Legal opinions do not often invoke Toni Morrison. But, last week, a federal judge relied on Morrison’s words in a rousing conclusion to the case on Harvard University’s use of race in admissions. “Race is the least reliable information you can have about someone. It’s real information, but it tells you next to nothing,” Morrison told Time, in 1998. The judge, Allison Burroughs, said that when this wisdom is accepted it will “ultimately make race conscious admissions obsolete.” But that hasn’t happened yet. The clearest message to emerge from the evidence in the Harvard case is that élite universities are very far from being able to achieve racially diverse student bodies using only race-neutral methods.
The case began in 2014, when Students for Fair Admissions, a group founded by Edward Blum, a conservative activist who opposes affirmative action, filed a lawsuit alleging that Harvard’s undergraduate-admissions process violated civil-rights laws that prohibit race discrimination. The most sweeping of Blum’s claims was that using race as a factor at all in admissions is unlawful, a count that was easily dismissed before trial, based on Supreme Court rulings permitting affirmative action. (That issue can be revisited on appeal.) The remaining counts at trial, which took place in 2018, in federal district court in Boston, alleged that Harvard actually deviated from the Supreme Court’s requirements, by engaging in “racial balancing”—that is, seeking a specified racial composition of a class rather than using race merely as a “plus” factor in evaluating an individual—and by using race as a factor in admissions even though there were available race-neutral alternatives. Harvard responded that its “whole person” review of each applicant considers race as only one factor among many, in order to achieve diversity, and denied the claim of racial balancing. Most controversially, the plaintiff alleged that Harvard’s admissions practices intentionally discriminated against Asian-Americans, a charge that Harvard rejected.
Last week, nearly a year after the trial, Judge Burroughs issued a judgment in favor of Harvard. The school’s admissions program survived “strict scrutiny,” meaning that Burroughs found that its use of race was necessary to serve the school’s compelling interest in diversity, particularly its interest in enrolling a critical mass of underrepresented minorities, such as African-American and Latino students. The court found that Harvard did not employ an impermissible quota or racial balancing, and that no available race-neutral alternatives would suffice to serve Harvard’s diversity goals. The “whole person” method of considering race as one factor among many in evaluating applicants along many dimensions has, for a half-century, tracked with the Supreme Court’s instructions on how to use race permissibly in admissions. This was the easy portion of the case, on which there was not significant suspense.
The difficult part, and the bulk of the trial, involved the allegation that Harvard intentionally depressed the rate of admission of Asian applicants relative to those of similarly qualified white applicants. The focal point was evidence that admissions officers gave Asian applicants higher extracurricular and academic ratings than they did white applicants. Therefore, Burroughs wrote in her decision, “Asian Americans would likely be admitted at a higher rate than white applicants if admissions decisions were made based solely on the academic and extracurricular ratings.” But admissions officers also gave Asian applicants lower “personal” ratings than they did white applicants—ranging from “outstanding” to “bland or somewhat negative or immature”—which brought down Asian applicants’ “whole person” scores and the group’s over-all admissions rate. The question, as Burroughs put it, was, “Why do Asian American applicants score lower on the personal rating?” The court considered two possibilities: that Asians “did not possess the personal qualities that Harvard is looking for at the same rate as white applicants” or that “there is overt discrimination or implicit bias at work.”
The Court found the reason for the disparity in the personal ratings unclear, speculating that undistinguished recommendations from high schools may have led admissions officers to give lower ratings. But the judge didn’t seem terribly troubled, because she didn’t think “the disproportionate strength of a racial group in one area necessarily implies that the same racial group should be strong in all areas.” (This was, to say the least, questionable reasoning; it is hard to imagine being unperturbed if admissions officers had systematically given weaker “personal ratings” to African-American applicants.) The court concluded, based on Harvard officials’ trial testimony, that there was no evidence of “discriminatory animus or conscious prejudice” on the school’s part, and that “any race-related discrimination against Asian American applicants relative to white applicants is unintentional.” Unintentional discrimination is not illegal discrimination in this context.
Burroughs emphasized the Harvard officials’ testimony, but she also noted that statistical analyses from two duelling expert economists on the role of race in admissions was “perhaps the most important evidence” in the case. The plaintiff’s expert understandably chose to exclude the personal rating in his statistical model, based on the idea that it was influenced by race; as a result, he found discrimination against Asians relative to whites. Harvard’s expert, also understandably, chose to include the personal rating in his model, because it contained information that was actually important to admissions, such as a compelling life story or admirable character; as a result, he found no discrimination against Asians. In this way, from the get-go, each side’s purportedly objective numbers already had its favored (and self-serving) assumptions baked in. In the end, the court said, “although the statistics perhaps tell ‘what,’ they do not tell ‘why,’ and here the ‘why’ is critically important.” In other words, yes, there was data, but it was the interpretation of it that mattered.
This tension between counting things and interpreting them, and between quantitative and qualitative knowledge, uncannily echoed the central issue in the discrimination claim against Harvard: the admissions process involved numbers, including scores for extracurriculars and academics, in which Asian applicants did well. But they were dragged down by the personal rating, which, “by its very nature,” included things that “cannot be fully captured by the statistical data,” such as “integrity, helpfulness, courage, kindness.” Harvard said such information was crucial to its admissions goals. The background stereotype that associates Asians with math and number-crunching inevitably presented an implied contrast to Harvard’s valorization of the unquantifiable. The numbers might be said to indicate discrimination against Asians, but that effect diminished to almost nothing when Harvard included a number that it said reflected less quantitative virtues. Unfortunately, it smacked of Harvard saying (to Asians), You just don’t understand what’s important, because you’re too focussed on numbers.
The quandary between hard and soft also reflects a larger paradox within affirmative-action law. If a school’s admissions were wholly data-driven, using numerical benchmarks, say, with the goal of tracking the racial population of the United States, that approach would run afoul of the Supreme Court’s prohibition of racial quotas. Aiming to select a class that is any particular percentage black or white is illegal. But the holistic approach, which the Supreme Court has deemed a permissible way to consider race, allows for consideration of qualities that aren’t always captured in numerical information. Considering someone’s experience of past discrimination or of overcoming hardship, for example, is permissible. It is also true, however, that if statistics revealed that using soft factors resulted in severe, rather than slight, racial disparities in admissions, a court would be more likely to find unlawful discrimination. Therefore, schools must pay some attention to numerical outcomes, but not too much, if they are to avoid, simultaneously, the appearance of a quota system and a statistical anomaly that seems to reveal discrimination. It’s reasonable to think the law has effectively induced schools to obscure racial preferences in nebulous and euphemistic categories like “personal” ratings.
The court recognized that Harvard’s admissions outcomes “may reflect some implicit biases” against Asians that, “while regrettable, cannot be eliminated in a process that must rely on judgments about individuals.” To address this, the court suggested that the admissions office conduct implicit-bias trainings and keep track of race-related statistical disparities in ratings. Urging Harvard to check its possible bias without actually finding discrimination was, perhaps, an acknowledgement of a problem that didn’t seem severe enough to warrant the heavy hand of a judicial order. Or perhaps the court understood that a finding of illegal discrimination against Asians here would invite a nuclear threat of a remedy—the end of affirmative action as we know it. If so, any finding that Harvard mistreated Asians, whether or not it had, was untenable. Thus, the court’s solution, of finding for Harvard but also telling it clearly to examine its own implicit bias, showed practical wisdom.
The court’s refusal to let “a few identified imperfections” amount to unlawful discrimination was also likely bolstered by fixes to admission criteria that Harvard made in 2018, while the case was pending, such as clarifying for the first time in a written policy that race should not play a role in the personal rating or any place other than the “overall rating,” where race can be considered as one factor among many. (The plaintiff pointed out that this change should cut the other way, as evidence of self-correction for past discrimination.) The court also suggested that “any implicit biases against Asian Americans dissipated or were eliminated” for the 2019 entering class, after Harvard grappled with the lawsuit’s allegations. More than a quarter of the students accepted identified as Asian this year, the highest number ever, up from twenty-three per cent in 2018 and twenty per cent in 2014, when the suit was filed. The number of white students, who have been a minority of the admitted class for several years running, will likely continue to dwindle. All of this implies that the impact of a high-profile case lies beyond the content of a resulting judicial order.
Harvard was victorious, and the losing plaintiff immediately filed an appeal in the First Circuit, starting the next stage of a slog of a litigation that could eventually wind up at the Supreme Court. And there the practice of race-conscious admissions is running up against a countdown clock. In Grutter v. Bollinger, which upheld the University of Michigan Law School’s affirmative-action program, the Court stated, in 2003, “We expect that 25 years from now, the use of racial preferences will no longer be necessary.” Under the governing doctrine, if the use of race were unnecessary to achieve diversity, then it would be discriminatory to use it. Justice Clarence Thomas, dissenting in Grutter, characterized universities’ diversity goals as “racial aesthetics” and approvingly characterized the majority as holding that “racial discrimination in higher education admissions will be illegal in 25 years.” As that time approaches, Thomas is positioned to lead the present conservative majority to affirmative action’s demise.
That heightens the urgency of the question: Is there in fact no race-neutral means of sustaining racial diversity? In the Harvard case, Burroughs noted Harvard’s evidence that race is currently a “determinative” factor “for approximately forty-five percent of all admitted African American and Hispanic applicants,” and that if race were not considered “more than one third of the admitted Hispanics and more than half of the admitted African Americans, would most likely not be admitted.” But Burroughs also pointed out that maintaining diversity through race-neutral admissions might nonetheless be possible, if Harvard were to take a range of drastic measures—including eliminating test scores and all preferences given to athletes, legacies, children of donors, and children of faculty and staff. These groups together comprise one-third of each class and are admitted at eight times the rate of other applicants. But abandoning such preferences, which, Burroughs wrote, currently “disproportionately benefit white applicants,” would have difficult costs for Harvard—namely foregoing “desired benefits from relationships with its alumni and other individuals who have made significant contributions to Harvard.” It would certainly be financially disadvantageous.
Indeed, the most important revelations of the Harvard trial were the vast privileges afforded to socially and economically advantaged, mostly white applicants within the “whole person” approach. Those privileges were further highlighted this spring, when the same Boston federal courthouse became the scene of an explosive admissions scandal, involving dozens of wealthy parents indicted for bribing exam administrators and college coaches—helping to sour trust regarding even perfectly legal means of gaining advantage. For example, in recent months, we’ve seen a gathering call to end legacy admissions outright, including through state and federal bills. In the end, outrage at advantages for the privileged may overshadow affirmative action in igniting general hostility about colleges’ admissions processes—which may, if left unchecked, obscure through double-speak and hide a multitude of sins.
