Richard Sander is Professor of Law at UCLA and the co-author of Mismatch.
In 2003, the last time the Supreme Court took up the issue of racial preferences in higher education (the case of Grutter v. Bollinger), Justice Kennedy was in dissent. Unlike his more conservative colleagues, he did not completely reject the idea that racial preferences (in that case, preferences extended by the University of Michigan Law School) could pass constitutional muster. But in his view, Justice O’Connor (who wrote the deciding opinion in favor of the university) was giving the Law School far too much deference. While O’Connor held that a racial preference program must be “narrowly tailored” to achieve the university’s interest in diversity, and laid out a series of tests that narrow tailoring implied, she deferred to the university’s judgment and “good faith” to a remarkable degree, giving the university the benefit of the doubt, on one issue after another, that its racial preferences really were a temporary, last resort to achieve educational diversity. Kennedy argued that strict scrutiny really must be strict and exacting.
Today, Kennedy’s holding in Fisher v. University of Texas makes precisely this point. In vacating the Fifth Circuit’s holding for the university, Kennedy wrote that the university is not entitled to any deference in showing that its racial preference program is narrowly tailored to achieve the university’s interest in a diverse educational environment. Kennedy was able to command a broad majority in the case by holding that while the Court would not overrule Grutter, it would insist on a Grutter with teeth – a Grutter that takes its own standards for acceptable racial preferences seriously.
This is not the sweeping repudiation of racial preferences that many conservatives hoped for, and that the higher education establishment greatly feared. But it is a break with the Court’s earlier decisions on affirmative action in higher education. In these earlier cases (Bakke, Gratz, and Grutter), the Court struck down specific, rigid sorts of racial preferences (a racial quota in Bakke, a system that gave points for race in Gratz) but essentially winked to universities that racial admissions preferences would be tolerated if they were not overtly aggressive and inflexible. And after those earlier decisions, the actual size and scale of racial preferences in higher education changed very little. Justice Kennedy’s opinion invites – though certainly by itself does not achieve — a new era of much tighter regulation of preferences.
Kennedy’s opinion is a rebuke, though a gentle, subtle one, to Justice O’Connor’s Grutter holding. The Fifth Circuit relied directly on O’Connor’s opinion in explaining why it is giving so much deference to the University of Texas. In vacating the Fifth Circuit opinion, Kennedy was able to criticize O’Connor without mentioning her directly, but there is no mistaking that her widely criticized language in Grutter was his actual target. Justice Ginsburg, in her lone dissent to Kennedy’s opinion, fairly points out that the Fifth Circuit was following Grutter; Kennedy’s point, and one that he brought a broad cross-section of the Court to agree upon, was that the application of Grutter’s standards was too vague to meaningfully regulate university conduct. The implication of Kennedy’s ruling – though here Kennedy provided few details –- is that a university must prove that specific educational benefits are achieved by the school’s use of racial preferences; that there is no other (race-neutral) way to achieve those benefits; that individuals receiving preferences are evaluated based on their specific diversity contributions; and that the university has some concrete plan for phasing racial preferences out over time.
Many observers will be surprised by the near-unanimity of the Court on an issue that has usually closely divided it. I think there are a couple of reasons for this. First, the Fisher plaintiffs did not challenge the constitutionality of Grutter; they argued instead that the University of Texas was violating some of Grutter’s specific requirements (such as the requirement that universities not engage in “racial balancing”). Kennedy specifically noted in his opinion that the Court did not consider whether to overrule Grutter because the plaintiffs had not asked it to. Second, I think many of the Justices found O’Connor’s opinion in Grutter (which was widely criticized at the time) too mushy; the seven-to-one vote suggests a consensus that the Court needs to have much clearer standards in this area. Third, the whole drift of debate on affirmative action over the past ten years has been from the question of “is it right?” to the question of “does it work?” The Court, in demanding that the university bear the burden of proving that its racial preference program does the things it purports to do, and is better at doing them than any race-neutral alternative, is essentially embracing this empirical approach to racial preferences.
In our book Mismatch, Stuart Taylor and I had predicted that Justice Kennedy would write an opinion requiring a much “stricter” application of strict scrutiny – and to that extent, we were right. But we had hoped that Kennedy would lay out in detail just how a university would carry its burden, and would suggest or require mechanisms (such as greater university transparency about admissions and student outcomes) that would make it much easier for the public to evaluate how racial preferences operate. This did not happen. It is possible – though this is sheer speculation – that Fisher’s long gestation period (eight months from oral argument to decision) reflected an attempt by Kennedy to win five votes for a more detailed prescription, but that liberals on the Court did not wish to commit themselves without more facts on the case, and conservatives on the Court, who would have preferred a broad ban on racial preferences, did not want to suggest a detailed formula for legitimizing them.
So, what will happen on remand? The Fifth Circuit will seek more facts. It might do this by requiring supplemental briefs from the parties on specific questions about what racial preferences at the University of Texas have actually accomplished, and what race-neutral alternatives would have accomplished. Or it might remand the case for a full trial (the case thus far has been decided entirely on motions and briefs). Kennedy’s decision gives the Fifth Circuit a fair amount of latitude to flesh out the exact empirical burden strict scrutiny places on universities. It is quite likely that, once the Fifth Circuit has done so, the case will come back to a (possibly quite different) Supreme Court for another review.
But Kennedy’s decision is also likely to spawn other challenges to racial preference programs. It makes it very clear that plaintiffs in such suits face a fairly low burden: they must show that a university is in fact engaged in using racial preferences. Universities must then demonstrate why their programs should survive the strict scrutiny of a court. While today’s opinion in Fisher is unlikely to substantially change the size and scale of racial preferences, it will probably push universities to document the operation and effects of those preferences much more carefully. And scholars will continue to scrutinize more closely whether university claims about these programs are really valid. The debate will continue to increasingly focus on whether racial preferences work, and whether alternatives can be found that work as well or better.