Internal Harvard Review Showed Disadvantage for Asian Applicants

Harvard’s internal research office concluded the College’s admissions policies produce “negative effects” for Asian Americans in a series of confidential reports circulated among top administrators in 2013, according to court documents filed early Friday morning in an ongoing lawsuit against the University.

In the reports, which were never made public, Harvard’s Office of Institutional Research also concluded the College’s admissions process advantages legacy students and athletes more than it does low-income students.

In one 2013 report, OIR employees wrote that “Asian high achievers have lower rates of admission.” In others, OIR found that Asian American applicants earned consistently lower "personal" ratings from Harvard admissions officers than did applicants of other races despite earning consistently higher rankings for their academic records and tests scores.

Both Harvard and advocacy group Students for Fair Admissions provided accounts of the institutional review—as well as hundreds of pages of other information related to Harvard’s admissions process—in court filings over the course of the day Friday. The University and SFFA filed the documents as part of SFFA’s ongoing lawsuit alleging Harvard discriminates against Asian Americans in its admissions process.

SFFA pointed to the OIR reports as evidence supporting its discrimination charges—charges Harvard has repeatedly denied. Edward Blum, president of SFFA, wrote in an emailed statement that “today’s court filing exposes the startling magnitude of Harvard’s discrimination against Asian applicants.”

In its own filings, Harvard strongly contested SFFA’s version of events and called the internal review inconclusive and incomplete.

“SFFA will point to documents prepared by individuals in Harvard’s Office of Institutional Research, which in SFFA’s view suggest that Asian-American applicants were disadvantaged in the admissions process,” Harvard lawyers wrote. “But the analysis in those documents was not designed to evaluate whether Harvard was intentionally discriminating and reached no such conclusion.”

The lawyers further wrote the OIR analysis was “incomplete, preliminary, and based on limited inputs.” The analysis did not control for unspecified vital information Harvard uses to evaluate applicants, the lawyers wrote. Models that account for the “full range of observable information” considered during Harvard’s admissions process show “no negative effect” for Asian Americans during the admissions process, the lawyers noted.

The office’s analyses formed part of broader study modeling the roles various factors—including gender and socioeconomic status—play in the College’s admissions procedures.

SFFA in its filings stated that Robert Iuliano, Harvard’s general counsel, requested the OIR review around late 2012 partly in response to allegations made by Harvard alumnus Ron K. Unz ’83 that Harvard displayed “an anti-Asian admissions bias." But Harvard spokesperson Anna Cowenhoven wrote in an emailed statement that this account of events is untrue.

"Harvard's Office of General Counsel did not request any of the three Office of Institutional Research (OIR) reports referenced in and attached to SFFA's filings," Cowenhoven wrote.

Between Dec. 29, 2012 and Feb. 12, 2013, members of OIR and the Admissions Office exchanged over 100 emails as the office gathered information on the College’s admissions procedures, according to court filings.

High-up Harvard administrators reviewed the OIR findings, according to court filings. On separate occasions, Dean of the College Rakesh Khurana, Dean of Admissions and Financial Aid William R. Fitzsimmons ’67, and senior members of the Admissions Office all examined the reports.

But the report itself argues against sharing its findings with the public.

“We imagine that sharing any analysis of admission weights will draw attention to the variety of factors that compete with one another in the admission process,” one of the OIR reports reads.“While we find that low income students clearly receive a ‘tip’ in the admissions process, our descriptive analysis and regression models also shows that the tip for legacies and athletes is larger and that there are demographic groups that have negative effects.”

The only demographic group which saw “negative effects” were Asian Americans, according to the OIR analyses.

Following a period of information gathering in late 2012 and early 2013, OIR wrote a report titled “Admissions and Financial at Harvard College.” In addition to examining issues of gender and early action admissions, the report was specifically meant to address the question: “Does the admissions process disadvantage Asians?”

Using 10 years of admissions demographic data and logistic regression models, OIR created a model that estimated the probability of admission for individuals based on certain characteristics.

This model included estimated demographic breakdowns of classes admitted given different weighting of various characteristics used to evaluate applicants. One of the breakdowns considered the demographics of a class that would be admitted if Harvard judged only by rankings and ratings of academics success.

Under this scenario, “the percentage of Asians would more than double to 43 percent,” according to SFFA’s Friday filings. SFFA’s document alleges representatives from OIR met with Fitzsimmons to present its findings and that Fitzsimmons took little—if any—further action to address the report.

“Following this presentation, Dean Fitzsimmons did not request any additional work from OIR into whether Asian-American applicants were being disadvantaged in response to the February 2013 Report,” the document states. “Dean Fitzsimmons did not share or discuss the February 2013 Report with anyone else in the Admissions Office or any senior leaders outside the Admissions Office.”

In early 2013, following the first report, OIR produced a second document titled “Admissions Part II” which specifically focused on differences in admission rates between Asian American and white applicants. The report solely compared admissions rates for “non-legacy, non-athlete” students.

The report found that Asian American applicants performed significantly better in rankings of test scores, academics, and overall scores from alumni interviews. Of 10 characteristics, white students performed significantly better in only one—rankings of personal qualities, which are assigned by the Admissions Office.

The report also found that, for students with comparable academic rankings or SAT scores, white students were generally admitted at higher rates than were Asian American students. The second report was also presented to Fitzsimmons and also spurred little further action, according to the SFFA filings.

A third OIR report—commissioned to gauge how low-income students fare in the admissions process—found a slight negative association between being Asian American and earning a spot at the College.

Ten days after taking office as Dean of the College on July 1, 2014, Rakesh Khurana met with four representatives from the OIR to discuss the three reports, according to the filings.

In previous court testimony reported in the filings, Khurana and other top Harvard officials questioned the validity of the OIR findings.

Khurana recalled thinking about “a lot of limitations to what are called fitted models like this” that led him to believe the analysis was not “done appropriately,” according to court testimony. SFFA filings state Khurana “took no steps to find out why OIR prepared these reports” nor did he “determine how the analysis could be done more ‘appropriately’” or pursue further investigation into the OIR’s findings.

Faust similarly questioned the legitimacy of the OIR reports in court testimony, testifying the documents were “preliminary, for discussion.”

“So I would say this is an exercise,” Faust testified. “I would not give it more credibility than being an exercise.”

Current and former OIR employees who completed analysis for the reports gave widely varying recollections of their involvement in the department’s work during court testimony.

Erin Driver-Linn and Erica Bever, who worked on the reports, said they recalled very little about their involvement producing the reviews “despite numerous emails and drafts documenting” their participation in the process, according to filings.

Bever did not recall ever completing such analysis and drew “a complete blank on this particular topic,” while Driver-Linn said she did not know whether the OIR was asked to complete such an investigation. However, each criticized the validity of the reports’ conclusions, with Bever stating they “oversimplified” Harvard admissions and Driver-Linn calling them “reductive.”

But one former Harvard employee, Mark Hansen, said he recalled working on the OIR’s investigation into potential racial bias and said he believes the reports comprise “evidence that Asians are disadvantaged in the admissions process at Harvard.”

Fitzsimmons recalled “how incomplete the analysis was” because it lacked certain data from the admissions office, according to his testimony. He asserted that Harvard “would always be vigilant about any suggestion of discrimination against any person” and confirmed that he believed his decision to neither discuss the conclusions of these reports with others in his office nor follow up with the OIR was consistent with this stance.

Update: June 16, 2018

This article has been updated to reflect the fact that Harvard contests Students for Fair Admissions' allegation that the University's Office of General Counsel requested the Office of Institutional Research's examination of the College's admissions process.

—Staff writer Delano R. Franklin can be reached at delano.franklin@thecrimson.com. Follow him on Twitter @delanofranklin_.

—Staff writer Samuel W. Zwickel can be reached at samuel.zwickel@thecrimson.com. Follow him on Twitter @samuel_zwickel.