Yale University

Class News

Dick Berk '64 quoted on affirmative action

Does Affirmative Action Do What It Should?

by Dan Slater

The New York Times

March 16, 2013

What's more important to how your life turns out: the prestige of the school you attend or how much you learn while you're there? Does the answer to this question change if you are the recipient of affirmative action?

From school admissions to hiring, affirmative action policies attempt to compensate for this country's brutal history of racial discrimination by giving some minority applicants a leg up. This spring the Supreme Court will decide the latest affirmative action case, weighing in on the issue for the first time in 10 years.

The last time around, in 2003, the court upheld the University of Michigan Law School's affirmative action plan. A divided court ruled, 5 to 4, that "student body diversity is a compelling state interest that can justify the use of race in university admissions." Writing for the majority, Justice Sandra Day O'Connor said, "We expect that 25 years from now, the use of racial preferences will no longer be necessary to further the interest approved today."

In the intervening period, scholars have been looking more closely at how affirmative action works in practice. Based on how they interpret the data that have been collected, some of these scholars have come to believe that affirmative action doesn't always help the students it's supposed to. Why? Because some minority students who get into a top school with the help of affirmative action might actually be better served by attending a less elite institution to which they could gain admission with less of a boost or no boost at all.

The idea that affirmative action might harm its intended beneficiaries was suggested as early as the 1960s, when affirmative action, a phrase introduced by the Kennedy administration, began to take hold as government and corporate policy. One long-simmering objection to affirmative action was articulated publicly by Clarence Thomas years before he joined the Supreme Court in 1991. Mr. Thomas, who has opposed affirmative action even while conceding that he benefited from it, told a reporter for The New York Times in 1982 that affirmative action placed students in programs above their abilities. Mr. Thomas, who was then the 34- year-old chairman of the Equal Employment Opportunity Commission, didn't deny the crisis in minority employment. But he blamed a failed education system rather than discrimination in admissions. "I watched the operation of such affirmative action policies when I was in college," he said, "and I watched the destruction of many kids as a result."

Scholars began referring to this theory as "mismatch." It's the idea that affirmative action can harm those it's supposed to help by placing them at schools in which they fall below the median level of ability and therefore have a tough time. As a consequence, the argument goes, these students suffer learningwise and, later, careerwise. To be clear, mismatch theory does not allege that minority students should not attend elite universities. Far from it. But it does say that students — minority or otherwise — do not automatically benefit from attending a school that they enter with academic qualifications well below the median level of their classmates.

The mismatch theory, if true, would affect many kids. According to a 2009 book, No Longer Separate, Not Yet Equal: Race and Class in Elite College Admission and Campus Life, by Thomas J. Espenshade and Alexandria Walton Radford, a black student with an otherwise similar application to a white student receives the equivalent of a 310- point bump in SAT scores.

Mismatch theory attracted little attention until 2005, when a law professor at UCLA, Richard H. Sander, published a provocative article in the Stanford Law Review, which focused on how affirmative action affected law students. Mr. Sander claimed that "a student who gains special admission to a more elite school on partly nonacademic grounds is likely to struggle more" and contended that "if the struggling leads to lower grades and less learning, then a variety of bad outcomes may result: higher attrition rates, lower pass rates on the bar, problems in the job market. The question is how large these effects are, and whether their consequences outweigh the benefits of greater prestige."

In other words, do the benefits of the connections made at, say, UCLA School of Law, and the weight UCLA carries in the job market, outweigh the cost of struggling academically there? Based on his reading of the data, Mr. Sander concluded that they did not.

Law school, as it turns out, is a somewhat natural, though imperfect, environment for studying mismatch effects. Law students have their knowledge tested in a fairly uniform way, first on the LSAT and then again, after graduation, on a state licensing exam, the bar.

Much of the squabble over mismatch centers on differing interpretations of the Bar Passage Study. The B.P.S. was commissioned by the Law School Admission Council in 1989 to determine whether blacks and Hispanics had disproportionately poor bar-passage rates. In 1991, more than 27,000 incoming law students — about 2,000 of them black — completed questionnaires for the B.P.S. and gave permission to track their performance in law school and later on the bar.

Among other things, the questionnaire asked students (a) whether they got into their first-choice law school, (b) if so, whether they enrolled at their first choice, and ( c) if not, why not.

Data showed that 689 of the approximately 2,000 black applicants got into their first-choice law school. About three-quarters of those 689 matriculated at their first choice. The remaining quarter opted instead for their second-choice school, often for financial or geographic reasons. So, of the 689 black applicants who got into their first choice, 512 went, and the rest, 177, attended their second choice, presumably a less prestigious institution.

This data presented a plausible opportunity to gauge mismatch. The fact that 689 black students got into their first-choice law school meant that all 689 were similar in at least that one regard (though possibly dissimilar in many other ways). If mismatch theory held any water, then the 177 students who voluntarily opted for their second-choice school — and were therefore theoretically better "matched" — could be expected, on average, to have better outcomes on the bar exam than their peers who chose the more elite school. Mr. Sander's analysis of the B.P.S. data found that 21 percent of the black students who went to their second-choice schools failed the bar on their first attempt, compared with 34 percent of those who went to their first choice.

The experiment is far from ideal. Mismatch opponents argue that there are many unobservable differences between second-choice and first-choice students and that those differences, because they're unknown, cannot be accounted for in a formula. In the case of the B.P.S. data, maybe the second-choice students tended to have undergraduate majors that made them particularly well suited to flourish in the classroom and on the bar, regardless of which law school they attended. "All this work on mismatch assumes you know enough to write an algebraic expression that captures what's really going on," says Richard A. Berk, a professor of criminology and statistics at the University of Pennsylvania. "Here, there's so much we don't know. Besides, the LSAT is a very imperfect measure of performance in law school and thereafter, as is the bar exam."

Daniel E. Ho, a law professor at Stanford, also disputes the mismatch hypothesis. In a response to Mr. Sander's 2005 law review article, Mr. Ho wrote in the Yale Law Journal that "black law students who are similarly qualified when applying to law school perform equally well on the bar irrespective of what tier school they attend."

Political changes in the '90s created another opportunity to study mismatch. In 1996, California voters passed Prop 209, a ban on affirmative action. Critics of Prop 209 expected black and Hispanic enrollment at top University of California schools, like U.C.L.A. and Berkeley, to plummet — and it did, for a while. But these schools eventually saw increases in minority enrollment, particularly among Hispanics, as sophisticated new outreach programs kicked in. Enrollment has not, however, gotten back to pre-Prop 209 levels.

Recently, economists from Duke studied the effects of Prop 209, comparing undergraduate graduation rates for blacks, Hispanics, and American Indians before and after the ban. In a paper being considered for publication by The Quarterly Journal of Economics, the Duke economists conclude that mismatch effects are strongest for students in so-called STEM majors — science, technology, engineering and math. These subjects proceed in a more regimented way than the humanities, with each topic and class building on what came before. If you don't properly learn one concept, it's easier to get knocked off track.

The Duke economists say that lower-ranked schools in the University of California system are better at graduating minority students in STEM majors. For example, they conclude that had the bottom third of minority students at Berkeley who hoped to graduate with a STEM major gone to Santa Cruz instead, they would have been almost twice as likely to earn such a degree.

"Prior to California's ban on affirmative action," Peter Arcidiacono, one of the study's authors, told me, "what Berkeley did well was switch relatively ill-prepared minority students out of the sciences and into majors where credentials are relatively less important."

Soon the Supreme Court will decide Fisher v. University of Texas. The case is complex, but essentially boils down to a young woman's claim that U.T. violated the Constitution's equal protection clause by denying her admission because she is white. The justices, save perhaps Clarence Thomas, are unlikely to address mismatch in their opinions. But the court could force schools to be more transparent about the racial preferences they use in admissions, and even to track the consequences for their students. For now, social scientists debate what can be gleaned from flawed data sets. They continue to argue over whether mismatch even exists and the extent of the harm it causes if it does. This raises another question: Do some of the more concrete if intangible benefits of affirmative action, like prestige and the superior connections one makes attending a fancier school, outweigh the potential cost? If affirmative-action admits are less likely to pass the bar after going to a certain type of school, or less likely to follow through with their field of choice, then the cost is potentially considerable. But are we really going to tell a kid embarking on adult life that he's better off attending a less prestigious school?

"The real question is what we want affirmative action to achieve," says Richard Brooks, a law professor at Yale. "Are we trying to maximize diversity? Engagement in the classroom? Whatever it is, I don't think the purpose of affirmative action is for everyone to have average grades." Mr. Brooks believes that mismatch exists. But he rejects the idea that it's as insidious as others claim and says that some mismatch might even be a good thing. Striving alongside people more capable than we are is a key ingredient for growth of all kinds.

In the Fisher case, Messrs. Brooks, Berk, and Ho signed a friend-of-the-court brief disputing a brief on mismatch that was co-written by Mr. Sander and the legal journalist Stuart Taylor Jr. Messrs. Sander and Taylor also wrote a book together that was published last fall, Mismatch: How Affirmative Action Hurts Students It's Intended to Help, and Why Universities Won't Admit it.

"Mismatch angers affirmative-action supporters because it quantifies a downside without weighing it against potential upsides," says Theodore Eisenberg, a law professor at Cornell, "such as the benefit of a diverse classroom, and the reality that some people who do attend better schools because of affirmative action are more successful in life as a result and help other minorities thrive." Mr. Eisenberg is on the board of The Journal of Empirical Legal Studies, which in June is publishing a paper by E. Douglass Williams, an economist at Sewanee: The University of the South, that appears to corroborate Mr. Sander's "first-choice/second-choice" analysis of the B.P.S. data.

In essence, affirmative action is about how to fairly distribute opportunity after our long history of racial discrimination. Whether it "works" is as much an issue for school administrators as for policy makers. That is, before we tell a student to choose School B over School A, it's worth asking what schools can do to improve the experiences of students, particularly those pursuing STEM majors, who arrive less well prepared.

The upside of affirmative action might be harder to quantify. But part of the problem with the current affirmative action regime is how its supporters define the goal, what the Supreme Court calls the "compelling state interest": classroom diversity. Meanwhile, little regard is given to the actual forms of adversity that disadvantaged students of all races must overcome. If affirmative action continues — either until Justice O'Connor's 2028 horizon or beyond — then the results from California and the Bar Passage Study suggest it's worth a closer, numbers-based look at the consequences, for everyone.


Dan Slater is a lawyer and the author of Love in the Time of Algorithms: What Technology Does to Meeting and Mating.