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

Filing 416

DECLARATION re 412 MOTION for Summary Judgment by Students for Fair Admissions, Inc.. (Attachments: # 1 Exhibit Expert Report, # 2 Exhibit Rebuttal Expert Report, # 3 Exhibit Supplemental Report)(Consovoy, William) (Additional attachment(s) added on 6/18/2018: # 5 Exhibit A- unredacted version) (Montes, Mariliz).

EXHIBIT A

IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF MASSACHUSETTS

STUDENTS FOR FAIR ADMISSIONS, INC.,
Plaintiff,
v.

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

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

EXPERT REPORT OF RICHARD D. KAHLENBERG

i

TABLE OF CONTENTS
I.

Professional Qualifications ...................................................................................................................... 1

II. Purpose ....................................................................................................................................................... 3
III. Summary of My Opinions ....................................................................................................................... 4
IV. Experience and academic research show that colleges and universities can maintain or
increase diversity through race-neutral alternatives without sacrificing academic quality. ........... 5
A. Experience at selective public universities shows that race-neutral strategies can produce
racial, ethnic, and socioeconomic diversity. ..................................................................................... 5
B. Academic research shows that selective universities (including elite private universities)
can employ effective race-neutral strategies. .................................................................................. 10
C. Well-crafted race-neutral strategies do not compromise academic quality. .............................. 13
V. Harvard failed to fully consider any of the numerous race-neutral alternatives that could
achieve the educational benefits of diversity. ..................................................................................... 15
A. Harvard could increase socioeconomic preferences. .................................................................... 17
1.

Socioeconomic factors such as income and wealth are highly correlated with race. .......... 17

2.

Harvard’s socioeconomic diversity is deeply lacking. .............................................................. 20

3.

Harvard could make critical socioeconomic data available to admissions officers. ............ 23

4.

Harvard could increase the weight it gives to socioeconomic factors................................... 25

B. Harvard could increase financial aid................................................................................................ 29
C. Harvard could reduce or eliminate preferences that favor non-minorities. .............................. 31
1.

Legacy preferences ........................................................................................................................ 31

2.

Donor preferences ......................................................................................................................... 33

3.

The Z-list......................................................................................................................................... 34

4.

Faculty and staff preferences ....................................................................................................... 36

D. Harvard could adopt admissions policies utilizing geographic diversity, including
percentage plans and the use of zip codes. .................................................................................... 36

ii

E. Harvard could increase its recruitment efforts. ............................................................................. 39
F. Harvard could increase its admission of community college transfers. ..................................... 41
G. Harvard could end early admissions................................................................................................ 42
VI.

Simulations of Harvard’s data show that workable race-neutral alternatives exist. ................. 45

A. A careful simulation indicates that Harvard could achieve the educational benefits of
racial, ethnic, and socioeconomic diversity without sacrificing academic quality. ................... 45
B. Through inclusion of additional socioeconomic factors and better recruiting of
low-income students, the simulation could predict even greater racial and ethnic diversity.. 51
VII.

Conclusion ........................................................................................................................................... 53

VIII. Appendices .......................................................................................................................................... 55
A. Appendix A – Curriculum Vitae ...................................................................................................... 55
B. Appendix B – Documents Relied Upon or Considered in Forming Opinion ......................... 73
C. Appendix C – Simulations ................................................................................................................ 75

iii

I.

Professional Qualifications
My name is Richard D. Kahlenberg. I am a senior fellow at The Century Foundation, a non-

profit, non-partisan research organization founded in 1919. The views expressed in this report are
solely my own, and this report is submitted on my own behalf and not on behalf of any organization.
I am the author or co-author of six books and the editor of ten books. (For the full list, see
my Curriculum Vitae in Appendix A.) Most relevant here, I am the author of The Remedy: Class, Race,
and Affirmative Action (Basic Books, 1996), which was described by Harvard University’s William Julius
Wilson in the New York Times as “by far the most comprehensive and thoughtful argument thus far
for . . . affirmative action based on class.”1 The book was named one of the best books of the year by
the Washington Post.2
In 2003, Diverse Issues in Higher Education, a widely read industry magazine on diversity
issues, called me “arguably the nation’s chief proponent of class-based affirmative action in higher
education admissions.”3 In 2013, The New York Times identified me as “perhaps the most prominent
self-described progressive with doubts about the current version of affirmative action.”4 And in 2016,
reflecting on my time researching and writing about higher education, William G. Bowen, the former
president of Princeton University, and Michael S. McPherson, the former president of Macalester

1
William Julius Wilson, “Class Consciousness,” New York Times Book Review, July 14, 1996.
2

Norman Ornstein, “Social Issues,” Washington Post Book World, December 8, 1996.

3

Ronald Roach, “Class-Based Affirmative Action,” Diverse Issues in Higher Education, June 19,
2003.
4

David Leonhardt, “The Leading Liberal Against Affirmative Action,” New York Times, March 9,
2013.

1

College, wrote that I deserve “more credit than anyone else for arguing vigorously and relentlessly for
stronger efforts to address disparities by socioeconomic status.”5
I am also the editor of four books that address, in part or in whole, race-neutral affirmative
action strategies:
•

America’s Untapped Resource: Low-Income Students in Higher Education (Century Foundation,
2004);

•

Rewarding Strivers: Helping Low-Income Students Succeed in College (Century Foundation, 2010);

•

Affirmative Action for the Rich: Legacy Preferences in College Admissions (Century Foundation,
2010); and

•

The Future of Affirmative Action: New Paths to Higher Education Diversity after Fisher v. University
of Texas (Century Foundation/Lumina Foundation, 2014).

My law review articles on race-neutral alternatives to racial preferences include:
•

“Getting Beyond Racial Preferences: The Class-Based Compromise,” 45 American
University Law Review 721 (February 1996);

•

“Class-Based Affirmative Action,” 84 California Law Review 1037 (July 1996); and

•

“Reflections on Richard Sander’s Class in American Legal Education,” 88 Denver University
Law Review 719 (September 2011).

I also have researched and published numerous articles on race-neutral alternatives to racial
preferences in prominent publications, including The New York Times, The Wall Street Journal, The
Washington Post, and The New Republic. (See all publications in Appendix A). Over the years, I have
served on numerous conference panels giving me an opportunity to interact with college admissions
officers at a number of selective colleges.
Before coming to The Century Foundation, I was a Fellow at the Center for National Policy,
a visiting associate professor of constitutional law at George Washington University, and a legislative

5
William G. Bowen & Michael S. McPherson, Lesson Plan: An Agenda for Change in American
Higher Education (Princeton University Press, 2016), p. 35.

2

assistant to Senator Charles S. Robb (D-VA). I graduated from Harvard College and Harvard Law
School.
I also serve on the advisory boards of the Pell Institute and the Albert Shanker Institute, as
well as the Research Advisory Panel of the National Coalition for School Diversity. In 2013, I was the
winner of the William A. Kaplin Award for Excellence in Higher Education Law and Policy
Scholarship.
II.

Purpose
In 2014, I was retained in this matter by Students for Fair Admissions, Inc. (SFFA) to provide

an opinion regarding the availability and feasibility of race-neutral alternatives to Harvard’s use of race
as a factor in undergraduate admissions. In particular, I was asked to examine whether Harvard could
implement workable race-neutral alternatives that would produce the educational benefits of diversity.
The rate for my services in this matter is $295 an hour.
In making my opinions, I draw first upon my extensive knowledge of the history and study of
race-neutral alternatives. See Section I, supra, and Appendix A. I have authored, co-authored, edited,
or reviewed virtually every major study or analysis on race-neutral alternatives from the past 20 years.
I have also reviewed substantial portions of the voluminous evidence that has been produced by
Harvard in this case, including numerous deposition transcripts and several internal reports from
Harvard’s Office of Institutional Research (OIR). A full list of the documents and transcripts I
reviewed is provided at Appendix B. Finally, I have reviewed and had access to the admissions data,
analysis, and conclusions from SFFA’s other expert witness, Duke Professor Peter Arcidiacono.
It is also important to understand what I have not reviewed. I did not have access to some of
the data that I would have liked to review—either because Harvard has declined to maintain the
information or because the Court permitted Harvard to withhold the information from SFFA. This

3

includes, among other things, precise data about student income, wealth, and zip code of residence.
These data would have been helpful to me, as they would have allowed me to consider additional raceneutral strategies and evaluate whether they would be workable as possible replacements for Harvard’s
use of race in admissions decisions. Nevertheless, I am confident about the opinions I am able to state
below.
I have not testified as an expert at trial or deposition in the past four years.
III.

Summary of My Opinions
The U.S. Supreme Court has long stated that student body diversity—by race and also by

socioeconomic status—offers important educational benefits.6 But because of the heavy costs
associated with using race in governmental decision making, the Fourteenth Amendment “forbids the
use even of narrowly drawn racial classifications except as a last resort.”7 In Fisher v. University of Texas,
therefore, the Supreme Court held that colleges cannot employ racial preferences unless “no workable
race-neutral alternatives would produce the educational benefits of diversity.”8 Indeed, in pursuing the
compelling goal of diversity, universities bear “the ultimate burden of demonstrating, before turning
to racial classifications, that available workable race-neutral alternatives do not suffice.”9
With these guideposts in mind, I am prepared to give testimony on three main opinions to a
reasonable degree of professional certainty.

6
Grutter v. Bollinger, 539 U.S. 306, 330 (2003).
7

City of Richmond v. J.A. Croson Co., 488 U.S. 469, 519 (1989) (Kennedy, J., concurring in part
and concurring in the judgment).
8

133 S. Ct. 2411, 2420 (2013).

9

Id.

4

First, there is extensive empirical evidence and academic research documenting the myriad
(and innovative) ways in which colleges and universities such as Harvard can use race-neutral
alternatives to produce the educational benefits of diversity.
Second, it is apparent from my review of the deposition testimony and relevant evidence
produced that, in the years between Fisher I and the filing of this lawsuit, Harvard failed to fully
consider any of the numerous available race-neutral alternatives that could achieve the educational
benefits of diversity. These include:
•

Increasing socioeconomic preferences;

•

Increasing financial aid;

•

Reducing or eliminating preferences for legacies, donors, and relatives of faculty and staff;

•

Adopting policies using geographic diversity, including percentage plans and the use of zip
codes;

•

Increasing recruitment efforts;

•

Increasing the admission of community college transfers; and

•

Eliminating the Early Action admissions option.
Finally, after reviewing Harvard’s admissions data and other relevant socioeconomic data, I

have concluded that there are race-neutral alternatives available that could provide Harvard with the
educational benefits of diversity without the use of racial preferences.
IV.

Experience and academic research show that colleges and universities can maintain
or increase diversity through race-neutral alternatives without sacrificing academic
quality.
A.

Experience at selective public universities shows that race-neutral strategies
can produce racial, ethnic, and socioeconomic diversity.

For years, supporters of racial preferences argued that no workable alternatives existed for
creating racial diversity. In the words of Justice Blackmun in his 1978 Bakke opinion, “I suspect that
it would be impossible to arrange an affirmative action program in a racially neutral way and have it

5

successful. To ask that this be so is to demand the impossible. In order to get beyond racism, we must
first take account of race. There is no other way.”10
Since then, however, numerous universities have proven him wrong. In 2012, my colleague
Halley Potter and I examined ten leading universities where racial preferences had been banned and
found that seven of the ten—the University of Texas at Austin, Texas A&M, the University of
Washington, the University of Florida, the University of Georgia, the University of Nebraska, and the
University of Arizona—had used race-neutral alternatives to meet or exceed the racial diversity levels
they had obtained in the past using racial preferences.11 These schools obtained such results through
a variety of approaches, including creating plans to encourage geographic and socioeconomic diversity,
bolstering financial aid policies, adopting programs that could attract disadvantaged students from
underrepresented demographics with the promise of financial support, and building partnerships with
K-12 schools to increase the pool of college-ready applicants.12
Many of these colleges had been adamant that race-neutral alternatives could never succeed.
For example, in 1998, the University of Washington was forced to abandon racial preferences after a
ballot initiative was passed banning the practice. At the time, Richard McCormick, the president of
the University of Washington, spoke out strongly against the referendum and made dire predictions
about its effect on racial diversity. But the University ultimately crafted new approaches to achieve
diversity, including recruiting at predominantly minority high schools, expanding financial aid, and

10
Regents of University of California v. Bakke, 438 U.S. 265, 407 (1978) (Blackmun, J., concurring).
11

Richard D. Kahlenberg & Halley Potter, A Better Affirmative Action: State Universities that Created
Alternatives to Racial Preferences (Century Foundation), pp. 26-61.
12

Id. at 76.
6

considering such factors as “personal adversity” and “economic disadvantage” in its admissions
decisions. By 2004, “the racial and ethnic diversity of the UW’s first-year class had returned to its pre1999 levels,” when race was still considered in admissions, and the new admissions policy also
increased economic diversity among the student body.13
Similarly, in 2000, the University of Georgia adopted a number of race-neutral strategies after
a federal court struck down the university’s use of race in admissions.14 In particular, the university
began using a number of socioeconomic factors in its admissions process, including parental education
and high school environment, began admitting the valedictorian and salutatorian from every high
school class, and stopped giving preference to children of alumni. Although alumni opposed the end
of legacy admissions, the university “has not encountered noticeable fundraising challenges as a result
of the change.”15 Although minority enrollment initially dropped after the ban on using race in
admission, it has since moved upward and “the years since 2000 have shown the university moving in
the right direction, toward increased racial, ethnic, socioeconomic, linguistic, and geographic diversity
on campus.”16
The other three universities we examined—the University of Michigan, UCLA, and the
University of California Berkeley—had not reached their prior levels of racial diversity. As an initial

13
Richard L. McCormick, “Converging Perils to College Access for Racial Minorities: Examples of
Responses that Work from Washington State and New Jersey,” in The Future of Affirmative Action:
New Paths to Higher Education Diversity after Fisher v. University of Texas, ed. Richard D.
Kahlenberg (New York: Century Foundation/Lumina Foundation, 2014), supra, p. 118.
14

See Johnson v. Board of Regents, 106 F. Supp. 2d 1362 (S.D. Ga. 2000).

15

Nancy G. McDuff & Halley Potter, “Ensuring Diversity Under Race-Neutral Admissions at the
University of Georgia,” in The Future of Affirmative Action, supra, p. 126.
16

Id. at 123.
7

matter, the data on African-American enrollment at Michigan are problematic. In 2010, the
Department of Education changed its methodology for categorizing students by race and ethnicity,
requiring colleges to report separately students who are members of two or more races. “So a drop in
the number of black students reported at a university from 2009 to 2010,” a Chronicle of Higher
Education article noted, “doesn’t necessarily mean that there were actually fewer black students.”17 In
fact, when the new “mark one or more” races methodology was proposed, civil rights groups raised
concerns that it would result in an artificial decline in African-American and Hispanic representation
in government statistics.18
To the extent that race-neutral alternatives have not been fully effective at these universities,
however, it is mostly because of their failure to utilize them fully.19 Michigan still gives preferences in
admission to the children of alumni (who, at selective colleges, tend to be disproportionately non-

17
Jonah Newman, “What Does the Education Dept. Know About Race?” Chronicle of Higher
Education, April 28, 2014. Consider, also, the case of the University of Virginia (UVA), which is not
subject to a voter-imposed ban on racial preferences and continues to use race as a factor in
admissions. In 2008, before students could use the multi-race category, UVA enrolled 1,199 AfricanAmerican students. By 2012, after the change in categories was put in place, the number of African
Americans was 946, suggesting a dramatic 21.1 percent drop. But when the 2012 data includes the 206
students who identified as African American and some other ethnicity (for a grand total of 1,152
African Americans under the old methodology), the drop was 3.9 percent. In other words, about 80
percent of the apparent decline in black enrollment at UVA was due to reporting changes. McGregor
McCance, “Analysis of U.Va.’s Incoming Class Shows Consistent Quality with Dynamic Change,”
UVA Today, May 16, 2013. Harvard has also seen changes in its methodology of reporting race and
ethnicity. See Fitzsimmons deposition, pp. 93-101.
18

See Kim M. Williams, Mark One or More: Civil Rights in Multicultural America (University of
Michigan Press, 2008).
19

U.C. Berkeley, UCLA, and the University of Michigan have also faced a special disadvantage in
recruiting minority students because they have a national pool of applicants and restrictions on using
race that were imposed by a state referendum rather than a federal court. As a result, out-of-state
competitors could continue to use racial preferences.

8

minority)20 and still provides substantial “merit” aid to wealthy students, thereby diverting funds from
need-based aid.21 U.C. Berkeley and UCLA currently employ only family income as the primary
determinant of economic disadvantage and thus are not using more accurate measures of
socioeconomic disadvantage.22 As discussed further below, using wealth alongside income would
better capture economic disadvantage than does income alone and could lead to greater racial diversity.
It is significant to note that these types of race-neutral approaches also produce much higher
levels of socioeconomic diversity than do race-based admissions.23 The enhancement of
socioeconomic diversity that flows from these plans is critical from an educational and legal
perspective, because the educational benefits of diversity arise from the interchange of ideas and
experiences with those from different financial circumstances just as surely as those from different
racial backgrounds.24

20
John Brittain & Eric L. Bloom, “Admitting the Truth: The Effect of Affirmative Action, Legacy
Preferences, and the Meritocratic Ideal on Students of Color in College Admissions,” in Affirmative
Action for the Rich: Legacy Preferences in College Admissions, ed. Richard D. Kahlenberg (Century
Foundation Press, 2010), pp. 127-32.
21

Richard D. Kahlenberg, “A Fresh Chance to Rein in Racial Preferences,” Wall Street Journal,
October 13, 2013.
22

Richard Sander, “The Use of Socioeconomic Affirmative Action at the University of California,” in
The Future of Affirmative Action, supra, p. 101 (that U.C. campuses look at parental education and
income).
23

See Matthew N. Gaertner, “Advancing College Access with Class-Based Affirmative Action: The
Colorado Case,” in The Future of Affirmative Action, supra, p. 181, Table 14.3; Anthony P. Carnevale,
Stephen J. Rose, & Jeff Strohl, “Achieving Racial and Economic Diversity with Race-Blind
Admissions Policy,” in The Future of Affirmative Action, supra, p. 192, Table 15.2.
24

See Grutter, 539 U.S. 306, 324 (2003); Bakke, 438 U.S. 265, 316 (1978). Harvard admissions Dean
Bill Fitzsimmons also testified that “economic background is a vital piece of putting together a class
that is educationally beneficial for all its members.” Fitzsimmons deposition, pp. 421-24.

9

In California, for example, students from economically disadvantaged backgrounds were
significantly more likely to be admitted to universities in California after the State banned racial
preferences.25 Likewise, when UCLA Law School adopted a socioeconomic affirmative action
program, the proportion of students who were the first in their families to attend college roughly
tripled.26
It seems hardly an accident, therefore, that the University of California dominates the list of
schools “doing the most for low-income students” in the New York Times’ “College Access Index” in
2015.27 Similarly, of the top seven institutions for social mobility, six were from the UC system, and
the seventh, the University of Florida, has also implemented race-neutral strategies in the face of a
racial preference ban.28
B.

Academic research shows that selective universities (including elite private
universities) can employ effective race-neutral strategies.

In the wake of Supreme Court rulings on affirmative action, think tanks and the academic
community have been examining in earnest the use of race-neutral strategies to promote racial, ethnic,
and socioeconomic diversity on campuses. For example, the Lumina Foundation teamed up with The
Century Foundation to produce a 299-page volume (which I edited) that brought together both
supporters and skeptics of racial preferences to consider the meaning of the Supreme Court’s rulings

25
See Kate Antonovics & Ben Backes, “The Effect of Banning Affirmative Action on College
Admissions Policies and Student Quality,” The Journal of Human Resources 49, no. 2 (Spring 2014):
p. 306.
26

Sander, “The Use of Socioeconomic Affirmative Action,” supra, p. 105.

27

David Leonhardt, “California’s Upward-Mobility Machine,” New York Times, September 16, 2015.

28

Id.; Kahlenberg & Potter, A Better Affirmative Action, supra.

10

and to examine the efficacy of race-neutral strategies.29 The College Board’s Access and Diversity
Collaborative produced papers on race-neutral policies, including “The Playbook: A Guide to Assist
Institutions of Higher Education in Evaluating Race-and Ethnicity-Neutral Policies in Support of the
Mission-Related Diversity Goals.”30 And the American Council on Education surveyed 338 colleges
on their use of race-neutral strategies.31
As a result, valuable research has emerged identifying concrete ways in which universities can
increase racial diversity through race-neutral means. For example, in 2014, Professors Anthony
Carnevale, Stephen Rose, and Jeff Strohl of Georgetown University examined how socioeconomic
affirmative action programs, percentage plans, or a combination of the two, could work at the nation’s
most selective 193 institutions.32 The authors found that if these schools used class-based affirmative
action—which would include a mix of socioeconomic considerations (such as parental education,
income, savings, and school poverty concentrations)—the combined African-American and Hispanic
representation would rise from 11% to 13%—all without the use of racial preferences. Under a
different simulation (in which the top 10% of test takers in every high school was among the pool

29
Kahlenberg (ed), The Future of Affirmative Action, supra.
30

See, e.g., Arthur L. Coleman, Teresa E. Taylor, & Katherine E. Lipper, “The Playbook: A Guide to
Assist Institutions of Higher Education in Evaluating Race- and Ethnicity-Neutral Policies in Support
of the Mission-Related Diversity Goals,” College Board and Education Counsel, October 2014,
http://educationcounsel.com/wp-content/uploads/2015/06/ADC%20Playbook%20October%
202014%20(for%20posting%20to%20website).pdf.
31

Lorelle L. Espinosa, Matthew N. Gaertner, & Gary Orfield, “Race, Class, and College Access:
Achieving Diversity in a Shifting Legal Landscape” American Council on Education, 2015,
http://www.acenet.edu/news-room/Documents/Race-Class-and-College-Access-AchievingDiversity-in-a-Shifting-Legal-Landscape.pdf.
32

Carnevale, Rose, & Strohl, “Achieving Racial and Economic Diversity with Race-Blind Admissions
Policy,” in The Future of Affirmative Action, supra; see also David Leonhardt, “If Affirmative Action
Is Doomed, What’s Next?” New York Times, June 17, 2014.

11

admitted to this collection of schools) the authors found that African-American and Hispanic
representation would rise from 11% to 17%. Under each of these scenarios, socioeconomic diversity
and mean SAT scores would also rise.33
Similarly, in 2014, Matthew Gaertner examined admissions at the University of Colorado at
Boulder and found that a sophisticated socioeconomic affirmative action plan that gave considerable
weight to economic disadvantage could achieve even more racial diversity than using racial preferences.
Based on national research, the University of Colorado devised an index of socioeconomic
disadvantage that looked at a number of factors, including “the applicant’s native language, singleparent status, parents’ education level, family income, the number of dependents in the family, whether
the applicant attended a rural high school, the percentage of students from the applicant’s high school
eligible for free or reduced-price lunch (FRL), the school-wide student-to-teacher ratio, and the size
of the twelfth-grade class.” Under the hypothetical program, the university gave socioeconomically
disadvantaged students a preference in admissions that was larger than what African-American and
Hispanic students had been provided in the past. When simulations were run, Gaertner found that
not only would socioeconomic diversity increase, but the acceptance rates of underrepresented

33
Carnevale, Rose, & Strohl, “Achieving Racial and Economic Diversity with Race-Blind Admissions
Policy,” in The Future of Affirmative Action, supra, p. 192, Tables 15.1, 15.2. The study’s breakdown
is as follows: Status quo (4% African American, 7% Hispanic; 14% from the bottom socioeconomic
half; 1230 mean SAT); Admissions by test score (1% African American, 4% Hispanic; 15% bottom
socioeconomic half; 1362 mean SAT); Socioeconomic affirmative action (3% African American, 10%
Hispanic; 46% from bottom socioeconomic half; 1322 mean SAT); Top 10% of test takers from every
high school (6% African American, 11% Hispanic; 31% from bottom socioeconomic half; 1254 mean
SAT). Id.

12

minority applicants would also increase—from 56 percent under race-based admissions to 65 percent
under class-based admissions.34
In addition, in a 2015 study, Professor Sigal Alon found that if the most selective 115 American
universities instituted broad reform—including effectively eliminating35 legacy, athletic,and racial
preferences—a socioeconomic boost “could not only replicate the current level of racial and ethnic
diversity at elite institutions but even increase it.”36 Professor Alon’s model looked at three variations:
(1) a “socioeconomic status” model, which looks at family-based economic disadvantages; (2) a
“structural” model, which looks at neighborhood-based economic disadvantages; and (3) a
“multidimensional” model, which looks at both. Professor Alon found that racial diversity would meet
or exceed current admissions and socioeconomic diversity would increase under all three models.
Meanwhile, because mean SAT scores would remain steady, “all this could be done without
jeopardizing academic selectivity.”37
C.

Well-crafted race-neutral strategies do not compromise academic quality.

Critics may argue that race-neutral alternatives will reduce academic standards. But experience
and research refute that claim. For example, after UCLA Law School adopted a socioeconomic

34
Gaertner, “Advancing College Access with Class-Based Affirmative Action,” supra, p. 181, Table
14.3.
35

Alon effectively eliminates athletic, legacy, and racial preferences by replacing those students in the
weakest academic quartile—whom she presumes includes those for whom preferences were
decisive—with the most academically competitive economically disadvantaged students of all races.
36

Sigal Alon, Race, Class, and Affirmative Action (Russell Sage Foundation, 2015), pp. 254-56.

37

Id. at 256.

13

preferences program, the school’s California bar exam passage rate rose to an all-time high.38 Likewise,
in a national simulation, Professors Carnevale and Rose found that top universities could nearly
quadruple the proportion of students from the bottom socioeconomic half (from 10% of all students,
the level they found in their research, to 38%) without any change in graduation rates.39
These studies are buttressed by a growing body of research on “undermatching,” in which
highly qualified students do not apply to selective colleges. Professor Caroline Hoxby of Stanford and
Professor Christopher Avery of Harvard have found that 35,000 low-income students are high
achieving, but that only one-third apply to one of the country’s 238 most selective colleges. Of those
low-income, high-achieving students, roughly 2,000 are African American and 2,700 are Hispanic.40
Additional research has found that 43 percent of students who are academically qualified to gain
admission to selective colleges undermatch, and that many are Hispanic and African American.41 In
raw numbers, that translates into 4,000 Hispanic and 2,000 African-American SAT takers who have
the strongest academic credentials yet do not attend a highly selective school.42 This research indicates
that there is enormous potential to increase socioeconomic and racial diversity without in any way

38
Sander, “The Use of Socioeconomic Affirmative Action at the University of California,” supra,
p. 107.
39

Anthony P. Carnevale & Stephen J. Rose, “Socioeconomic Status, Race/Ethnicity, and Selective
College Admissions,” in America’s Untapped Resource: Low-Income Students in Higher Education,
ed. Richard D. Kahlenberg (The Century Foundation Press, 2004), pp. 148-49.
40

Caroline M. Hoxby & Christopher Avery, “The Missing ‘One-Offs’: The Hidden Supply of HighAchieving, Low Income Students,” NBER Working Paper no. 18586, December 2012, p. 34.
41

Alexandria Radford & Jessica Howell, “Addressing Undermatch: Creating Opportunity and Social
Mobility,” in The Future of Affirmative Action, supra, p. 134,
42

Id.
14

sacrificing academic quality if colleges were aggressively to recruit high-achieving, low-income
students.
V.

Harvard failed to fully consider any of the numerous race-neutral alternatives that
could achieve the educational benefits of diversity.
The Supreme Court’s instructions regarding race-neutral alternatives are clear. Colleges must

prove that “no workable race-neutral alternatives would produce the educational benefits of
diversity.”43 This requirement has been widely discussed in the academic community.44 Indeed, in a
2013 article in the Chronicle of Higher Education, Harvard professor Thomas Kane and current
Harvard Education School Dean James Ryan noted that the Fisher decision means that “[t]o consider
race in admissions . . . institutions must prove to courts that race-neutral alternatives—such as relying
on socioeconomic status or where students live—will not work.”45 The two Harvard faculty members
then suggested ways in which schools could assess the availability of race-neutral alternatives.46 They
warned that “few universities and colleges are prepared to answer the questions that courts will soon
be asking. If they fail to prepare convincing answers, they will lose. And, having been put on notice,

43
Fisher, 133 S. Ct. 2411, 2420 (2013).
44

See, e.g., Arthur L. Coleman & Teresa E. Taylor, “Emphasis Added: Fisher v. University of Texas
and Its Practical Implications for Institutions of Higher Education,” in The Future of Affirmative
Action, supra, 50-51.
45

Thomas J. Kane and James E. Ryan, “Why ‘Fisher’ Means More Work for Colleges,” Chronicle of
Higher Education, July 29, 2013.
46

They noted: “To quantify such costs, colleges could review their admissions folders (or at least a
representative sample of them) and have admissions officers flag the family-background factors that
are potential race-neutral alternatives. Analysts could then estimate how much those factors would
have to be weighted (and other factors diminished) in order to produce the outcomes now produced
with race-conscious admissions. They could then compare the results of race-conscious and raceneutral policies on individual dimensions—like test scores or high-school grades—or on combinations
of traits such as academic indices.” Id.

15

responsibility for that loss will be with our college and university leaders, not our courts.”47 Evidence
in this case shows that the professors sent a draft of this article to Dean Fitzsimmons seeking his input
in advance of publication.48
Despite all this, it appears that Harvard—one of the nation’s great research universities—
conducted no investigation to see whether race-neutral strategies could yield the educational benefits
of diversity, as required by law. Dean Fitzsimmons testified that Harvard has conducted no formal
analysis of how race-neutral alternatives might work at the college or what would happen if the school
transitioned to a race-blind system.49 Similarly, Director McGrath testified that she could not recall
any discussion of the possibility of going to a race-blind admissions system.50 And Sally Donahue,
Director of Financial Aid, testified that she could not recall ever discussing alternatives to using race
in admissions, such as socioeconomic status.51 It appears that the sum total of Harvard’s investigation
into race-neutral alternatives is a disbanded committee led by Dean James Ryan and the establishment
of a new committee with just three members (Dean Fitzsimmons, Dean Michael Smith, and Dean
Rakesh Khurana) that as of August 3, 2017 had met only once.52

47
Id.
48

Harvard documents, 7799-7802.

49

Fitzsimmons deposition, pp. 137, 168, 180-81.

50

McGrath deposition, pp. 255, 369. See also Faust deposition, pp. 50-51; Smith deposition, pp. 6264.
51

Donahue deposition, pp. 70-72.

52

Fitzsimmons deposition, pp. 213-14, 224-26; Harvard’s Response to Plaintiff’s Second Set of
Interrogatories, pp. 18-19. See also Harvard documents, 72381 (Dean Smith memo calling the
formation of the committee on June 9, 2017).

16

Throughout this period of time, there were numerous race-neutral alternatives available that
have the potential to obtain the educational benefits of diversity, and which Harvard certainly could
have considered and potentially adopted. I discuss these options below.
A.

Harvard could increase socioeconomic preferences.
1.

Socioeconomic factors such as income and wealth are highly
correlated with race.

Well-crafted race-neutral alternatives, while not providing a racial preference, are nevertheless
cognizant of the ways in which past and present racial discrimination shapes opportunities in America.
Race-neutral alternatives based on socioeconomic factors work to produce racial diversity because
economic disadvantage is often influenced by the legacy of racial discrimination. This helps explain
why African Americans and Hispanics on average have lower incomes and smaller savings than whites
do, and why even middle-class blacks live in neighborhoods with higher poverty rates than low-income
whites.53
Research finds that when socioeconomic affirmative action programs are constructed using a
wide variety of variables—not just parental income, but factors such as wealth/net worth, and
neighborhood and school levels of poverty that are correlated with race—they can produce substantial
racial and ethnic diversity, because this wider array of socioeconomic factors better captures the
economic impact of ongoing and past racial discrimination than does income (or race) alone.
For example, Professor Dalton Conley of New York University finds that a family’s wealth
(rather than income) better reflects the nation’s legacy of slavery and segregation because wealth is

53
John R. Logan, “Separate and Unequal: The Neighborhood Gap for Blacks, Hispanics and Asians
in Metropolitan America,” US2010 Project, July 2011, p. 5.

17

handed down from generation to generation.54 African Americans typically have incomes that are 70%
of white incomes, but African-American wealth is just 10% of white wealth.55 Moreover, parental
wealth and education are far more powerful predictors of college completion than race or income,
Conley finds.56 Wealth matters more than income because “educational advantages are acquired
through major capital investments and decisions,” such as purchasing a home in a neighborhood with
good public schools.57 And because concentrated poverty is highly correlated with race, a student’s
family’s wealth will capture the lasting effects of discrimination in the housing market.58 Colleges that
give a preference to students growing up in concentrated poverty will acknowledge the challenges that,
in the aggregate, poor minority children face much more often than poor white children.
UCLA Law School is an exemplar of an institution that examined factors such as wealth and
concentrated poverty to obtain racial diversity. In the fall 2011 entering class, African Americans were
11.3 times as likely to be admitted under the socioeconomic status (SES) program as other programs,
and Latinos were 2.3 times as likely to be admitted. African Americans constituted 20.4% of those
admitted under the SES program (22 of 108) compared with 0.8% of admissions for non-SES

54
Dalton Conley, “The Why, What, and How of Class-Based Admissions Policy,” in The Future of
Affirmative Action, supra, p. 209. See also Lisa J. Dettling, Joanne W. Hsu, Lindsay Jacobs, Kevin B.
Moore, & Jeffrey P. Thompson, “Recent Trends in Wealth-Holding by Race and Ethnicity: Evidence
from the Survey of Consumer Finances,” Federal Reserve FEDS Notes, September 27, 2017 (Black
median family wealth was 10.3% of white median family wealth in 2016, and Hispanic wealth was
12.1% of white wealth. Meanwhile, black median family income was 57.8% of white median family
income and Hispanic income was 62.9% of white income.).
55

Conley, “The Why, What, and How of Class-Based Admissions Policy,” supra, p. 209.

56

Id. at 206.

57

Id. at 207.

58

Logan, “Separate and Unequal,” supra, pp. 4-6.

18

programs (12 of 1,363). Likewise, Hispanics constituted 35.2% of SES admits (38 of 108) compared
with 5.5% for non-SES admits (75 of 1,363). Even though the SES program admitted 108 students,
compared with 1,363 under non-SES, the absolute number of African Americans admitted under the
SES program (22) exceeded the number admitted under other programs (12).59 Similarly, Professor
Richard Sander and Aaron Danielson of UCLA found in a 2014 analysis that richer measures of
socioeconomic status, above and beyond income to include factors such as wealth and neighborhood
poverty levels, significantly increased the correlation between race and socioeconomic status and the
racial dividend of class-based affirmative action.60
Some criticize race-neutral alternatives as subterfuges seeking a desired racial result covertly.
But this thinking has it exactly backwards because the beneficiaries are a very different subset of
African-American and Hispanic students than those who usually benefit from racial preferences. The
new beneficiaries are more likely to be working-class and actually to live in segregated neighborhoods.
As Georgetown University Law Professor Sheryll Cashin notes, place-based approaches help “those
who are actually disadvantaged by structural barriers” rather than enabling “high-income, advantaged
blacks to claim the legacy of American apartheid.”61
Class-based preferences also avoid two important costs associated with racial preferences: a
reinforcement of negative stereotypes and an increase in racial and ethnic antagonism.62 Polls find that

59
Kahlenberg & Potter, “A Better Affirmative Action,” p. 14, supra.
60

Richard Sander & Aaron Danielson, “Thinking Hard About ‘Race-Neutral’ Admissions,” 47
University of Michigan Journal of Law Reform 967, 990-991 (2014).
61

Sheryll Cashin, Place Not Race: A New Vision of Opportunity in American (Boston: Beacon Press,
2014), p. 78.
62

Bakke, 438 U.S. 265, 298-99.
19

most Americans (including a majority of black respondents) oppose the use of race or ethnicity as a
factor in college admissions, but large majorities favor the consideration of economic disadvantage.63
Because students of all races who have overcome economic disadvantage are seen as deserving of
special consideration, such students are unlikely to face the stigma or resentment that has been directed
toward recipients of racial preferences.64
2.

Harvard’s socioeconomic diversity is deeply lacking.

Both external studies and internal data from Harvard suggest that Harvard’s student body is
deeply lacking in socioeconomic diversity. Most notably, in 2017, Professor Raj Chetty (a former
Harvard professor now at Stanford University) and colleagues examined a unique data set of 30 million
college students and financial data from the IRS and concluded that “approximately 3% of children
at Harvard in the 1980-82 birth cohorts come from the lowest income quintile of families, compared
with more than 70% from the top quintile.”65 Put differently, Harvard had 23 times as many highincome students as low-income students. The authors further found that “15.4% of students at
Harvard come from families in the top 1% of the income distribution—about the same number as

63
Scott Jaschik, “Poll: Public Opposes Affirmative Action,” Inside Higher Ed, July 8, 2016 (citing
Gallup poll finding 63%-36% opposition to race as a factor in college admissions, but 61%-39%
support for considering family economic circumstances in admissions).
64

Paul M. Sniderman & Thomas Leonard Piazza, The Scar of Race (Harvard University Press, 1993),
pp. 102-04. See also Robert P. Jones, Daniel Cox, Betsy Cooper, & Rachel Lienesch, “Anxiety,
Nostalgia and Mistrust: Findings from the 2015 American Values Survey,” Public Religion Research
Institute, November 17, 2015, p. 5 (finding resentment associated with racial preferences).
65

Raj Chetty, John N. Friedman, Emmanuel Saez, & Nicholas Turner, and Danny Yagan, “Mobility
Report Cards: The Role of Colleges in Intergenerational Mobility,” July 2017, p. 14, http://
www.equality-of-opportunity.org/papers/coll_mrc_paper.pdf.

20

from the bottom three quintiles combined.”66 In other words, about as many students at Harvard
came from the top 1% of the income distribution as the bottom 60%. Similarly, looking at data from
the cohort born in 1991 (most of whom ended up in the class of 2013), Chetty found that more
Harvard students came from the most affluent 10% of the population than came from the bottom
90%.67
Harvard testimony and evidence in this case reinforces these findings. According to Sally
Donahue, the number of students who qualify for Harvard Financial Aid Initiative (with family
incomes at or below $80,000) has remained at roughly the 25% level for a decade.68 This means that
roughly the bottom half of the income distribution (among 45-54 year olds)69 has just a 25%
representation at Harvard—and this proportion has remained constant over time. According to
Harvard documents, in the class of 2011, 302 students came from families making less than $60,000,
and a total of 430 below $80,000 in a class of more than 1600.70 In addition, Harvard’s documents
show that in the freshman year class of 2011, 38% of Harvard students were “full pay,” meaning they
came from families in the top 4% of the income distribution.71

66
Id.
67

See “Economic Diversity and Student Outcomes at Harvard University,” New York Times,
http://www.nytimes.com/interactive/projects/college-mobility/harvard-university (summary of
Chetty data).
68

Donahue deposition, pp. 102-03.

69

See Bernadette D. Proctor, Jessica L. Semega, & Melissa A. Kollar, “Income and Poverty in the
United States: 2015,” U.S. Census Bureau, September 2016, pp. 6-7, http://www.census.gov/
newsroom/press-releases/2016/cb16-158.html. (For those headed by individuals ages 45-54—a
typical age for parents with college aged students—the median income was $73,857 in 2015.)
70

Harvard documents, 77540.

71

Harvard documents, 77593, 77605.

21

By comparison, at top flagship public universities such as U.C. Berkeley and UCLA, Chetty’s
data show only about a quarter of the percentage of students came from the top 1% by income as at
Harvard. A larger percentage came from the bottom 20% of the income distribution and a
substantially lower percentage from the top 20%.72
Harvard’s own data also show that the proportion of students who are first generation hovered
around just 10% for the classes of 2007-2016. Stunningly, in every one of those classes, the number
of legacy students outnumbered the number of first-generation students.73 This is remarkable in a
nation where there are 382 times as many American adults age 25 and older without a college degree
(143 million) as adults in the world with a Harvard degree (375,000).74 If African-American students
were as underrepresented in Harvard’s population as first-generation college students currently are,
blacks would constitute just 2.25% of the undergraduate student body—something Harvard would
presumably find intolerable.75
Another way to consider socioeconomic diversity is eligibility for the federal Pell grant for
students needing financial aid to pay for college. Using federal data, U.S. News & World Report found

72
See “Economic Diversity and Student Outcomes at the University of California, Berkeley,”
New York Times, http://www.nytimes.com/interactive/projects/college-mobility/university-ofcalifornia-berkeley (summary of Chetty data); See “Economic Diversity and Student Outcomes at the
University of California, Los Angeles,” New York Times, http://www.nytimes.com/interactive/
projects/college-mobility/university-of-california-los-angeles (summary of Chetty data).
73

Harvard documents, 65583.

74

Camille L. Ryan & Kurt Bauman, “Educational Attainment in the United States: 2015,” U.S. Census
Bureau, March 2016, p. 2, Table 1; Christina Pazzanese, “Harvard’s Alumni Impact,” Harvard Gazette,
December 8, 2015.
75

Ryan & Bauman, “Educational Attainment in the United States: 2015” (68% of adults age 45-64
lack a bachelor’s degree, compared with 10% of Harvard undergraduates who are first generation
college students. This 15% representation rate, if applied to African Americans (who make up 15%
of the population) would yield a student body that is 2.25% black.)

22

that the proportion of Harvard undergraduates receiving Pell grants in the 2015-2016 school year was
18%. By comparison, at U.C. Berkeley, 33% received Pell grants, and at UCLA the figure was 37%,
more than double Harvard’s proportion. (While some Ivy League colleges had even fewer Pell
recipients than Harvard, one member of the Ivy League, Columbia University, had considerably more:
32% of students received Pell grants.)76
3.

Harvard could make critical socioeconomic data available to
admissions officers.

Because Harvard has adopted “need-blind” admissions, it has placed a firewall between the
admissions and financial aid offices that prevents admissions officers from knowing the family income
or wealth of applicants.77 This policy creates an enormous barrier to implementing a central raceneutral strategy used at numerous other colleges: one that provides a preference in admissions to lowincome and low-wealth applicants.
When asked about implementing “race-blind” admissions, President Faust dismissed the idea
because “reducing information in a file does not help advance an understanding of the whole
person.”78 Yet in the case of socioeconomic status, admissions officers lack a full picture of the
students and so must piece together clues about whether a student is economically disadvantaged.
Accordingly, admissions officers try to make educated guesses by examining whether a student

76
“Economic Diversity: National Universities,” US News & World Report, http://www.usnews.com/
best-colleges/rankings/national-universities/economic-diversity. Harvard reports on its website that
“16% of the roughly 6,600 current undergraduate students are Pell Grant recipients.” Harvard College
Griffin Financial Aid Office, “Fact Sheet,” http://college.harvard.edu/financial-aid/how-aidworks/fact-sheet.
77

Donahue deposition, pp. 36 -38.

78

Faust deposition, p. 44.

23

According to Sally Donahue, Harvard has never discussed providing this information to
admissions officers.82 But withholding critical information about a student’s specific family income
and assets makes it impossible for Harvard to implement a sophisticated socioeconomic affirmative
action program as a race-neutral alternative for attaining the educational benefits of racial, ethnic, and
socioeconomic diversity. Because HFAI eligibility extends to families making up to $80,000 a year,
knowing a student is HFAI-eligible does not allow an admissions officer to discern whether the

79
Donahue deposition, p. 47; Harvard’s Response to Plaintiff’s Second Set of Interrogatories, p. 21.
80

Donahue deposition, p. 41. See also Fitzsimmons deposition, p. 203; Harvard’s Response to
Plaintiff’s Second Set of Interrogatories, p. 22; Kaitlin Howrigan deposition, p. 207.
81

Donahue deposition, pp. 40-41. By contrast, Harvard does not have to guess whether
underrepresented minority students are “likely” to be African American or Hispanic.
82

Donahue deposition, pp. 57-58.
24

student comes from a very disadvantaged family making less than $10,000 a year, compared with one
making as much as $80,000,83 which is above the national median.84 Harvard’s current admissions
system thus asks admissions officers to treat the bottom half of the U.S. income distribution as an
undifferentiated mass.
Moreover, admissions officers have no solid information about a family’s assets. As discussed
above, that is a critical omission because wealth is an important determinant of opportunity. Indeed,
for the purposes of race-neutral analysis, wealth has a much higher correlation with race than does
income, which means the potential racial dividend of using wealth is substantially greater than it is for
using income.85
4.

Harvard could increase the weight it gives to socioeconomic factors.

Harvard has long claimed to “give significant favorable consideration” to economically
disadvantaged students in pursuit of socioeconomic alongside racial and ethnic diversity.86 But
statistical analyses from both Harvard and SFFA’s expert refute that claim.

83
There is some ambiguity regarding the HFAI admissions “flag.” The HFAI program is for students
that come from families making $80,000 or less, but Director McGrath testified that Harvard considers
families making $85,000 or less to be low-income. McGrath deposition, p. 25.
84

See Proctor, Semega, and Kollar, “Income and Poverty in the United States: 2015,” supra, pp. 5-6,
Table 1. In 2015, the median household income in the United States was $56,516. Id. at 5. For those
households headed by individuals ages 45-54—a typical age for parents with college-aged students at
Harvard—the median income was $73,857. Id. at 6, Table 1. Households headed by individuals ages
45-54 is the metric Harvard has used in national comparisons. See Harvard documents, 77544.
85

Conley, “The Why, What, and How of Class-Based Admissions Policy,” supra, p. 209.

86

See “Brief of Harvard University, Brown University, The University of Chicago, Dartmouth College,
Duke University, The University of Pennsylvania, Princeton University, and Yale University as Amicus
Curiae Supporting Respondents,” U.S. Supreme Court in Grutter v. Bollinger and Gratz v. Bollinger,
February 18, 2003, p. 22, n.13.

25

First, in May 2013, Harvard’s Office of Institutional Research (OIR) examined the preferences
in admissions provided to students in the classes of 2009-2016. Looking at data from 192,359
applicants, the Harvard researchers ran a logistic regression to estimate the size of the probability of
admissions for such factors as athletic rating, legacy status, race, and income. Setting aside ratings that
were based on academic, extracurricular, or personal characteristics (which most would consider
legitimate criteria for admission),87 preferences mattered in the following order of size as denoted by
the coefficient (the larger the coefficient, the bigger the size of the preference):
Preference
Athletic rating of 1
Legacy
African American
Native American
Hispanic
Income less than $60,000
International
Asian

Coefficient
6.33
2.40
2.37
1.73
1.27
0.98
0.24
-0.37

Given these findings, the authors understandably conclude that “[c]ompared to athletes and
legacies, the size of the advantage for low income students is relatively small.”88
In addition, OIR also noted that another way of expressing the relative weights provided for
various preferences is the increased chances of admissions, controlling for academic ability. Among
students receiving the highest academic ratings (1 or 2),89 OIR concluded that

87
Personal rating includes teachers, guidance counselor and alumni interview evaluations along with
the student essay. Harvard documents, 65755.
88

Harvard documents, 23549.

89

According to testimony from Director McGrath, an academic ranking of 1 is rare and roughly
translates into SATs in the 700s, excellent high school grades, and award-winning distinctions. An
academic rank of 2 is typically SATs in the 680s to high 700s. An academic rating of 3 is typically SATs

26

•

Recruited Athletes received a 67-percentage point boost (83% chance of admission vs.
16% for nonathletes).

•

Legacies received a 40-percentage point boost (55% chance of admission vs. 15% for
nonlegacies).

•

Low-income students received a 9-percentage point boost (24% chance of admission vs.
15% for non-low income).

Second, SFFA’s expert witness, Peter Arcidiacono of Duke University, provides similar (and
confirming) analysis. Arcidiacono reviewed data from the 216,334 applicants from the class of 2014
to the class of 2019 admissions cycles, of which 150,701 were identified as an appropriate dataset. He
provides logit estimates of admission (with the largest numbers suggesting the largest boost). In rank
order of importance, his results (using his Model 6) show the relative weight of certain preferences in
Harvard’s admissions:
Preference

Logit Estimate of
Admission
7.849
2.659
1.84
1.704
1.419
1.282
1.083
0.023
-0.271

Recruited Athlete
African American
Legacy
Faculty/Staff Dependent
Hispanic
Early Action
Disadvantaged
First-Generation
Asian

Moreover, Arcidiacono finds that while Harvard provides a preference for disadvantaged
applicants, that preference is smaller for Hispanics, who already receive a large bump, and non-existent
for African Americans. A perverse effect of giving such a big preference for race is that it negates the

in the 600s and B grades. An academic rating of 4 is 500 SATs or below. McGrath deposition, pp.
167-69.

27

incentive to give minorities a socioeconomic preference. Indeed, Arcidiacono’s analysis of the six-year
period covering the classes of 2014-2019 finds that 70.5% of underrepresented minority students are
advantaged.
Finally, it is worth noting that these findings are in line with prior studies examining similar
schools (including Harvard). Empirical research—from four sets of supporters of racial preferences—
suggest that universities do not in fact provide much of a leg up to economically disadvantaged
students, at least so long as direct racial preferences are available to them.
•

In a 2004 study of the nation’s most selective 146 institutions, Georgetown professors
Anthony Carnevale and Stephen Rose found that race-based preferences on average triple
the representation of blacks and Hispanics students compared to admission based on
grades and test scores, but that universities do nothing to boost socioeconomic
representation.90 In fact, the representation of poor and working class students is slightly
lower than if grades and test scores were the sole basis for admissions, the researchers
found.91 Harvard was among the institutions included.

•

In a 2005 study of highly selective institutions, the Mellon Foundation’s William Bowen
and colleagues found that being an underrepresented minority increases one’s chance of
admissions by 27.7 percentage points; that is, an applicant with a 40% chance of
admissions has a 68% chance if she is African American, Hispanic, or Native American.
By contrast, being in the bottom income quartile (relative to the middle quartiles) has no
positive effect.92 Harvard was among the 13 colleges examined.93

•

A 2009 analysis by Thomas Espenshade of Princeton and Alexandria Radford finds that,
at highly selective private institutions, the boost provided to African-American applicants
is worth 310 SAT points (on a 1600 scale), compared with 130 points for poor students,
70 points for working-class applicants, and (distressingly) 50 points for upper-middle class

90
Carnevale & Rose, “Socioeconomic Status, Race/Ethnicity, and Selective College Admissions,”
supra, p. 135.
91

Id. at 142.

92

William G. Bowen, Martin A. Kurzweil, & Eugene M. Tobin, Equity and Excellence in American
Higher Education (University of Virginia Press, 2005), p. 105, Table 5.1.
93

Id. at 289, Appendix Table 5.1.
28

students, relative to middle-class pupils.94 The authors did not disclose the identity of the
institutions studied.95
•

A 2015 study of 40 selective colleges by Sean Reardon of Stanford and colleagues using
2004 data concludes that “racial affirmative action plays (or played, in 2004) some role in
admissions to highly selective colleges but SES-based affirmative action did not.”96

In the end, these regression analyses indicate that Harvard is dramatically undervaluing
socioeconomic status compared with race.
B.

Harvard could increase financial aid.

In 2007, Harvard made headlines by announcing it would require no parental contribution for
families making less than $60,000 and would cap tuition at 10% of income for families making between
$80,000 and $180,000.97 Since then, however, Harvard has “scaled back” its commitment to families
making between $150,000 and $180,000, as Sally Donahue testified.98 Likewise, the cutoff for
Harvard’s financial aid program (which now stands at $80,000) has not been updated for several years,
failing to keep up with inflation, even as the “free college” movement has become more prominent
nationally.
These reductions in aid obviously matter for socioeconomic diversity but they also matter for
racial diversity. Unaided students come from the wealthiest families in the country, so it is relevant to

94
Thomas J. Espenshade & Alexandria Walton Radford, No Longer Separate, Not Yet Equal
(Princeton University Press, 2009), p. 92, Table 3.5.
95

Id. at 411.

96

Sean F. Reardon, Rachel Baker, Matt Kasman, Daniel Klasik, & Joseph B. Townsend, “Can
Socioeconomic Status Substitute for Race in Affirmative Action College Admissions Policies?
Evidence From a Simulation Model,” Educational Testing Service, 2015, p. 6.
97

“Harvard Announces Sweeping Middle-Income Initiative,” Harvard Gazette, December 10, 2007,
http://news.harvard.edu/gazette/story/2007/12/harvard-announces-sweeping-middle-incomeinitiative.
98

Donahue deposition, p. 84.
29

note that whites constitute 96.2% of the nation’s top 1% of earners and African Americans just 1.4%.99
Overall, Harvard has seen a decline in the number of aided freshman students since a peak for the
class of 2012, when 68% of students received financial aid. By the freshman class scheduled to
graduate in 2017, the proportion of aided students declined to 55%.100 Reductions in financial aid were
projected to save Harvard $46 million for freshmen entering in the class of 2022.101
Harvard may claim that increasing financial aid would be too expensive. But Harvard officials
repeatedly testified that the institution could handle the expenses related to providing financial aid for
more economically disadvantaged students. Sally Donahue testified that Harvard has no maximum or
cap on the amount of money available for financial aid. Indeed, she said, it would not present a
problem if the number of HFAI students doubled.102 (Doubling would take HFAI students from 25%
to 50% of the class.) Likewise, Dean Fitzsimmons testified that Harvard’s charge is to recruit and
admit “the best, most interesting students,” and that there is no economic restraint on the number of
students Harvard can admit through its financial aid initiative.103

99
Shartia Brantley, “Who Are the Black ‘1 Percent’?” The Grio, November 21, 2011 (based on
calculations from Federal Reserve data).
100

Harvard documents, 7807.

101

Harvard documents, 7810.

102

Donahue testimony, pp. 43-44. See also id. at 95-96.

103

Fitzsimmons deposition, pp. 46, 203. See also Howrigan deposition, p. 204.

30

That Harvard has the financial means to significantly increase its financial aid awards is not
surprising. Harvard’s endowment is $37.1 billion.104 This is the largest endowment in the nation,
exceeding the GDP of more than half the countries in the world.105
C.

Harvard could reduce or eliminate preferences that favor non-minorities.

Harvard also insists on retaining a number of preferential programs that disproportionately
benefit wealthy and white students—policies whose elimination would increase socioeconomic and
racial diversity. These include preferences for the children of alumni, donors, and faculty and staff. I
discuss each in turn.
1.

Legacy preferences

Harvard has a long and shameful history of using legacy preferences for the offspring of
alumni. As Peter Schmidt of the Chronicle of Higher Education has noted, Harvard began using legacy
preferences for the children of alumni as a strategy for reducing the admissions of Jewish students.106
To this day, legacy preferences disproportionately benefit white students to the detriment of AsianAmerican, African-American, and Hispanic students. Recall that Harvard’s own analysis shows that
legacies received a 40% boost in their chances of admissions (compared with just a 9% boost for low-

104
N.P. “Narv” Narvekar, “Message from the CEO,” Harvard Management Company, Inc.,
September 2017, 1, p. 1, http://www.hmc.harvard.edu/docs/Final_Annual_Report_2017.pdf.
105

Adam Vaccaro, “Harvard’s Endowment Is Bigger Than Half the World’s Economies,”
Boston.com, September 25, 2014, http://www.boston.com/news/business/2014/09/25/harvardsendowment-is-bigger-than-half-the-worlds-economies.
106

See Peter Schmidt, “A History of Legacy Preferences and Privileges,” in Affirmative Action for the
Rich: Legacy Preferences in College Admissions, ed. Richard D. Kahlenberg (New York: Century
Foundation press, 2010), p. 42.

31

income students).107 And Arcidiacono’s analysis finds legacy status provided a bigger boost than other
preferences except for African Americans and recruited athletes.
Harvard persists in promoting legacy preferences despite ample evidence that doing so
undermines its efforts to promote racial and socioeconomic diversity. As the former chief counsel for
the Lawyers Committee for Civil and Human Rights, John Brittain, and his coauthor Eric Bloom have
noted, at Harvard, only 7.6% of legacy admits in 2002 were under-represented minorities, compared
with 17.8% of all students. The authors note that “affirmative action does not offset legacy preference:
the use of legacy preference, in fact, requires college admission officers to rely more heavily on
affirmative action.”108 Likewise, Harvard’s own data show that legacies are disproportionately wealthy.
In the six years for the classes of 2009-2014, 12% of the class consisted of legacies, and of legacies, an
astonishing 76% did not apply for financial aid.109 Arcidiacono’s modeling suggests eliminating legacy
would have a small, but positive influence on African-American representation and on Hispanic
representation.
Finally, it should be noted that eliminating legacy preferences is a workable race-neutral
strategy. Among the top 10 universities in the widely-cited Shanghai rankings, four (Caltech, U.C.
Berkeley, Oxford, and Cambridge) do not employ legacy preferences.110 Research also finds that the
existence of legacy preferences does not increase alumni donations to an institution. In an examination
of the top 100 universities in U.S. News & World Report, Chad Coffman of Winnemac Consulting

107
Harvard documents, 23550.
108

Brittain & Bloom, “Admitting the Truth,” supra, p. 132.

109

Harvard documents, 77643.

110

Richard D. Kahlenberg, “Introduction,” in Affirmative Action for the Rich, supra, p. 8.

32

and colleagues found “no evidence that legacy preference policies themselves exert an influence on
giving behavior.”111
In his testimony, Dean Fitzsimmons suggested that providing legacy preferences is “essential
to Harvard’s well-being.”112 Nevertheless, Fitzsimmons could not point to any empirical evidence that
legacy preferences had boosted alumni giving at Harvard.113 Dean Khurana’s testimony on the
purported educational benefits of legacy preferences seemed particularly strained. He suggested it was
important for Harvard to favor the children of alumni in order to bring students who “have more
experience with Harvard” together with “others who are less familiar with Harvard.” The ability of
these different groups to “exchange perspectives, points of view,” he claimed, would make “them
more effective citizens and citizen leaders for society.”114 I am aware of no research to support this
claim.
2.

Donor preferences

Harvard also generally provides preferential admissions to the children of major donors.
Director McGrath testified that applicants related to substantial donors are noted by admissions
officers.115

111
Chad Coffman, Tara O’Neil, & Brian Starr, “An Empirical Analysis of Legacy Preferences on
Alumni Giving at Top Universities,” in Affirmative Action for the Rich, supra, p. 113.
112

Fitzsimmons deposition, p. 189.

113

Fitzsimmons deposition, pp. 192-94. By contrast, Dean Smith testified that he did not think of
legacy admissions as tied to alumni giving. Smith deposition, p. 236.
114

Khurana deposition, p. 247.

115

McGrath deposition, p. 207.

33

According to Daniel Golden’s book, The
Price of Admission, Harvard has formalized this system for the wealthiest donors, those giving $1
million or more, through the creation of Harvard’s Committee on University Resources (COUR).
Golden found that 336 children of COUR members have gone on to attend Harvard, “an astonishing
enrollment rate of one child per major donor.”117 Ending admissions preferences to the children of
donors obviously would help increase socioeconomic diversity.
3.

The Z-list.

Relatedly, Harvard operates a unique form of admissions known as the “Z-list” in order to
admit a group of mostly advantaged students, including legacies, who would probably not be admitted
through the ordinary course.118 The Z-list is a method by which Harvard admits a select group of
students but only on the condition that they take a year off before enrolling at Harvard. As Golden
explained, “If wealthy or well-connected aren’t admitted to Harvard in the standard fashion, they need
not despair... They may be placed on the ‘Z-list’—a Harvard admissions office term for a little-known
policy that compromises standards in the interest of alumni and donors, enabling their children to
enter America’s most famous university by a side door.”119
Harvard’s admissions data confirm Golden’s analysis:

116
Fitzsimmons deposition, pp. 275-78.
117

Daniel Golden, The Price of Admission: How America’s Ruling Class Buys Its Way into Elite
Colleges—and Who Gets Left Outside the Gates (Crown Publishers, 2006), p. 26.
118

Id.

119

Id. at 37.

34

In the six years covered in Arcidiacono’s database, Harvard admitted on average almost 60
students each year through the Z list. While Harvard boasts of having a majority-minority student
population overall, 70% of Z-list students were white, 14% were Asian, 4% were Hispanic, and just
2% black. A whopping 58.8% were on the special “Dean’s Interest” list and 46.5% were legacies. By
contrast, just 1.2% were disadvantaged, and 1.8% first generation. Grace Cheng, a former Harvard
admissions officer, explained the Z-list is not designed for students of modest means. “We had
concerns of lower income, underresourced students not knowing what to do for a year, for a year off

35

of traditional schooling.”120 Academically, Z-list admits fell far below the typically admitted student;
indeed, their academic records on average fall about as close to rejected students as they do to admitted
students. Eliminating this preferential program for largely white, wealthy, and well-connected students
would be an important way to increase Harvard’s racial and socioeconomic diversity.
4.

Faculty and staff preferences

Harvard also provides a substantial preference in admissions to the offspring of faculty and
staff. According to Arcidiacono’s analysis, the boost provided for this favored category of applicants
is bigger than that provided to Hispanics, those applying early admission, disadvantaged students, and
first-generation students. Arcidiacono also observes that the applicants receiving the faculty and staff
are disproportionately white, and that African-American and Hispanic applicants are underrepresented
compared to their share in the applicant pool as a whole..
D.

Harvard could adopt admissions policies utilizing geographic diversity,
including percentage plans and the use of zip codes.

Harvard says it seeks geography diversity in its student body, but the commitment appears to
be weak, which in turn undercuts its efforts to promote student body diversity. For example, the 2010
U.S. Census finds that 37% of Americans (and 55% of African Americans) live in the South.121
Nevertheless, in the class of 2021, just 18.8% of Harvard students came from the South.122

120
Cheng deposition, p. 141-42.
121

U.S. Census Bureau, “2010 Census Shows Black Population has Highest Concentration in the
South,” September 29, 2011, http://www.census.gov/newsroom/releases/archives/2010_census/
cb11-cn185.html; U.S. Census Bureau, “Population by Region,” http://www.census.gov/popclock/
data_tables.php?component=growth.
122

Harvard University, “Admissions Statistics,” http://college.harvard.edu/admissions/admissionsstatistics.

36

Dean Fitzsimmons testified that Harvard has not conducted a formal analysis into the use of
precise geographic factors, including zip codes.123 This omission is remarkable given that the raceneutral zip code strategy has been championed by Harvard Professor Danielle Allen, who holds one
of the school’s prestigious university professorships, Harvard’s “highest faculty honor.”124 According
to Professor Allen, programs that enhance geographic diversity (and thus leverage the unfortunate
reality of residential and high school segregation by race and class for a positive purpose) can promote
integration in higher education. Professor Allen has noted that zip codes provide an important way
for national universities to provide geographic diversity and also contribute to racial, ethnic, and
socioeconomic diversity.125 Allen has described how “[g]eographically based structures for seeking
talent are tried and true” and “the pursuit of geographic diversity in admissions is our best hope of
merging the goals of diversity and excellence.”126 Such geographic diversity could “be taken to the
level of ZIP codes and, in particular, to the level of the ZIP+4 system, which divides the United States
into geographic units as small as a city block or group of apartments.”127 Professor Allen suggests that
a university might sort students through a “geographic diversity algorithm” and then “review the
identified admits, case-by-case, confirming or disconfirming [each] selection.”128 A university might

123
Fitzsimmons deposition, p. 184.
124

Anna Steinbock, “Danielle Allen Named University Professor,” Harvard Gazette, November 14,
2016, http://news.harvard.edu/gazette/story/2016/11/danielle-allen-named-university-professor/.
For a related idea, see Cashin, Place Not Race, supra.
125

See Danielle Allen, “Talent is Everywhere: Using ZIP Codes and Merit to Enhance Diversity,” in
The Future of Affirmative Action, supra.
126

Id. at 147.

127

Id.

128

Id. at 148.

37

also “determine the combination of SAT score and GPA that would constitute its entrance threshold”
and then choose the highest performing applicants within specific ZIP codes.129 Given the increasing
number of “ethnic census tracts,” in which certain minority groups constitute more than 25% of the
tract population, Professor Allen expects that “at selective colleges and universities a stronger
orientation toward geographic diversity could well support diversification of student populations by
ethnicity, thereby permitting us to slip free of the contested terrain of affirmative action.”130
Such methods have already been put into action. For example, Halley Potter and I have written
about public charter schools in San Diego, California, which have used zip codes to ensure
socioeconomic and racial diversity.131 Such geographic and socioeconomic diversity can succeed
because, unfortunately, concentrated poverty is often highly correlated with race. African Americans
and Hispanics are much more likely to live in neighborhoods with concentrated poverty than whites.
For example, while 6% of young whites live in neighborhoods with more than 20% poverty rates,
66% of African Americans live in such neighborhoods.132 Indeed, Carnevale’s simulation, noted above,
finds that a comparable approach—admitting high test scorers within schools—promotes
socioeconomic and racial diversity.133 Harvard’s failure to examine excellence within zip codes—what

129
Id. at 147.
130

Id. at 155-56.

131

See Richard D. Kahlenberg & Halley Potter, A Smarter Charter: Finding What Works for Charter
Schools and Public Education (Teachers College Press, 2014), p. 186.
132

See Patrick Sharkey, Stuck in Place: Urban Neighborhoods and the End of Progress Toward Racial
Equality, Figure 2.1 (University of Chicago Press, 2013), p. 27.
133

Carnevale, Rose, & Strohl, in “Achieving Racial and Economic Diversity with Race-Blind
Admissions Policy,” supra.

38

might be considered a private university alternative to the Texas top 10 percent plan—represents a
major missed opportunity.
E.

Harvard could increase its recruitment efforts.

Although Harvard touts its recruitment of underrepresented minorities, it exerts far less effort
to recruit economically disadvantaged applicants, many of whom are underrepresented minorities. For
example, Dean Fitzsimmons testified that while Harvard purchases information from institutions such
as the College Board to recruit high achieving underrepresented minority students (those scoring
between 1100-1600 on the SAT), it does not appear to make a similar effort to purchase information
about high achieving but socioeconomically disadvantaged students.134 Harvard fails to purchase this
information despite Fitzsimmons’ testimony that there are socioeconomically disadvantaged students
who could succeed at Harvard who are not currently in the applicant pool.135
Harvard does an especially poor job of recruiting into its applicant pool students whose
parents do not have a college degree. For the classes of 2007-2011, such students comprised just
12.5% of all applicants. For the class of 2016, the percentage was 12.7%.136 By comparison, as noted
above, 68% of adults age 45-54 lack a college degree.137 Looking at very high achieving students (SAT
scores of 1450 and higher), nearly half (43%) of those students from families making more than

134
Fitzsimmons deposition, pp. 59-67.
135

Fitzsimmons deposition, p. 207.

136

Harvard documents, 31726.

137

Ryan & Bauman, “Educational Attainment in the United States: 2015,” supra, p. 2, Table 1.

39

$100,000 applied to become members of the Harvard class of 2009. By contrast, less than a quarter
of those very high test scorers making less than $100,000 applied.138
Part of Harvard’s problem is that it recruits disproportionately from areas that have few highachieving low-income students. For example, the Midwest and Mountain states produce 21.2% of
high-achieving low-income students, which is six times as many as exist in New England (3.5%).139
Yet in the class of 2021 Harvard enrolled more students from New England (16.5%) than in all of the
Midwest and Mountain states combined (13.4%).140
More specifically, Harvard relies heavily on a relatively small number of “feeder” schools to
fill a significant part of its class. For the classes of 2007-2016, 20.3% of matriculates and 12.9% of
applicants, came from schools that represent just 0.6% of American high schools.141 Viewed through
the lens of the College Board’s high school cluster designations, which sort schools by demographics,
Harvard’s data produced in this case reveals that it recruits 54.5% of domestic students from just two
of 29 clusters.142
This failure to recruit high-achieving, low-income students, including thousands who are
African American and Hispanic, is an enormous missed opportunity. As discussed above, there is a
very large reservoir of such students whom Harvard, the nation’s oldest and wealthiest university, is
not recruiting.

138
Harvard documents, 77538.
139

Cashin, Place Not Race, supra, p. 47.

140

“Harvard Statistics,” https://college.harvard.edu/admissions/admissions-statistics

141

Harvard documents, 32982.

142

For background on College Board’s 29 high school clusters, see College Board, “Segment Analysis
Service: An Educationally Relevant Geodemographic Tagging Service,” 2011, p. 4, http://
media.collegeboard.com/mSSS/media/pdf/segment-analysis-service-overview.pdf.

40

F.

Harvard could increase its admission of community college transfers.

Harvard also fails to provide the opportunity for a meaningful number of high-achieving
community college students to transfer to Harvard—a strategy used by many selective public and
private colleges to promote socioeconomic and racial diversity in their student bodies. Community
colleges have many more African-American, Hispanic, and low-income students than selective fouryear colleges.143 According to the American Association of Community Colleges, “the majority of
Black and Hispanic undergraduate students in this country study at [community] colleges.”144
While other colleges began ramping up community college transfers, Harvard has for years
lagged in this arena.145 For the Classes of 2014 through 2019, only two community college students
transferred to Harvard. By contrast, Amherst College, a highly competitive institution, enrolls between
12 and 15 community college transfer students annually to its undergraduate population of 1790.146 If
Harvard, which has an undergraduate population of 6700, adopted a similarly scaled program, it would
include between 44 and 55 community college transfer students per year. This rate would translate
into about 300 community college transfers over a six-year period, rather than the two actually

143
See Bridging the Higher Education Divide: Strengthening Community Colleges and Restoring the
American Dream – Report of The Century Foundation Task Force on Preventing Community
Colleges from Becoming Separate and Unequal (Century Foundation Press, 2013), pp. 18-21.
144

American Association of Community Colleges, “Students at Community Colleges,”
http://www.aacc.nche. edu/AboutCC/Trends/Pages/studentsatcommunitycolleges.aspx.
145

See Arianna Markel, “Harvard Lags in Community College Recruitment,” The Harvard Crimson,
December 12, 2007.
146

Jennifer Glynn, “Opening Doors: How Selective Colleges and Universities Are Expanding Access
for High-Achieving, Low-Income Students,” (Jack Kent Cooke Foundation, April 2017), p. 37; see
www.jkcf.org/assets/1/7/JKCF_Opening_Doors_Executive_Summary.pdf.

41

admitted.147 The failure represents another missed opportunity to add racial and socioeconomic
diversity to Harvard.
G.

Harvard could end early admissions.

Finally, Harvard could increase racial, ethnic, and socioeconomic diversity if it were to drop
its “early admissions” program that disproportionately benefits wealthy and white students. Early
admissions is a practice in which schools allow students to submit their application in the early Fall if
they apply to only one school. For several years, Harvard eliminated early admissions, citing the
unfairness to low-income and minority students. But then, in 2011, it abruptly reversed course,
reinstating a policy that advantages the already advantaged. According to Arcidiacono’s model,
applying early has a positive logit estimate of 1.282 at Harvard—meaning that students who apply
early receive a bigger boost than Harvard provides to first-generation and disadvantaged students.
Early admission programs, like Harvard’s program, usually benefit wealthier and betterinformed students because these students have the resources to submit their application early and do
not need to hold out for the prospect of financial aid.148 By contrast, low-income students and
minorities face a disadvantage under early admissions because they often receive inadequate
information and counseling and lack the economic resources to commit to a school so early in the
process. According to a 2011 study by Julie J. Park of Miami University and M. Kevin Eagan of the
UCLA Higher Education Research Institute, students who applied early-action to 290 colleges and

147
Harvard at a Glance, https://www.harvard.edu/about-harvard/harvard-glance.
148

See Alan Finder & Karen W. Arenson, Harvard Ends Early Admission, New York Times,
September 12, 2006, http://www.nytimes.com/2006/09/12/education/12harvard.html.

42

universities across the country are more economically advantaged and more likely to be white than
those who did not apply early.149
Because early admissions undermine the chances of low-income and minority applicants,
Harvard terminated early admissions in 2006. According to Harvard’s then-President Derek Bok: “We
hope that doing away with early admission will improve the process and make it simpler and fairer.
Early admission programs tend to advantage the advantaged. . . . Students from more sophisticated
backgrounds and affluent high schools often apply early to increase their chances of admissions, while
minority students and students from rural areas, other countries and high schools with fewer resources
miss out.”150 Similarly, Dean Fitzsimmons supported ending early admissions, noting: “An early
admission program that is less accessible to students from modest economic backgrounds operates at
cross-purposes with our goal of finding and admitting the most talented students from across the
economic spectrum.”151 Nevertheless, in 2011, Harvard reinstated early admissions. Harvard did so
because it was concerned that its yield rate (those accepting their offer of admission) had declined.
The decline occurred among white and Hispanic students, while yield among African-American
students remained steady and increased for Asian students.
Part of the rationale for restoring early admission, according to internal documents, was that
Harvard was losing out to other colleges in the competition for students from certain high schools.
“At some of Harvard’s most productive ‘feeder’ high schools . . . up to 90% of students apply early

149
Julie J. Park & M. Kevin Eagan, “Who Goes Early? A Multi-Level Analysis of Enrolling via Early
Action and Early Decision Admissions,” Teachers College Record, 2011.
150

Harvard documents, 77577-77578.

151

Harvard documents, 77579.

43

somewhere. By not offering an admission option, therefore, Harvard has placed itself at odds with
trends in the market.”152 Yet Harvard’s data showed that even without early admissions, it was winning
the competition for students who were also admitted to Yale, Stanford, and Princeton two-thirds of
the time.153 Regardless, it appears that Harvard’s concerns about its place in the market outweighed its
earlier stated concerns about equity.
In 2013, after the introduction of early admissions, Harvard analyzed its admissions process
and found that significant differences remained between the types of students who applied for early
versus regular admissions. White students and legacy students were more likely to apply early than a
typical student. African-American students, Hispanic students, international students, those whose
parents had no college degree, and those seeking a fee waiver were all more likely to apply regular
admission than the typical student.154 Harvard could increase student body diversity by eliminating
early admissions as several other selective colleges have.155

152
Harvard documents, 77565.
153

Harvard documents, 77587.

154

Harvard documents, 65579. The differences were statistically significant in all these cases in 2016
except for African Americans and Hispanics. In 2016 African Americans represented 8.9% of the
early action pool and 9.9% of the regular pool. Hispanics represented 9.0% of the early action pool
and 10.3% of the regular pool. Harvard documents, 65609.
155

Christopher Avery & Jonathan Levin, “Early Admission at Selective Colleges,” Stanford Institute
for Economic Policy Research, March 2009, p. 4 (noting that the University of Michigan, and four
top University of California colleges did not employ early admissions).

44

VI.

Simulations of Harvard’s data show that workable race-neutral alternatives exist.
A.

A careful simulation indicates that Harvard could achieve the educational
benefits of racial, ethnic, and socioeconomic diversity without sacrificing
academic quality.

To simulate the likely results of adopting race-neutral strategies at Harvard, Professor
Arcidiacono tested the results of a few race-neutral options using the admissions data provided by
Harvard. At my request, he conducted simulations of multiple race-neutral alternatives to forecast the
likely outcomes thereof.156 These simulations, and the underlying assumptions, are set forth in the
charts set forth in detail in Appendix C.157 For discussion purposes in this report, I will focus on
Simulation 4.
To replicate as closely as possible Harvard’s existing system of admissions, Arcidiacono began
by using the model he developed that accounts for numerous criteria for admission and characteristics
of the applicants, including Harvard’s four-part rating system, which rates applicants on in four areas:
academics, extracurricular activities, athletics, and personal ratings.158 I asked Arcidiacono to exclude

156
I have worked in the past with researchers such as Anthony Carnevale at Georgetown University
to measure the effectiveness of race-neutral alternatives through similar simulations. See supra Section
IV.B.
157

Some of the simulations include an “expanded pool,” in which I assume that Harvard—through
increased recruiting efforts focused on socioeconomically disadvantaged areas and high schools—
could double the number of disadvantaged students in its applicant pool. This assumption is
reasonable, given data that suggest that in the past, Harvard had twice as many high-achieving, highincome students apply as high-achieving, lower-income students. Harvard documents, 77538 (Among
high achieving students (SAT scores of 1450 and higher), nearly half (43%) of those students from
families making more than $100,000 applied to become members of the Harvard class of 2009. By
contrast, less than a quarter of those very high test scorers making less than $100,000 applied.) The
expanded pool scenarios assume that the expanded pool would contain students with characteristics
identical to Harvard’s current socioeconomically disadvantaged pool.
158

This is based upon Model 6 illustrated in Table B.8.2 in the Arcidiacono report, with two
adjustments. In addition to race interacted with year, the model also contains an interaction between
disadvantaged and year. Second, the overall rating is excluded as a control.

45

from the model Harvard’s overall rating, which may reflect admissions preferences. The advantage
associated with the various preferences were then “turned off”—specifically, the preferences for race,
legacy, recruited athlete, early decision, faculty and student children, the Dean/Director list, fee waiver,
first generation, and financial aid recipients. With those preferences off, admissions probabilities could
be generated and the applicants could be ranked in order of strength under the remaining aspects of
Harvard’s admissions process. This approach allows for simulating the effects of a variety of raceneutral options on racial diversity, socioeconomic diversity, and academic readiness.
Before beginning Simulation 4, Arcidiacono turned Harvard’s existing preferences for
recruited athletes back “on.” He did this at my direction, because I have found that removing athletic
preferences in connection with race-neutral alternatives is sometimes perceived as radical. This
particular simulation thus avoids any concern that eliminating recruited athletes is unworkable or
otherwise inappropriate when seeking a race-neutral alternative.
The first step in the simulation involved providing a preference to students that Harvard has
designated as economically “disadvantaged.” The disadvantaged tag is one applied by admissions
counselors through a somewhat subjective process: it would include those students whose parents do
not have a bachelor’s degree, whose family income is likely to be $80,000 or lower, who live in a
neighborhood admissions officers deem disadvantaged, or who requested a fee waiver on the
application.159

159
See supra Section V.A.3. Because the disadvantaged tag doesn’t identify where family income falls
in the range from $0 to $80,000 or the net worth/wealth of applicants, the broad designation does not
allow for consideration of the available economic data that is most highly correlated with race. As a
result, the simulations likely form a lower bound estimate of the racial dividends of these strategies.
Better data could produce higher levels of racial diversity.

46

The magnitude of the preference for disadvantaged students in the simulation is somewhat
larger than the one currently provided to African-American students, but only half the size provided
to recruited athletes (who currently receive the largest preference). The simulated preference is larger
than that provided to African Americans because evidence suggests that socioeconomic obstacles to
academic achievement are greater in magnitude than racial obstacles. An economically disadvantaged
student who managed to overcome hurdles may have a more promising future than her academic
profile on paper.160 Moreover, as William Bowen, the former President of Princeton University, has
noted, SAT scores do not over-predict the college grades of low-income students as they do those of
African-American students.161 At the same time, the preference is calibrated to be smaller than those
provided to recruited athletes. Even though Harvard recruited athletes are, on average, fairly
wealthy,162 they receive a very large preference that, if provided in a simulation to economically
disadvantaged students, would overwhelm the freshman class with such students.
This first step by itself underestimates the potential of Harvard to create race-neutral strategies
to promote diversity because it does not directly consider geographic residence of applicants. As noted
above, a number of leading state universities have created racial diversity by employing “percentage

160
See, e.g., Anthony P. Carnevale & Jeff Strohl, “How Increasing College Access Is Increasing
Inequality, and What To Do About It,” in Rewarding Strivers 170, Table 3.7 (Century Foundation,
2010), p. 170, Table 3.7 (estimating the SAT scores socioeconomically disadvantaged students on
average are 399 points below socioeconomically advantaged students, while for African American
students, the expected score is 56 points lower).
161

Bowen, Kurzweil, & Tobin, Equity and Excellence in Higher Education, supra, p. 118 (SAT’s do
not over-predict college grade point average for low-income students); and William Bowen and Derek
Bok, The Shape of the River, p. 77 (SAT’s over-predict college grade point average for African
American students.)
162

See Harvard documents, 77643. (In the six-year period for the classes of 2009-2014, 53% of varsity
athletes did not apply for financial aid.)

47

plans” that enroll top students in a variety of high schools.163 Harvard could take top students from
various geographic locales, as Harvard professor Danielle Allen has suggested.164 Although Harvard
refused to provide zip code data of applicants, it ultimately did provide data on whether applicants
reside in one of 33 “Educational Neighborhood Clusters,” provided by the College Board.165 This
approach is consistent with Harvard’s stated commitment to socio-geographic diversity.
The second step of Simulation 4 combines a commitment to geographic diversity, which
admits top-ranked students within each neighborhood cluster, and a preference for socioeconomically
disadvantaged students. After disadvantaged students are given a boost equivalent to half of that given
to athletes, applicants are then sorted according by rank and year within a neighborhood cluster. An
even number of the top applicants are then selected from each cluster, so that the total number of
admits in each year approximates the observed number of admits in each year.166 This combination of
approaches provides a sizable increase in socioeconomic diversity and maintains racial diversity to a
remarkable degree, yielding a net plus in overall diversity, while maintaining academic excellence. See
Appendix C for the full results (Simulation 4). But take the most recent class of 2019 as an example:

163
See supra Section IV.B.
164

See supra Section V.D.

165

See College Board, “Segment Analysis Service: An Educationally Relevant Geodemographic
Tagging Service,” 2011, p. 4; College Board, “Descriptor PLUS: Cluster Description Guide—
Educational Neighborhood Clusters; High School Clusters,” 2011.
166

A small percentage (3.5%) of the applicants considered in this model were not listed with a
neighborhood cluster. To ensure full representation of these students, they were merged with the
smallest cluster.

48

Harvard – Admitted Class of 2019
Status Quo
Race-Based Admissions
White
40.4%
African American
13.6%
Hispanic
12.9%
Asian American
23.7%
Other minority
9.3%
Disadvantaged
17.4%
Advantaged
82.6%
Academic Index
227.8

Simulation
Race-Neutral Admissions
White
39.6%
African American
10.1%
Hispanic
13.5%
Asian American
27.6%
Other minority
9.2%
Disadvantaged
54.3%
Advantaged
45.7%
Academic Index
225.9

Several observations about these findings are worth highlighting.
First, under the simulation, socioeconomic diversity would increase considerably, with the
“advantaged” proportion of the admitted class dropping from 82.6% to 45.7%. In considering these
numbers, it is important to recognize that “disadvantaged” in the context of Harvard’s extremely
wealthy student body would be thought of as a “typical American family” in any other context. The
“disadvantaged” designation covers more than two-thirds of American families headed by individuals
between the ages of 45 and 54 because simply having parents who lack a bachelor’s degree earns the
disadvantaged tag. This is in addition to those students with a parent holding a college degree making
less than $80,000 (roughly the median family income for adults age 45-54) or who live in a
disadvantaged neighborhood.167At 54.3% of the admitted class, these “disadvantaged” students would
still be underrepresented compared with the general population, but far less than under the status quo.
Second, under the simulation, overall racial and ethnic diversity would hold basically stable,
with Hispanic students rising from 12.9% to 13.5% of the class and African-American students
declining from 13.6% to 10.1% of the class. (The “other minority” category would remain basically

167
See supra Section V.E.

49

even.) The small decline in African-American representation could be addressed by using a wealth
measure (see discussion below). But it is important to recognize that this level of African-American
representation is not unusual for Harvard or in relation to other schools. A 10.1% representation is
greater than the representation found at Harvard throughout most of the affirmative-action era.
Harvard has prided itself for decades on providing the educational benefits of diversity, boasting that
“diversity is the hallmark of the Harvard experience.”168 Yet in the classes graduating in the 1970s and
early 1980s, Harvard’s African-American representation ranged from 7-8%, which seemed to satisfy
Harvard’s claimed interest.169 Even accounting for yield rates, this simulation would result in
comparable representation. In any event, Harvard President Drew Faust testified (repeatedly) that the
particular racial composition of the Harvard class is of no interest to Harvard officials.170 Moreover,
this simulation would greatly increase the share of disadvantaged African-American and Hispanic
students.
Third, in looking at the educational benefits of diversity, the Supreme Court—and Harvard
officials—have repeatedly suggested that both racial and socioeconomic diversity are important.171
While media reports often focus solely on the racial impact of alternatives, the critical measure is the
net impact on socioeconomic and racial diversity taken together. Given the large increase in

168
Elizabeth Bangs, “Harvard Adds Another ‘Hallmark’ ‘Distinction and Diversity’ Now Central to
Your College Experience,” Harvard Crimson, September 16, 1994 (noting Harvard touting diversity
“for years”).
169

Alan M. Dershowitz & Laura Hanft, “Affirmative Action and the Harvard College DiversityDiscretion Model: Paradigm or Pretext?” 1 Cardozo Law Review 379, 383 (1979).
170

Faust deposition, pp. 25-26, 91, 107-08.

171

See Grutter, 539 U.S. 306, 324 (2003); Bakke, 438 U.S. 265, 316 (1978); Fitzsimmons deposition,
pp. 421-23; McGrath deposition, p. 231; Faust deposition, p. 196; Khurana deposition, pp. 66, 75;
Smith deposition, p. 48.

50

socioeconomic diversity and the mixed effect on racial diversity, the simulation suggests a substantial
net increase in the educational benefits of diversity.
Fourth, it is important to put the impact on the “academic index” finding—a modest drop
from 227.8 to 225.9—in context. The 225.9 academic index is well within Harvard’s range of what it
considers to be academically qualified: it is comparable to the average academic index of Hispanic
students for the classes of 2014-2019 (225.9) and exceeds the average academic index of AfricanAmerican students and athletes (221.5 and 213.5 respectively). Indeed, Harvard officials repeatedly
testified that all students at Harvard—including the lowest scoring—are academically qualified.
President Faust, for example, said none of the current students were academically mismatched because
Harvard “would not want to have any students on the campus who were not able to thrive and succeed
in our academic environment.”172
This simulation is not the only way that Harvard could achieve its goals without the use of
race. But it confirms—using information about Harvard’s current process, and data already available
to Harvard—that such alternatives are both available and workable. Harvard no doubt could identify
alternative methods, if it were committed to doing so.
B.

Through inclusion of additional socioeconomic factors and better recruiting of
low-income students, the simulation could predict even greater racial and
ethnic diversity.

As noted above, this simulation could have achieved a more robust racial dividend if I had
access to additional information about critical socioeconomic factors the Court permitted Harvard to

172
Faust deposition, pp. 225-26, 228, 291-93, 295-96, 304. See also Khurana deposition, pp. 194-95
(that academic mismatch is not a problem at Harvard College.)

51

withhold from SFFA or which Harvard could not make available, or if Harvard had recruited more
aggressively.
More accurate income data. I was limited to using Harvard’s economic disadvantage flag,
which includes roughly the bottom half of the income distribution among 45-54 year olds nationally.
This metric, however, masks considerable variation. For instance, I could not model providing a bigger
boost in the analysis to a remarkable student who performed well academically despite coming from
a very disadvantaged household compared to a student near the median income nationally. This
limitation has important implications for the racial dividend of class-based affirmative action because
African-American and Latino students are far more likely to be in the bottom quarter of the country’s
income distribution. In 2016, the median income of white families was $61,200 but median family
income for black families was $35,400 and for Hispanic families was $38,500.173
Wealth data. Second, I did not have access to data on the wealth of applicants. As discussed
earlier, these data have enormous implications for the racial dividend of class-based affirmative action.
While African Americans make roughly 60% of what whites make in annual income, the median
wealth of African Americans is just 10% the median wealth of whites.174
Better recruitment. Third, the simulation necessarily understates the racial and
socioeconomic dividend of the alternatives studied because it was limited to the existing pool of
applicants even though evidence suggests that Harvard does a poor job of recruiting disadvantaged

173
See Dettling, Hsu, Jacobs, Moore, & Thompson, “Recent Trends in Wealth-Holding by Race and
Ethnicity,” supra.
174

Id.
52

students to apply.175 Recall that for the class of 2009, for example, Harvard found that roughly twice
as many very high achieving applicants from families making more than $100,000 applied to Harvard
than similar students from families making less than $100,000.176 (There are more than 20,000 very
high achieving low-income applicants who do not attend any of the most selective 238 colleges, much
less Harvard.)177 A more robust applicant pool thus likely could have increased the racial divided in
our simulations.
VII.

Conclusion
Under the Fourteenth Amendment, Harvard bears “the ultimate burden of demonstrating,

before turning to racial classifications, that available, workable race-neutral alternatives do not
suffice.”178 Harvard officials have claimed the college is “doing everything possible” to use race-neutral
strategies to promote socioeconomic and racial diversity, and has engaged in “a massive good-faith
effort to do everything that seemed to be effective.”179
The record refutes that assertion. Experience and research demonstrates that there are
numerous ways that universities can achieve the educational benefits of racial and socioeconomic
diversity without using race. Despite all of its financial and academic resources, Harvard has failed to
take even the most rudimentary steps to determine whether there are workable race-neutral strategies

175
See supra Section V.A.1.
176

Harvard documents, 77538.

177

Hoxby & Avery, “The Missing ‘One-Offs,’” supra, p. 35 (finding that two-thirds of 35,000 high
achieving low income students do not attend a selective colleges); see David Leonhardt, Better
Colleges Failing to Lure Talented Poor, New York Times, March 16, 2013,
http://www.nytimes.com/2013/03/17/education/scholarly-poor-often-overlook-bettercolleges.html?pagewanted=all.
178

Fisher, 133 S. Ct. 2411, 2420.

179

Fitzsimmons deposition, pp. 172, 175. See also id. at 176, 179, 204.

53

available. Moreover, a careful investigation of Harvard’s admissions data and practices confirms that
Harvard has at its disposal viable race-neutral alternatives that would provide a net increase in racial
and socioeconomic diversity without requiring the use of race.

Dated: October 16, 2017

s/ Richard D. Kahlenberg
Richard D. Kahlenberg

54

VIII. Appendices
A.

Appendix A – Curriculum Vitae

55

RICHARD D. KAHLENBERG
Senior Fellow
The Century Foundation
2040 S Street, N.W.
Washington, D.C. 20009
202-683-4883
[e-mail]
EDUCATION
1986-1989

Harvard Law School, Cambridge, Massachusetts.
J.D., cum laude, June 1989.

1985-1986

University of Nairobi School of Journalism, Nairobi, Kenya.
Certificate, Mass Communications, June 1986.
Rotary International Fellowship.

1981-1985

Harvard College, Cambridge, Massachusetts.
A.B. in Government, magna cum laude, June 1985.
Senior Honors Thesis “Coalition Building and Robert Kennedy’s 1968 Presidential
Campaign”

EMPLOYMENT HISTORY
1998-

The Century Foundation (formerly Twentieth Century Fund), Washington, D.C.
Senior Fellow. Coordinating programs involving elementary, secondary and higher
education and organized labor.

1996-1998

Center for National Policy, Washington, D.C.
Fellow. Coordinated project on New Strategies to Promote Equal Opportunity.

1994-1995

Professorial Lecturer and Independent Writer, Washington, D.C.
Taught Cases in Public Policy, George Washington University Department of Public
Administration and completed book on affirmative action.

1993-1994

George Washington University National Law Center, Washington, D.C.
Visiting Associate Professor of Law. Taught Constitutional Law.

1989-1993

Senator Charles S. Robb, Washington, D.C.
Legislative Assistant. Advised Senator on issues relating to Crime, Energy,
Environment, Judicial Appointments, Campaign Finance, and Civil Rights.

56

PUBLICATIONS
I. BOOKS

A Smarter Charter: Finding What Works for Charter Schools and Public Education
(coauthored with Halley Potter) (Teachers College Columbia University Press, 2014). The Washington
Post called A Smarter Charter, “A remarkable new book...Wise and energetic advocates such as
Kahlenberg and Potter can take the charter movement in new and useful directions.”
Why Labor Organizing Should Be a Civil Right: Rebuilding a Middle-Class Democracy by
Enhancing Worker Voice (coauthored with Moshe Z. Marvit) (Century Foundation Press, 2012).
The book was called “a must read” by NAACP President and CEO Benjamin Todd Jealous and “a
persuasive roadmap for extending the protections of the Civil Rights Act to workers who want to
organize a union” by American Federation of Teachers President Randi Weingarten.
Tough Liberal: Albert Shanker and the Battles Over Schools, Unions, Race and Democracy
(Columbia University Press, 2007). The Wall Street Journal called the book “a well researched and
engaging biography,” and Slate labeled it a “stirring account.” The book has also been reviewed in
The Nation, The American Prospect, The Weekly Standard, Newsday, New York Sun, City Journal,
Publishers Weekly, and The Washington Monthly. The book was written with the support of the
Hewlett, Broad and Fordham foundations. It was named one of the Five Best Books on Labor in
the Wall Street Journal
All Together Now: Creating Middle Class Schools through Public School Choice (Brookings
Institution Press, 2001). The book, labeled “a clarion call for the socioeconomic desegregation of
U.S. public schools” by Harvard Educational Review, was said by the Washington Post to make “a
substantial contribution to a national conversation” on education. The book was also reviewed in
Teachers College Record, Education Next, and National Journal. One author called Kahlenberg
“the intellectual father of the economic integration movement.”
The Remedy: Class, Race, and Affirmative Action (Basic Books, 1996). The book was named
one of the best of the year by the Washington Post and William Julius Wilson’s review in the New
York Times called it “by far the most comprehensive and thoughtful argument thus far
for...affirmative action based on class.” The book was also reviewed in The American Lawyer, The
New Yorker, The Progressive, The Washington Monthly, The Detroit News, National Review,
Legal Times, The Atlanta Journal-Constitution, and Publishers Weekly
Broken Contract: A Memoir of Harvard Law School (Hill & Wang/Farrar, Straus & Giroux,
1992). The book, which details the way in which idealistic liberal law students are turned to
corporate law, was called “a forceful cri de coeur” by the L.A. Times. The book was reviewed in
The New York Times, The Washington Post Book World, The Harvard Law Review, The
Washington Monthly, Legal Times, The Boston Globe, The Hartford Courant, The Baltimore
Evening Sun, The St. Petersburg Times, The Detroit News, The Cleveland Plain Dealer, The Dallas
Morning News, and Publishers Weekly. In 1999, the book was reissued by University of

57

Massachusetts Press with a new afterword. The book has also been translated into Japanese and
Chinese.
Editor, The Future of Affirmative Action: New Paths to Higher Education Diversity after
Fisher v. University of Texas (Century Foundation Press, 2014). Chapters include, “Defining the
Stakes,” by Nancy Cantor and Peter Englot; “Promoting Economic Diversity for College
Affordability,” by Sara Goldrick-Rab; “Fisher v. University of Texas and Its Practical Implications
for Institutions of Higher Education,” by Arthur L. Coleman and Teresa E. Taylor; “New Rules for
Affirmative Action in Higher Education,” by Scott Greytak; “Transitioning to Race-Neutral
Admissions,” by Halley Potter; “Striving for Neutrality,” by Marta Tienda; “The Use of
Socioeconomic Affirmative Action at the University of California,” by Richard Sander; “Converging
Perils to College Access for Racial Minorities,” by Richard L. McCormick; “Ensuring Diversity
Under Race-Neutral Admissions at the University of Georgia,” by Nancy G. McDuff and Halley
Potter; “Addressing Undermatch,” by Alexandria Walton Radford and Jessica Howell; “Talent is
Everywhere,” by Danielle Allen; “Reducing Reliance on Testing to Promote Diversity,” by John
Brittain and Benjamin Landy; ‘Advancing College Access with Class-Based Affirmative Action,” by
Matthew N. Gaertner; “Achieving Racial and Economic Diversity with Race-Blind Admissions
Policy,” by Anthony P. Carnevale, Stephen J. Rose, and Jeff Strohl; “The Why, What, and How of
Class-Based Admissions Policy,” by Dalton Conley; “A Collective Path Upward,” by Richard
Sander; and “Increasing Socioeconomic Diversity in American Higher Education,” by Catharine
Hill.
Executive Director (and primary author and editor), Bridging the Higher Education Divide:
Strengthening Community Colleges and Restoring the American Dream (Century Foundation
Press, 2013.) The task force on community colleges, cochaired by Anthony Marx and Eduardo
Padron, included John Brittain, Walter Bumphus, Michele Cahill, Louis Caldera, Patrick Callan,
Nancy Cantor, Samuel Cargile, Anthony Carnevale, Michelle Asha Cooper, Sara Goldrick-Rab,
Jerome Karabel, Catherine Koshland, Felix Matos Rodriguez, Gail Mellow, Arthur Rothkopf,
Sandra Schroeder, Louis Soares, Suzanne Walsh, Ronald Williams, and Joshua Wyner. In addition,
the volume included background papers by Sandy Baum and Charles Kurose; Sara Goldrick-Rab
and Peter Kinsley; and Tatiana Melguizo and Holly Kosiewicz.
Editor, The Future of School Integration: Socioeconomic Diversity as an Education Reform
Strategy (Century Foundation Press, 2012). Chapters include, “Housing Policy is School Policy:
Economically Integrative Housing Promotes Academic Success in Montgomery County, Maryland,”
by Heather Schwartz; “Socioeconomic Diversity and Early Learning: The Missing Link in Policy for
High-Quality Preschools,” by Jeanne L. Reid; “The Cost-Effectiveness of Socioeconomic School
Integration,” by Marco Basile; “The Challenge of High-Poverty Schools: How Feasible is
Socioeconomic School Integration?” by An Mantil, Anne G. Perkins, and Stephanie Aberger; “Can
NCLB Choice Work? Modeling the Effects of Interdistrict Choice on Student Access to HigherPerforming Schools,” by Meredith P. Richards, Kori J. Stroub, and Jennifer Jellison Holme; “The
Politics of Maintaining Balanced Schools: An Examination of Three Districts,” by Sheneka M.
Williams; and “Turnaround and Charter Schools that Work: Moving Beyond Separate but Equal,”
by Richard Kahlenberg.

58

Editor, Affirmative Action for the Rich: Legacy Preferences in College Admissions (Century
Foundation Press, 2010). Chapters include “Legacy Preferences in a Democratic Republic,” by
Michael Lind; “A History of Legacy Preferences,” by Peter Schmidt; “An Analytical Survey of
Legacy Preferences,” by Daniel Golden; “An Empirical Analysis of the Impact of Legacy
Preferences on Alumni Giving at Top Universities,” by Chad Coffman, Tara O’Neil and Brian Starr;
“Admitting the Truth: The Effect of Affirmative Action, Legacy Preferences, and the Meritocratic
Ideal on Students of Color in College Admissions,” by John Brittain and Eric Bloom; “Legacy
Preferences and the Constitutional Prohibition of Titles of Nobility,” by Carlton Larson; “Heirs of
the American Experiment: A Legal Challenge to Preferences as a Violation of the Equal Protection
Clause of the Constitution and the Civil Rights Act of 1866,” by Steve Shadowen and Sozi Tulante;
“Privilege Paving the Way for Privilege: How Judges Will Confront the Legal Ramifications of
Legacy Admissions to Public and Private Universities,” by Boyce F. Martin Jr. with Donya Khalili;
and “The Political Economy of Legacy Admissions, Taxpayer Subsidies, and Excess ‘Profits’ in
American Higher Education: Strategies for Reform,” by Peter Sacks.
Editor, Rewarding Strivers: Helping Low-Income Students Succeed in College (Century
Foundation Press, 2010). Chapters include: “The Carolina Covenant,” by Edward B. Fiske, and
“How Increasing College Access is Increasing Inequality and What to do About It,” by Anthony P.
Carnevale and Jeff Strohl. William Fitzsimmons called the book part of Century’s “trailblazing
mission to prevent the tragic waste of human talent that threatens America’s future,” while Anthony
Marx declared, “Kahlenberg again gathers the best thinkers on how to challenge this status quo;
what to do, what works, and what does not.”
Editor, Improving on No Child Left Behind: Getting Education Reform Back on Track
(Century Foundation Press, 2008). Chapters include: an analysis of the under-funding of the No
Child Left Behind Act, by William Duncombe, John Yinger and Anna Lukemeyer; a discussion of
the rights of students in low performing schools to transfer to better performing public schools
across district lines, by Amy Stuart Wells and Jennifer Holme; and an exploration of how to improve
the accountability provisions of the act, by Lauren Resnick, Mary Kay Stein, and Sarah Coon. Diane
Ravitch called Improving on No Child Left Behind “the best of the books on this topic.”
Editor, America’s Untapped Resource: Low-Income Students in Higher Education (Century
Foundation Press, 2004). The chapters include: “Socioeconomic Status, Race/Ethnicity, and
Selective College Admissions,” Anthony P. Carnevale and Stephen J. Rose; “Improving the
Academic Preparation and Performance of Low-Income Students in American Higher Education,”
by P. Michael Timpane and Arthur M. Hauptman; and “Low-Income Students and the Affordability
of Higher Education,” by Lawrence E. Gladieux. Carnevale and Rose’s finding, that 74% of
students at selective colleges come from the top socioeconomic quartile and 3% from the bottom
quartile is widely cited.
Editor, Public School Choice vs. Private School Vouchers (Century Foundation Press, 2003).
The volume consists of a compilation of new and previously published materials, including articles
by Edward B. Fiske, Helen F. Ladd, Sean F. Reardon, John T. Yun, Amy Stuart Wells, Richard Just,
Ruy Teixeira, Thad Hall, Gordon MacInnes, Richard C. Leone, and Bernard Wasow.

59

Executive Director (and primary author and editor), Divided We Fail: Coming Together
Through Public School Choice. The Report of The Century Foundation Task Force on the
Common School, (Century Foundation Press, 2002). The task force on school integration, chaired
by Lowell Weicker, included Joseph Aguerrebere, Ramon Cortines, Robert Crain, John Degnan,
Peter Edelman, Christopher Edley, Kim Elliott, Jennifer Hochschild, Helen Ladd, Marianne
Engelman Lado, Leonard Lieberman, Ann Majestic, Dennis Parker, Felipe Reinoso, Charles S.
Robb, David Rusk, James Ryan, Judi Sikes, John Brooks Slaughter, Dick Swantz, William Trent,
Adam Urbanski, Amy Stuart Wells, and Charles V. Willie. In addition, the volume included
background papers by Duncan Chaplin, David Rusk, Edward B. Fiske, William H. Freivogel,
Richard Mial, and Todd Silberman.
Editor, A Notion at Risk: Preserving Public Education as an Engine for Social Mobility
(Century Foundation Press, 2000). The book identifies individual sources of inequality and
proposes concrete public policy remedies. The chapters include: “Summer Learning and Home
Environment” by Doris Entwisle, Karl Alexander and Linda Olson of Johns Hopkins; “Equalizing
Education Resources for Advantaged and Disadvantaged Children” by Richard Rothstein of the
Economic Policy Institute; “High Standards: A Strategy for Equalizing Opportunities to Learn?” by
Adam Gamoran of the University of Wisconsin; “Inequality in Teaching and Schooling: Supporting
High-Quality Teaching and Leadership in Low Income Schools” by Linda Darling-Hammond and
Laura Post of Stanford; “Charter Schools and Racial and Social Class Segregation: Yet Another
Sorting Machine?” by Amy Stuart Wells, Jennifer Jellison Holme, Alejandra Lopez, and Camille
Wilson Cooper of UCLA; “Student Discipline and Academic Achievement” by Paul Barton of the
Educational Testing Service; and “Critical Support: The Public View of Public Education,” by Ruy
Teixeira of the Century Foundation
II. BOOK CHAPTERS
“The Bipartisan, and Unfounded, Assault on Teachers’ Unions,” in Michael B. Katz and Mike Rose
(eds.), Public Education Under Siege (Philadelphia: University of Pennsylvania Press, 2013.)
“Socioeconomic Integration and Segregation,” in James A. Banks (ed.), Encyclopedia of Diversity in
Education (Thousand Oaks, CA: Sage Publications, 2012).
“Socioeconomic School Integration: Preliminary Lessons from More than 80 Districts,” in Erica
Frankenberg and Elizabeth DeBray-Pelot (eds.), Integrating Schools in a Challenging Society: New Policy and
Legal Options for a Multiracial Generation, (Chapel Hill, N.C.: University of North Carolina Press, 2011)
“Combating School Segregation in the United States,” in Guido Walraven, Dorothee Peters, Eddie
Denessen and Joep Bakker (eds.), International Perspectives on Countering School Segregation (Dutch
National Knowledge Centre for Mixed Schools, 2010).
“Levelling the School Playing Field: A Critical Aim for New York’s Future,” in Jonathan P. Hicks
and Dan Morris (eds.), From Disaster to Diversity: What’s Next for New York City’s Economy? (New York:
Drum Major Institute, 2009).

60

“Higher Education Access,” in RobertMcKinnon (ed), Actions Speak Loudest (Guilford, CT: Globe
Pequot Press, 2009)
“Socioeconomic School Integration,” in Marybeth Shinn and Hirokazu Yoshikawa (eds), Toward
Positive Youth Development: Transforming Schools and Community Programs (New York: Oxford University
Press, 2008).
“The History of Collective Bargaining Among Teachers,” in Jane Hannaway and Andrew J.
Rotherham (eds) Collective Bargaining in Education: Negotiating Change in Today’s Schools (Cambridge, MA:
Harvard Education Press, 2006).
“Socioeconomic School Integration: A Symposium,” in Chester Hartman (ed), Poverty and Race in
America: The Emerging Agendas (New York: Rowman and Littlefield, Publishers, 2006).
“The Return of ‘Separate but Equal,’” in James Lardner and David Smith (eds), Inequality Maters: The
Growing Divide in America and Its Poisonous Consequences (New York: New Press, 2005).
“Economic School Integration,” in Stephen J. Caldas and Carl L. Bankston III (eds), The End of
Desegregation? (New York: Nova Science Publishers Inc., 2003).
“President Clinton’s Race Initiative: Promise and Disappointment,” and “How to Achieve One
America: Class, Race, and the Future of Politics,” in Stanley A. Renshon (ed), One America? Political
Leadership, National Identity and the Dilemmas of Diversity (Washington DC: Georgetown University
Press, 2001).
III. LAW REVIEW ARTICLES
“‘Architects of Democracy’: Labor Organizing as a Civil Right,” (with Moshe Marvit) 9 Stanford
Journal of Civil Rights & Civil Liberties 213 (June 2013).
“Reflections on Richard Sander’s Class in American Legal Education,” 88 Denver University Law
Review 719 (September 2011).
“Socioeconomic School Integration,” 85 North Carolina Law Review 1545 (June 2007).
“Remarks: Symposium – Brown v. Board of Education at Fifty: Have We Achieved Its Goals?” 78
St. John’s Law Review 295 (Spring 2004).
“Socioeconomic School Integration Through Public School Choice: A Progressive Alternative to
Vouchers,” 45 Howard Law Journal 247 (Winter 2002).
"Class-Based Affirmative Action," 84 California Law Review 1037 (July 1996).

61

.

"Getting Beyond Racial Preferences: The Class-Based Compromise," 45 American University Law
Review 721 (February 1996).
IV. PERIODICAL ARTICLES
Have written articles in the popular press for the American Educator, American Prospect, American
School Board Journal, Atlantic Monthly, Baltimore Sun, Boston Globe, Boston Review, Chicago
Sun Times, Christian Science Monitor, Chronicle of Higher Education, Civil Rights Journal,
Education Next, Education Week, Educational Leadership, Forward, Inside Higher Education,
Jurist, Journal of Blacks in Higher Education, Journal of Commerce, Legal Affairs, Legal Times,
New Labor Forum, Nation, New Republic, New York Daily News, New York Times, Orlando
Sentinel, Philadelphia Inquirer, Political Science Quarterly, Poverty and Race, Principal Magazine,
Slate, Wall Street Journal, Washington Monthly, Washington Post and Wilson Quarterly.
Articles on Affirmative Action:
4/3/95
Author, “Class, Not Race: A Liberal Case for Junking Old-Style Affirmative Action
in Favor of Something that Works,” The New Republic (cover story).
7/17/95

Author, “Affirmative Action by Class,” Washington Post, A19

7/17/95

Author, “Equal Opportunity Critics: Class vs. race, round 2,” New Republic.

2/96
Author, “Getting Beyond Racial Preferences: The Class-Based Compromise,”
American University Law Review.
6/2/96
Author, “Bob Dole’s Colorblind Injustice: On Affirmative Action, He Caves to Big
Business,” Outlook Section, Washington Post.
7/96

Author, “Class-Based Affirmative Action,” California Law Review.

8/23/96

Author, “The Sound of Affirmative Action,” The Forward.

9/13/96

Author, “Dishonest Defenders of Racial Preferences,” Wall Street Journal.

10/7/96

Author, “Goal Line,” (re Jack Kemp and affirmative action), The New Republic.

11/4/96

Author, “Need-based affirmative action,” Christian Science Monitor.

12/96
Author, “Defend It, Don’t Mend It: Clinton’s affirmative action man has little bad to
say about racial preferences,” The Washington Monthly.
12/2/96

Author, “A Sensible Approach to Affirmative Action,” The Washington Post.
62

4/20/97

Author, “Need-based affirmative action in the spotlight,” Orlando Sentinel.

1/19/98
Author, “Affirmative Action? Yes: But let’s base it on need rather than on race,”
Philadelphia Inquirer.
Spring ’98

Author, “Class-Based Affirmative Action: A Natural for Labor,” New Labor Forum.

6/98

Author, “In Search of Fairness: A Better Way,” The Washington Monthly.

11/98

Author, “Style, not Substance,” The Washington Monthly, pp. 45-48.

1/19/99

Author, “Class-based affirmative action,” The Boston Globe.

9/21/99

Author, “The Colleges, the Poor, And the SATs” Washington Post, A19.

7-8/00
Author, “Class Action: The good and the bad alternatives to affirmative action,” The
Washington Monthly, 39-43.
9/15/01
Author, “President Clinton’s Racial Initiative: Promise and Disappointment,”
(Chapter 4); and “How to Achieve One America: Class, Race, and the Future of Politics,” (Chapter
11), in Stanley A. Renshon (ed) One America? Political Leadership, National Identity, and the Dilemmas of
Diversity (Georgetown University Press)
Spring/02
144-145.

Author, Review of John David Skrentny “Color Lines,” Political Science Quarterly, pp.

9/9/03
Author, “The Conservative victory in Grutter and Gratz,” Jurist (symposium with
Derick Bell, Peter Schuck, Susan Low Bloch and others).
1/14/04
Today, p.7D.

Author, “Q&A: Low-income college students are increasingly left behind,” USA

3/19/04
Education.

Author, “Toward Affirmative Action for Economic Diversity,” Chronicle of Higher

5/05
Monthly

Author, “Class Action: Why education needs quotas for poor kids,” Washington

11/10/06
Author, “Time for a New Strategy,” [re the Michigan affirmative action vote] Inside
Higher Education.
3/07
Author, “Invisible Men: Race is no longer the unacknowledged dividing line in
America. Class Is,” The Washington Monthly.

63

2/4/08
Slate.

Author, “Obama’s RFK Moment: How he could win over working class whites,”

5/12/08

Author, “Barack Obama and Affirmative Action,” Inside Higher Education.

5/23/08

Author, “A touch of class” (Obama and affirmative action), Guardian America.

11/6/08

Author, “What’s Next for Affirmative Action?” The Atlantic.

9/30/09
Author, “The Next Step in Affirmative Action: Class-based systems can skirt court
and ballot defeats – and do a better job of addressing socioeconomic diversity” Washington Monthly
Online. [Referenced in Steve Benen, “Political Animal,” Washington Monthly.Com 9/30/09]
12/16/09
Author (along with Julian Bond, Lee Bollinger, Jamie Merisotis and others),
“Reactions: Is It Time for Class-Based Affirmative Action?” The Chronicle of Higher Education.
3/3/10
Author, “Disadvantages,” [review of Thomas Espenshade and Alexandria Walton
Radford, No Longer Separate, Not Yet Equal], New Republic.
4/2/10

Author, “The Affirmative Action Trap,” The American Prospect

5/23/10
Post, p. B3 [

Author, “Five myths about college admissions,” Outlook Section, The Washington

5/30/10

Author, “Toward a New Affirmative action,” Chronicle of Higher Education Review.

6/10/10
Author, “A Response to the Critics of Class-Based Affirmative Action,” Innovations
Blog, Chronicle of Higher Education.
6/18/10

Author, “Rewarding Strivers,” Innovations Blog, Chronicle of Higher Education.

7/7/10
Author, “The French Twist on Affirmative Action,” Innovations Blog, Chronicle of
Higher Education.
7/20/10
Author, “Ross Douthat, White Anxiety and Diversity,” Innovations Blog, Chronicle of
Higher Education.
7/28/10
Author, “Next Week’s Court Hearing on Affirmative Action,” Innovations Blog, The
Chronicle of Higher Education.
9/17/10
Author, “Colorado’s Affirmative Action Experiment,” Innovations Blog, Chronicle of
Higher Education.
9/22/10
Author, “10 Myths about Legacy Preferences in College Admissions,” Chronicle of
Higher Education.

64

9/24/10
Author, “A Response to Supporters of Legacy Preferences,” Innovations Blog,
Chronicle of Higher Education.
11/3/10
Education.

Author, “Arizona’s Affirmative Action Ban,” Innovations Blog, Chronicle of Higher

11/22/10
Author, “New Ways to Achieve Diversity in California,” Innovations Blog, Chronicle
of Higher Education.
11/24/10
Author, “South Africa’s Affirmative Action Debate,” Innovations Blog, Chronicle of
Higher Education.
11/29/10
Author, “Does it Matter Where You Go to College? Numbers Favor Top Schools,”
Room for Debate, The New York Times.
12/10/10
Author, “Oxford’s Research-Based Affirmative Action,” Innovations Blog, Chronicle
of Higher Education.
1/6/11
Author, “Do Legacy Preferences Count More than Race?” Innovations Blog,
Chronicle of Higher Education.
1/27/11
Author, “The Next Big Affirmative-Action Case,” Innovations Blog, Chronicle of
Higher Education.
2/11/11
Author, “Nick Clegg’s Attack on Social Segregation in Higher Education,”
Innovations Blog, Chronicle of Higher Education.
3/3/11
Author, “Are Legacy Preferences ‘Defensible Corruption?’” Innovations Blog,
Chronicle of Higher Education.
3/10/11
Author, “Who Benefits Most from Attending Top Colleges?” Innovations Blog,
Chronicle of Higher Education.
4/5/11
Author, “The ‘Reverse Discrimination Sentiment,’” Innovations Blog, Chronicle of
Higher Education. [cited in “Attitudes Toward Access to Higher Education Affected by Race, Study
Shows,” Huffington Post, 4/6/11]
4/29/11
Author, “The Decline of Legacy Admissions at Yale,” Innovations Blog, Chronicle of
Higher Education.
5/11/11
Author, “Purchasing Seats at Top British Universities,” Innovations Blog, The
Chronicle of Higher Education.

65

5/26/11
Author, “Restoring LBJ’s Original Vision of Affirmative Action,” Innovations Blog,
The Chronicle of Higher Education.
6/21/11
Author, “Is Affirmative Action Headed Back to the Supreme Court?” Innovations
Blog, Chronicle of Higher Education.
7/5/11
Author, “Steps Forward and Back on Affirmative Action, Innovations Blog, Chronicle
of Higher Education.
8/4/11
Author, “Achieving Racial Diversity Without Using Race,” Innovations Blog,
Chronicle of Higher Education.
8/17/11
Author, “Race, Class and the New ACT Results,” Innovations Blog, Chronicle of
Higher Education.
9/13/11
Education.

Author, “An Affirmative Action Success,” Innovations Blog, Chronicle of Higher

9/27/11
Author, “Reflections on Richard Sander’s Class in American Legal Education,”
Denver University Law Review.
9/28/11
Author, “Economic Segregation in American Law Schools,” Innovations Blog,
Chronicle of Higher Education.
10/3/11
Author, “The First Monday in October,” [re Fisher v. Texas], Innovations Blog,
Chronicle of Higher Education.
10/17/11
Education.

Author, “A Third Path on Affirmative Action?” Innovations Blog, Chronicle of Higher

11/2/11
Author, “The Amicus Briefs on Affirmative Action,” Innovations Blog. Chronicle of
Higher Education [re Sander and Taylor brief]
11/13/11
Author, “Affirmative Action for the Rich,” (with Stephen Joel Trachtenberg, John
Brittain, Peter Sacks, Michele Hernandez, Terry Shepard and Debra Thomas), “Why Do Top
Schools Still Take Legacy Applicants?” Room For Debate Blog, New York Times.
11/17/11
Author, “Legacy Preferences at Private Universities,” Innovations Blog, Chronicle of
Higher Education.
11/21/11
Author, “What Should Obama Do on Affirmative Action?” Innovations Blog,
Chronicle of Higher Education.
11/29/11
Author, “The Days of Legacy Admissions May be Numbered,” Minding the Campus
Blog.

66

12/5/11
Education

Author, “Obama’s New Guidance on Diversity,” Innovations Blog, Chronicle of Higher

1/8/12
Author, “The Broader Significance of Fisher v. Texas,” Innovations Blog, Chronicle of
Higher Education.
2/9/12

Author, “Waiting on Fisher v. Texas,” Innovations Blog, Chronicle of Higher Education.

2/21/12
Slate.

Author, “Fisher v. Texas: How Obama Should Talk About Affirmative Action,”

2/22/12
Author, “Will the Supreme Court Kill Diversity?” Innovations Blog, Chronicle of
Higher Education.
3/29/12
Author, “Three Myths about Affirmative Action,” Innovations Blog, Chronicle of
Higher Education.
4/20/12
Education.

Author, “Does the Texas Top-10%-Plan Work?” Innovations Blog, Chronicle of Higher

5/10/12
Author, “A Bad Week for Elizabeth Warren - and Affirmative Action,” Chronicle of
Higher Education.
5/29/12
Author, “Overturning or Modifying ‘Grutter v. Bollinger’?” Innovations Blog,
Chronicle of Higher Education.
6/1/12
Author, “Asian Americans and Affirmative Action,” Innovations Blog, Chronicle of
Higher Education. [cited in Asian American Educational Foundation, 6/4/12
6/25/12
Author, “Should Colleges Consider Legacies in the Admissions Process? No: It
Hurts the Deserving,” (debate with Stephen Joel Trachtenberg), Wall Street Journal.
7/11/12
Author, “Transparency About Legacy Preferences,” (re MIT), Innovations Blog,
Chronicle of Higher Education.
8/8/12
Author, “The University of Texas’s Weak Affirmative-Action Defense,” Innovations
Blog, Chronicle of Higher Education.

67

8/10/12
Author, “President Obama’s Affirmative Action Problem and What He
Should Do About It,” The New Republic.
8/16/12
Author, “Obama’s Affirmative-Action Brief,” Innovations Blog, Chronicle of
Higher Education
9/4/12

Author, “Fisher Symposium: Race-neutral alternatives work,” SCOTUSblog.

9/11/12
Author, “In defense of race-neutral alternative jurisprudence,” Fisher
Symposium, SCOTUSblog,
9/17/12
Author, “3 views on whether US still needs affirmative action: A middle way
- Use affirmative action to help economically disadvantaged students of all races,” Christian
Science Monitor.
10/3/12
Author (with Halley Potter), “A Better Affirmative Action: State Universities
that Created Alternatives to Racial Preferences,” The Century Foundation.
10/3/12
Author, “A New Kind of Affirmative Action Can Ensure Diversity,”
Chronicle of Higher Education.
10/10/12

Author, “A Liberal Critique of Racial Preferences,” Wall Street Journal, A17.

10/10/12
New Republic.

Author, “The Race to the Flop – The Problem with Affirmative Action,” The

10/11/12
Author, “The Achilles Heel of Affirmative Action,” Conversation Blog,
Chronicle of Higher Education
10/22/12
Author, “Diversity or Discretion? Essay questions motives of U. Of Texas
in affirmative action case,” Inside Higher Education.
11/7/12
Author, “Another Nail in Affirmative Action’s Coffin,” The Conversation
Blog, Chronicle of Higher Education.
11/9/12
Author, “Economic Affirmative Action,” The Washington Post, A27. [
12/13/12
Author, “Supreme Court Double Header: The Arguments for Gay Marriage
Undermine Affirmative Action,” Slate.
12/19/12
Author (with John Brittain), “When Wealth Trumps Merit,” in Room for
Debate (along with Ron Unz, S.B. Woo and others), “Fears of an Asian Quota in the Ivy
League,” New York Times.
1/17/13
Author, “Where Sotomayor and Thomas Agree on Affirmative Action,”
Conversation Blog, Chronicle of Higher Education.
3/12/13
Author, “Presidents in denial on use of race-based admissions preferences,”
Inside Higher Ed.
68

3/19/13
Author, “The Untapped Pool of Low-Income Strivers,” The Conversation
Blog, Chronicle of Higher Education.
5/13/13
Author, “Addressing the Economic Divide,” in “Diversity Without
Affirmative Action?” Room for Debate (with Patricia Williams, Richard Vedder, Marta
Tienda, and John Brittain), New York Times.
6/2/13
Author, “End race-based affirmative action? Yes: Class matters much more,”
New York Daily News.
6/24/13
Author, “The Next Affirmative Action? Universities Should Respond to the
Supreme Court Ruling by Giving a Bigger Admissions Boost to Low-Income Students,”
Slate.
6/24/13

Author, “A new affirmative action based on class,” USA Today.

6/25/13
Prospect.

Author, “The Class-Based Future of Affirmative Action,” The American

6/26/13
Author, “Why Everyone Is Wrong about Fisher vs. University of Texas,”
Washington Monthly9/2/13
Author, “A Refreshingly Honest Book About
Affirmative Action,” The New Republic.
9/27/13
Author, “The Misleading Administrative Guidance on Affirmative Action,”
Chronicle of Higher Education. [
10/14/13
Author, “A Fresh Chance to Rein in Racial Preferences: The Supreme
Court’s Fisher decision last spring has been largely ignored. Now the justices can strengthen
it.” Wall Street Journal, A15.
11/21/13
Author, “In defense of proxies,” (symposium with Sigal Alon, John Skrentny
and others), Contexts: American Sociological Association, Fall 2013.
3/11/14
Author, “No Longer Black and White: Why Liberals Should Let California’s
Affirmative Action Ban Stand,” Slate.
4/10/14
Author, “Good News for Low-Income Students: A campaign to challenge
racial-preference policies at three universities should move higher education toward
affirmative action based on class,” Conversation Blog, Chronicle of Higher Education.
4/22/14
Author, “Did the Supreme Court Just Kill Affirmative Action? No. But it’s
clearly on its deathbed. That might not be such a bad thing,” Politico.
4/27/14
Author, “Affirmative Action Fail: The Achievement Gap By Income is
Twice the Gap by Race,” The New Republic.
69

4/27/14
Coauthor (with Halley Potter), “Focus on Class Instead,” in Room for
Debate, “Should Affirmative Action Be Based on Income?” New York Times.
6/17/14
Author, “What Sotomayor Gets Wrong About Affirmative Action,” Chronicle
of Higher Education.
7/17/14
Author, “Affirmative-Action Ruling Could Be Pyrrhic Victory for UTAustin,” Chronicle of Higher Education.
9/12/14
Author (with Peter Dreier), “Making Top Colleges Less Aristocratic and
More Meritocratic,” The Upshot Section, The New York Times.
11/20/14
Author, “Achieving College Diversity Without Discriminating by Race,” Wall
Street Journal, p. A17.
12/2/14
Author, “Why Labor Should Support Class-Based Affirmative Action,” New
Labor Forum; and “Richard D. Kahlenberg Responds” (to Julie Park), New Labor Forum.
2/13/15
Author, “Affirmative Action for the Advantaged at UT-Austin,” The
Conversation Blog, Chronicle of Higher Education.
5/18/15
Author, “For the Sake of Working-Class Students, Give ‘Fisher’ Another
Chance,” Chronicle of Higher Education.
6/4/15
Author, “Race-Based Admissions: The Right Goal, but the Wrong Policy”
(re LBJ 50th anniversary of affirmative action), The Atlantic.
7/23/15
Author, “How a New Report May Hasten the End of Racial Preferences in
Admissions,” Chronicle of Higher Education.
12/8/15
Author, “Texas’ college admissions policies give the well-to-do a leg up,” Los
Angeles Times.
12/8/15
Author, “The Future of Affirmative Action: How a conservative decision at
the Supreme Court could lead to a liberal outcome,” The Atlantic.
12/11/15
Author, “Right-wing judge for working-class kids: In praise of Samuel
Alito’s stand on affirmative action in higher education,” New York Daily News.
12/14/15
Author, “Scalia’s Rant and Alito’s Reasoning: What will influence Anthony
Kennedy and determine the fate of affirmative action in Fisher?” Slate.
12/24/15
Author (with Anthony Carnevale and Jeff Strohl) “Should Race Be a Factor
in College Admissions?” Letter to Editor (re Sigal Alon oped), New York Times, A18.
1/11/16
Coauthor (with Jennifer Giancola), “True Merit: Ensuring Our Brightest
Students Have Access to Our Best Colleges and Universities, Jack Kent Cooke Foundation.
70

3/14/16
Author, “Racial Diversity Without Racial Preferences: The growing case for
class-based affirmative action in college admissions,”, Washington Monthly.
6/23/16

Author, “A win for wealthy students,” Fisher II Symposium, Scotusblog.

7/1/16
Author, “How the Legal Victory on Affirmative Action Undermines the
Progressive Coalition: The University of Texas’ policies make it harder to build an enduring
cross-racial class-based coalition in American politics,” The Washington Monthly.
1/4/17
Author, “How to Protect Diversity During Trump’s Presidency: Liberals
should expand the concept to include socioeconomic status,” The New Republic.
4/14/17
Author, “Harvard’s Class Gap: Can the academy understand Donald
Trump’s ‘forgotten’ Americans?” Harvard Magazine, May-June 2017, 35-39.
8/3/17
Author, “The right fix to affirmative action: Progressives should answer the
President’s apparent plans with their own reforms” New York Daily News.
V. ACADEMIC/PUBLIC POLICY APPEARENCES
Have spoken before audiences in numerous settings: government (U.S. Commission on
Civil Rights; U.S. Department of Education); academic associations (American Educational
Research Association; Association for Public Policy Analysis and Management); colleges and
universities (American, Amherst, Centre, Columbia, Flagler, George Washington,
Georgetown, Harvard, Howard, Marymount, Middlebury, Missouri Western, National
Defense University, New York University, Oberlin, Pitzer, Rutgers, St. Johns, St. Louis,
Stanford, Stetson, Suffolk, University of Chicago, University of Maine, University of
Maryland, University of North Carolina, University of Pennsylvania, University of
Richmond, University of Southern California, University of Virginia, West Chester, William
and Mary, Yale); and public policy forums (American Association of Community Colleges,
American Enterprise Institute, Brookings Institution, Cato Institute, Center for American
Progress, Chautauqua Institution, College Board, Committee for Economic Development,
Council for Opportunity in Education, Economic Policy Institute, Demos, Education Law
Association, Education Sector, Ethics and Public Policy Center, Fordham Institute,
Hechinger Institute, KnowledgeWorks Foundation, National Academy of Sciences Board on
Testing and Assessment, National Alliance for Public Charter Schools, National Council of
Educational Opportunity, New America Foundation, New York Historical Society, New
York Public Library, Pioneer Institute, Progressive Policy Institute, William T. Grant
Foundation, and Woodrow Wilson Center).
VI. AWARDS
William A. Kaplin Award for Excellence in Higher Education Law and Policy Scholarship,
Stetson Law School National Conference on Law & Higher Education (2013).
VII. EXPERIENCE CONSULTING WITH SCHOOL DISTRICTS
71

Chicago Public Schools (Illinois) (2008-2010). Helped school district create a
socioeconomic school integration plan for magnet and selective enrollment schools.
Charlotte-Mecklenburg Schools (North Carolina) (2016). Helped school district create a
socioeconomic school diversity plan.
New Haven Public Schools (2017). Helping school district implement a socioeconomic
diversity plan for magnet schools.
Pasadena Educational Foundation (California). (2006 and 2016). Prepared reports for
educational foundation associated with Pasadena Unified School District recommending
adoption of socioeconomic diversity policies.

72

B.

Appendix B – Documents Relied Upon or Considered in Forming
Opinion

73

In addition to the academic literature cited above, I considered the following facts and data
specific to this case:
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

SFFA Complaint
Harvard Answer to SFFA Complaint
SFFA Notice of Subpoena to U.S. Department of U.S. Department of Education
Harvard’s Responses to SFFA’s Second Interrogatories
Deposition Transcript of Roger Banks and Exhibits Thereto
Deposition Transcript of Grace Cheng and Exhibits Thereto
Deposition Transcript of Sally Donahue and Exhibits Thereto
Deposition Transcript of Erin Driver-Linn and Exhibits Thereto
Deposition Transcript of Drew Faust and Exhibits Thereto
Deposition Transcript of William Fitzsimmons and Exhibits Thereto
Deposition Transcript of Kaitlin Howrigan and Exhibits Thereto
Deposition Transcript of Rakesh Khurana and Exhibits Thereto
Deposition Transcript of Christopher Looby and Exhibits Thereto
Deposition Transcript of Marlyn McGrath and Exhibits Thereto
Second Deposition Transcript of Marlyn McGrath and Exhibits Thereto
Deposition Transcript of Lucerito Ortiz and Exhibits Thereto
Deposition Transcript of Tia Ray and Exhibits Thereto
Deposition Transcript of Michael Smith and Exhibits Thereto
Deposition Transcript of Brock Walsh and Exhibits Thereto
HARV00002730
HARV00007396-7398
HARV00007799-7802
HARV00007803-7864
HARV00022931-22936
HARV00023547-23555
HARV00031687-31772
HARV00032957-33022
HARV00065570 -65612
HARV00065741-65774
HARV00072381
HARV00075704-75715
HARV00077521-77822

74

C.

Appendix C – Simulations

75

Simulation One: Half-Athlete Preference
Turn off Preferences: Legacy, Athlete, Race, Early Decision, Staff Kid, Faculty Kid, Dean Director, Disadvantaged,
Waiver, First Gen, Fin Aid Applicant
Assign Disadvantaged Applicants Half-Athlete Preferences
Estimation is based on Extended Sample and Excludes Overall Rating
Exclude Heterogeneity in Demographic Effects
2014

2015

2016

2017

2018

2019

White
African American
Hispanic
Asian
Other

0.505
0.122
0.099
0.219
0.055

0.474
0.130
0.124
0.217
0.055

0.455
0.112
0.103
0.228
0.101

0.431
0.128
0.115
0.219
0.106

0.443
0.132
0.132
0.214
0.080

0.404
0.136
0.129
0.237
0.093

Non-Disadvantaged
Disadvantaged

0.821
0.179

0.815
0.185

0.840
0.160

0.846
0.154

0.865
0.135

0.826
0.174

Academic Index
Extracurricular Rating
Personal Rating

227.2
2.425
2.253

227.2
2.410
2.270

227.9
2.440
2.270

227.8
2.390
2.280

228.5
2.430
2.260

227.8
2.390
2.300

0.478
0.063
0.094
0.325
0.040

0.445
0.068
0.103
0.335
0.048

0.417
0.061
0.105
0.307
0.110

0.414
0.063
0.091
0.317
0.115

0.412
0.057
0.108
0.341
0.081

0.383
0.066
0.114
0.340
0.097

Non-Disadvantaged
Disadvantaged

0.637
0.363

0.622
0.378

0.625
0.375

0.652
0.348

0.653
0.347

0.540
0.460

Academic Index
Extracurricular Rating
Personal Rating

228.56
2.40
2.28

228.53
2.35
2.28

228.54
2.41
2.29

229.28
2.35
2.29

229.53
2.37
2.27

228.74
2.38
2.31

Status Quo

Assign Half-Athlete Preferences
White
African American
Hispanic
Asian
Other

Simulation Two: Half-Athlete Preference, Expanded Pool
Turn off Preferences: Legacy, Athlete, Race, Early Decision, Staff Kid, Faculty Kid, Dean Director, Disadvantaged, Waiver, First Gen, Fin Aid
Applicant
Assign Disadvantaged Applicants Half-Athlete Preferences
Estimation is based on Extended Sample and Excludes Overall Rating
Exclude Heterogeneity in Demographic Effects
When running counterfactual, double number of disadvantaged applicants
2014

2017

2018

2019

White
African American
Hispanic
Asian
Other

0.505
0.122
0.099
0.219
0.055

0.474
0.130
0.124
0.217
0.055

0.455
0.112
0.103
0.228
0.101

0.431
0.128
0.115
0.219
0.106

0.443
0.132
0.132
0.214
0.080

0.404
0.136
0.129
0.237
0.093

0.821
0.179

0.815
0.185

0.840
0.160

0.846
0.154

0.865
0.135

0.826
0.174

Academic Index
Extracurricular Rating
Personal Rating

Assign Half-Athlete Preferences

2016

Non-Disadvantaged
Disadvantaged

Status Quo

2015

227.2
2.425
2.253

227.2
2.410
2.270

227.9
2.440
2.270

227.8
2.390
2.280

228.5
2.430
2.260

227.8
2.390
2.300

White
African American
Hispanic
Asian
Other

0.428
0.082
0.123
0.328
0.039

0.390
0.091
0.131
0.338
0.050

0.373
0.079
0.135
0.308
0.104

0.369
0.083
0.117
0.320
0.111

0.370
0.071
0.136
0.345
0.078

0.346
0.080
0.137
0.345
0.091

Non-Disadvantaged
Disadvantaged

0.415
0.585

0.398
0.602

0.401
0.599

0.434
0.566

0.438
0.562

0.319
0.681

Academic Index
Extracurricular Rating
Personal Rating

227.52
2.42
2.26

227.29
2.36
2.28

227.19
2.44
2.26

228.06
2.38
2.27

228.41
2.40
2.27

227.65
2.40
2.29

Simulation Three: Neighborhood Percentage Rule
Turn off Preferences: Legacy, Athlete, Race, Early Decision, Staff Kid, Faculty Kid, Dean Director, Disadvantaged, Waiver, First Gen,
Fin Aid Applicant
Estimation is based on Extended Sample and Excludes Overall Rating
Exclude Heterogeneity in Demographic Effects
Assign even number of admits from each neighborhood cluster
2015
0.480
0.079
0.109
0.290
0.043

2016
0.445
0.073
0.093
0.277
0.111

2017
0.439
0.077
0.102
0.276
0.107

2018
0.448
0.074
0.105
0.295
0.078

2019
0.437
0.077
0.100
0.291
0.095

0.812
0.188

0.810
0.190

0.801
0.199

0.834
0.166

0.840
0.160

0.801
0.200

228.41
2.41
2.30

228.50
2.34
2.28

228.84
2.37
2.31

229.28
2.33
2.30

229.80
2.36
2.27

229.84
2.34
2.29

White
African American
Hispanic
Asian
Other

0.460
0.095
0.119
0.285
0.041

0.433
0.103
0.131
0.288
0.045

0.409
0.098
0.128
0.269
0.096

0.402
0.095
0.122
0.276
0.104

0.417
0.087
0.126
0.298
0.072

0.379
0.096
0.137
0.295
0.093

Non-Disadvantaged
Disadvantaged

0.550
0.451

0.530
0.470

0.531
0.469

0.557
0.443

0.559
0.441

0.431
0.570

Academic Index
Extracurricular Rating
Personal Rating

Status Quo

2014
0.508
0.071
0.099
0.284
0.038

Academic Index
Extracurricular Rating
Personal Rating

Half Athlete Boost

White
African American
Hispanic
Asian
Other
Non-Disadvantaged
Disadvantaged

NO SES Boost

226.20
2.47
2.32

226.45
2.41
2.33

226.51
2.45
2.33

227.03
2.40
2.33

227.66
2.43
2.31

227.00
2.45
2.34

White
African American
Hispanic
Asian
Other

0.505
0.122
0.099
0.219
0.055

0.474
0.130
0.124
0.217
0.055

0.455
0.112
0.103
0.228
0.101

0.431
0.128
0.115
0.219
0.106

0.443
0.132
0.132
0.214
0.080

0.404
0.136
0.129
0.237
0.093

Non-Disadvantaged
Disadvantaged

0.821
0.179

0.815
0.185

0.840
0.160

0.846
0.154

0.865
0.135

0.826
0.174

Academic Index
Extracurricular Rating
Personal Rating

227.2
2.425
2.253

227.2
2.410
2.270

227.9
2.440
2.270

227.8
2.390
2.280

228.5
2.430
2.260

227.8
2.390
2.300

Simulation Four: Half-Athlete Preference, Neighborhood Percentage Rule, Athlete Preference Turned On
Turn off Preferences: Legacy, Race, Early Decision, Staff Kid, Faculty Kid, Dean Director, Disadvantaged, Waiver, First Gen, Fin Aid Applicant
Keep athlete preferences
Estimation is based on Extended Sample and Excludes Overall Rating
Exclude Heterogeneity in Demographic Effects
Assign even number of admits from each neighborhood cluster

White
African American
Hispanic
Asian
Other Minority

2014
0.522
0.076
0.100
0.263
0.040

2015
0.497
0.083
0.109
0.264
0.048

2016
0.462
0.079
0.094
0.257
0.109

2017
0.454
0.085
0.099
0.256
0.106

2018
0.474
0.079
0.098
0.275
0.074

2019
0.456
0.084
0.100
0.269
0.091

Non-Disadvantaged
Disadvantaged

0.815
0.185

0.814
0.186

0.801
0.199

0.840
0.160

0.843
0.157

0.805
0.196

201
146

205
161

174
139

196
132

195
115

192
130

Academic Index
Extracurricular Rating
Personal Rating

226.92
2.49
2.33

226.97
2.41
2.30

227.41
2.44
2.34

227.82
2.39
2.33

228.29
2.43
2.31

228.35
2.41
2.31

White
African American
Hispanic
Asian
Other Minority

0.475
0.099
0.117
0.267
0.042

0.451
0.102
0.134
0.265
0.047

0.421
0.104
0.127
0.253
0.096

0.424
0.103
0.119
0.252
0.102

0.440
0.093
0.122
0.276
0.070

0.396
0.101
0.135
0.276
0.092

Non-Disadvantaged
Disadvantaged

0.565
0.435

0.552
0.448

0.549
0.451

0.577
0.423

0.570
0.430

0.457
0.543

119
307

118
334

100
318

117
278

117
259

85
328

Academic Index
Extracurricular Rating
Personal Rating

NO SES Boost

224.97
2.54
2.34

225.26
2.47
2.34

225.41
2.50
2.36

225.90
2.46
2.35

226.30
2.50
2.35

225.92
2.50
2.34

White
African American
Hispanic
Asian
Other Minority

0.505
0.122
0.099
0.219
0.055

0.474
0.130
0.124
0.217
0.055

0.455
0.112
0.103
0.228
0.101

0.431
0.128
0.115
0.219
0.106

0.443
0.132
0.132
0.214
0.080

0.404
0.136
0.129
0.237
0.093

Non-Disadvantaged
Disadvantaged

0.821
0.179

0.815
0.185

0.840
0.160

0.846
0.154

0.865
0.135

0.826
0.174

294
144

323
165

274
120

315
124

360
108

333
133

227.2
2.425
2.253

227.2
2.410
2.270

227.9
2.440
2.270

227.8
2.390
2.280

228.5
2.430
2.260

227.8
2.390
2.300

URM Non-Disadvantaged
URM Disadvantaged

Half Athlete Boost

URM Non-Disadvantaged
URM Disadvantaged

Status Quo

URM Non-Disadvantaged
URM Disadvantaged
Academic Index
Extracurricular Rating
Personal Rating

Simulation Five: Neighborhood Percentage Rule, Expanded Pool
Turn off Preferences: Legacy, Athlete, Race, Early Decision, Staff Kid, Faculty Kid, Dean Director, Disadvantaged, Waiver, First Gen,
Fin Aid Applicant
Estimation is based on Extended Sample and Excludes Overall Rating
Exclude Heterogeneity in Demographic Effects
Assign even number of admits from each neighborhood cluster
2015
0.473
0.084
0.112
0.288
0.043

2016
0.441
0.082
0.101
0.267
0.109

2017
0.425
0.081
0.111
0.281
0.102

2018
0.440
0.078
0.109
0.293
0.080

2019
0.427
0.081
0.107
0.293
0.093

0.702
0.298

0.711
0.289

0.696
0.304

0.737
0.263

0.745
0.255

0.696
0.304

228.66
2.39
2.25

228.49
2.30
2.24

228.77
2.34
2.25

229.33
2.31
2.26

229.83
2.34
2.23

229.89
2.33
2.25

White
African American
Hispanic
Asian
Other

0.427
0.105
0.140
0.284
0.044

0.404
0.108
0.147
0.294
0.048

0.390
0.112
0.141
0.267
0.090

0.382
0.112
0.135
0.271
0.100

0.392
0.101
0.143
0.292
0.072

0.361
0.106
0.146
0.300
0.088

Non-Disadvantaged
Disadvantaged

0.349
0.652

0.344
0.656

0.347
0.653

0.364
0.636

0.365
0.635

0.258
0.742

Academic Index
Extracurricular Rating
Personal Rating

Status Quo

2014
0.492
0.076
0.105
0.286
0.040

Academic Index
Extracurricular Rating
Personal Rating

Half Athlete Boost

White
African American
Hispanic
Asian
Other
Non-Disadvantaged
Disadvantaged

NO SES Boost

225.94
2.47
2.28

225.91
2.40
2.29

225.91
2.44
2.28

226.68
2.41
2.29

227.33
2.44
2.29

226.81
2.44
2.30

White
African American
Hispanic
Asian
Other

0.505
0.122
0.099
0.219
0.055

0.474
0.130
0.124
0.217
0.055

0.455
0.112
0.103
0.228
0.101

0.431
0.128
0.115
0.219
0.106

0.443
0.132
0.132
0.214
0.080

0.404
0.136
0.129
0.237
0.093

Non-Disadvantaged
Disadvantaged

0.821
0.179

0.815
0.185

0.840
0.160

0.846
0.154

0.865
0.135

0.826
0.174

Academic Index
Extracurricular Rating
Personal Rating

227.2
2.425
2.253

227.2
2.410
2.270

227.9
2.440
2.270

227.8
2.390
2.280

228.5
2.430
2.260

227.8
2.390
2.300