With the Supreme Court poised to reduce or even eliminate affirmative action in college admissions, a recent study has offered a unique window into the magnitude of racial preferences in America’s elite colleges.
The paper, part of a series of studies conducted in the wake of high-profile litigation against Harvard and the University of North Carolina, shows that Hispanic and African American applicants to both colleges enjoy substantial advantages relative to whites and Asian Americans. Their chances of acceptance are drastically higher than they would be in the absence of affirmative action, but with a somewhat counterintuitive addendum: preferential treatment is relatively weaker for minority applicants from poor and working-class backgrounds than it is for their peers from more affluent families.
Those findings, and those of the preceding papers, are built on data that was made publicly available during the discovery phase of two lawsuits — Students for Fair Admissions v. President and Fellows of Harvard College and Students for Fair Admissions v. University of North Carolina — that were consolidated for oral argument before the court. Peter Arcidiacono, an economist at Duke University and the studies’ lead author, has provided expert testimony on behalf of the plaintiffs, who claim that the storied institutions have systematically discriminated against Asian applicants. In an interview with The 74, he said he hoped his work would help clarify the public debate over one of the most divisive issues in American politics.
“So much of the debate about affirmative action is happening in this binary where you’re either for it or against it,” Arcidiacono observed. “But there’s a large range of possibilities, from just [using] race as a tiebreaker to fully equal outcomes. So in order to get a sense for whether affirmative action has gone too far or has not gone far enough, you have to understand the role that race plays currently, and you can’t do that without the data.”
Those questions are becoming more concrete by the month. Opening briefs have been filed in the case, which will be heard in the 2022-23 term. With plaintiffs asking the nation’s highest court to bar the consideration of race and ethnicity as a factor in the college application process, and Republicans in Congress pursuing legislation that would force colleges to publicize their use of non-academic characteristics in admissions, the stage is being set for a major rollback of affirmative action as it has been practiced for half a century. According to Arcidiacono’s latest study, a significant reversal could shrink the percentage of African American students admitted to Harvard by more than two-thirds.
Georgetown economist Harry Holzer finds those projections plausible. A proponent of race-based affirmative action, he signed a 2018 brief (alongside multiple Nobel laureates) defending Harvard’s policies that was filed in a lower-court iteration of the suit. Arcidiacono’s line of research “makes reasonable points,” Holzer said, while arguing that it does not invalidate the use of racial considerations by admissions officers.
“It doesn’t change my support for affirmative action to see his numbers, though I certainly don’t disagree with the research findings. It shows that when very elite schools practice affirmative action in admissions, which they do, it does effectively raise the admission rate for people of color — especially African Americans — by a lot.”
Race & class
Arcidiacono and his coauthors dug into admissions records for over 300,000 domestic applicants to the admissions classes of 2014–2019, of which roughly 142,000 applied to Harvard, 57,000 applied to UNC as in-state candidates, and 105,000 applied to the same college from out of state. Applicant-level information included demographic attributes such as race and socioeconomic status, as well as richly detailed academic records covering high school grades, standardized test scores, and individual ratings from admissions officers across a range of academic and non-academic indicators.
Combining high school GPA, SAT scores, and scores on SAT II subject tests, the research team created academic ratings for each applicant and ranked them by decile (a statistical measurement dividing data into 10 equal parts). The lowest-performing students were grouped into the bottom 10 percent and the strongest performers grouped into the top 10 percent. Most African Americans fell into the bottom 20 percent of all applicants to both Harvard and UNC, but they were admitted at the highest rate for almost every performance decile, followed by Hispanic, white, and Asian applicants.
The acceptance gaps between categories are largest around the middle of the spectrum for academic qualifications, with African Americans applying to Harvard being accepted at a rate double that of Hispanics — and 12 times greater than Asian Americans — at the fifth decile (i.e., between the 41st and 50th percentile of qualifications). For out-of-state applicants to UNC, African Americans at the fifth decile were almost 33 times more likely to be accepted than Asian Americans and 14 times more likely than whites.
Overall, Harvard’s policies roughly quadrupled the likelihood that an African American applicant would be accepted relative to a white student with similar academic qualifications, while multiplying the likelihood of admissions 2.4 times for Hispanics. For out-of-state applicants to UNC, the force of racial preferences multiplied African Americans from 1.5 percent of admitted students to 15.6 percent, a tenfold increase. Black applicants applying in-state to Chapel Hill gained a smaller advantage from affirmative action, becoming 70 percent likelier to win admission.
Beyond these general calculations, the authors noticed a peculiar interaction between race and class. While white applicants from lower-income families appear to receive an advantage in admissions relative to wealthier whites with similar academic profiles, disadvantaged African Americans and Hispanics do not. Affluent applicants of color therefore receive a comparatively larger boost over affluent white applicants than poor and working-class students of color enjoy over poor and working-class whites.
Holzer said the substantive arguments in favor of affirmative action — particularly the educational value of maintaining a racially diverse campus — “don’t require that the specific recipient face bias or barriers in the past.” Still, he asserted that low-income students of color deserve “extra credit” not only for their race but also their class background.
“If the bump for just being African American is really large, you could imagine that maybe [admissions officers] think, ‘We’re already taking care of that problem,’” Holzer said. “But for someone like me, who thinks that class really matters a lot, you want to make sure that lower-income students of color get consideration.”
Arcidiacono argued that the large edge claimed by some wealthy students, even if they come from historically excluded groups, risks eroding public faith in the fairness of admissions altogether. Disillusionment already exists not only due to long-running patterns of underrepresentation, he added, but also newer blots such as the Varsity Blues scandal, which saw moneyed parents conniving with college employees and private admissions counselors to game the system.
“This is in the context of a system that completely favors people who come from richer backgrounds,” Arcidiacono said, listing the factors already favoring upper-class applicants: ready access to college counselors, special weight placed on extracurricular activities, and recruitment for sports like sailing and golf.
“To me, one of the arguments for affirmative action would be that you’re trying to build trust in the system for a group that has been traditionally disenfranchised. But the way you do that matters, and it’s not really hitting the poorer African Americans — they’re not the ones benefiting the most.”
The ‘narrowly tailored’ standard
The Supreme Court will consider Students for Fair Admissions v. Harvard this fall, eight years after it was first filed in federal district court. When a decision is finally reached, the case could fundamentally alter the practice of affirmative action in college admissions and, with it, the racial composition of some of the country’s most prestigious schools.
Existing Court precedent was set in the 2003 Grutter vs. Bollinger case, in which a plaintiff alleged that preference systems at public graduate schools — in that instance, the University of Michigan Law School — illegally disadvantaged white students on racial grounds. A majority led by Justice Sandra Day O’Connor found instead that “the narrowly tailored use of race in admissions decisions” was constitutionally permissible, while adding that the issue would be ripe for reexamination within 25 years. That deadline has nearly elapsed, and the Harvard litigants seek to reopen the question of whether affirmative action has been “narrowly tailored” to begin with.
The case, along with the corresponding UNC suit, centers on the accusation by a group of Asian American students that Harvard’s policies unfairly disfavored them relative to applicants from every other racial group. Although the admissions candidates with the highest GPAs and test scores are disproportionately Asian American — 2019 data from the College Board showed that about one-quarter of Asian high school graduates scored above a 1400 on the SAT, compared with 8 percent of whites, 2 percent of Hispanics, and 1 percent of African Americans — they were consistently graded lower according to Harvard’s personality scores. A 2020 study released by Arcidiacono and his colleagues suggested that, absent the subjective penalty that Asian applicants face, they would be admitted at a rate 19 percent higher than currently prevails.
The findings instantiated a theory that first gained widespread attention a decade ago, when one right-wing commentator alleged that Asian Americans were tacitly being held back by admissions quotas lest they grow to dominate Ivy League campuses. Although Asian high schoolers were routinely among the top-performing students in the United States, their numeric presence on elite campuses peaked around 1990 and remained roughly the same over the next 20 years.
Trends in Asian American college enrollment were the focus of a report released in April by the Manhattan Institute. Author Robert VerBruggen, a journalist and fellow at the conservative think tank, noted that the unmistakable stagnation in representation — which occurred even as Asians were continually growing as a percentage of all college aspirants — began to lift about about a decade ago. Whether that was connected to the growing focus on apparent discrimination is unclear.
“It’s an interesting question why that’s happened, but it’s certainly consistent with the narrative that everybody started making a stink about it, lawsuits were filed, and schools got a lot more careful about what they were doing,” Verbruggen said.
The release of the schools’ data allowed researchers to investigate trends in admissions beyond purported anti-Asian bias. In a widely covered paper released in 2019, Arcidiacono and his co-authors calculated that 43 percent of white students admitted to Harvard between 2014 and 2019 were either legacies, recruited athletes, children of university faculty and staff, or on the dean’s interest list (i.e., relatives of potential high-dollar donors). Another postulated that the school, which has for years trumpeted its efforts to diversify, may deliberately recruit African American students who stand virtually no chance of gaining admission.
“In some sense, there’s this uneasy compromise that works to the detriment of Asian Americans and poor whites,” Arcidiacono said. “You’ve got the racial preferences helping underrepresented minorities get into certain colleges, and you’ve got the legacy and athlete preferences helping rich, disproportionately white kids get into college.”
A Supreme Court ruling favorable to the plaintiffs could leave that system profoundly changed, upsetting the demographic mix that elite schools have worked hard to cultivate. By Arcidiacono’s accounting, the proportion of African Americans admitted to Harvard over the period he studied would have been less than 1 percent if acceptance was offered on the basis of academic qualifications alone; those admitted to UNC in-state or out-of-state would sink to 4.3 percent and less than 2 percent, respectively. At the same time, the percentage of Asian Americans would have risen substantially — to over 50 percent of all admitted students, in Harvard’s case.
While oral arguments in Students for Fair Admissions v. Harvard won’t come for months, the recently announced recusal of soon-to-be Justice Ketanji Brown Jackson is a promising sign for the plaintiffs. With the departure of Anthony Kennedy and the death of Ruth Bader Ginsburg, the Court has lost two members who had previously ruled in favor of race-based affirmative action in postsecondary education.
Whatever the legal outcome, VerBruggen said that Arcidiacono’s work offered considerable value simply by shining a light on the internal admissions processes in two highly competitive universities.
“Schools are so tight-lipped about their affirmative action policies that we don’t have a lot of data on them,” he remarked. “If you want to know what’s going on in a school, you basically have to sue them.”