Matters of Mismatch: The Debate Over Affirmative Action's Effectiveness

The debate over affirmative action and race preferences in university admissions is among the most bitter and intractable in American politics. In principle, affirmative action is a straightforward case of compensatory justice. Black and Hispanic students, supporters claim, suffer from structural injustices that may cause them to perform worse than they would if they were white. The unfair leg up that white students get from societal forces, supporters reason, is best overcome by giving minority students an advantage in university admissions. Critics of affirmative action argue that such injustice does not exist, or that universities’ race preferences add another injustice that does not fix historical inequities. Racist laws, they claim, are more than a half-century in the past, and race preferences punish white students for a state of affairs that they had no hand in forming.
This debate, touching on fraught issues of fairness, justice, history, and race relations, is extraordinarily vexed. But both camps accept a tacit assumption that affirmative action actually helps minority students who are able to attend more prestigious universities with their peers from other backgrounds. Their principal argument is whether this help justifies the potential effects caused to students who do not benefit from race preferences.
A controversial theory by University of California, Los Angeles, School of Law professor Richard Sander challenges this assumption. Sander’s interest in affirmative action began in the early 1990s, when he began working on the UCLA Law admissions committee. Sander suspected that affirmative action admitted black and Hispanic students with weaker preparation than white students, but he was surprised at the size of the disparity. Affirmative action programs at the most competitive law schools appeared to siphon off the best black and Hispanic applicants to Harvard and Yale, leaving UCLA with an entering class of black students whose LSAT scores and undergraduate grades fell far below the average of UCLA’s white students.
Would this disparity harm the black students once they matriculated at these elite schools? Yes it did, Sander argued in a provocative 117-page article published in the Stanford Law Review in 2004, “A Systemic Analysis of Affirmative Action in American Law Schools.” In the article, Sander claimed that he sympathized with the moral arguments for affirmative action, not least because he had a biracial son, and “favor[ed] race-conscious strategies in principle, if they can be pragmatically justified.” Such a pragmatic justification, he concluded, was lacking; in fact, affirmative action was actively harmful to its intended beneficiaries.
The Problem with Affirmative Action
Sander’s hypothesis was simple: affirmative action admits black students (and other minorities, though his 2004 article considered only black and white students) to law schools with more rigorous programs than those for which they are equipped, in which they must compete with better-prepared peers. Instead of making up lost ground, Sander guessed, they fall to the bottom of the class ranks, and emerge more poorly prepared for a career than if they had gone to a university for which they were better qualified.
Sander’s original 2004 paper used data collected from the students who matriculated in 1991 at 163 American law schools, including several elite schools. Forty-one point four percent of the first-year black students and 42.5 percent of the graduated students fell in the lowest tenth of their classes by GPA; at elite schools, the first-year figure was 51.6 percent. This was not the result of anti-black bias in law school, Sander claimed: the statistical relationship between black students’ performance and their academic indices—weighted sums of undergraduate GPA and LSAT scores used for law school admissions—was the same as for white students. Sander found similar trends for passage of the bar exam. About 88 percent of law school graduates passed the bar exam on their first attempt, but only 61.4 percent of black graduates did so. But the bar did not seem to be racially biased, either, since its results paralleled differences in the law school GPAs of black and white students.
Sander’s most controversial finding, however, came when he compared law school graduates’ bar-passage rates to their academic indices. Black students failed the bar far more often than white students with similar academic indices. This, Sander proposed, actually harmed their career prospects. Sander cited surveys of young black attorneys to argue that law firms considered applicants’ GPA more carefully than the selectivity of the school which they attended as undergraduates, and proposed that attending a more prestigious school at which they would receive worse grades may actually hurt law students’ careers.
Even in 2004, this “mismatch” hypothesis—that affirmative action harms students by admitting them to schools with more rigorous curricula than they are prepared for—was not altogether new, and Sander’s 2004 article cites several earlier researchers on issues of race and education who made similar arguments. Black conservative economist Thomas Sowell, for example, had expressed similar ideas in a 1974 essay. Dartmouth psychology professor Rogers Elliott had found that black Ivy League students were less likely to stick with majors in scientific fields than white and Asian students, but that this discrepancy was accounted for not by their pre-college preparation per se, but rather by their preparation relative to the other students at their college.
Pushing Back on the Mismatch Hypothesis 
Sander’s views have often embroiled him in controversy. In 2007, a number of black students at UCLA Law protested being assigned to his sections, claiming that they felt unwelcome and discriminated against. Sander disputes these charges, and has pointed out that black students did better on average in his classes than they did elsewhere.
His conclusions have also been challenged. For example, in a review of Sander and journalist Stuart Taylor, Jr.’s 2012 book Mismatch: How Affirmative Action Hurts Students It’s Intended to Help, and Why Universities Won’t Admit It for the Los Angeles Review of Books, and in a later article cowritten with University of Iowa law professor Angela Onwuachi-Willig in the Texas Law Review, William Kidder, the assistant executive vice chancellor of University of California, Riverside, criticized several of Sander’s methodological choices, including combining data on law school admissions rates from 1991 with bar-passage rates from a decade afterward. Kidder quoted several previous papers that argued against Sander’s dramatic conclusion that without affirmative action, there would be more black lawyers. Sander’s statistical analysis, Kidder claimed, ignored that many black law students under affirmative action would simply not attend law school at all without it.
Proponents of race preferences could also quite reasonably object that Sander’s theory misses one distinct benefit and purpose of affirmative action. Minorities who attend an elite college or law school, even if they do not receive a better education than they would otherwise, do receive access to elite networking opportunities and a community of academic-minded individuals who often provide validation and affirmation for the intellectual endeavors of these students. And in many cases, a diploma with a prestigious name on it does in fact open doors for post-graduate employment. Many of these benefits cannot be measured quantitatively or reduced to the discreteness of test scores and percentages.
Despite all of this, Sander has argued that the benefits of attending a prestigious law school, if not nonexistent, are at least far less than is commonly believed. Mismatch claims that top law firms draw from a much larger array of law schools than they did 50 years ago, reducing the benefit of attending Harvard or Yale over less competitive law schools. In a paper cowritten with Jane Bambauer of the law school of the University of Arizona, he argued that the effect of law schools’ rankings on their white graduates’ careers is far weaker than the effect of the students’ grades. Sander and Bambauer concluded that “there is little empirical basis for the overwhelming importance students assign to ‘eliteness’ in choosing a law school.”
An Undergraduate Mismatch?
Understanding undergraduate education is more complicated, partly because college graduates pursue much more diverse careers than law school graduates and because of a paucity of data on graduates’ job performance. Several studies have shown that if two high-school students have the same high-school grades and SAT scores, the one who attends a more selective college can expect a more lucrative career. This suggests that moving a student from a less selective to a more selective college could improve her lifetime earnings.
A 2011 study from the National Bureau of Economic Research by Stacy Dale and Alan Krueger, however, casts doubt on this finding. Students who want to attend very selective colleges, they reasoned, differ in other ways from students with the same grades and SAT scores with lesser ambitions. They may be more ambitious or unafraid of challenges, traits that could help them in the job market. To try to adjust for this, Dale and Krueger controlled not only for students’ high-school grades and SAT scores but also for the list of colleges to which they applied, and they found that the apparent financial returns to college selectivity vanished. Two high-school students with the same qualifications who both applied to Harvard and Pennsylvania State University, for example, could expect the same salary upon graduating. Just as remarkably, however, there were a few types of students for whom this did not hold. For black, Hispanic, and first-generation students, college selectivity correlated with increased earnings even after applying these statistical controls. Dale and Krueger hypothesize that this is because such students benefit more from improved networking opportunities as well as assimilation to professional-class culture.
The Dale and Krueger study is imperfect. The universities that they studied were all at least somewhat selective, and the students who apply to Harvard and Penn State and ultimately attend Harvard can differ in other unobserved characteristics. Sander told the HPR that he finds the study interesting and well done, but emphasizes that such results on the “indirect” effects of academic mismatch on the job market do not contradict his findings on the “direct” effects on the quality of students’ education. He further hypothesizes that Dale and Krueger’s results for black and Hispanic students could still be the result of affirmative action.
Sander’s mismatch theory raises several considerations for personal decision-making. When this article goes to press, several million high-school students who have received their college acceptance letters will be deciding where to attend. Many of them, for one of many reasons not necessarily related to race, have been accepted to schools where they would be less prepared than the average student. To tell such students that they should not attend the college of their dreams, and instead attend one at which they would be better matched, may feel monstrous. But if Sander is right, such an approach could offer better odds of academic success and a remunerative career
Image Source: Flickr/jojolae