. But on the whole even this coverage has spread confusion.
The size of the preferential treatment is all-important. Mismatch problems almost always result from very large preferences—ones that give applicants the equivalent of, say, a 200-point SAT boost. Some studies that claim to provide evidence against mismatch turn out to involve small preferences, perhaps the equivalent of a 50-point SAT boost. My own view is that relatively small preferences (based, for example, on socioeconomic disadvantage) are often a good thing. Giving a slight benefit of the doubt to ambitious students trying to rise out of poverty, and placing them with peers who are slightly better prepared, can push them to greater achievement.
Much of the controversy about mismatch arises because scholars or pundits talk past one another. After sidestepping the noise, there is a surprising level of consensus in the literature. There are now five unrebutted peer-reviewed studies—for instance, one by Frederick Smyth and John McArdle, published in 2004 by Research in Higher Education—concluding that aspiring scientists who receive large admissions preferences drop off the STEM track at up to twice the usual rate. A study by three labor economists, to be published next year in the American Economic Review, finds that large preferences substantially depressed the rate at which minorities achieved science degrees at the University of California before racial preferences were banned in 1996.
Law schools are another case: I estimate that only one in three African-Americans entering law school graduates and passes the bar on the first try, compared with two in three whites. All four of the peer-reviewed articles on the subject—such as Doug Williams’s 2013 study in the Journal of Empirical Legal Studies—find strong evidence that mismatch helps explain this gap. None of the critics of law-school mismatch have managed to publish their rebuttals in a peer-reviewed journal, and their claims have not stood up well to scrutiny.
When we turn to “second-order” effects, such as whether a student granted preferences ultimately graduates, the evidence of mismatch is weaker. This is partly because Ivy League schools now have graduation rates of close to 100%, so a mismatched student who struggles at Brown or Yale, perhaps switching out of a difficult curriculum and ending with mediocre grades, will still almost certainly wind up with a degree.
At the University of Texas, the gaps in academic preparation are wide: Among freshman admitted outside the state’s top 10% system in 2009, the mean SAT score for whites was 390 points higher than for blacks. That indicates a large preference indeed. Mismatch very likely occurs at Texas, but not enough data has been made public to study the matter properly. In any event, the relevance of the mismatch debate to Fisher goes deeper.
Scholars who find evidence of mismatch almost always phrase their findings cautiously and do not claim to have “proven” that preferences are invariably harmful. Defenders of affirmative action, on the other hand, almost universally claim that mismatch never exists. This week the New York Times published a letter from two former college presidents, William Bowen of Princeton and Michael McPherson of Macalester College. The pair claim that “there is an abundance of empirical evidence refuting the mismatch hypothesis and no credible evidence supporting it” (emphasis theirs). This is the line taken in Fisher by nearly all the relevant briefs on the university’s side, whose authors ignore the studies staring them in the face.
The pattern of denial is telling. To admit that mismatch is even sometimes a problem would concede the crucial point: that universities should be more transparent; that they should make available enough data about admissions, learning and student outcomes to enable independent scholars (not to mention applicants) to evaluate the effects of preferences. This the universities are extremely reluctant to do, not least because they fear the reaction of campus activists to the conclusions that this data might suggest.
The fundamental significance of mismatch for Fisher, then, is not about precisely nailing down whether 20%, 50% or 80% of the students who receive large preferences would have been better off at a less elite school. What’s important is the way that universities have taken the absolutist position of denying there is any effect at all. In this way universities have shown that they cannot be trusted to be good fiduciaries for their most vulnerable students.
In the past the Supreme Court has granted deference to universities to give them, in effect, a special dispensation to take race into account in admissions. The mismatch debate has shown that universities should receive the opposite of deference.
Mr. Sander is an economist and law professor at UCLA and the author, with Stuart Taylor Jr., of “Mismatch: How Affirmative Action Hurts Students It’s Intended to Help, and Why Universities Won’t Admit It” (Basic Books, 2012).
Comments are closed.
