CHAPTER EIGHTEEN

CONCLUSION

WE THINK THAT WE HAVE DEMONSTRATED that the present system of racial admissions preferences has grave problems and has shown a remarkable incapacity to heal itself. Given the established inability of our elected officials to confront mismatch and other problems with preferences as well as the sometimes perverse unintended consequences of ballot measures such as Proposition 209 in California, the US Supreme Court seems to be the only hope for serious and stable reform.

But the Supreme Court decisions that have purported to put limits on racial preferences in admissions, in 1978 and 2003, have, as we have seen, been sadly ineffective. The question this raises is whether, as reasonable critics of racial preferences argue, the only alternative to the rotten system we have now is immediate, total abolition—that is, a Supreme Court ruling that the Constitution and civil rights laws simply prohibit any consideration of race in college admissions.

We think that immediate abolition might go too far, unless it turns out that there is no other way expeditiously to escape the very bad status quo. More important, five of the Supreme Court’s current justices have said that they are not prepared to abolish racial preferences. We suspect that these justices’ main reason, like ours, is that abolition might produce an alarmingly sudden and drastic plunge in black and Hispanic presence at our selective schools. In addition, such a ruling would be a sharp and sudden break from a substantial line of Supreme Court constitutional precedents, a break that should be entertained only as a last resort, after more modest adjustments of legal doctrine have failed. We also think that the Supreme Court can do much better than it has in the past.

But what is the right path to reform? Traditionally, moderate critics of the racial preference system have found themselves facing the dilemma that economist Glenn Loury eloquently captured when, in 1996, he wrote about the merits of Prop 209:

I do not think the color-blind absolutism which undergirds advocacy for Prop 209 is a logically coherent or morally correct position. . . . I think the proposition is too extreme. Race is a reality in this society; its social meaning is so powerful that sometimes the state cannot discharge its essential responsibilities, like educating the young, or maintaining order in the cities, without taking race into account. . . . [But while] I am genuinely torn about this . . . the bottom line is that I would probably vote for the initiative [if I were a Californian].

For all of its problems, Loury said, a complete ban on racial preferences might be preferable to “the maintenance of a corrupt and excessive status quo.” He explained:

Affirmative action desperately needs to be reformed, [but] the political realities are such that significant change is not likely to come from the bureaucrats who develop and administer such programs, absent dramatic, outside intervention of the sort which CCRI represents. . . . For many years now reasoned criticism of particular affirmative action programs—like an admissions policy at UC Berkeley which results in a three hundred point gap in SAT scores, or a race-based procurement policy which drives white men out of certain lines of business—have been met not with reasoned responses, but with hysterical charges that the critics are racists.

The choice Proposition 209 presented to Loury and other thinkers was all or nothing: Ban racial preferences or see them perpetuated and extended indefinitely. But the Supreme Court has more flexible tools for engineering serious reform short of an immediate ban. And we have some specific proposals for how the justices might use those tools to resolve the dilemmas racial preferences pose in a way that, though far from perfect, would be a vast improvement on the status quo. In this final chapter we shall briefly review what we have learned, outline our reform proposals, and explain why the pending Supreme Court case of Fisher v. University of Texas provides an opportunity for the Court to start us down this better path.

We have suggested that there are really three reasonable sides to the debate over racial preferences: the principled argument for preferences, the principled argument against them, and the so-far-less-visible empirical assessment of the preference system, which in our view is critical as we look to the future. The two arguments based on principle are often expressed in stark, one-dimensional terms. But suppose we found an honest insider on each side of the debate. Consider what they might say:

The defender of racial preferences: “It is true that affirmative action programs are flawed. But consider what they have accomplished. In the space of two generations we have gone from a society of glaring racial castes to one with a genuinely integrated leadership class. America has changed radically, and its ability to change peacefully and to operate effectively as a multiracial society without paralyzing racial conflict is an extraordinary achievement, both by historical standards and by comparison with almost any country in the world. We can never know how important affirmative action has been to this progress, but even if it was only partly responsible, that by itself justifies our efforts.

“Affirmative action today keeps open to minorities a vital path of access. Without preference programs to integrate our elite universities, many fewer minorities will end up in the management and leadership ranks of other social institutions. That is why corporate and military leaders join university and civil rights leaders in virtually unanimous support for current programs.

“Critics complain about the undesirable aura of secrecy around preference programs. But the American public has never supported the idea of large racial preferences, and the signs of its hostility have only grown over time. This is unfortunate but not by itself a reason to eliminate preferences. Does the average voter understand the severity of the racial disparities in academic preparation? Do they realize how the tenuous level of integration at top schools would plunge if we eliminated preferences? The Supreme Court’s rulings have essentially told us to dissemble. So we have dissembled, and the nation has been better for it.

“Moreover, we have shown a capacity for reform. As researchers have demonstrated an absence of socioeconomic diversity at elite colleges, at least some of these schools have made themselves more accessible. Harvard, Yale, Princeton, and Stanford are now essentially tuition-free for anyone with an income below $60,000, and Columbia has sharply increased its actual proportion of working-class and lower-income students in recent years. But given our understandable fear of being demagogued on our well-intentioned use of preferences, the path to reform is to work constructively on the inside, not to tear things down from the outside.”

We see merit to this argument. But many critics of racial preferences argue with equal eloquence in support of pure and simple abolition.

The critical view of racial preferences: “Affirmative action—a misleading euphemism for racial preferences—may have been justified in the late 1960s, when it arguably helped to knock down barriers and open doors that needed to be opened, both as a matter of basic fairness and as a safety valve at a time when racial tensions in American society seemed building to some sort of explosion. But it was always no more than a highly imperfect fix to a challenging social problem; by now racial preferences have become more a fig leaf for bureaucrats than an enlightened way of opening the doors of opportunity for students. Hidden preferences enable college presidents to pose as leaders of integrated communities that, just under the surface, are actually segregated. Far from breeding tolerance, preferences breed resentment and negative stereotypes among Asians and whites, who feel that they are being discriminated against—and they are. Far from breeding a sense of access and social equality among blacks and Hispanics, preferences put these students in environments where they will struggle academically, and this generates more resentment, a perception of discrimination, and a sense of racial grievance and entitlement.

“The real revolution that occurred in the 1960s was an embrace of equality across racial lines and a society in which advancement is based on objective merit, not skin color. Scholars have provided strong reasons to doubt that racial preferences played much part in the advance of many blacks and Hispanics into our leadership classes, including the large percentage of these leaders who do not come from elite universities. There is a good reason why most Americans oppose racial preferences; they directly undercut the post-1960s social contract. They elevate the importance of skin color at a time when young Americans increasingly see that as an irrelevant distinction. They increasingly single out Asian Americans—a group that has suffered historical discrimination and includes millions of struggling individuals who do not fit the ‘model minority’ stereotype—for particularly invidious discrimination, to the point that “the ‘Asian penalty’” in university admissions has become a term of art among experts. They systematically and enormously favor affluent blacks and Hispanics over working-class Asians and whites. They encourage students with mixed racial backgrounds to present themselves as being part of a group that will maximize their admissions preference, and they provide incentives for every black and Hispanic student to play up, as much as possible, any past experience that can validate her status as a victim deserving compensation.

“Racial preference advocates argue that without preferences, some universities will only have a ‘token’ black presence, but they seem to have forgotten what ‘tokenism’ means. Placing students into classrooms where they will struggle to compete, all for the purpose of demonstrating a university’s open-mindedness and making sure that the supposed ‘point of view’ of every racial group is heard in the classroom is the quintessence of tokenism as well as racial stereotyping and essentialism. Preferences stigmatize the recipients in the eyes of classmates, teachers, and themselves, and they throw into perpetual doubt the value of the degrees the recipients have earned. Far from preventing tokenism, racial preferences foster a lifetime of it.

“The racial preference regime also creates institutional pressures to discriminate in grading and make grading systems and graduation standards less rigorous. It shifts the preeminent mission of universities from the pursuit of excellence to the pursuit of diversity. It fosters cynical games of campus racial politics. And it distracts us all from the real problem that we must address: the racial gaps in K-12 academic achievement that tremendously handicap Hispanics and, especially, blacks.”

What light has our empirical journey shed on this debate? On the whole, it has tended to seriously undercut many of the racial preference proponents’ claims. There is no serious dispute that large preferences greatly undermine paths to success for blacks in STEM fields and in academia. They substantially increase the risk that black and Hispanic law students will fail bar exams. The University of California experience under Prop 209 strongly suggests that shrinking the largest preferences improves black and Hispanic graduation rates. To the extent we can get at the ultimate question of whether large preferences generate learning mismatch, the evidence is limited but growing and, thus far, entirely consistent with the hypothesis that it does.

The general claim that boosting blacks and Hispanics up to more elite institutions is essential for their long-term success relies on outdated assumptions and falls apart on close examination. Perhaps our best data on this issue comes from the field of law. Fifty years ago elite law firms hired overwhelmingly from elite law schools, so it made sense to think that the best way to integrate these firms racially was to place minorities in elite schools. But hiring patterns have changed radically since then and are now driven by a search for cognitive horsepower and success in law school. Large preferences have become an obstacle rather than a catalyst toward top positions in the law. Many methodologically sound studies of postcollege careers, like that of Loury and Garman, strongly suggest that the same thing is true for undergraduates: Performance trumps elite credentials.

Empirical research has also undermined the “social cohesion” arguments for racial preferences. The “warming effect” at the University of California suggests that many if not most black and Hispanic students are eager to avoid the stigma of racial preferences. Large academic gaps among classmates of different races undermine cross-racial friendships and cluster blacks in particular in non-STEM fields and near the bottom of the class. The available evidence suggests that preferences aggravate rather than ameliorate minority self-image problems like stereotype threat, and the failure of the noisy field of diversity research to tackle the obvious issue of how preferences affect stereotyping speaks volumes.

Our review of institutional behavior in Part III identifies plenty of malfeasance and powerful institutional mechanisms that block meaningful efforts to address the mismatch problem. And Chapters Sixteen and Seventeen show how far we have to go—and how much potential for progress there is—on both K-12 reform and the improvement of socioeconomic opportunity in higher education.

Nearly all of this evidence weighs in on the side of the abolitionists. Indeed, it is this growing body of evidence that has caused the present authors to slowly drift over the past twenty-five years toward greater sympathy for the abolitionists. And from a purely legal standpoint, one can imagine that if the Supreme Court were faced with a case that squarely posed the mismatch problem, its own jurisprudence on preferences (“narrow tailoring . . . requires that a race-conscious admissions program not unduly harm members of any racial group”) might compel a decision prohibiting preferences in the absence of clear evidence that they are helping rather than hurting recipients.

Yet abolitionism has some serious difficulties. The three most serious are that (1) ending racial preferences (were that possible) would lead to drastic declines—probably more drastic than in the unique circumstances of the University of California after proposition 209—in black, Native American, and, to some extent, Hispanic enrollments at scores of the nation’s top universities; (2) because of the peculiar workings of the cascade effect, the case against preferences is weakest at the most elite schools; and (3) because of universities’ determination to circumvent any ban, outlawing preferences would not end them but rather would lead—and has led—to universities evading bans, thus possibly making mismatch worse, not better.

First, in a hypothetical world totally purged of racial preferences, the proportion of blacks at the most elite universities and professional schools could fall dramatically. Our “cascade” simulation in Chapter Two implies that black enrollment at the most elite schools could, if they admit students predominantly on academic criteria, fall as low as 1 percent. (This would be true, for example, at law schools.) Other simulations that take into account athletic preferences or the current mild socioeconomic preferences predict smaller but still rather dramatic declines for blacks (drops of 50 percent or more from current levels) as well as smaller but still substantial declines for Hispanics (drops of 30 percent or more from current levels).

Second, due to the peculiar workings of the cascade effect, explained in Chapter Two, the racial gaps in academic credentials—and thus race-related manifestations of mismatch—are smaller at the very top universities than at institutions even a little below them on the eliteness scale. Most or all of the black and Hispanic students admitted to these top schools are so smart and able that a substantial minority might be admitted without preferences, and the rest with preferences smaller than those that less elite schools use. In addition, the institutional resources to help any struggling students succeed are abundant. Another advantage is that preferentially admitted students at these top schools will get to know many classmates who will be future leaders. After graduation, the reputational and networking benefits of a super-elite degree are in some ways unique. So all graduates of these very top institutions have good odds of finding niches where they will thrive.

For all these reasons, a black or Hispanic student receiving a preference to Harvard or Yale is probably a good deal less likely than her counterparts at less elite schools to end up in the bottom tenth of her class—and even if she does, she will probably do just fine, or much better than fine, after graduation. In short, abolishing racial preferences would reduce the black and Hispanic presence most dramatically at the very schools where the effect is most benign.

Third, the post-209 experience at the University of California (UC), like the somewhat similar post-Hopwood experience at the University of Texas, illustrates vividly how hard it is to enforce an absolute ban on racial preferences and how damaging universities’ evasive tactics can be. Where university leaders are ideologically convinced that the ban is a bad thing, where important political actors continue to push for higher minority headcounts, and where minority students feel aggrieved, the pressures for ill-conceived workarounds and outright (if usually covert) violation of the law become very strong and often irresistible.

One of us (Taylor) thinks that some evasion would have its attractions if the effect were to make preferences smaller without making them disappear and, thus, to avoid the drastic drop in minority presence in the top schools that we just discussed. But the lesson of experience is that some of the ostensibly race-neutral proxies for racial preferences have brought in students who encounter even greater mismatch problems than under the racial-preference regime. Such has proven to be the case at both the University of California and the University of Texas. After Prop 209, UC greatly favored students attending the weakest high schools in the state, gave large weight to other proxies linked to race, and, at professional schools, engaged in blatant evasions of the law. These measures blunted the effect of Prop 209 in reducing mismatch, and as greater evasions have returned, mismatch will foreseeably be as serious a problem at the UCs as it was before Prop 209. After the 1996 Hopwood decision discussed in Chapter Seven, the Texas legislature automatically admitted to its flagship university the top 10 percent of students from every high school in the state, including many who were at greater risk of academic mismatch than recipients of racial preferences had been.

If a ban on racial preferences were broad enough to punish evasion and block use of such proxies, with a new enforcement agency that aggressively monitored admissions practices, then racial preferences really would fade away. But we know of no initiative that has included such features and cannot imagine any court taking such steps.

The main tool now available to enforce a legal ban on preferences—private lawsuits by disappointed applicants—is very ineffective, as we discussed in Chapter Thirteen. Few college applicants want to start their adult lives by becoming plaintiffs against universities. If they do, there is a good chance they will graduate from some other college long before achieving any relief and see their lawsuits dismissed for mootness or lack of standing. Proving that schools are using disguised racial preferences is difficult and complex. Judges are wary of overturning the practices of highly respected universities.

Is there another way? We think there is. Below we outline three reforms, each of which would address a serious problem in the current system; together, they could set in motion a virtuous cycle of further improvements and reforms. We then discuss the pending Supreme Court case, Fisher v. University of Texas, that could advance these reforms. We do not think that our mix of recommendations is well tailored as a basis for the Court’s decision in the current Texas case, given the importance of judicial self-restraint and the fact that the sole remaining plaintiff, Abigail Fisher, has not raised the mismatch problem at all during the case and has not asked for anything remotely like the remedies that we propose. But we believe that the Court could—and we hope that it will—point the way toward these or similar reforms.

REFORM ONE: THE POWER OF TRANSPARENCY

Nearly a decade ago the late, legendary liberal Senator Ted Kennedy proposed that universities be required to disclose the size and extent of any legacy preferences they use in admissions. He argued that legacy preferences were unfair throwbacks to an earlier time. If universities had to reveal the economic status and race of those admitted through legacy preferences, the sunlight shone on these practices would encourage schools to end them or to at least reconsider their policies.

Kennedy’s bill did not pass, but we think (and one of us wrote at the time) that he was on the right track. The single-most important step toward reforming affirmative action is to adopt a comprehensive system of disclosure. The Supreme Court could mandate that any university that wishes to take students’ race into account in admissions must do so in a way that makes both the use of preferences and their consequences transparent to applicants and the public. This would have three specific components:

 

 

First, each school that uses racial preferences should so state in its admissions materials, should disclose the size of the preferences it uses, and should provide each admitted student with information about the academic outcomes of past enrollees with comparable entering credentials. The data disclosed should be sufficiently aggregated to protect the privacy of past enrollees but detailed enough to allow interested observers to discern and report on the weight given in admissions to race and other relevant factors (including, per Kennedy, legacy status). Admitted students should also be given broader data disclosing the overall process of creating the student body, its characteristics, and its outcomes.

 

 

Second, schools should provide any information they have available or can reasonably obtain on learning outcomes for past students similar to the admitted student. For example, a student admitted to a college with a given SAT score and high school GPA should receive its best estimate of the past graduation rates of comparable students, their college GPAs, and their rates of attrition from intended majors. Ideally, schools should also be encouraged (perhaps required as part of the accreditation process) to gather and make available data on students’ postcollege outcomes: What proportion of past students similar to the applicant ended up attending graduate school? What proportion were employed five years after graduation, and what were the median earnings of those graduates? For law schools this would include data on bar passage rates or, preferably, bar scores. For medical schools this would include data on the outcomes of national boards and eventual rates of licensing in chosen fields. For colleges this would include scores on exams taken in anticipation of graduate school (e.g., GRE scores) and any results from participation in exams like the Collegiate Learning Assessment.

 

 

Third, these data on outcomes as well as on the size of all admissions preferences should be made publicly available so that researchers, legislators, the media, and all other citizens can evaluate the accuracy and completeness of the information provided to applicants.

Universities are well equipped to gather such data through mechanisms including national networks that create samples of alumni to track outcomes. If any complain that these duties are burdensome, the answer is that, for the sake of their own students, they should have been gathering and analyzing such data for many years.

Transparency would create many positive-feedback cycles, in contrast to the perverse negative-feedback cycles racial preferences create, such as the cascade effect and the campus dynamics of unequal performance. Consider the following:

Disclosure would empower minority students to make their own assessments of the mismatch risks of enrolling in a particular school based on a preference of known size. Networks would quickly spring up to promote discussions among applicants, their parents, current students, graduates, high school guidance counselors, and others. These conversations would be far deeper and more useful than today’s choice-of-school conversations, which are dominated by the simple rankings generated by U.S. News and its kin.

The universities themselves would be drawn into these conversations. No longer able to deny the fact or scale of preferences, administrators would suddenly find themselves accountable for the preferences that they use and also for the struggles of students whose academic problems are foreseeable based on their incoming grades and scores. More importantly, the competition for students would no longer be limited to glossy brochures and scholarship offers but rather actual outcomes. Schools would have very substantial incentives to minimize these mismatch problems or risk being shunned by many high school seniors.

With results exposed to sunlight, some colleges and universities would probably choose to reduce preferences significantly. Others would try to develop more effective academic support programs. One way or another, mismatch effects would likely be addressed and steadily eroded.

The most likely objection to a transparency requirement is that it would, by exposing the poor outcomes that often result from preferences, undercut the academic confidence of vulnerable students. But, as documented in Chapter Six, students’ grades tend to reflect their preparation levels relative to their classmates, not their expectations when they enroll. Administrators’ assurances that every admitted student is equipped to flourish are not true; they only mislead the students and their parents, often to the students’ detriment. Cole and Barber have richly documented the tremendous loss of academic self-confidence that disproportionately plagues students receiving large preferences. In a more transparent regime students will have a better sense of the odds they are up against and be able to make more realistic and, certainly, more informed choices.

Happily, a transparency mandate is already implicit in the Supreme Court’s rulings on racial preferences. Grutter, along with the logic of a more recent ruling on K-12 schools, require universities to prove that their consideration of race in admissions is “narrowly tailored” to promote a healthy diversity. Among other things, “narrow tailoring . . . requires that a race-conscious admissions program not unduly harm any racial group” (as noted above) and that schools must engage in “serious, good faith consideration of workable race-neutral alternatives” to minimize such harm.

As both logic and experience under Grutter have shown, these narrow-tailoring requirements are largely meaningless without full disclosure of the operation and effects of preferences. How can anyone know which schools have considered or incorporated race-neutral alternatives without better knowledge of what they are individually doing? How can one minimize the harms of racial preferences to the minorities that receive them without providing information that allows prospective students to understand their likely outcomes at schools that do and do not use racial preferences? How can unsuccessful applicants assess whether they might have a valid discrimination claim worth pursuing? And how can courts monitor and enforce compliance with the constitutional narrow-tailoring requirement? In short, secret admissions can’t possibly be narrow tailoring.

To be sure, as we described in Chapter Thirteen, the Gratz decision and Justice O’Connor’s opinion for the Grutter majority created such strong incentives for universities to keep their veils of secrecy that it seems likely that was the way she wanted it. But only majority holdings—and certainly not any justice’s private purposes—are relevant to future cases. Although no decision has ever squarely addressed the question of whether narrow tailoring requires transparency, no precedent stands in the way of imposing such a requirement, and the logic of the most relevant precedents argues strongly in favor of doing so.

REFORM TWO: TARGETING ECONOMIC NEED BEFORE RACIAL IDENTITY

Despite the Supreme Court’s explicit assertion that racial preferences must not unduly harm members of any racial group—and the fact that undue harm is a function of how large the preferences are—the Court has never imposed a specific size limit. And it’s easy to imagine that the difficulty of determining in any principled way how large is too large may have deterred the Court from doing so.

One solution to this dilemma—which would also have the highly beneficial side effect of returning affirmative action to its original social-mobility goal—would be to require that the racial preferences a university uses be no larger than the average size of preferences based on an individual applicant’s financial need or socioeconomic status (SES). As Chapter Sixteen demonstrates, socioeconomic diversity is every bit as compelling an interest as racial diversity, but most selective schools have grossly neglected the former.

Creating an “SES cap” on racial preferences would accomplish several goals: It would ensure that pursuit of campus diversity includes meaningful consideration of each applicant’s individual circumstances rather than just her skin color. It would ensure that universities actually pay attention to race-neutral ways of pursuing diversity and, thus, would inhibit unconstitutional racial balancing.

An SES cap on racial preferences would also foster simple justice to the economic have-nots who have been so sorely neglected by our selective universities. As discussed in Chapter Sixteen, racial minorities have much better access to our universities than do Asians and whites of modest means. Indeed, blacks are about 30 percent more likely to enter college than are whites with similar socioeconomic backgrounds and academic credentials. (They are less likely than comparable whites to graduate—in part, we are confident, because of mismatch.) In contrast, students from the bottom quarter of the SES spectrum are about 80 percent less likely to enter college than are students with similar academic credentials from the top fifth of the spectrum. This is a shocking state of affairs, especially at a time of rising concern about economic inequality, and our proposed remedy would help change it.

Note that socioeconomic criteria can (and should, we think) be cognizant of and capture the manifold ways that race and disadvantage intersect in our society. For example, blacks with family incomes of $40,000 are likely to live in significantly poorer neighborhoods than are whites at the same income level. Even blacks with higher incomes tend to have much less wealth than do whites with similar incomes. A careful SES measure can take into account such nuances.

How would one actually implement this cap on racial preferences? Peter Schuck and commentator Michael Kinsley have memorably warned that SES preferences would be complicated, messy, and prone to abuse. But Sander actually devised and implemented a sophisticated system of SES preferences at UCLA after the passage of Proposition 209, and the system worked exceedingly well. Audits against financial aid statements showed little abuse; the preferences substantially changed the social makeup of the class and never, to our knowledge, prompted complaints of unfairness. Compared, for instance, to the complexity of Supreme Court rules on racial gerrymandering, our proposed SES mandate would be straightforward.

The most significant challenge in this proposal, from an enforcement point of view, is the task of actually comparing the size of socioeconomic versus racial preferences. But as detailed in the endnotes, we think this is quite doable. This sort of reliance on the specific weight assigned to race (as well as to SES status) may seem to run afoul of the Supreme Court’s decision in Gratz to ban any admissions systems that assign a specific number of bonus points to each preferred-minority applicant. But we think—and we suspect that at least five justices think—that the Gratz rule has proven unworkable. As Chapter Thirteen shows, Gratz was based on false assumptions about how college admissions works, as all admissions decisions involve implicit assignments of relative weights to the various criteria, including race, even if only in the subconscious of the decision maker. The Gratz ban on making that weighting process explicit perversely makes narrow tailoring all but impossible.

REFORM THREE: OUTLAWING RACE-BASED AID AWARDS

Our third proposed reform would reinforce the effectiveness of the second and is eminently warranted in its own right: The Court should, in a proper case, prohibit state schools from using racial preferences in awarding financial aid and scholarships. Such awards produce little if any diversity benefit because they do not increase by much, if at all, the number of preferred-minority university students. Rather, the main function of race-based financial aid is to fuel zero-sum bidding wars among competing campuses for the limited supply of blacks with strong academic qualifications. Because a large proportion of those blacks are from well-off families, they need financial aid much less—but are much more likely to get it—than better-qualified, less well-off white and Asian students. Race-exclusive scholarships are already clearly unconstitutional, and if the Supreme Court takes the narrow-tailoring requirement at all seriously—as at least five of the current justices clearly do—it would strike down in a heartbeat all consideration of race in financial aid awards.

Our proposed reforms would reinforce one another. For example, one objection to creating an incentive for schools to make more use of SES preferences might be that this would just bring the mismatch effect in through another door. Certainly that is a danger, and we would not favor this step if it were not accompanied by a complete transparency that allows both students and outside experts to monitor whether the SES preferences are excessive and doing damage. But SES preferences used for their own sake (and not as an evasion of racial restrictions) are likely to be modest. As we noted in Chapter Sixteen, simply making current admissions processes more fair and better directing financial aid to those with real need would improve SES diversity significantly. Because of the large, untapped supply of academically strong, low-SES students, modest SES preferences can produce a dramatic increase in current levels of SES diversity. (Moreover, schools using very aggressive SES preferences would likely find the demands for financial aid and the demands on their own resources quite daunting, and this would also encourage moderation.)

The combined effect of our proposed reforms would thus be to give schools incentives both to increase their SES diversity and to avoid large preferences—and the mismatch they cause—while also requiring healthy transparency. Admitted applicants and their parents could assess for themselves the mismatch risks, if any, posed by this or that college. A very elite, very rich school that produces good outcomes for nearly all of its students could continue using substantial racial preferences as long as its SES preferences were just as substantial. And schools would have a new incentive to develop superb academic support programs to improve their appeal as places where preferred-minority students need not fear being left behind.

FISHER AND THE OPPORTUNITY FOR REFORM

As we finish this book, the Supreme Court is receiving briefs in Fisher v. University of Texas, its first case since Grutter and Gratz to deal with racial preferences in higher education. The Court has set the oral argument for October 10, 2012. It is expected to decide the case by the end of June 2013. The decision could be momentous, or it could be based on fairly narrow grounds. But for reasons we discuss below, we are hopeful that the Court’s decision will mark a notable and constructive break with the patterns we criticized in Chapter Thirteen.

Abigail Noel Fisher, of Sugar Land, Texas, applied in 2008 for freshman admission to the state’s flagship campus, the University of Texas at Austin (UT). Because she did not finish in the top 10 percent of her high school class, she did not qualify for automatic admission under the state’s “Top Ten Percent” law, which the legislature had adopted in 1998. Instead, she competed for admission against other in-state candidates for the 19 percent of seats in the next entering class that the top 10 percent admits and out-of-state students did not fill. Though her academic credentials exceeded those of many admitted black and Hispanic applicants, UT denied her application. Fisher filed suit in 2008, challenging UT’s use of race in undergraduate admissions as contrary to the Fourteenth Amendment guarantee of “equal protection of the laws,” as interpreted by Grutter and other precedents. She lost in federal district court and appealed. A three-judge panel of the Fifth Circuit Court of Appeals ruled unanimously against her. One member, Judge Emilio Garza, harshly criticized UT’s preferences and said the Supreme Court should overrule Grutter and strike them down. But Garza added that, as a lower court judge, he was bound to follow Grutter, which left him no choice but to uphold UT’s preferences. The full Fifth Circuit, sitting en banc, declined by 9–7 to rehear the panel’s decision, with the votes of the seven dissenters showing that they disagreed with Garza and saw UT’s preferences as contrary to Grutter.

Fisher then filed a petition for Supreme Court review, arguing both that the UT preferences were contrary to Grutter and that if the Court disagreed, it should overrule Grutter as suggested by Judge Garza. The justices announced in February 2012 that they would hear Fisher’s case. Court watchers could immediately see that—because of changes in the Court’s membership since Grutter—this might become the biggest case ever on racial preferences in university admissions.

Four of the nine current Justices—including Samuel Alito, whom President Bush appointed to Justice O’Connor’s seat when she retired—are on record saying that most or all state-sponsored racial preferences are unconstitutional. A fifth, Anthony Kennedy, strongly dissented in Grutter on the ground that the then-majority’s purported narrow-tailoring requirement was toothless. Many saw the Court’s decision to hear the Texas case as suggesting that Kennedy would provide the fifth vote to strike down UT’s racial preferences. To be sure, the justices sometimes surprise. And the state’s well-crafted brief, filed on August 6, 2012, was clearly written to persuade Kennedy that, even under the logic of his Grutter dissent, he should uphold the Texas preferences.

Critics see at least three characteristics of UT’s preference program as making it vulnerable even under Grutter. First, UT is one of the few elite universities that already had a facially race-neutral system in place for creating racial diversity—the top 10 percent law, which has given the university a higher percentage of black and Hispanic students than most state schools—before it added its current racial preferences to that system. (There have been modifications since 2008, when Fisher applied.) Because some Texas high schools are heavily black or Hispanic, the top 10 percent system generates much more diversity than the typical college admissions system. In 2008, for example, nearly a quarter of UT’s entering students were black or Hispanic even without counting the relatively small additional number who were admitted via conventional direct racial preferences. Fisher’s lawyers and others argue that there can be no constitutional justification for using racial preferences to produce at most a marginal increase in diversity at a school that is already unusually diverse.

Second, UT has explicitly set a goal of moving toward enrolling a student body with roughly the same proportion of each racial group as the state’s population. But this seems hard to distinguish from the “racial balancing” that Grutter declared to be flatly unconstitutional. And third, UT has also set a goal of using racial preferences until there are black and Hispanic students in every classroom. But that could take many, many decades because the reason so many tough classes now have no or few black or Hispanic students is that students with relatively weak preparation avoid such classes.

And the racial gaps in academic qualifications at UT are very large indeed. Among freshmen entering UT in 2009 who were admitted outside the top-ten-percent system, for example, Asians had a mean SAT score that would put them at the 93rd percentile nationwide; whites were at the 89th percentile, Hispanics at the 80th percentile, and blacks at the 52nd percentile. A staggering 467 points (on a 2,400-point scale) separated the mean SAT scores of Asians and blacks admitted with explicit preferences. (Note that the black-white gaps at Texas are similar to those estimated for Tier 3 or 4 schools in our cascade effect model from Figure 2.1, p. 24.)

We do not suggest that there is no coherent legal argument for UT’s preferences, but we doubt that there is an argument that Justice Kennedy will find persuasive. He has been a pragmatic skeptic of racial preferences; he has never declared that all such preferences are unconstitutional, but he has also never voted to uphold such a program. Instead, he has suggested that any racial preference program is unconstitutional unless it can pass some fairly rigorous tests, which no racial preference yet to come before Kennedy has ever passed.

We therefore suspect—and hope—that Justice Kennedy will be engaged in his review of UT’s practices in a process similar to our own more general review of racial preferences in higher education. He will not take an absolutist position either for or against preferences but instead will try to give substance and specificity to the Court’s standards of “strict scrutiny” and “narrow tailoring.” In the process of doing that, it would be both plausible and logical for him to articulate some type of transparency requirement and some limit on the size of racial preferences.

Indeed, Justice Kennedy could cite Justice Ruth Bader Ginsburg, perhaps the most liberal current justice, in support of transparency. Recall her assertion in her Gratz dissent that “if honesty is the best policy, surely [an] accurately described, fully disclosed . . . affirmative action program is preferable to achieving similar numbers through winks, nods, and disguises.”

Fisher does not directly pose the problem of mismatch. The plaintiff is a white woman who was not admitted and who made no argument that racial preferences harmed anyone except rejected applicants from nonpreferred groups like her own. But the mismatch issue lurks in the background. For example, part of the reason that UT’s preferences have not produced classroom diversity is, as noted above, the tendency of mismatched students to drop or avoid tough courses in which competing is particularly difficult. Simply increasing the aggressiveness of preferences is not a plausible way of solving this problem. And for the reasons we have detailed, preferences on the scale used by UT are almost certain to backfire on the students they purport to help.

The alternatives to a Supreme Court requirement of muscular reforms seem stark: either perpetuation of large and, thus, harmful racial admissions preferences into the far-distant future, or immediate abolition of any and all consideration of race. We think there is a better way. And there is reason to hope that the Court’s deliberations in Fisher—and the broader national deliberation that will accompany it—will start us down that better path.