The fight over race-sensitive admissions is usually framed as pitting fairness to white students against the societal benefits of diversity. Because that's a collision between rival world-views, not an empirical argument, it's nearly impossible to strike a balance that both sides can accept. But when opponents of affirmative action contend that race-conscious admissions rules are also unfair to the very minorities who are supposed to be being helped, they are making a different kind of argument—one that can be tested empirically.
The so-called "minority mismatch" argument goes like this: non-whites who wouldn't otherwise be admitted to selective universities find themselves competing with better-prepared students, and that's a prescription for failure. They would be better off at a school a rung or two lower on the prestige ladder, where the academic standards are less rigorous and expectations are lower. Affirmative action advocates dismiss this claim as patronizing—and plain wrong. Which side is correct? Can research affect the character of the persisting affirmative action debate?
In their landmark 2001 study, "The Shape of the River," William Bowen and Derek Bok, the former presidents of Princeton and Harvard, respectively, introduced empirical rigor to this debate. They found that minority students at elite universities who had gotten a leg up in admissions because of their race did considerably less well academically than white students—indeed, their performance was worse than what their SAT scores predicted. Score one for the "minority mismatch" side. However, once these students graduated, many of them joined the ranks of the elite. They went to graduate school; entered the top echelon of business, medicine and law; earned big bucks; and became pillars of their communities.
What's to be made of those findings? Bowen and Bok read the results as proof that affirmative action is good for these students; and because the Establishment is rendered more diverse, it's also good for the country. But the critics see things differently. To them, "The Shape of the River" confirms the fact that an Ivy League diploma is a ticket to success—and that means white students with better academic qualifications are truly disadvantaged if they are turned away.
A 2008 study of the "ten-percent" admissions policy in Texas, reported recently in Inside Higher Ed, mounts a stronger challenge to the anti-affirmative action position. The Texas policy works like this: High school students who graduate in the top tenth of their class are assured a place at one of the state's most competitive universities, the University of Texas and Texas A&M. Their SAT scores don't matter; and whether they come from a leafy Dallas neighborhood or a Brownsville barrio doesn't matter either. Eighty percent of the undergraduate slots at these two universities are filled in this way; the rest of the students—those promising quarterbacks, saxophone soloists and budding Shakespeareans—are picked by the admissions office.
Unlike affirmative action admissions policies, this plan is formally color-blind. But since many high schools in Texas are overwhelmingly Hispanic or black, the ten-percent rule has boosted minority enrollment; and because those students typically have lower SAT test scores, experience there affords a good test of the "mismatch" theory. If the theory is correct, then black and Latino students at the flagship campuses should do worse than if they had attended a less selective school. In fact the opposite is true—these students are 21 percent more likely to earn their bachelor's degree than are students with similar qualifications who opt to enroll in one of the less selective universities. The losers are minority students whose class rank isn't quite in the top ten percent, but who, because they graduated from top high schools, would previously have been admitted to one of the flagship universities—their graduation rates decline.
A natural experiment at the University of California bolsters these findings. UC admits a sizeable number of students who transfer from community colleges in their junior year. Despite the fact that many of them would never have made it as freshmen on the basis of their high school grades, they do as well academically as students with more stellar high school records who arrive as freshmen. What's more, the community college transfers take less time to graduate.
It seems that students—minority students in particular—rise or fall to the level of expectations. Although that conclusion might surprise the anti-affirmative action crowd, any professor knows that expectations matter. At Harvard or Princeton, flunking out is a rarity, and those on the verge of failing get kid-gloves treatment; the assumption is that if a student is good enough to be admitted, she is good enough to graduate. By contrast, at Big U—the mass universities where almost all comers get in—students are left largely to their own devices, and it is taken for granted that many of them will drop out or flunk out. While Big U academics will blanch at the thought, they are tacitly embracing Charles Murray's mistaken claim, recently propounded in his 2008 book "Real Education," that only a minority of undergraduates can actually do college-level work. The result, at Big U, is a self-fulfilling prophesy: high dropout rates, a waste of talent and a loss of human capital.
Big U—and Charles Murray, for that matter—could learn a lot from studying the Accelerated Schools model, developed a generation ago by Columbia Teachers College professor Henry Levin and widely used in public schools. The premise of Accelerated Schools is that "at-risk" students, many of them poor and non-white, can thrive if they are introduced to the brand of idea-filled instruction typically reserved for "gifted" children, instead of suffering through skill-and-drill teaching that supposedly is geared to their lower abilities. Evaluations have largely borne out this claim—higher expectations really can generate greater achievement.
The "stereotype threat" line of social psychology research on college students' academic performance, conducted by Stanford professor Claude Steele and others, points to a similar conclusion: Expectations can spell success or failure. Minority students are especially vulnerable—when they are exposed to negative stereotypes, their test scores plummet. "Conditions designed to make black subjects stereotype vulnerable," like telling minority students who are about to take a sample Graduate Record Exam that the GRE tests their ability, "depress their performance relative to white subjects," said Steele. The good news is that when such stereotyping is eliminated—when undergraduates are told they are taking a problem-solving test—not an ability test—black students do dramatically better.
These disparate findings—the Bowen-Bok study of elite universities, the University of Texas "ten-percent rule" research, the natural experiment in California, the positive results from Accelerated Schools, and the social psychologists' conclusions about the effect on minority students of exposure to "stereotype threats" and positive role models—tell a common story. Both positive and negative stereotypes affect how non-white students do on tests; change those expectations, and academic performance can change as well. These findings should put defenders of the "mismatch" theory on the defensive. What's more, if minority students actually do better in a high-expectation academic environment, then there's good reason to question the bedrock argument against affirmative action-that race-conscious admission policies are unfair to white students.
Goodbye ideology, hello evidence? That's too optimistic, for it's hard to persuade ideologues to confront inconvenient truths. Still, this research should embolden defenders of affirmative action—it shows that race-conscious admissions is not, as the critics insist, a sop to political correctness. It is smart human capital policy.
David L. Kirp, professor of public policy at the University of California, Berkeley, and a regular contributor to National CrossTalk, worked on education policy issues as a member of the presidential transition team.