The Slatest

Admitting Bias

Harvard had proof its admissions process was hurting Asian Americans. How will its dean explain why he did nothing about it?

A Harvard University building in Cambridge, Massachusetts on Aug. 30.
A Harvard University building in Cambridge, Massachusetts on Aug. 30. Photo illustration by Slate. Photo by Scott Eisen/Getty Images.

BOSTON—The long-awaited trial over Harvard’s admissions policies began Monday in a packed federal courtroom in Boston with perhaps the university’s most important witness taking the stand to defend its policies. The trial’s opening day also telegraphed critical questions that witness—Harvard’s Dean of Admissions William Fitzsimmons—is likely to face Tuesday about a key piece of evidence that, at first blush, looks fairly damning for the university.

The case, filed in 2014 by a conservative interest group called Students for Fair Admissions, was once billed as the next big test for the survival of affirmative action, but Judge Allison Burroughs has already ruled in Harvard’s favor on the question of whether the use of race in admissions is constitutional. (The issue could re-emerge on appeal.)

The trial will instead focus on allegations that Harvard illegally discriminates against Asian Americans in its admissions process, particularly in relation to white applicants. As Adam Mortara, the lawyer for the plaintiff, said in his opening argument, “The future of affirmative action is not on trial over the next three weeks.”

The plaintiff’s questioning of its first witness, Fitzsimmons, largely followed in this vein. Of particular interest Monday was the university’s outreach efforts to more sparsely populated areas of the country. Harvard sends invitations to apply to high school students who score well on a standardized test called the PSAT. Internal documents showed that the cutoff score for white students to receive the invitations was 1310, while the cut off was 1350 for Asian women and 1380 for Asian men. Under questioning, Fitzsimmons largely fell back on platitudes about getting “Harvard on the map” for disadvantaged students and a desire to “have a diverse class.”

Fitzsimmons, who has held a central role in Harvard admissions for decades, will have a tougher time dodging questions about his apparent lack of concern with three internal reports that tell a similar story. It’s unclear how exactly the documents, produced by Harvard’s Office of Institutional Research in 2013, will factor into the legal questions raised by the case. However, the reports strongly suggest that Harvard has exhibited a pattern of willful disregard for evidence that its admissions process suppresses the number of Asian-American applicants it accepts.

The most consistent and striking findings from the three reports was that white applicants have a disproportionate advantage over Asian Americans, and that preferences for legacies and athletes account for much of that racial inequity. The plaintiffs allege, and Harvard partly admits, that Fitzsimmons chose not to look further into the matter when presented with the findings. In fact, it looks as if he and other university officials buried at least one of the studies in part because of what it revealed about Asian-American applicants.

In February 2013, the OIR produced a report titled “Admissions and Financial Aid at Harvard College.” The report addressed three questions:

1.    What is the effect on our applicant pool and yield of reintroducing early action?

2.    Is the shift in the gender balance at Harvard College due to increased interest and recruitment for SEAS?

3.    Does the admissions process disadvantage Asians?

To answer the third question, the OIR devised four theoretical admissions systems based on 10 years of admissions data, each considering different applicant criteria. The analysis produced the chart below:

Projected admitted student pools

If academic ratings were the only factor in admissions, Asian Americans would make up 43 percent of the incoming class. Once you consider legacy and athlete preferences, however, that proportion drops to 31 percent. Also taking into account extracurricular and personal ratings further drops Asian representation to 26 percent. Adding demographic factors to the equation then depresses Asian admissions to 18 percent. (The actual admissions percentage for Asian Americans that year was 19 percent.)

After coming to these conclusions, Harvard’s OIR subsequently conducted another study, summarized in a report entitled “Admissions Part II.” The study narrowed the analysis to compare Asian admissions trends to white admissions trends and further found that “athletes and legacies explain the difference in raw admit rates for Asian and white applicants.” This study also compared Asian-American and white applicants who were not legacies or athletes, and nevertheless found that these Asian-American applicants performed better on SAT II, SAT I, academic, and extracurricular ratings. The only area where Asian Americans significantly fell behind white applicants was in the personal rating, which in part judges the essays, teacher recommendations, and qualities such as courage and kindness. The personal rating is not supposed to take race into account, but OIR’s studies demonstrate that the scores tend to be racialized.

Difference in average test scores and ratings for white and asian applicants

Despite these discrepancies, white admit rates still far outstripped Asian ones.

Admit rates by academic index for White and Asian applicants, classes of 2007–2016

In April 2013, the OIR conducted a third study in response to a scholarly article that found a majority of high-achieving, socioeconomically underprivileged students were not applying to elite colleges, which had stirred public criticism of Harvard. Fitzsimmons asked the OIR to investigate the admissions office’s handling of low-income applicants, and the resulting study found that Harvard was in fact giving a “tip,” or advantage, to those students. However, the study also concluded that legacy and athlete applicants had an even bigger tip, and that there was a 6 percent “negative effect” for Asian applicants compared to those of other races.

In an email summarizing the study, OIR officers consider the prospect of making the study public to refute recent criticisms, but ultimately decide against it. The email reads, “To draw attention to the positive benefit that low income students receive, may also draw attention to the more controversial findings around Asians, or the expected results around legacies and athletes.”

The findings of at least two of these reports should not have surprised Fitzsimmons. In the 1980s, a series of news articles questioned the differences in admissions rates for white and Asian-American applicants. In response to this public attention, Fitzsimmons, who was dean of admissions even back then, issued a statement that read, in part, “While Asian-Americans are slightly stronger than whites on academic criteria, they are slightly less strong on extracurricular criteria. In addition, there are very few Asian-Americans in our applicant pool who are alumni/ae children or prospective varsity athletes. When all these factors are taken into account, the difference in admissions rates for the two groups disappears.”

In its rebuttal to the plaintiffs, Harvard contended its studies used incomplete data and were not reliable. As the university’s lawyers wrote, “The OIR employees who created the documents had a limited understanding of the admissions process, and they neither examined nor reached any conclusion on the question whether the admissions process discriminates against any group.” Harvard’s lawyer, Bill Lee, also argued in his opening statement on Monday that the report was riddled with errors—citing, for example, that the date on the first slide listed the wrong year.

Burroughs previously cited Harvard’s refusal to follow up on the OIR reports in her decision denying the university’s attempt to get the case thrown out of court. In her opinion, she wrote, “Determining whether Harvard’s explanation for its response to the OIR reports is credible, or as SFFA submits, an implausible post-hoc justification in light of this lawsuit, requires the Court to assess the credibility of Harvard’s witnesses and to consider expert testimony regarding the OIR reports.”

Prioritizing athletes and legacies at the inadvertent expense of Asian Americans is not technically illegal; this practice is not at issue at the current trial and has little to do with affirmative action. (Scoring Asian-Americans lower in personal rating, on the other hand, would be legally suspect.) But these studies demonstrate the Harvard admissions office’s repeated assertion that Asian-American applicants are not facing greater odds is highly misleading. The university has in effect known for three decades that these two factors heavily skew acceptance chances to the detriment of Asian Americans.

Fitzsimmons will take the stand again on Tuesday, and the plaintiffs will almost certainly dig into those reports. It’s a shame that his explanations for the racial tinge of legacy and athlete admissions won’t likely be legally relevant to this case, but he’ll nevertheless have to explain why he didn’t feel the need to examine the disparity any further.