Students for Fair Admissions, Inc. v. President and Fellows of Harvard College et al

Filing 569

Opposition re 547 MOTION in Limine filed by Students for Fair Admissions, Inc.. (Attachments: # 1 Exhibit 1, # 2 Exhibit 2, # 3 Exhibit 3, # 4 Exhibit 4, # 5 Exhibit 5, # 6 Exhibit 6, # 7 Exhibit 7, # 8 Exhibit 8, # 9 Exhibit 9, # 10 Exhibit 10, # 11 Exhibit 11, # 12 Exhibit 12, # 13 Exhibit 13, # 14 Exhibit 14, # 15 Exhibit 15, # 16 Exhibit 16)(Mortara, Adam) (Attachment 6 replaced on 10/1/2018) (McDonagh, Christina). (Additional attachment(s) added on 10/2/2018: # 17 Unredacted Memorandum in Opposition (FILED UNDER SEAL), # 18 Exhibit 4 (Filed Under Seal)) (McDonagh, Christina).

Exhibit 13

P434
SFFA v. Harvard

EXPERT REPORT OF PETER S. ARCIDIACONO
Students for Fair Admissions, Inc. v. Harvard
No. 14-cv-14176-ADB (D. Mass)

exhibitsticker.com

TRIAL EXHIBIT

For each rating measure, more highly rated applicants are more likely to be
admitted. This can be seen because the fraction of admits assigned to the lowest
category (<3-) in each racial/ethnic group is almost always smaller than the fraction
of total applicants assigned to the lowest category, while the fraction of admits
assigned to the highest category (>3+) are always higher than the fraction of total
applicants assigned to the highest category. For some of the rating categories in the
baseline dataset, the probabilities are incredibly small—if not zero—if the applicant
is rated in the lowest category. The share of admits is 0.1% or less for those who are
in the lowest category for the academic, personal, either teacher rating, or the
counselor rating.
Consistent with the objective measures in both the baseline and expanded datasets,
Asian-American applicants rank higher than any other group based on their
academic rating. For example, in the baseline dataset, 58.6% of Asian-American
applicants are in the highest category (>3+), compared with 44.7% of whites, 14.7%
of Hispanics, and 7.3% of African Americans. Almost 93% of Asian-American admits

36

were in the highest academic rating, compared to 88% of whites, 62% of Hispanics,
and 58% of African Americans.
Asian-American applicants are substantially stronger in other dimensions as well.
Compared

to

white

applicants,

Asian-American

applicants

have

better

extracurricular ratings and overall alumni ratings, similar teacher 1 ratings, but
slightly lower ratings than whites on counselor, teacher 2, and alumni personal
ratings. Asian-American applicants are stronger than African-American and
Hispanic applicants on all these dimensions except two: the athletic and personal
ratings). As shown in Section 2.4., the athletic rating is relatively unimportant.
For Harvard’s personal rating, however, the difference is more striking and
consequential. Asian-American applicants have the lowest share of applicants
receiving 2- or better on the personal rating. These scores diverge significantly from
the personal rating scores given by alumni interviewers, where Asian-American
applicants fared better than African-American and Hispanic applicants and only
slightly worse than white applicants. They also are inconsistent with testimony
from Harvard’s own admissions personnel, who firmly rejected the idea that AsianAmerican applicants were somehow lacking in personal qualities compared to other
applicants.47
It is worth pausing to note that the opportunity for racial penalties and preferences
is least present in academic and extracurricular ratings for two reasons. First, both
are easily measured. For the academic rating, Harvard’s files contain information
on the test scores of the students, their grades, number of AP exams taken and the
scores on these AP exams, etc. For the extracurricular rating, lists of activities are
included that specify the type of activity, the years the student participated in that
activity, and the number of hours per week devoted to the activity. Second, they are
specific, reflecting how an applicant scored on a particular set of tasks.
This is in contrast to the personal rating, which is difficult to measure directly, and
the various ratings that reflect agglomerations of another individual’s rating of a
candidate along many dimensions (e.g., the counselor and teacher ratings, as well as
47

See, e.g., Fitzsimmons Depo. at 347:10-348:2; Donahue Depo. at 165:17-167:12.

37

the overall ratings of the reader and the alumni interviewer). Harvard’s Reader
Guidelines illustrate why it would be easy to manipulate the personal rating. While
the guidelines provide detailed instructions for the various other ratings, for the
personal rating, the guidelines list only the following: “1. Outstanding. 2. Very
strong. 3. Generally positive. 4. Bland or somewhat negative or immature. 5.
Questionable personal qualities. 6. Worrisome personal qualities.”48
Harvard’s OIR researchers in fact recognized racial differences in the assignment of
personal ratings in 2013. Using data over ten years, they found that Harvard’s
admissions officers assigned substantially lower personal ratings to Asian-American
applicants versus white applicants, especially when compared to the ratings
assigned by teachers, counselors, and alumni interviewers.49
These component ratings all contribute to the separate overall rating Harvard
assigns to each applicant.50 Here, I am using the ratings assigned by the last reader
of the applicant file. Unlike the component ratings, Harvard’s data also provide
more detailed overall ratings for all years that include any pluses and minuses. For
the purposes of this descriptive analysis, I aggregate the overall ratings of the final
reader into four groups: 3- or less, 3, 3+, all 2’s, and 1.
Table 4.2 shows the share of each racial/ethic group that received a particular
overall rating and, conditional on that rating, the probability of being admitted for
the baseline and expanded dataset. Higher overall ratings are associated with
higher probabilities of admission. Those who have an overall score of 3- or worse are
almost always rejected: the admit rates for each group are below 0.03% in both the
baseline and expanded datasets. In contrast, those who receive an overall rating of
a 1 are always accepted (in both datasets).

48

See HARV00000803-04.

49

See HARV00065745.

50

See McGrath Depo. at 159:2-5.

38