Students for Fair Admissions, Inc. v. President and Fellows of Harvard College et al
MEMORANDUM in Opposition re 64 MOTION to Compel Production filed by President and Fellows of Harvard College. (Ellsworth, Felicia)
UNITED STATES DISTRICT COURT FOR
THE DISTRICT OF MASSACHUSETTS
STUDENTS FOR FAIR ADMISSIONS, INC.,
PRESIDENT AND FELLOWS OF HARVARD
COLLEGE (HARVARD CORPORATION),
Civil Action No. 1:14-cv-14176-ADB
HARVARD’S MEMORANDUM IN OPPOSITION TO SFFA’S MOTION TO COMPEL
Making good on its promise to seek “broad sweeping discovery” in this case (Tr. of July
21, 2015 Status Conference at 26, 28 (Dkt. 84)), Plaintiff Students for Fair Admissions, Inc.
(“SFFA”) has moved to compel Defendant President and Fellows of Harvard College
(“Harvard”) to produce (a) 6,400 complete student application files, (b) the names of, and
contact information for, more than 5,000 alumni interviewers who are not parties to this action
and who play a limited role in the admissions process, and (c) detailed information about transfer
The motion should be denied. In support of its first demand, SFFA has argued that it
requires the applicant files to conduct a statistical analysis described by its expert. However,
SFFA does not need the sensitive personal information contained in hundreds of thousands of
pages’ worth of applicant files to conduct the statistical analysis that it proposes to undertake.
That analysis can be conducted with information contained in the Admissions Office database—
If the Court grants Harvard’s pending Motion to Stay (Dkt. 58), it can defer ruling on this
Motion (or deny the Motion without prejudice), except as necessary in delimiting the scope of
any limited discovery the Court permits during the pendency of the stay.
an approach that would be less costly for the parties and more useful for both the parties and the
Court, as it would not require statistical sampling of the files at all. Because information from
the database will give SFFA what it needs for the statistical analysis it proposes, the “burden or
expense” of producing such a large number of applicant files (and redacting them to shield
students’ identities) would significantly “outweigh” the “likely benefit” of doing so, Fed. R.
Civ. P. 26(b)(2)(C). And because the files are protected by the Family Educational Rights and
Privacy Act (“FERPA”), 20 U.S.C. § 1232g, SFFA bears a “‘significantly heavier’” burden to
justify their production. See, e.g., Black v. Kyle-Reno, 2014 WL 667788, at *2 (S.D. Ohio Feb.
20, 2014). Especially given the personal, indeed intimate, nature of the information in the files,
SFFA cannot meet that burden.
The Court should also deny SFFA’s request for the personal contact information of more
than 5,000 non-party alumni interviewers, who are of marginal relevance to this litigation and
who would be subjected to “annoyance, embarrassment, or oppression,” Fed. R. Civ. P. 26(c)(1),
if haled into it. These alumni each volunteer a few hours each year to interview a handful of
applicants to Harvard College; they should not be dragged into this litigation as a cost of that
assistance. Moreover, any information that SFFA might obtain from alumni interviewers is of
little relevance to the central issue in this case, which concerns Harvard’s Admissions Office’s
practices. Alumni interviewers do not participate in that Office’s admissions decisions; rather,
they assign numerical ratings to the applicants, which are recorded in the database, and file
narrative reports. Thus, the database—together with the 160 files Harvard has agreed to provide
and any depositions of Admissions Office personnel SFFA is permitted to take—will provide
SFFA with a full understanding of the role played by alumni interviews in the admissions
Finally, the Court should deny SFFA’s request for information about the transfer
admission process because it is irrelevant to this action, which challenges Harvard’s decisionmaking in admitting its freshman classes. SFFA has not challenged any decision by Harvard to
deny transfer admission, and none of SFFA’s members has applied for transfer admission.
Moreover, because only a very small number of applicants are admitted as transfer students in
any year, the transfer process does not materially affect the composition of the overall class.
Any connection between the transfer admission process and this action is therefore too
attenuated to justify the discovery that SFFA seeks.
SFFA’s Request For A Large Number Of Applicant Files Should Be Denied
The Court should deny SFFA’s request for the production of 6,400 applicant files for two
reasons. First, with respect to SFFA’s contention that it needs applicant files in order to conduct
statistical analysis, SFFA does not need any applicant files to conduct the statistical analysis its
expert proposes; rather, information from the database will enable SFFA to conduct that
statistical analysis. Second, the “burden or expense” of producing those files would significantly
“outweigh” the “likely benefit” of doing so, Fed. R. Civ. P. 26(b)(2)(C), particularly in light of
the need to redact the files, which contain intensely personal information that applicants have
entrusted to Harvard in the utmost confidence. Indeed, applicant files are protected from
disclosure under both federal and Massachusetts law, and SFFA has not met the high threshold
of justifying their wholesale production in this litigation.
SFFA Does Not Need Applicant Files To Conduct Its Proposed Analysis
SFFA proposes to use the applicant files to conduct “a statistically defensible analysis of
the extent to which race is playing a factor” in Harvard’s admissions process. Pls.’ Mem. in
Support of Mot. to Compel 10, ECF No. 65 (“Pls.’ Mem.”). In particular, SFFA proposes that
“this analysis can be done by assigning values to Harvard’s criteria for scoring applicants and
then determining, through regression analysis, the extent to which race, rather than the other
criteria, appears to be influencing the admissions process.” Id. at 10-11. SFFA’s expert, Dr.
Peter Arcidiacono, explains: “With the raw files in hand, we can code the various factors
Harvard describes as important in determining application subscores (for example, creating an
indicator variable for whether the student was a valedictorian). We can then use regression
analysis to see, for example, whether Asians received lower subscores conditional on the factors
that Harvard describes as important for that subscore.” Declaration of Patrick Strawbridge, Ex.
A ¶ 28, Dkt. No. 66-1 (“Arcidiacono Decl.”).
Harvard disputes the notion that SFFA’s proposed methodology is statistically
appropriate or that it could demonstrate any improper use of race in Harvard’s admissions
process. But whatever the substantive merits of the analysis described by SFFA’s expert, SFFA
does not need applicant files to conduct it, because the Admissions Office database contains
information sufficient for that purpose. Indeed, because the database contains detailed
information about every applicant to Harvard, review of information in the database is superior
for the purposes of this analysis to the statistical-sampling method that SFFA proposes. See
Declaration of Justin McCrary (“McCrary Decl.”). For example, Dr. Arcidiacono suggests (at
¶ 28) that “[w]ith the raw files in hand,” SFFA could “creat[e] an indicator variable for whether
the student was a valedictorian.” But SFFA can create that variable just as easily, if not more so,
by looking at the database, which will include information reported to Harvard about applicants’
class rank, as well as applicants’ self-reported honors and awards. McCrary Decl. ¶ 34.2 Dr.
Professor Arcidiacono’s reference to valedictorian status as a metric is confusing,
because students are typically named valedictorians when they graduate from high school—that
is, after their application has been submitted and in most cases after admission decisions have
been made. To the extent that an applicant might disclose valedictorian status among his or her
honors and awards, however, the database would capture that information.
Arcidiacono also suggests (at ¶ 34) that “SFFA will need to ‘score’ the extracurricular activities
of numerous applicants along various dimensions.” Again, SFFA does not need the files for that
exercise. The database contains information about extracurricular activities that SFFA can use
for the analysis that Dr. Arcidiacono describes: the title of the applicant’s position, the duration
of the applicant’s participation (grade levels, hours per week, and weeks per year), an indicator
of whether the applicant plans to continue the activity in college, and the applicant’s narrative
description of the activity. McCrary Decl. ¶ 35.
In addition, the database contains a depth and richness of information about each
applicant that will assist SFFA in conducting its expert’s proposed analysis. The database
contains hundreds of fields, including, for example:
Applicants’ demographic information, including race and ethnicity
Applicants’ intended careers and academic concentrations
Applicants’ GPAs and standardized test scores
Applicants’ academic honors
The occupations and educational background of the applicants’ parents
Information indicating that the applicant may be financially disadvantaged
Applicant ratings assigned by admissions officers and alumni interviewers
McCrary Decl. ¶ 36. Dr. Arcidiacono identifies nothing of relevance to the analysis he proposes
that SFFA could glean from full application files but not from the database.
Harvard has agreed to produce information from the Admissions Office’s database in
response to SFFA’s Request for Production No. 1. That is more than sufficient for SFFA to
conduct the statistical analysis its expert proposes. Notably, SFFA’s expert himself
acknowledges that an analysis based on a database covering the full universe of freshman
applicants is “preferable” to an analysis based on a mere sample of files. See Arcidiacono Decl.
¶ 19 (“Analyzing the whole population is preferable but tends to be costly.”). There is no need
for the second-best option of sampling where SFFA will have access to extensive information
about every applicant in any class to conduct the analysis that Dr. Arcidiacono proposes.
McCrary Decl. ¶ 37.
SFFA criticizes Harvard for the supposed inconsistency between Harvard’s discovery
response and its proposal in the Joint Rule 26(f) Statement (at 13) to “produc[e] a statistically
significant sampling of redacted applications.” But there is no inconsistency. At the time that
the parties filed their Joint Rule 26(f) statement, Harvard’s counsel had not examined the
database in detail and was not aware of the scope of the information contained within the
database. As the extent of the information in the database became clear, it also became clear that
that information would be sufficient for the statistical analyses in this case and that 160 files
would be sufficient to provide context for those analyses, and for other purposes: to help explain
how Harvard conducts its admissions process, to describe the holistic review undertaken by the
Admissions Office, and to provide a picture of the diverse and exceptional group of applicants
that apply to Harvard every year, only a few of whom can be offered admission. Indeed,
referring to more than 160 files for these purposes would likely be unmanageable. Harvard
therefore has proposed to produce 160 files, with 80 chosen by Harvard and 80 by SFFA,
selected in whatever manner each party prefers. Harvard’s use of the phrase “statistically
significant” in the Rule 26(f) statement may have been infelicitous, but the Court was not
confused; it read that phrase quite correctly to refer simply to a sample large enough to allow the
parties to understand and describe the admissions process. See Tr. of April 30, 2015 Scheduling
Conference 5-6 (Dkt. 43) (initial comments of the Court, describing Harvard’s proposal to
conduct “representative samplings of those applications”).3 Those objectives will be satisfied by
producing information from the database along with 160 applicant files, without resorting to the
more burdensome, more intrusive, and less useful approach of producing an even larger number
of applicant files.
SFFA Cannot Justify The Production Of A Large Number Of Files Even If
They Are Marginally Relevant To Its Proposed Analysis
Any marginal benefit that applicant files might offer to SFFA’s analysis would be heavily
outweighed by the cost and burden of producing the files—a burden to be borne both by Harvard
and by the applicants whose private information would be revealed. See Fed. R. Civ. P.
26(b)(2)(C)(iii). It is particularly extraordinary that SFFA thinks the 6,400 files it seeks would
be only a “preliminary” sample, as it made clear in its Requests for Production and repeatedly
throughout its brief (at 2, 3, 8, 11-13, 15).4 The Court should not authorize this unnecessary
The burden and expense of producing a large number of files,
including the intrusion on student privacy, would outweigh the benefit
Applicant files are replete with personal information of the utmost sensitivity—not just
grades and test scores but deeply personal essays and recommendation letters that touch on
financial hardships, medical conditions and treatments of applicants and their families,
Citing one footnote of an unpublished, out-of-circuit district court opinion, SFFA argues
that “statements in [a] 26(f) report constitute judicial admissions.” Pls.’ Mem. 3 n.3 (citing
Carter v. Reiner, Reiner & Bendett, P.C., No. 06-cv-988, 2007 WL 2221432, at *1 n.1 (D. Conn.
July 30, 2007)). But the footnote in Carter concerns a factual statement in a Rule 26(f) report.
Even if parties are bound by the facts they state in a Rule 26(f) report, that hardly implies—as
SFFA suggests—that no party may ever revisit a discovery proposal made in a Rule 26(f) report
at the very outset of litigation.
SFFA cites various cases (at 12-13) in which courts have approved the sampling of a
comparable proportion of the overall population. But in this case, sampling is not necessary at
all for the purposes of the statistical analysis that SFFA and its expert propose.
information about troubled familial relationships, and the applicants’ most deeply held beliefs
and aspirations. That is, on its own, a reason to limit the production of applicant files to the
minimum necessary for the parties to litigate this case. See, e.g., Fed. R. Civ. P. 26(c)(1) (court
may restrict discovery “to protect a party or person from annoyance, embarrassment, [or]
oppression”); Briddell v. Saint Gobain Abrasives Inc., 233 F.R.D. 57, 61 (D. Mass. 2005)
(“Courts have routinely limited discovery to protect the privacy interests of parties and nonparties alike.”); Whittingham v. Amherst Coll., 164 F.R.D. 124, 127-128 (D. Mass. 1995)
(“[W]hile discovery is usually broad, Plaintiff has not demonstrated that the [employee] files he
seeks, even if marginally relevant, outweigh the privacy interests of these individuals.”).
Moreover, Harvard’s obligations to protect the privacy of its applicants and students
requires it to make significant redactions of each application file produced. Harvard estimates
that, just to remove information directly identifying students and others (such as students’ family
members), each application file produced will require, on average, more than 100 separate
redactions. That is reason enough for the Court to limit the number of files that Harvard is
required to produce. See, e.g., United States ex rel. Gelfand v. Special Care Hosp. Mgmt. Corp.,
No. 05-cv-6079, 2010 WL 2399693, at *2 (E.D.N.Y June 10, 2015) (limiting discovery of
sensitive substance abuse treatment records in recognition that the “heavy burden of redaction
and production … rests primarily with the defendant institution”).
The production of any files, even with redactions, injures the privacy interests of
Harvard’s past applicants and could well chill the candor of future applicants, who might
reasonably fear that their highly personal self-reflections could themselves be turned over to
opposing counsel in litigation. To mitigate these concerns, the Court should limit SFFA to the
discovery of the minimum number of files necessary to litigate this case—an objective fully
satisfied by the 160 files Harvard has agreed to produce.
FERPA sets an especially high bar for the production of educational
An explicit purpose of FERPA is “‘to protect [students’] rights to privacy by limiting the
transferability of their records without their consent.’” Frazier v. Fairhaven School Comm., 276
F.3d 52, 67 (1st Cir. 2002) (quoting 120 Cong. Rec. 39858, 39862 (1974) (joint statement of
Sens. Pell and Buckley)). FERPA codifies “the strong public policy of protecting the privacy of
student records.” Moeck v. Pleasant Valley Sch. Dist., 2014 WL 4988274, at *2 (M.D. Pa. Oct.
7, 2014); see also Brown v. Univ. of Kansas, 2012 WL 612512, at *2 (D. Kan. Feb. 27, 2012)
(“[T]he law recognizes an important privacy interest for students.”); Smyth v. Lubbers, 398 F.
Supp. 777, 796 (W.D. Mich. 1975) (“Congress has … recognized the extreme importance of a
student’s ‘record’ by closely regulating the circumstances under which student records may be
released by school authorities.”).5
Harvard’s application files both constitute and contain educational records protected by
FERPA. Application files for enrolled students are themselves educational records maintained
by Harvard. See 20 U.S.C. § 1232g(a)(6) (defining “student” to include “any person with respect
to whom an educational … institution maintains education records or personally identifiable
information”); 34 C.F.R. § 99.3 (similar). SFFA acknowledges this, but argues (Pls.’ Mem. 15)
that FERPA does not apply to “a large percentage of the produced files” because they will be
The application files are also arguably subject to the protections of the Massachusetts
Privacy Act, which provides that “[a] person shall have a right against unreasonable, substantial,
or serious interference with his privacy.” Mass. Gen. Laws c. 214, § 1B; see Menten v.
Starbucks Corp., No. 09-cv-11545, 2012 WL 1466780, at *9 (D. Mass. Apr. 26, 2012) (Act
applies to private entities); Wagner v. City of Holyoke, 241 F. Supp. 2d 78, 100 (D. Mass. 2003)
(Act applies to disclosures that are not “to the broad public”); Cefalu v. Globe Newspaper Co., 8
Mass. App. Ct. 71, 77 (1979) (Act applies to “manuscripts” and “private letters”).
“for students who never enrolled at Harvard.” SFFA is incorrect. The application files of nonenrolled applicants do contain records covered by FERPA, including, for example, high-school
transcripts. FERPA’s implementing regulations make clear that its restrictions apply not only to
the initial disclosure of information (for example, Harvard’s disclosure of information from a
Harvard student’s educational record), but also to the redisclosure of information that an
educational institution has received (for example, Harvard’s redisclosure of a transcript it has
received from a high school). See 34 C.F.R. § 99.33(a)(1) (an educational institution “may
disclose personally identifiable information from an education record” only under certain
conditions); see also id. § 99.33(b)(2) (noting that redisclosures may be made only under certain
circumstances); Letter from LeRoy S. Rooker, Dir., Family Policy Compliance Office, U.S.
Department of Education (Aug. 16, 2007), available at http://www2.ed.gov/policy/gen/
guid/fpco/ferpa/library/vasexoffenderlaw081607.html (“[A]n institution that receives
information (such as a transcript) on an applicant from a high school or from another
postsecondary institution is required to protect that information and may not redisclose the
information except in according with … the FERPA regulations.”).6
The gravity with which federal and state law treats the disclosure of records like these is
ample evidence that the Court should not indulge SFFA’s blunderbuss request for 6,400 records
that it does not need to conduct the analysis that its expert proposes. Courts have long recognized
that FERPA “places a heavy burden on a party seeking access to student records to demonstrate a
genuine need which outweighs the student’s privacy interest.” Smith v. Duquesne University,
612 F. Supp. 72, 80 (W.D. Pa. 1985); see also, e.g., Moeck, 2014 WL 4988274, at *2 (explaining
that in determining whether to require disclosure of FERPA-covered records, “[c]ourts balance
Counsel for Harvard informed counsel for SFFA of this guidance, and provided a copy of
Director Rooker’s letter, before SFFA filed its motion to compel.
the potential harm to the privacy interests of students with the importance and relevance of the
sought information to resolving the claims before the court,” and that “policy concerns
sometimes require the withholding” of even “relevant information … over disclosure to a party
who argues that the information is needed for his case”); Black, 2014 WL 667788, at *2 (“Courts
have imposed a ‘significantly heavier burden’ on a party requesting the discovery of educational
records to show its interests in obtaining the records outweighs ‘the significant privacy interest of
the students.’”); Zaal v. State, 326 Md. 54, 72 (Md. 1992) (“[W]hen the issue before the court is
whether to allow disclosure of education records covered by the FERPA, a trial judge … must
conduct a balancing test in which the privacy interest of the student is weighed against the
genuine need of the party requesting the information.”); Rios v. Read, 73 F.R.D. 589-598
(E.D.N.Y. 1977) (“Congressional policy expressed in [FERPA] places a significantly heavier
burden on a party seeking access to student records to justify disclosure than exists with respect
to discovery of other kinds of information, such as business records.”).
SFFA cannot meet this high burden by making a sweeping, unfocused request for the
production of 6,400 files—again, just as a “preliminary” matter. Rather than identifying
precisely the number of files that are truly essential to its case, SFFA asks the Court to order the
production of 6,400 applicant files in the hope that it might turn up something of relevance in
those files. But FERPA does not allow “demands … like a bulldozer that levels an entire hill in
the hopes of finding some specks of gold,” particularly demands “to bulldoze nonparty academic
records.” Alig-Mielcarek v. Jackson, 286 F.R.D. 521, 527 (N.D. Ga. 2012). SFFA has not
pointed to anything contained in the files, and not contained in the database, that it needs for the
analysis that Dr. Arcidiacono proposes. But even if SFFA can identify some bits of information
contained in applicant files (and not contained in the database) that might to some degree inform
Dr. Arcidiacono’s proposed statistical analysis, it cannot bear its heavy burden to show that its
need for more than 160 files outweighs the applicants’ privacy interests.
SFFA is not entitled to limitless discovery because it has sued under
SFFA is wrong in suggesting that it is entitled to discovery of boundless scope and
breadth simply because it has brought suit under Title VI. Courts routinely place reasonable
limits on discovery in civil rights cases. In the employment discrimination context, for example,
the First Circuit has held that parties “ought not to be permitted to use broadswords where
scalpels will suffice, nor to undertake wholly exploratory operations in the vague hope that
something helpful will turn up.” Mack v. Great Atl. & Pac. Tea Co., 871 F.2d 179, 187 (1st Cir.
1989); see also Woodward v. Emulex Corp., 714 F.3d 632, 636-637 (1st Cir. 2013) (district court
permissibly limited discovery to avoid “a fishing expedition into possibly barren waters”).
Discovery may be limited in civil rights cases as in any other type of case “to avoid unduly
burdensome requests,” Tyler v. Suffolk Cnty., 256 F.R.D. 34, 37 (D. Mass. 2009) (prisoner
rights), to “protect the privacy interests of . . . individuals,” Whittingham, 164 F.R.D. at 127
(employment discrimination), and to reduce the hardship faced by defendants “when weighed
against the absence of facts alleged to support . . . liability,” Santiago v. Fenton, 891 F.2d 373,
379-80 (1st Cir. 1989) (wrongful and abusive arrest). Indeed, although SFFA cites Raza v. City
of New York, 998 F. Supp. 2d 70, 82 (E.D.N.Y. 2013), for the proposition that discovery in civil
rights cases should be broad, the court in that case “limited discovery only to those documents
necessary for Plaintiffs to prove their case because of the sensitive nature of the requested
discovery.” Thus, the fact that SFFA is suing under Title VI does not entitle it to intrusive
discovery where “the burden or expense of the proposed discovery outweighs its likely benefit”
to the adjudication of this case. Fed. R. Civ. P. 26(b)(2)(C).
SFFA’s Request For The Names And Contact Information Of Alumni Interviewers
Should Be Denied
SFFA has demanded that Harvard produce the names and contact information of all
Harvard alumni who have conducted admissions interviews since September 1, 2011, in eleven
different, heavily populated states and regions. Producing this information for just the last two
admissions cycles would require Harvard to provide SFFA with the names and contact
information of more than 5,000 alumni. Declaration of Marlyn E. McGrath (“McGrath Decl.”)
¶ 11. This request should be denied because granting the request could subject these individuals
to “annoyance, embarrassment, or oppression.” Fed. R. Civ. P. 26(c); see, e.g., Cusumano v.
Microsoft Corp., 162 F.3d 708, 717 (1st Cir. 1998) (“[C]oncern for the unwanted burden thrust
upon non-parties is a factor entitled to special weight in evaluating the balance of competing
needs.”). SFFA’s request is particularly unwarranted because any information SFFA might
obtain from contacting alumni interviewers is of marginal relevance to this case.
Harvard’s alumni interviewers are volunteers who agree to take time from their busy
schedules to interview a handful of students in a particular year and then to prepare an online
report on the interview. To ensure that alumni remain willing to participate in the interview
process, Harvard requests only a modest time commitment of them. Alumni interviewers are not
employees of Harvard and certainly do not expect to be entangled in litigation of this nature
when they agree to conduct interviews. If alumni knew that, by agreeing to conduct interviews,
they might be involuntarily injected into litigation, they might well hesitate in the future to
participate in this important process. McGrath Decl. ¶ 12.
Further, any information that might be obtained from alumni interviewers is of, at most,
limited relevance to this case. Alumni interviewers do not review any part of the application
files and do not participate in the Admissions Office’s discussion of applicants. Moreover, the
alumni interview is only one part of the many facets of the application considered by the Office.
Therefore, any one interviewer’s report necessarily has only a very small effect on the ultimate
composition of a Harvard class.
SFFA fails to explain why alumni interviewers “are likely to have relevant information
about … the extent to which race drives admissions decision[s]” (Pls.’ Mem. 18). If SFFA is
seeking information about Harvard’s admissions policies, priorities, or practices that is
communicated to alumni interviewers, that information can be obtained through discovery
directed at Harvard, without the need to entangle thousands of alumni who have no obligation to
participate in this litigation. If, upon review of the discovery produced by Harvard, SFFA
determines that it must identify specific alumni interviewers (though Harvard cannot conceive of
a reason why it would), then SFFA can return to Harvard and the Court with a targeted request
and explanation of that need. But the Court should deny SFFA’s current broad and unnecessary
request that will intrude upon the lives of non-parties to the litigation who have volunteered their
time to assist their alma mater. See United States v. Comley, 890 F.2d 539, 544 (1st Cir. 1989)
(recognizing that unduly intrusive requests of private organizations can have chilling effects that
impinge upon First Amendment associational rights).
SFFA’s Request For Information About Transfer Admissions Should Be Denied
SFFA’s efforts to seek discovery into Harvard’s transfer admissions processes should be
denied because transfer admissions are irrelevant to the Complaint, which alleges that a specific
student was unfairly denied admission to Harvard’s freshman class. SFFA has not alleged that
any of its purported members was denied, or even applied for, transfer to Harvard from another
institution. Moreover, in response to a document request from Harvard, SFFA stated that it has
no documents indicating that any of its current members have any intent to apply for transfer to
Harvard. SFFA fails to explain how any of its members could benefit from any inquiry into
Harvard’s transfer process.
SFFA has asserted in its Complaint that one of its members, who was allegedly denied
admission to Harvard in the 2013-14 cycle, would be willing to apply for transfer to Harvard if
Harvard “ceases the use of race or ethnicity” in undergraduate admissions. Complaint, Dkt. 1 at
¶ 24. But that speculative possibility cannot justify the inquiry that SFFA seeks. SFFA has not
sought any preliminary injunctive relief in this case (nor could it satisfy the requirements for
such relief), and there is no reason to believe Harvard will cease its entirely lawful consideration
of race while this alleged member might be eligible to transfer. The possibility of transfer is
even more attenuated for other SFFA members referred to in the Complaint, who have not even
yet applied to Harvard as freshmen, much less been denied admission. See Compl. ¶¶ 25-26.
Finally, discovery into transfer admissions should be denied because transfers play a tiny
role in the composition of any Harvard undergraduate class. Very few transfer slots are available
at Harvard because the vast majority of Harvard undergraduates complete their studies there.
Harvard’s undergraduate classes have approximately 1,660 students, but in the 2013-2014 and
2014-2015 admissions cycles, Harvard admitted only 15 and 21 transfer applicants respectively.
See McGrath Decl. ¶¶ 7-8. Gaining admission as a transfer student is exceptionally difficult, and
the rate at which transfer applicants are offered admission (just over 1 percent) is not comparable
to the rate for freshman applications. See McGrath Decl. ¶¶ 6-8. Discovery into the transfer
process, therefore, will not yield any information that will assist the parties or the Court in
evaluation the core allegations in this case, which focus on the use of race in Harvard’s freshman
For the foregoing reasons, Harvard respectfully requests that SFFA’s Motion to Compel
/s/ Felicia H. Ellsworth
Felicia H. Ellsworth (BBO #665232)
WILMER CUTLER PICKERING
HALE AND DORR LLP
60 State Street
Boston, MA 02109
Tel: (617) 526-6687
Fax: (617) 526-5000
Seth P. Waxman (pro hac vice)
Paul R.Q. Wolfson (pro hac vice)
WILMER CUTLER PICKERING
HALE AND DORR LLP
1875 Pennsylvania Ave. NW
Washington, D.C. 20006
Tel: (202) 663-6800
Fax: (202) 663-6363
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: (212) 295-6717
Fax: (212) 230-8888
Dated: July 30, 2015
Counsel for Defendant President and
Fellows of Harvard College
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 and
that paper copies will be sent to those indicated as non-registered participants on July 30, 2015.
/s/ Felicia H. Ellsworth
Felicia H. Ellsworth
Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.
Why Is My Information Online?