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

Filing 484

REPLY to Response to 417 MOTION for Summary Judgment filed by President and Fellows of Harvard College. (Waxman, Seth)

UNITED STATES DISTRICT COURT FOR
THE DISTRICT OF MASSACHUSETTS
BOSTON DIVISION
STUDENTS FOR FAIR ADMISSIONS, INC.,
Plaintiff,

Civil Action No. 1:14-cv-14176-ADB

v.
PRESIDENT AND FELLOWS OF HARVARD
COLLEGE (HARVARD CORPORATION),
Defendant.

HARVARD’S REPLY MEMORANDUM IN SUPPORT OF ITS MOTION FOR
SUMMARY JUDGMENT ON ALL REMAINING COUNTS

TABLE OF CONTENTS
ARGUMENT ...................................................................................................................................1
I.

SFFA LACKS STANDING .......................................................................................................1

II.

HARVARD DOES NOT DISCRIMINATE AGAINST ASIAN-AMERICAN APPLICANTS ..................4
A.

There Is No Documentary Or Testimonial Evidence Of Discrimination ................4
1.
2.

B.

SFFA’s reliance on the OIR documents is unavailing .................................4
SFFA’s reference to statistical trends is erroneous ......................................7

There Is No Statistical Evidence Of Discrimination................................................8

III.

HARVARD DOES NOT ENGAGE IN RACIAL BALANCING ......................................................12

IV.

HARVARD CONSIDERS RACE IN THE MANNER PERMITTED BY SUPREME COURT
PRECEDENT .........................................................................................................................17

V.

HARVARD COULD NOT ACHIEVE ITS EDUCATIONAL OBJECTIVES WITHOUT
CONSIDERING RACE ............................................................................................................22

CONCLUSION ..............................................................................................................................25

-i-

TABLE OF AUTHORITIES
Page(s)
CASES
American Postal Workers Union v. Frank, 968 F.2d 1373 (1st Cir. 1992) .....................................2
Cavalier ex rel. Cavalier v. Caddo Parish School Board, 403 F.3d 246 (5th Cir.
2005) ........................................................................................................................................13
Fisher v. University of Texas at Austin, 136 S. Ct. 2198 (2016) ...........................17, 18, 21, 22, 24
Gill v. Whitford, 138 S. Ct. 1916 (2018)..........................................................................................2
Gratz v. Bollinger, 539 U.S. 244 (2003) ....................................................................................3, 21
Grutter v. Bollinger, 539 U.S. 306 (2003) .........................................................3, 13, 14, 19, 20, 21
Harrington v. Aggregate Industries Northeast Region, 668 F.3d 25 (1st Cir. 2012) ......................7
Hunt v. Washington State Apple Advertising Commission, 432 U.S. 333 (1977) ...........................2
Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992) ..................................................................1, 2
Parents Involved in Community Schools v. Seattle School District No. 1, 551 U.S.
701 (2007) ..........................................................................................................................12, 13
Perrea v. Cincinnati Public Schools, 709 F. Supp. 2d 628 (S.D. Ohio 2010) ...............................13
Purkett v. Elem, 514 U.S. 765 (1995) ..........................................................................................6, 7
Refuse & Environmental Systems, Inc. v. Industrial Services of America, 120
F.R.D. 8 (D. Mass. 1988) ...........................................................................................................4
Regents of the University of California v. Bakke, 438 U.S. 265 (1978) ........................................14
Trump v. Hawaii, 138 S. Ct. 2392 (2018)........................................................................................2
Washington Legal Foundation v. Leavitt, 477 F. Supp. 2d 202 (D.D.C. 2007) ..............................4
RULES
Fed. R. Civ. P. 23(a)(4) ....................................................................................................................4

- ii -

SFFA’s opposition brief is laden with invective, mischaracterizations, and in some cases
outright misrepresentations of the record, but its overcharged rhetoric fails to compensate for the
lack of evidence supporting SFFA’s claims. Harvard’s admissions practices are entirely
consistent with the law and necessary to build the diverse community critical to the success of its
students. No reasonable factfinder could conclude that Harvard intentionally discriminates
against Asian-American applicants, engages in racial balancing, considers race in a manner
beyond what the Supreme Court has permitted, or could achieve its educational objectives
without considering race. And SFFA’s last-minute attempts to salvage its standing highlight
why it has never been a proper plaintiff. The Court should enter summary judgment for Harvard.
ARGUMENT
I.

SFFA LACKS STANDING
There are two independent reasons why SFFA lacks standing to pursue this action. First,

SFFA is not a genuine membership organization that can sue on behalf of its members; rather, it
is a litigation vehicle designed to advance the ideological objectives of its founder, Edward
Blum. Second, none of SFFA’s “standing members” would have standing themselves to bring
this case. SFFA’s recent filing illustrates even more clearly the deficiency of its theory of
standing and demonstrates that this Court lacks subject-matter jurisdiction.
SFFA tries to circumvent these concerns by arguing (at 7) that this Court should simply
“confirm that the Article III issue has been resolved and it is not a matter for trial.” That is
wishful thinking. The Court’s ruling on Harvard’s motion to dismiss did not “definitively
resolve[] the issue of Article III standing” for all time, as SFFA argues; it resolved only whether
SFFA could proceed past the pleading stage. As the plaintiff, SFFA bears the burden of showing
the elements of standing, and because those elements “are not mere pleading requirements but
rather an indispensable part of [its] case, each element must be supported in the same way as any
-1-

other matter on which the plaintiff bears the burden of proof, i.e., with the manner and degree of
evidence required at the successive stages of the litigation.” Lujan v. Defenders of Wildlife, 504
U.S. 555, 561 (1992); see also, e.g., Gill v. Whitford, 138 S. Ct. 1916, 1931 (2018). Where, as
here, discovery makes clear that a plaintiff does not have standing, the defendant is entitled to
summary judgment even if the plaintiff’s allegations sufficed at the pleading stage. See Lujan,
504 U.S. at 561. And in any event, courts at all stages have an independent duty to assure
themselves of jurisdiction. E.g., Trump v. Hawaii, 138 S. Ct. 2392, 2415-2416 (2018). 1
SFFA’s latest filing shows why it is so keen to avoid judicial scrutiny of its standing.
SFFA concedes that most of its prior standing members do not have a redressable injury because
they are no longer eligible to transfer. See Pltf.’s Resp. to Deft.’s Statement of Undisputed
Material Facts (Dkt. 452), Resps. to ¶¶ 260-265. It quibbles (at 6-7) with Harvard’s arguments
about the testimony of the only two prior standing members who could arguably have retained
their eligibility to transfer. But even if (as SFFA unpersuasively argues) those students were
willing to transfer at the time of their depositions, that was more than a year ago, when they had
not begun their studies elsewhere. It is even less plausible now that either would seek to transfer,
and SFFA adduces no evidence to that effect. Mere conclusory allegations are not sufficient to
meet SFFA’s burden to establish standing.

1

SFFA tries to flip the burden of proof by characterizing the question as one of mootness,
not standing. That is incorrect. SFFA’s standing to maintain this litigation depends on its having
(at all times) members who have suffered cognizable injuries that would be redressed by a
judgment in its favor. See Hunt v. Wash. State Apple Advert. Comm’n, 432 U.S. 333, 343 (1977)
(association can have standing only if members would “have standing … in their own right”);
Lujan, 504 U.S. at 561 (standing requires injury that would “be redressed by a favorable
decision” (internal quotation marks omitted)); see also American Postal Workers Union v.
Frank, 968 F.2d 1373, 1376-1377 (1st Cir. 1992). An individual who no longer has the desire or
ability to transfer could not have his or her asserted injury redressed by the forward-looking
relief that SFFA seeks.
-2-

In an effort to paper over those fatal deficiencies, SFFA tacks on boilerplate declarations
from an additional seven standing members, none of whom Harvard has had the opportunity to
depose, and six of whom joined SFFA only in June 2018—presumably after SFFA realized its
standing was in jeopardy in light of Harvard’s summary-judgment brief. Connolly Exs. 277,
278, 280, 281, 282, 283. Before either party sought summary judgment, in response to
Harvard’s inquiry whether SFFA had any changes to its roster of standing members, SFFA
represented that it did not. Yet now, SFFA not only seeks to rely on these seven new
individuals; it has refused even to identify in response to Harvard’s request whether these seven
are the only new standing members. The result is that Harvard knows next to nothing about the
individuals whose claims are supposedly being litigated. It has not had the opportunity to test
whether the new standing members would genuinely seek to transfer if SFFA were to prevail.
And for all Harvard knows, SFFA may show up on the first day of trial and announce a whole
new set of never-before-identified standing members.
SFFA’s approach appears designed to evade the framework of Rule 23, which might have
allowed a plaintiff to sue for injunctive relief on behalf of a putative class, as in Grutter v.
Bollinger, 539 U.S. 306 (2003), and Gratz v. Bollinger, 539 U.S. 244 (2003). A class action
would have allowed the parties and the Court to handle this case in an ordered way, with welldefined processes for identifying the class and ensuring the propriety of class treatment. Instead,
Mr. Blum founded an organization that is no more than a mailing list and sued purportedly on
behalf of its members, even though those members exercise no control over the organization. 2

2

In a footnote (at 4 n.2), SFFA suggests that its standing members “communicate with [its]
leadership and have the ability to provide substantive input,” and that “SFFA communicates
regularly with its members.” But SFFA, having resisted Harvard’s attempts to test the extent of
such communications or elicit such information during depositions, see, e.g., Ellsworth Ex. 163
at 40:7-14, 63:20-64:4; Ex. 164 at 13:16-14:2, 24:7-20, 114:13-19; Ex. 165 at 36:17-21, 47:15-3-

Cf. Washington Legal Found. v. Leavitt, 477 F. Supp. 2d 202 (D.D.C. 2007) (similar association
lacked standing). And it is now clear why Mr. Blum chose that course: Given the amicus briefs
demonstrating support for Harvard’s use of race-conscious admissions among members of the
Asian-American community, including current Harvard students (Dkts. 440-1, 455-1), he could
not credibly have argued that a class action plaintiff would “fairly and adequately protect the
interests of” Asian-American applicants, Fed. R. Civ. P. 23(a)(4).
SFFA has been on notice since it filed this suit that Harvard would hold it to its burden of
establishing standing, see Tr. of Apr. 30, 2015 Status Conference (Dkt. 43) at 21:12-22:14, and
SFFA has time and again tried to evade that burden. Now that the problems wrought by SFFA’s
strategy have become too clear to ignore, the Court should rule that SFFA lacks standing.
II.

HARVARD DOES NOT DISCRIMINATE AGAINST ASIAN-AMERICAN APPLICANTS
A.

There Is No Documentary Or Testimonial Evidence Of Discrimination

There is no direct evidence in this case—either documentary or testimonial—that forty
members of Harvard’s Admissions Office engaged in a systemic effort to limit the number of
Asian Americans at Harvard. It is simply implausible that such a wide-ranging scheme could
have existed without a single document or a single piece of testimony attesting to it. SFFA’s
arguments all try—and fail—to close that yawning gap in its case.
1.

SFFA’s reliance on the OIR documents is unavailing

SFFA persists in arguing that Harvard’s response to the limited and incomplete analyses
prepared by several employees in the Office of Institutional Research (OIR) shows an intent to

48:16; Ex. 160 at 218:6-14; Ex. 162 at 24:16-25:3; Ex. 161 at 64:5-19, can hardly rely on
information that it prevented Harvard from testing in discovery. See, e.g., Refuse & Envtl. Sys.,
Inc. v. Indus. Servs. of Am., 120 F.R.D. 8, 12 (D. Mass. 1988) (“Defendants may not shield from
inquiry during the pretrial phase … communications on the ground of privilege, then pull these
communications out of the hat shortly before or during the trial, as a part of their defense.”).
-4-

discriminate. But the OIR analyses neither inquired whether Harvard was discriminating against
Asian-American applicants nor reached any conclusion that it was. See Mem. in Support of
Def.’s Mot. for S.J. (Dkt. 418) (“Harvard SJ Mem.”) 38; Def.’s Mem. in Opp. to Pltf.’s Mot. for
S.J. (Dkt. 435) (“Harvard SJ Opp.”) 19-23. And OIR lacked far too much information—as the
analysts themselves realized—to draw reliable conclusions about the admissions process. Id.
SFFA claims that Harvard’s characterization of how certain officials reacted to the OIR
documents is itself proof of intentional discrimination, deriding it as “implausible and
contradictory.” That argument is both factually and legally incorrect.
SFFA misstates the facts in numerous ways. SFFA initially claims (at 10-11) that
Harvard officials who saw the OIR documents did not criticize or question them at the time. The
record shows the opposite; Dean Fitzsimmons testified that when he saw OIR’s analyses, he
“explained to” the OIR analysts the “incomplete” nature of their work. Connolly Ex. 9 at
402:22-403:5. Dean Khurana did not see the analyses until long after they were created, under
circumstances in which it would have made little sense for him to criticize or question them. He
was shown them upon becoming Dean of Harvard College, see Connolly Ex. 13 at 253:21-254:4,
merely as an example of the work OIR had conducted with respect to Harvard College, see
Connolly Ex. 162, and did not believe they were addressing a question of bias in the admissions
process. Ellsworth Ex. 159 at 259:15-21.
SFFA next accuses Harvard (at 9-12) of relying in other contexts on OIR documents that,
like the documents SFFA invokes here, were labeled “preliminary.” But it is not just the fact
that the OIR analyses in question were labeled “preliminary” that demonstrates the
incompleteness of those analyses; the documents on their face make plain that they reflected only
a limited understanding of the admissions process—a point confirmed by testimony from the

-5-

analysts who worked on them. 3 SFFA makes no attempt to show, nor could it, that the other
analyses it points to had similar limitations. The facial disclaimers in the admissions-related OIR
documents also explain why Dean Fitzsimmons and Dean Khurana were able to perceive the
shortcomings of OIR’s analysis without any “further inquiry,” Opp. 11.
SFFA is also incorrect to suggest (at 11) that Dean Fitzsimmons fully embraced parts of
OIR’s admissions analysis even as he supposedly disregarded the part concerning race. For the
proposition that the Dean “found OIR’s work” on low-income applicants “to be trustworthy and
reliable,” SFFA relies not on any statement from Dean Fitzsimmons, but rather on an exchange
between then-OIR Director Erin Driver-Linn and a member of Harvard’s Public Affairs and
Communications office. Opp. 11 (citing Pltf.’s Statement of Undisputed Material Facts (Dkt.
414) (“SFFA SJ SMF”) ¶ 483). But that exchange shows (Connolly Ex. 163), and SFFA
acknowledges, that “OIR … had not yet shared its full analysis with Dean Fitzsimmons.” SFFA
SJ SMF ¶ 484. When the Dean saw the actual OIR analysis, he recognized it had “limitations.”
See Connolly Ex. 9 at 426:18-427:9.
SFFA’s legal argument is equally misguided. SFFA cites cases (at 9-10) saying that if a
decisionmaker offers an implausible or unsupported justification for an allegedly discriminatory
action, the weakness of the asserted justification can show the intent to discriminate. See, e.g.,
Purkett v. Elem, 514 U.S. 765, 767-768 (1995) (addressing prosecutor’s asserted race-neutral

3

See, e.g., Connolly Ex. 145 at HARV00065753-54 (“Conclusions” and “Possible
Explanations” left blank); id. at HARV00065757 (listing admissions factors missing from the
OIR models); Connolly Ex. 134 at HARV00031718 (“The following analysis is preliminary
and for discussion.”); id. at HARV00031722 (listing “factors that quantitative data is likely to
miss or ratings do not capture”); Connolly Ex. 112 at HARV00023548 (noting limitations of the
data sources); id. at HARV00023549 (describing “several limitations” of OIR’s approach); id.
(“Our analysis should not be considered exhaustive.”); Ellsworth Ex. 23 at 196:8-10 (testimony
of Erin Driver-Linn); Ellsworth Ex. 116 at 137:20-138:21 (testimony of Mark Hansen).
-6-

explanation for striking a juror); Harrington v. Aggregate Indus.-Ne. Region, Inc., 668 F.3d 25,
32-33 (1st Cir. 2012) (addressing employer’s asserted “facially legitimate, nonretaliatory reason
for discharging the” plaintiff). But Harvard’s explanation for how the OIR documents were
handled is not its justification for the admissions outcomes of Asian-American candidates;
indeed, Harvard’s point is that the OIR documents shed little light on that question. 4 SFFA’s
cases are therefore inapposite.
2.

SFFA’s reference to statistical trends is erroneous

SFFA also suggests (at 13) that Harvard must have discriminated by comparing the
proportion of Asian-American students in the Class of 2011 and the Class of 2018, and then
claiming the share of Asian-American admitted students rose for the Classes of 2019 through
2022, after this suit was filed. But in citing figures for the Classes of 2011 and 2018, SFFA
obscures the fact that over the period of years spanned by those two Classes, the proportion of
Asian-American admitted students fluctuated considerably, between 17.5% for the Class of 2013
and 20.5% for the Class of 2016—a swing of more than 17%. Connolly Ex. 231. The Class of
2016 figure is also almost identical to the first post-lawsuit figure for the Class of 2019
(21.0%)—fatally undermining SFFA’s theory that Harvard changed course upon being sued. 5
And the growth in the proportion of Asian-American admitted students in recent years continues
4

SFFA briefly points (at 12) to what it calls “documentary and testimonial evidence of
intentional discrimination”—scattered summary sheets and correspondence. But that evidence
does not come close to showing intentional discrimination, see Harvard SJ Opp. 23-27, which
explains why SFFA discusses it so halfheartedly.
5
SFFA also errs in arguing (at 13 n.4) that Harvard supposedly manipulated the proportion
of Asian-American admitted students during the Department of Education investigation in the
late 1980s (which found no discrimination). The proportion of Asian-American students rose
from 13.3% in the Class of 1992 (admitted in 1988, when the investigation began) to 17.9% in
the Class of 1994 (admitted in 1990, when it ended). Connolly Ex. 231. But that also reflected a
growth trajectory that had begun long before any investigation—from 3.4% in the Class of 1980,
to 7.5% in the Class of 1984, to 9.1% in the Class of 1988. Id. During the same time, the AsianAmerican proportion of the applicant pool grew sharply. Id.
-7-

a long upward trend, which is similar to trends for African-American and Hispanic students. See
Ellsworth Ex. 62. Those trends do not show discrimination; they show the success of Harvard’s
laudable efforts, including recruiting and financial aid, to build the learning community it values.
B.

There Is No Statistical Evidence Of Discrimination

Given the absence of documentary or testimonial evidence that could allow its intentional
discrimination claim to survive summary judgment, SFFA must rest instead on the analysis
submitted by its statistical expert, Dr. Arcidiacono. But Dr. Arcidiacono’s analysis is far too
unreliable to allow SFFA’s claim to proceed to trial. SFFA’s attempt to distract from the flaws
of Dr. Arcidiacono’s analysis by attacking Dr. Card cannot remedy those flaws.
First, SFFA argues (at 15) that Dr. Card had “no justification” for including in his model
a group the experts refer to as ALDC applicants—“‘recruited athletes, applicants whose parent or
parents attended Harvard or Radcliffe as an undergraduate, applicants whose names appeared on
a ‘Dean’s interest’ or ‘Director’s interest’ list, and children of Harvard faculty and staff.’” SFFA
argues that those applicants should be excluded because they are “not similarly situated to the
rest of the pool,” in the sense that they are admitted at higher rates than other applicants. That is
fundamentally incorrect. Applicants who fall into those groups may have a greater likelihood of
admission relative to otherwise similar applicants outside the groups—but, as Dr. Arcidiacono
conceded, so do other kinds of applicants whom he did not exclude. See Ellsworth Ex. 166 at
96:24-99:18. The correct way to account for the effect on an applicant’s likelihood of admission
of falling within one of the ALDC categories is to control for that fact (as Dr. Card does), not to
exclude all applicants in those categories (as Dr. Arcidiacono does). See Harvard SJ Opp. 7-8.
Dr. Arcidiacono’s decision to exclude ALDC applicants—but not applicants who possess
other characteristics that increase their likelihood of admission—is transparently directed toward
finding a negative effect of Asian-American ethnicity. For ALDC applicants, both experts agree,
-8-

Asian-American ethnicity has if anything a positive effect on the likelihood of admission, so
excluding ALDC applicants from the analysis causes Dr. Arcidiacono’s estimate of the effect of
Asian-American ethnicity to become more negative. See Ellsworth Ex. 35 at 69; Ellsworth Ex.
37 ¶ 98, ¶ 105 & Card Ex. 13; Harvard SJ Opp. 7. In effect, Dr. Arcidiacono was able to achieve
the result he desired by purposefully omitting from his analysis subsets of the applicant pool in
which Asian-American applicants perform especially well.
SFFA also fails to grapple with the fact that—even if one credits Dr. Arcidiacono’s
flawed methodological choices—his model still would not support an inference of intentional
discrimination based on race, absent some explanation of why that discrimination would be
manifested only among a subset of Asian-American applicants (i.e., non-ALDC applicants).
Neither SFFA nor Dr. Arcidiacono tries to explain that pattern. SFFA instead argues (at 15) that
it is unsurprising “[t]hat Harvard would treat the small number of Asian Americans who qualify
for this special treatment [as ALDC applicants] better than those who are not fortunate enough to
be legacies or athletes.” But that misses the point. The pattern SFFA needs to explain is not that
Asian-American applicants in an ALDC category are admitted at a higher rate than AsianAmerican applicants who are not in those categories; that is unsurprising, since ALDC applicants
generally have a higher rate of admission than non-ALDC applicants. Rather, the finding SFFA
cannot explain is that—according to both experts—Asian-American ALDC applicants are treated
at least as well as White ALDC applicants. See Ellsworth Ex. 35 at 69; Ellsworth Ex. 37 ¶ 98.
SFFA also ignores that Dr. Card took what Dr. Arcidiacono acknowledged was an alternative
approach to dealing with the possibility that the effect of ALDC attributes might vary by race:
He included interactions between race and the ALDC variables. See Ellsworth Ex. 37 ¶ 99.
When he did so, his results were materially unchanged. See id.; id. ¶ 107 & Card Ex. 14.

-9-

Second, SFFA attacks (at 16-18) Dr. Card’s inclusion of the personal rating, on the
flawed theory that the rating reflects bias against Asian-American applicants. SFFA bases that
argument largely on Dr. Arcidiacono’s finding of a statistical association between AsianAmerican ethnicity and lower personal ratings. But there is no basis to infer that that association
reflects bias, rather than merely the effects of unobserved non-academic factors. See Harvard SJ
Mem. 42-44; Harvard SJ Opp. 10-15. In modeling the academic and extracurricular ratings, for
example, Dr. Arcidiacono found a statistical association between Asian-American ethnicity and
higher ratings—but he attributed that effect not to pro-Asian bias but to unobserved factors.
There is no reason that correct interpretation should not also be attributed to his personal-rating
regression. See Harvard SJ Mem. 43-44; Harvard SJ Opp. 11-12. 6
SFFA tries (at 17) to justify Dr. Arcidiacono’s inconsistent interpretations by arguing that
“‘Asian-American applicants have observed characteristics associated with higher personal
ratings.’” As a result, SFFA argues, one can assume that Asian-American applicants are also
stronger than White applicants in unobserved characteristics associated with higher personal
ratings—and thus, it argues, Dr. Arcidiacono’s finding cannot be explained by unobserved
factors. But SFFA’s premise is faulty: As Dr. Card explains, the applications submitted by
Asian-American applicants were slightly weaker than those of White applicants across observed
non-academic measures of the sort that inform the personal rating. See Ellsworth Ex. 37 ¶ 46.
The data make that clear in several ways. First, Dr. Card showed that Asian-American
applicants were slightly less strong than White applicants, on average, across the full range of

6

SFFA misunderstands Harvard’s argument about Dr. Arcidiacono’s attempts to model the
academic, extracurricular, and personal ratings. According to SFFA (at 16), “Harvard claims
that Professor Arcidiacono’s attack on the personal rating is flawed because it would necessarily
mean that there is ‘bias in favor of Asian-Americans in academic and extracurricular ratings[.]’”
But Harvard’s actual point, as discussed above, is that none of the ratings reflects bias.
- 10 -

non-academic factors included in the model—even when he excluded any effect of the personal
rating or of ALDC attributes. Id. ¶ 53 & Card Ex. 8. Second, Dr. Arcidiacono’s own model of
the personal rating shows that, as non-academic factors are added to the model, the estimated
negative effect of Asian-American ethnicity shrinks—suggesting that, if the model were able to
control fully for non-academic factors, the supposed bias would disappear. Id. ¶ 46; Ellsworth
Ex. 33 ¶ 148. Third, Dr. Card showed Asian-American applicants received lower ratings than
White applicants, in aggregate, from alumni interviewers and from admissions officers assessing
the recommendation letters submitted by teachers and guidance counselors. See Ellsworth Ex.
37 ¶ 48 & Card Exs. 4, 5. That information is fundamental to the personal rating. Id. ¶ 49.
Finally, SFFA attacks Dr. Card (at 18-19) for his choice to model the admissions process
on a year-by-year basis. Again, SFFA is disregarding core statistical concepts. SFFA first
argues (at 18) that Dr. Arcidiacono’s model “captures any year-to-year variations in the
competitiveness of the applicant pool” because it “control[s] for [the] application cycle.” But
that half-measure does not solve the problem with Dr. Arcidiacono’s approach, because
controlling for the year does not fully capture year-to-year variations in the process; it still makes
the implausible assumption that the effect of each factor in the admissions process is the same
from year to year (for example, that Harvard values certain intended concentrations equally from
year to year). See Ellsworth Ex. 33 ¶¶ 104-105 (explaining the flaw in that assumption); see also
Ellsworth Ex. 37 ¶ 84 (explaining that Dr. Arcidiacono’s selective use of interaction terms does
not solve the problem). SFFA also wrongly argues (at 18-19) that pooling allows Dr.
Arcidiacono’s model to have greater statistical precision. Dr. Card’s approach—averaging the
results of his year-by-year models across all six years—actually achieves greater precision than
Dr. Arcidiacono’s. See Ellsworth Ex. 37 ¶¶ 81-82.

- 11 -

In sum, SFFA’s attacks on Dr. Card cannot rescue Dr. Arcidiacono’s flawed analysis.
SFFA lacks evidence sufficient to proceed to trial on its intentional discrimination claim.
III.

HARVARD DOES NOT ENGAGE IN RACIAL BALANCING
SFFA also fails to show how a reasonable factfinder could rule in favor of its racial

balancing claim. There is no evidence—documentary, testimonial, or statistical—that Harvard
seeks any particular racial composition of its admitted class.
1.

SFFA’s argument is replete with blatantly false statements—(1) that “Harvard

concedes that it sets racial targets and engineers its process … to ensure that it hits those targets,”
Opp. 19; (2) that Director McGrath “confessed that Harvard intentionally shapes its class in the
very way the Supreme Court and the First Circuit have condemned as illegal,” id. at 21; and (3)
that Dean Fitzsimmons “testified that Harvard sets racial targets and never misses them,” id. For
the first accusation, SFFA cites only its own brief; for the latter two, SFFA cites nothing.
The reason SFFA cites no evidence is that it has none. There is no evidence that even a
single employee of the Admissions Office, let alone anyone serving on the 40-person
Admissions Committee over many years, sought to admit a class with any preordained racial
composition. See Harvard SJ Opp. 30-34. To the contrary, the record shows beyond dispute that
the Admissions Committee has no racial targets. Director McGrath, for example, testified that
the Admissions Office seeks broad racial diversity and is “not ever looking for a particular
number or percentage.” Ellsworth Ex. 98 at 240:18-19; see also id. 252:4-10. The only targets
that exist are overall targets for the number of admitted students—an uncontroversial practice
necessary given the set number of beds on campus. See Harvard SJ Opp. 32.
SFFA insists (at 20) that it need not prove the Admissions Committee “aims for a precise
numerical target for each racial group.” In the cases on which SFFA relies, the racial goals at
issue were defined as numerical ranges. See Parents Involved in Cmty. Sch. v. Seattle Sch. Dist.
- 12 -

No. 1, 551 U.S. 701, 726 (2007); Cavalier ex rel. Cavalier v. Caddo Par. Sch. Bd., 403 F.3d 246,
248 (5th Cir. 2005); see also Perrea v. Cincinnati Pub. Sch., 709 F. Supp. 2d 628, 635 (S.D.
Ohio 2010). But they were still numerical goals, which required race to be considered in a
formulaic, outcome-determinative way, rather than the “flexible, nonmechanical way” used by
Harvard in accordance with Supreme Court precedent, Grutter, 539 U.S. at 334.
SFFA has supplied not a shred of evidence that Harvard sets racial targets, or even
targeted ranges. Dean Fitzsimmons and Director McGrath consider information regarding the
composition of the tentatively admitted class—in many dimensions, not just race, see Harvard SJ
Opp. 33—for various purposes. One is to estimate the overall proportion of admitted students
who are likely to accept their offers of admission, since students from different groups tend to
accept their offers at different rates. See id. at 32. Another is to see whether there has been “a
dramatic dropoff” from the prior year’s representation of any group. Connolly Ex. 16 at 269:11270:4. If there is “a dramatic change, not corresponding to a dramatic change in the number of
applicants” from that group, Dean Fitzsimmons or Director McGrath might “raise the question of
whether” the Committee was “giving good, careful consideration to those candidates from that
group that seemed to be doing less well.” Id. at 240:6-19; see also Ellsworth Ex. 98 at 201:19202:6; Ellsworth Ex. 120 at 313:5-6. That is entirely proper. Under no circumstances does the
Committee aim to achieve “a particular number or percentage” of students from the group in
question. Ellsworth Ex. 98 at 240:18-19.
The Supreme Court has recognized that “‘[s]ome attention to numbers’ … does not
transform a flexible admissions system into a rigid quota.” Grutter, 539 U.S. at 336. Indeed, the
Court observed—quoting the Harvard admissions plan appended to Justice Powell’s opinion in
Bakke—that “there is of course ‘some relationship between numbers and achieving the benefits

- 13 -

to be derived from a diverse student body, and between numbers and providing a reasonable
environment for those students admitted.’” Id. (quoting Regents of Univ. of Cal. v. Bakke, 438
U.S. 265, 323 (1978) (opinion of Powell, J.)). That is why the Grutter Court was untroubled by
the University of Michigan Law School’s practice of consulting reports that kept “track of the
racial and ethnic composition of the class” on a daily basis, id.—far more frequently than the
leaders of the Harvard Admissions Office consult “one pagers,” see Harvard SJ Mem. 19 n.13. 7
Their periodic consultation of statistics—in an effort to seek broad diversity through an
individualized admissions process that considers race flexibly as one among many factors—is
exactly the kind of “‘attention to numbers’” that Grutter approved, 539 U.S. at 336.
The Grutter Court also noted that the year-to-year variation in the racial composition of
the law school’s enrolled classes was “inconsistent with a quota.” 539 U.S. at 336. Again, the
same is true here: The racial composition of Harvard’s admitted classes fluctuated meaningfully
over 19 years. Ellsworth Ex. 33 ¶¶ 193-194 & Card Ex. 31-34. SFFA’s own expert admitted
that the data showed “changes in the fraction of admitted students by race/ethnicity over time”
and never suggested those changes were statistically insignificant. Ellsworth Ex. 35 at 58.
SFFA responds (at 21) that “the admitted share of each minority racial group was stable” across
“the six years for which SFFA received database information.” But SFFA’s statistics contradict

7

SFFA argues (at 19 n.8) that unlike in Grutter—where “the Law School’s admissions
officers testified without contradiction that they never gave race any more or less weight based
on the information contained in [the daily] reports,” 539 U.S. at 336—Dean Fitzsimmons and
Director McGrath use the one-pagers “extensively to give race more weight in order to achieve
the desired racial balance.” Continuing a pattern, the only source SFFA cites for that accusation
is its own Rule 56.1 statement, SFFA SJ SMF ¶¶ 246-255. But the assertions in that statement—
even if taken at face value, contra Harvard’s Response to SFFA’s Local Rule 56.1 Statement
(Dkt. 437)—reflect only Harvard’s flexible, nonmechanical consideration. They do not suggest
that Dean Fitzsimmons or Director McGrath—let alone the 40-member Admissions Committee
as a whole—uses the one-pagers to “give race more weight” or seek a “desired racial balance”
for the admitted class.
- 14 -

its claim. For example, SFFA notes that “the admitted share of Hispanics was always between
8.8% and 11.6%”—neglecting to acknowledge that that is a difference of more than 30%.
2.

SFFA then turns (at 21-25) to the only theory of racial balancing its expert did

attempt to support—namely that, during a particular three-year period, Harvard supposedly
sought to equalize the admission rate of African-American applicants, as defined by a particular
method for categorizing applicants by race (IPEDS), with the overall admission rate. SFFA has
as little evidence to support that theory as one would expect given its implausibility.
The foundation of SFFA’s flawed theory is the argument, offered by Dr. Arcidiacono,
that there is only a 0.2% probability that the observed alignment of admission rates occurred by
chance. SFFA argues (at 22) that, rather than “disput[ing] Professor Arcidiacono’s findings,”
Harvard “throws out various theories for why this was all due ‘simply to chance.’” To the
contrary, Harvard very much “dispute[s] Professor Arcidiacono’s findings,” as the likelihood
that the alignment of admission rates occurred by chance is far greater than 0.2%.
As Dr. Card explains, the problem with Dr. Arcidiacono’s analysis is that it focuses on
the likelihood of observing an alignment for a particular racial group, using a particular
definition of racial categories, in a particular three-year period. See Ellsworth Ex. 37 ¶ 160.
Given the divergence between Dr. Arcidiacono’s bizarre theory of racial balancing and the one
in SFFA’s complaint (Dkt. 1 ¶¶ 288-304), it is clear Dr. Arcidiacono’s theory arose when he
went hunting in search of any pattern that could arguably support SFFA’s claims. But given the
many places he could have looked, the likelihood of finding a statistical alignment in any one of
those places considerably exceeds 0.2%. For example, even focusing solely on rates of offers of
admission, Dr. Arcidiacono could have looked for an alignment over any three-year stretch, for
any racial group, by any definition of racial categories—and there are 92 such opportunities to

- 15 -

find the type of pattern he observes, radically increasing the odds of finding it by chance. See
Ellsworth Ex. 37 ¶ 161. Thus, there is no statistical basis to conclude that the pattern Dr.
Arcidiacono observed reflects deliberate manipulation as opposed to happenstance. Id.
SFFA responds (at 23) that the “92 opportunities” Dr. Card considers are not “equally
likely.” For example, SFFA says, there is no reason to think Harvard would adopt a “floor” in
the admission rate for White applicants. That argument by no means negates the point, however,
because the 92 opportunities Dr. Card discusses are but a small fraction of the actual number of
places in which Dr. Arcidiacono could have looked to find a supposedly suspicious pattern. As
Dr. Arcidiacono recognizes, if Harvard were manipulating its process by race, it might not seek
to equalize the admission rate for applicants of a given race with the overall admission rate. It
might instead, more straightforwardly, seek to admit a minimum number of students of that race,
or a minimum proportion of students of that race, or any of many other possibilities. See
Ellsworth Ex. 35 at 56; Ellsworth Ex. 37 ¶ 159. Any such pattern would be far more probative
evidence of racial balancing, yet Dr. Arcidiacono finds none.
SFFA’s other supposed evidence of racial balancing (at 22-24) fares no better. That
evidence—including Dean Fitzsimmons’s expression of concerns about the soundness of the
IPEDS methodology, and the inclusion of IPEDS statistics on the reports he periodically
received—comes nowhere close to showing racial balancing. Moreover, SFFA’s account of it is
inaccurate. Most notably, SFFA is incorrect in arguing (at 23) that “Harvard relaxed its
standards for admitting single-race African Americans beginning with the 2017 admissions
cycle.” Dr. Arcidiacono made that argument in his rebuttal report, Ellsworth Ex. 35 at 59-61, but
as Dr. Card explained, Dr. Arcidiacono’s analysis reflected “a critical calculation error … that,

- 16 -

when corrected, reverses his key finding.” Ellsworth Ex. 37 ¶ 164; see also id. ¶¶ 165-166.
Indeed, Dr. Arcidiacono conceded that his calculation had been erroneous. Dkt. 415, Ex. C. 8
SFFA’s theory ultimately makes no sense. If Harvard were interested in manipulating its
admissions process to achieve a particular racial balance, why would it focus on the admission
rate of applicants of a particular group, as opposed to—for example—the number or proportion
of students of that group? Why would it focus on a particular (narrow) definition of a racial
category that is used for mandated federal reporting, rather than the definition used to report the
racial composition of the class to the Harvard community and the public (and for that matter used
by Dr. Arcidiacono in the balance of his analyses)? And why, if this nefarious manipulation
actually took place, would there be not a single document evincing it and not one witness
testifying about it? SFFA has no answer to any of those questions. Given the lack of evidentiary
support—or logic—for its theory, SFFA’s racial balancing claim cannot survive.
IV.

HARVARD CONSIDERS RACE IN THE MANNER PERMITTED BY SUPREME COURT
PRECEDENT
1.

SFFA’s opposition reprises (at 26) its argument that “Harvard is not pursuing the

only conception of ‘student body diversity’ the Supreme Court has ever endorsed—enrolling a
‘critical mass’ of underrepresented minority students.” But as Harvard has explained (Harvard
SJ Opp. 35-36), the Supreme Court rejected that argument when it was proffered to contend that
the University of Texas’s admissions practices were unlawful because the university had not
sufficiently defined “the level of minority enrollment that would constitute a ‘critical mass.’”
Fisher v. Univ. of Tex. at Austin (Fisher II), 136 S. Ct. 2198, 2210 (2016). The Court explained
8

Trying to limit his concession, Dr. Arcidiacono argued the effects he found “remain[ed]
significant … at an 8.4% significance level and … at a 9.3% significance level.” Dkt. 415, Ex.
C. But that is not the conventional threshold of statistical significance, or the one Dr.
Arcidiacono used throughout his reports, Ellsworth Ex. 166 at 305:16-306:3. Dr. Arcidiacono
was able to characterize the effects as statistically significant only by moving the goalposts.
- 17 -

that “the compelling interest that justifies consideration of race in college admissions is not an
interest in enrolling a certain number of minority students,” but rather an interest in “obtaining
the educational benefits that flow from student body diversity.” Id. (internal quotation marks
omitted). The Court thus repudiated SFFA’s view that universities must define their objectives
in terms of critical mass.
Harvard is pursuing exactly the objective the Supreme Court has validated. For decades
Harvard has articulated the importance of diversity of all kinds, including racial diversity, to the
educational experience of its students. See Harvard SJ Mem. 7-9. And it recently reaffirmed
those educational judgments in a report produced by a committee of Harvard faculty members
and unanimously endorsed by the Faculty of Arts and Sciences. See id. at 8-9 & n.7. That is the
sort of “reasoned, principled” explanation for seeking diversity that the Supreme Court has held
is entitled to “deference.” Fisher II, 136 S. Ct. at 2208 (internal quotation marks omitted).
Moreover, the faculty report articulates exactly the sorts of objectives the Supreme Court deemed
sufficiently “concrete and precise” in Fisher II. See id. at 2211 (“the destruction of stereotypes,”
the “promotion of cross-racial understanding,” “the preparation of a student body for an
increasingly diverse workforce and society,” and the “cultivation of a set of leaders with
legitimacy in the eyes of the citizenry” were all “concrete and precise goals” (some internal
quotation marks and alterations omitted)).
SFFA accuses Harvard of having “adopt[ed] an institutional position that the use of race
always will be necessary.” Opp. 27 (emphasis SFFA’s). That is directly contrary to the record.
The only source SFFA cites for its assertion is its own summary-judgment brief (at 40). That
brief, in turn, cites SFFA’s Rule 56.1 statement (SFFA SJ SMF ¶ 830), which in turn cites the
following passage from Dean Fitzsimmons’s deposition:

- 18 -

[Q.]

Four years ago, could you imagine evidence that would change your mind as to
whether or not Harvard should stop using race in the admissions process?

…
A.

You know, I think—again, I think a reasonable person or reasonable institution
would always keep an open mind about anything. I mean, new information
comes in but despite massive efforts that people would reasonably call raceneutral efforts to make a difference, there’s still no case in our minds that, despite
all the things that we’ve done that we could achieve our goal of having a diverse
class and an effective educational experience without using race.

Connolly Ex. 9 at 178:21-179:11. Contrary to SFFA’s erroneous characterization, the actual
quotation shows Dean Fitzsimmons said that he had an “open mind” and that “new information”
could “come[] in,” but that he did not think at that time that Harvard could achieve its objectives
without considering race. On the basis of that answer, SFFA manages to accuse Harvard of
“adopting an institutional position that the use of race will always be necessary”—ignoring the
explicit recommendation by Harvard’s own Committee to Study Race-Neutral Alternatives in
Harvard College Admissions that Harvard reevaluate the need to consider race in five years,
Ellsworth Ex. 47 at HARV00097328, a fact that SFFA relegates to a footnote (at 27 n.10).
SFFA next accuses Harvard (at 27-28) of adopting “irreconcilable positions” on the use
of race. How could it be, SFFA asks with mock incredulity, that Harvard (1) considers race only
in a flexible and individualized way, without pursuing any numerical quota; but (2) pays enough
“attention to numbers” to seek a broadly diverse class, and believes “a significant decline in
African-American and Hispanic enrollment” would inhibit its educational goals? The answer, as
SFFA knows, is that Harvard is doing exactly what the Supreme Court has said universities may
and must do if they consider race in admissions. Universities may consider race only “flexibly
as a ‘plus’ factor in the context of individualized consideration of each and every applicant,” and
must not “establish quotas for members of certain racial groups,” Grutter, 539 U.S. at 334—but

- 19 -

“[s]ome attention to numbers, without more, does not transform a flexible admissions system
into a rigid quota,” id. at 336 (internal quotation marks omitted).
2.

SFFA next argues (at 28-31) that, “[e]ven if Harvard were pursuing a permissible

end,” its consideration of race would still be unlawful because race has too great an effect on
admissions outcomes. SFFA’s argument is again inconsistent with Supreme Court precedent.
The parties have thoroughly explored what the data do and do not show about the effect
of race in the admissions process. It is clear that race alone explains little about the overall
pattern of admissions outcomes, and that race has virtually no effect on most applicants’
likelihood of admission. See, e.g., Ellsworth Ex. 33 ¶¶ 177-181; Ellsworth Ex. 35 at 49-51;
Ellsworth Ex. 37 ¶¶ 137-146. It is also true that for certain applicants—namely, AfricanAmerican and Hispanic candidates who would be highly competitive regardless whether race
were considered—the consideration of race does increase the likelihood of admission. See
Ellsworth Ex. 33 ¶¶ 180-181 & Card Ex. 28; Ellsworth Ex. 37 ¶ 143 & Card Ex. 23. But that is
little different from the effect of other meaningful factors in Harvard’s admissions process, like
academic, extracurricular, or personal strength. See Ellsworth Ex. 37 ¶ 143 & Card Ex. 23. And,
for several reasons, it does not run afoul of Supreme Court precedent.
First, even if statistical modeling estimates that race has a substantial marginal effect on
the likelihood of admission for otherwise highly competitive candidates, see id., that does not
mean race dictates whether any such candidate is admitted. Even if an otherwise highly
competitive applicant’s likelihood of admission is estimated to rise materially as a result of the
consideration of race, such an applicant might very well have been admitted without
consideration of race, and might very well be denied admission notwithstanding the
consideration of race. As the Supreme Court held in Grutter, the fact that consideration of race

- 20 -

may tip the balance from denial to admission for some candidates is itself not problematic; after
all, “the same could be said of the Harvard plan discussed approvingly by Justice Powell in
Bakke, and indeed of any plan that uses race as one of many factors.” 539 U.S. at 339.
Second, and relatedly, race is just one of many factors that may tip a highly competitive
candidate over the line from denial to admission. See Ellsworth Ex. 37 ¶¶ 143-146. That would
also be true, for example, of an excellent academic or extracurricular rating. Id. ¶ 143 & Card
Ex. 23. SFFA finds that comparison problematic (Opp. 30-31), but it is hard to see why.
Harvard regards diversity of all forms as central to its educational objectives, just as it regards
academic excellence and a vibrant extracurricular community as central to its educational
objectives. As long as the consideration of race does not overwhelm other factors, and still
allows the applicant to be “evaluated as an individual,” it does not “make[] an applicant’s race or
ethnicity the defining feature of his or her application,” Grutter, 539 U.S. at 337. 9
Finally, if the consideration of race did not meaningfully affect the admission chances of
highly competitive candidates, it would have little effect on the diversity of the admitted class—
in which case SFFA would surely argue that the consideration of race would fail strict scrutiny
because it did not advance Harvard’s educational objectives. In Fisher II, for example, the
plaintiff argued “that considering race was not necessary because” it “had only a ‘minimal

9

SFFA argues (at 28-29) that “[t]he relative size of the preference” for applicants of
certain racial groups “was one of the chief reasons the University of Michigan’s point-based
system” for undergraduate admissions was held unconstitutional in Gratz v. Bollinger, 539 U.S.
244 (2003). That argument misstates the law. In Gratz, the deficiency in the admissions
program centered on the fact that the preference given on the basis of race was inflexible and
automatic. See id. at 270 (20 points were given for race “to every single ‘underrepresented
minority’ applicant”); id. at 271 (the program did not provide “individualized consideration”
because it “automatically distribute[d] 20 points to every single applicant from an
‘underrepresented minority’ group”). Here, by contrast, Harvard considers race only flexibly
and in an individualized manner. See Harvard SJ Mem. 21-25.
- 21 -

impact in advancing the University’s compelling interest’”—and that was in a case where the
consideration of race had increased Hispanic representation by 54% among the first-year
students admitted via “holistic review,” and increased African-American representation in the
same group by 94%. 136 S. Ct. at 2212 (some internal quotation marks and alterations omitted).
The Supreme Court rightly rejected the argument, explaining that “[t]hose increases … show[ed]
that consideration of race ha[d] had a meaningful, if still limited, effect on the diversity of the
University’s freshman class.” Id. The same is true here.
V.

HARVARD COULD NOT ACHIEVE ITS EDUCATIONAL OBJECTIVES WITHOUT
CONSIDERING RACE
SFFA has noticeably little to say about the conclusions of Harvard’s Committee to Study

Race-Neutral Alternatives in Harvard College Admissions. After seven months of careful study,
the Committee concluded that, at present, no combination of race-neutral measures could
advance Harvard’s interest in obtaining the educational benefits that flow from a student body
that is diverse in many ways and excellent in many dimensions. Unable to mount a serious
challenge to those judgments, SFFA resorts to collateral challenges, all of which are meritless.
First, SFFA charges (at 31) that “Harvard admits that it did not begin to consider raceneutral alternatives until 2017.” That is incorrect several times over. As an initial matter,
Harvard has long pursued diversity through many strategies other than the consideration of race.
See Harvard SJ Mem. 26-27. It has adopted a nation-leading financial aid policy to ensure that a
lack of funds does not keep any student from attending Harvard. It makes tremendous efforts to
recruit a diverse applicant pool. Admissions officers give particular consideration to lowerincome applicants and those in the first generation of their families to attend college. Once a
diverse group of students has been admitted, Harvard makes extensive efforts to encourage them

- 22 -

to matriculate. And Harvard regularly reviews its practices with an eye toward increasing
diversity; for example, it eliminated and then reinstated Early Action for that reason. Id. at 32.
If SFFA’s claim is that Harvard did not begin a formal study of the effects of ending
race-conscious admissions until 2017, that, too, is inaccurate. A university-wide committee
began such a review in 2014, shortly after the Supreme Court decided Fisher I, then paused its
work, reasonably, after SFFA brought this suit. Harvard SJ Mem. 8; Harvard SJ Opp. 41-42.
That committee’s work as to Harvard College was then resumed by two new committees,
beginning shortly thereafter. Harvard SJ Mem. 8-10; Harvard SJ Opp. 42.
In any event, because this lawsuit seeks only forward-looking relief, it is irrelevant
whether Harvard formally considered race-neutral alternatives at some point in the past. See
Harvard SJ Opp. 40-41. What matters is that, to the extent Harvard had an obligation to convene
such a formal study before considering race in the future, it has complied. SFFA (at 31) derides
the work of the three senior deans appointed to the committee, but its invective cannot refute the
rigor of the committee’s work and does not constitute actual evidence, much less evidence
sufficient to withstand summary judgment. The committee members spent many hours
discussing relevant materials before producing a thorough report. See Harvard SJ Opp. 43.
Second, to the extent SFFA does contest the committee’s findings (at 32-37), its
arguments are meritless. Those findings have ample support, especially in the reports of Dr.
Card, which the committee considered and which SFFA’s reports, also carefully considered by
the committee, did not rebut.
SFFA claims that the committee did not consider what would happen if, in addition to
eliminating the consideration of race, Harvard also did everything possible to promote diversity
in race-neutral ways. That again reflects a misstatement of the record and ignores much of the

- 23 -

committee’s report. The committee made clear that certain combinations of race-neutral
practices could allow Harvard to achieve a comparably diverse class, but that they would
unacceptably compromise other attributes of the class that are also essential to fulfilling
Harvard’s mission. See generally Ellsworth Ex. 47.
For example, the committee considered (among others) the three combinations of
possible practices that SFFA calls (at 32-33) Kahlenberg Simulations 6 and 7 and Card
Simulation 4x. All three, however, would sharply reduce the proportion of admitted applicants
who rate highly on attributes of excellence important to Harvard—in particular the proportion of
applicants with the highest academic, extracurricular, personal, and athletic ratings. Ellsworth
Ex. 37 ¶¶ 192-197 & Card Ex. 26. The committee explained why those changes would be
inimical to Harvard’s educational objectives. See Ellsworth Ex. 47 at HARV00097323.
SFFA declares (at 33) that those changes in the proportion of students receiving a top
academic rating “would be modest,” and that “Harvard’s ‘reputation for academic excellence’
would remain undiminished.” But SFFA provides no basis to contest the educational judgments
of senior Harvard officials responsible for student academic achievement that Harvard’s
academic excellence would suffer if the proportion of the most academically exceptional
students declined by 13-19%. Those officials, not SFFA, are responsible for stewardship of one
of the world’s leading institutions of higher learning, and the Supreme Court has repeatedly
stressed that universities cannot be forced “to choose between a diverse student body and a
reputation for academic excellence.” Fisher II, 136 S. Ct. at 2213.
SFFA then devotes several pages (at 34-36) to practices it claims inhibit diversity. SFFA
contends that Harvard did not study the effects of eliminating those practices in addition to
undertaking other race-neutral measures to pursue diversity. That is incorrect. In fact, Dr. Card

- 24 -

not only studied the effects of eliminating those practices as an independent step; he also
examined what would happen if Harvard eliminated the practices and undertook various other
race-neutral measures, and found that doing so would not allow Harvard to achieve a class that
was comparable in diversity and excellence to Harvard’s current classes. That was the purpose
of the “Simulation 4x” discussed above. See Ellsworth Ex. 33 ¶¶ 233, 238; see also Ellsworth
Ex. 47 at HARV00097322-324 (committee report referencing Dr. Card’s analysis).
SFFA next argues (at 34-36) that the challenged practices do not actually advance
asserted objectives other than the pursuit of diversity and educational excellence, such as the
maintenance of a strong alumni community, and that, if they did, those other objectives would
not justify the consideration of race. But those issues are irrelevant here, because no race-neutral
alternative practices—including the elimination of the challenged practices, combined with
increased consideration for applicants of lower socioeconomic status—would allow Harvard to
achieve comparable diversity without compromising the academic excellence of the class. See
Ellsworth Ex. 33 ¶ 243; Ellsworth Ex. 47 at HARV00097327.
Finally, SFFA asserts (at 36-37) that “Harvard could increase its recruiting efforts”;
“increase its financial aid commitment”; take “geography into consideration by, for instance,
using zip codes”; “admit more transfers from community colleges”; or “eliminate the preference
for those applying early action.” SFFA points to nothing to suggest those measures would be
workable or “would help increase student body diversity beyond current levels.” And SFFA’s
conclusory argument wholly fails to address the committee’s contrary conclusions.
CONCLUSION
For the reasons stated above, and in Harvard’s memorandum in support of its summary
judgment motion, the Court should grant summary judgment to Harvard on all remaining counts
of the Complaint.
- 25 -

Respectfully submitted,
/s/ Seth P. Waxman
Seth P. Waxman (pro hac vice)
Paul R.Q. Wolfson (pro hac vice)
Daniel Winik (pro hac vice)
WILMER CUTLER PICKERING
HALE AND DORR LLP
1875 Pennsylvania Ave. NW
Washington, D.C. 20006
Tel: [phone]
Fax: [phone]
[e-mail]
[e-mail]
[e-mail]
Debo P. Adegbile (pro hac vice)
WILMER CUTLER PICKERING
HALE AND DORR LLP
7 World Trade Center
250 Greenwich Street
New York, NY 10007
Tel: [phone]
Fax: [phone]
[e-mail]
William F. Lee (BBO #291960)
Felicia H. Ellsworth (BBO #665232)
Andrew S. Dulberg (BBO #675405)
Elizabeth Mooney (BBO #679522)
WILMER CUTLER PICKERING
HALE AND DORR LLP
60 State Street
Boston, MA 02109
Tel: [phone]
Fax: [phone]
[e-mail]
[e-mail]
[e-mail]
[e-mail]
Dated: August 27, 2018

Counsel for Defendant President and
Fellows of Harvard College

- 26 -

CERTIFICATE OF SERVICE
I hereby certify that this document filed through the CM/ECF system will be sent
electronically to the registered participants as identified on the Notice of Electronic Filing.
/s/ Seth P. Waxman
Seth P. Waxman