Posselt, Julie R - Inside Graduate Admissions - Merit, Diversity, and Faculty Gatekeeping-Harvard University Press (2016)

Inside Graduate Admissions
t
Merit, Diversity, and Faculty Gatekeeping
j u l i e r. p o s s e l t
Cambridge, Massachusetts
London, England
2016
Copyright © 2016 by the President and Fellows of Harvard College
All rights reserved
Printed in the United States of America
First printing
Library of Congress Cataloging-in-Publication Data

Posselt, Julie R.
Inside graduate admissions : merit, diversity, and faculty gatekeeping /
Julie R. Posselt.
pages cm
Includes bibliographical references and index.
ISBN 978-0-674-08869-6 (hardcover : alk. paper)
1. Universities and colleges—United States—Graduate work—Admission.
2. Discrimination in higher education—United States. 3. Minorities—Education
(Higher)—United States. 4. Universities and colleges—United States—Faculty.
5. Teacher participation in administration—United States. I. Title.
LB2371.4.P67 2015
378.1'55—dc23 2015015782
To Derek and Daniel
Contents
Preface ix
Introduction: Gatekeeping Reconsidered 1
one Decision Making as Deliberative Bureaucracy 20
two Meanings of Merit and Diversity 46
three Disciplinary Logics 74
four Mirror, Mirror 95
five The Search for Intelligent Life 116
six International Students and Ambiguities of

Holistic Review 133
Conclusion: Merit beyond the Mirror 154
Methodological Appendix 179

Notes 193
References 215
Acknowledgments 241
Index 243
Preface
It has been about twelve years since I first imagined conducting research
on graduate admissions, and almost exactly five years since I began de-
signing and conducting the study described in this book. In this time span,
I have transitioned from being a college administrator and lecturer, to a
doctoral student, to an assistant professor on the tenure track. My interest
in understanding doctoral admissions from the faculty point of view origi-
nated in a problem of practice—a simple question from an advisee that, at
the time, I was ill equipped to answer. I was new to my job with the Mc-
Nair Scholars Program, and she had just decided to apply to graduate school.
She wanted to learn what reviewers would be looking for in her application,
and I had little more for her than a boilerplate response about strong GRE
scores, letters of recommendation, research experience, and fit. I promised
to get back to her after reviewing the literature.
It surprised me to discover that the research literature offered little more
detail than I had offered my student. Scholars have been examining under-
graduate admissions from the institutional point of view since the early
1970s, but until this project there had been only one study—a chapter in
Robert Klitgaard’s 1985 text Choosing Elites—that offered an empirically
grounded explanation of what was important to admissions decision makers
at the graduate level and why. The seed for a research study on faculty
evaluation in graduate admissions was planted for me that day in 2003,
and it has been a privilege and pleasure to see it come to fruition through
x Preface
the support, high standards, and academic freedom extended to me by my

mentors.
Today, when I advise students about graduate admissions, I have a
better-informed answer about what admissions committees in leading doc-
toral programs are looking for in applicants. At one of the conferences
where I shared some findings of this project with faculty members, the
discussant started her remarks by admitting how vulnerable she felt while
reading the paper—how she squirmed to read the provocative quotes from
scholars in disciplines far afield from her own, only to realize the squirming
was because she herself had held similar thoughts or made similar com-
ments. She was uncomfortable, she said, but she couldn’t put down the paper
(which has since developed into Chapter 5 of this book) because it helped
her see assumptions that she didn’t even realize she had been holding. I
received these comments positively, for one of my principal aims with this
work has been the classic sociological task of “seeing the strange in the
familiar.” It can be hard to see our own assumptions, and harder still to let
them go in favor of new ways of thinking. Aside from the intellectual con-
tribution this book makes to the literature, I hope it will help faculty mem-
bers see familiar admissions work with fresh eyes.
My own moments of vulnerability have come recently, as the tables have
turned and I have participated in admissions, not as an outside observer,
but as one of the faculty decision makers. As a member of my depart-
ment’s master’s admissions committee and a participant in the late stages
of reviewing applications for the PhD program, I came to empathize in
new ways with my research participants. I saw for myself the complexity
of comparing applicants’ relative admissibility when they have diverse
strengths. I felt the pangs of saying no to highly qualified applicants due to
budget constraints, the fatigue brought on by constantly checking myself
for cognitive and other biases, and the tension of choosing carefully the
moments to speak up in committee meetings. This empathy hasn’t changed
my findings or the way I describe them. It has, however, redoubled my own
commitment to approaching admissions work as an area worthy of pro-
fessional development—one that can’t be separated from improving how
we recruit and mentor the students whom we do admit. I hope readers of this
book will join me in committing to ongoing learning about how we can se-
lect and serve graduate students more effectively and equitably.
t
Introduction
Gatekeeping Reconsidered
F or three hours the philosophy admissions committee had been

working in a cramped storage room that doubled as their meeting space.
They had been discussing applicants on their short list, one by one, but had
reached a point of deadlock over who should ultimately receive offers.
Breaking a long silence that betrayed the group’s exhaustion, their admin-
istrative assistant, Leon,1 spoke up for the first time. He noted that agreeing
to admit everyone who received an average rating of 1.8 or higher would
give them their desired cohort size: the lucky 13. Their work would be done.
No one jumped at the idea, but the committee chair, Liana, and a senior
professor, Olivia, expressed their support. Another senior professor, Gerald,
wasn’t so sure.
“People seem to be very confident about the line where admissible leaves
off and inadmissible picks up,” he said. “I have a hard time drawing lines
because wherever we draw it, it’s going to look arbitrary.”
Olivia emphatically responded, “Well, it is an artificial line!”
Continuing to push, Leon noted that drawing the line at a rating of
2.4 would finish the job even more quickly by giving them the optimal
number of admitted and wait-listed students. A long pause and a few sighs
and shoulder shrugs later, they agreed to use this threshold and started
packing up.2
There is a story behind every statistic—including the lucky 13, the 2.4
rating, and the 18 percent of applicants admitted to research doctoral
2 I N S I D E G R A D U AT E A D M I S S I O N S
programs nationally.3 This book tells the story of how faculty in ten top-
ranked doctoral programs draw the almost imperceptibly fine line between
those whom they admit and those whom they reject. Two years of ob-
serving and interviewing graduate admissions committees in core academic
disciplines—astrophysics, biology, classics, economics, linguistics, philos-
ophy, physics, political science, sociology—gave me a unique window into
the evaluation and selection processes that go into graduate admissions.
My research revealed faculty members’ nebulous, shifting ideals about
student quality; how departmental, disciplinary, and personal priorities are
woven into judgments of admissibility; and the implications of it all for eq-
uity and the health of the academy.
Changes in society, the applicant pool, and the labor market have fun-
damentally altered the markets for graduate education and for people with
graduate degrees, yet the criteria associated with admission to degree
programs have changed little. Of the three strongest determinants of access
to graduate education—college grades, Graduate Record Examination (GRE)
scores, and the reputation of a student’s undergraduate institution—the
latter two are part of a conventional notion of student quality that fails on
at least two counts.4 GRE scores and college prestige fail to reliably pre-
dict whether a student will complete the PhD, and disproportionately ex-
clude some of the very groups whom our mission statements and websites
claim we wish to attract. What is more, the structure of the academy in the
twenty-first century will not sustain many of the positions that admissions
decision makers themselves hold. If faculty do not adapt their mindsets to
meet changes in the academy, labor market, and society, they will select and
train students for jobs that do not exist. For graduate education to fulfill its
promise of developing leaders for today’s knowledge economy and diverse
democracy, many faculty will need to rethink how they evaluate prospec-
tive students and draw the line in admissions. Let’s consider these dy-
namics in greater detail.
Doctoral Students and Their Professional Pathways
Worldwide, pursuit of the PhD continues to grow. The PhD is not only the
central prerequisite for faculty careers; credential inflation has also rendered
graduate degrees necessary for access and promotion in many professions
that once required only a bachelor’s degree.5 Economic and technological
development outside the United States has also sparked interest in grad-
uate education from international students, whose share of doctorates
awarded in the United States has more than doubled in the last forty years.6
Introduction 3
Overall, the proportion of adult women (thirty to sixty years old) in the
United States with graduate degrees grew almost tenfold from 1965 to
2005, from 1.1 percent to 9.68 percent.7 And from 2000 to 2010 alone,
the number of master’s degrees earned by African American and Latino/a
students more than doubled.8
Yet aggregate statistics like these conceal considerable complexity about
the state of equity in graduate education. Gender and racial inequities are
persistent and pervasive in doctoral education, for example, despite the
progress in closing gaps in master’s degrees awarded. Women and U.S. resi-
dents of color remain less likely than men and whites to attend research
universities, and they continue to receive fewer doctorates than we would
expect given their shares of both the overall population and the popula-
tion of baccalaureates awarded.9 African Americans and Latinos comprised
13 percent and 16 percent, respectively, of the U.S. population in the 2010
Census, but received just 6 percent and 7 percent of the doctorates awarded
that year—numbers that reflect little change from the previous decade.
Meanwhile, Native American doctoral attainment has fallen to its lowest
point in twenty years.10
Gender and racial/ethnic representation also varies by academic field of
study.11 The well-known disparities in science, technology, engineering, and
math (STEM) disciplines are evident in many humanities and social science
fields as well. In the humanities, for example, only 3 percent of PhDs in 2009
were awarded to African American students and only 5 percent were
awarded to Latino students.12 Table 1 displays data from the Survey of
Earned Doctorates (SED) about the number of PhDs awarded in 2012, by
gender and race, for a selection of fields. Philosophy, which is not represented
in SED data, awarded only 29 percent of their PhDs to women in 2009. It
is notable that fields of study with continued racial and gender inequities also
have some of the lowest admission rates, nationally.13
Like the population of graduate students, the range of careers pursued by
persons who have earned a PhD has diversified.14 As a result, most graduate
programs are preparing, ad hoc, a much broader group of professionals
than the next generation of faculty researchers. For example, applied intel-
lectual and technical expertise is instrumental in today’s economy, and the
diffusion of public research into industry has created a whole sector of
PhD-level researchers outside the academy. Less than half of engineering
doctoral students now expect to enter academia.15 Specialized intellectual
inquiry is the heart of doctoral education, but viable alternatives to the aca-
demic track are necessary because there are far fewer tenure-track faculty
positions than PhDs looking for jobs. Just one academic track faculty
position is posted for every twelve PhDs produced in science, technology,
Table 1 Doctoral Degree Attainment by Gender and Race in Selected Disciplines, 2012
African Native
Academic Amer./ Amer. Indian/ Hispanic/ Asian Hawaiian/
Discipline Gender Black Alaska Native Latino White Amer. Pacific Island. Other Total
Astronomy Male *** *** *** 137 *** *** 10 190

Female *** *** *** 52 *** *** 8 85
Subtotal *** *** *** 189 53 *** 18 275
Physics Male *** *** *** 805 464 *** 132 1,504
Female *** *** *** 163 134 *** 45 363
No answer *** *** *** *** *** *** 4 4
Subtotal 35 *** 87 968 598 *** 181 1,871
Biological Sciences Male 119 *** 231 2,171 1,083 *** 324 3,936
Female 195 *** 269 2,385 1,268 *** 353 4,494
No answer *** *** *** *** 2 *** 8 10
Subtotal 314 15 500 4,556 2,353 17 685 8,440
Economics Male *** *** 86 416 276 *** 90 897
Female *** *** 26 179 184 *** 29 433
No answer *** *** *** *** *** *** 1 1
Subtotal *** *** 112 595 460 *** 120 1,331
Political Science/ Male *** *** 45 443 90 *** 65 696
Public Admin. Female *** *** 39 330 81 *** 57 549
No answer *** *** *** *** *** *** 1 1
Subtotal 90 *** 84 773 171 *** 123 1,246
Sociology Male *** *** 21 147 26 *** 26 235
Female 40 *** 45 244 55 *** 28 412
No answer *** *** *** *** *** *** 1 1
Subtotal *** *** 66 391 81 *** 55 648
Linguistics Male *** *** 9 64 21 *** 18 117
Female *** *** 10 77 39 *** 13 143
Subtotal *** *** 19 141 60 *** 31 260
Note: *** = suppressed to avoid disclosure of confidential information.

Source: NSF/NIH/USED/USDA/NEH/NASA 2012 Survey of Earned Doctorates.
Introduction 5
engineering, and mathematics. And in the humanities, where the glut of

PhDs relative to academic jobs regularly makes the news, many graduates end
up working in positions that do not require the PhD at all, much less in the
narrow specializations for which they received training.16
The professoriate itself is changing in ways that compel a fresh look at
recruitment, selection, and broader ideals of what makes an excellent scholar.
Adjunct, clinical, and other nontenured appointments, not tenure-track po-
sitions, now constitute the majority of faculty positions listed. In response
to these trends, a few PhD programs are shifting or broadening their focus
and degree requirements. Some institutions are scaling back the size of
their doctoral cohorts, while others are reevaluating their qualifying exams
and the structure of the dissertation.17 Largely absent, however, has been a
conversation about what the changes inside and outside the gates of grad-
uate school mean for who gets in.
Gatekeeping Reconsidered
Reform in doctoral education today must better align notions of student

quality with the diverse students and varied career pathways that doctoral
students pursue. Tenured faculty have both the influence and the responsi-
bility to respond to changes in student trajectories. Some, though, are un-
comfortable with students’ increasingly diverse identities and career paths
or feel stymied by political dimensions of the change process. Most profes-
sors in research universities, after all, are products of a system that gauges
program excellence by placing graduate students in faculty positions at re-
search universities, and that privileges theory over applied research. Rising
demand from a diversified population is leading some within academia to
circle the wagons around the PhD, striving to preserve its purity as a badge
of honor that signifies individuals with special aptitude to advance theory.
The urgency to reconsider gatekeeping is greater than ever, but these issues
have been building for more than a century. In his famous essay “The Ph.D.
Octopus,” Harvard philosopher and psychologist William James called his
colleagues to task. Worried that rising demand for the PhD as a college
teaching credential was degrading its character as a stimulus for scholarship,
James denounced “the increasing hold of the Ph.D. Octopus upon American
life” and graduate education’s development into “a tyrannical Machine with
unforeseen powers of exclusion and corruption.” He wrote:
America is thus as a nation rapidly drifting towards a state of things in which

no man of science or letters will be accounted respectable unless some kind of
badge or diploma is stamped upon him, and in which bare personality will be
a mark of outcast estate. It seems to me high time to rouse ourselves to con-
sciousness, and to cast a critical eye upon this decidedly grotesque tendency.
He charged that faculty and universities had been complicit in these trends,
allowing the patina of prestige and the vanity of titles to distract them from
the university’s educational mission. Elite doctoral programs had responded
to rising student demand by raising standards rather than expanding
enrollments, he concluded, which preserved their status but heightened
competition and created a mismatch between the degree requirements they
publicized and those they put into practice. James wrote:
We advertise our “schools” and send out our degree-requirements, knowing

well that aspirants of all sorts will be attracted, and at the same time we set
a standard which intends to pass no man who has not native intellectual
distinction . . . We dangle our three magic letters before the eyes of these pre-
destined victims, and they swarm to us like moths to an electric light.18
The competitive trends James identified—among doctoral programs for

status and among prospective students for admission to top programs—
continue in the twenty-first century.19
Among the barriers to expanding access are the strong incentives doc-
toral programs have to limit their size. Selectivity goes hand-in-hand
with prestige.20 Today, “proportion of applicants admitted” is one factor
used to calculate program rankings, such as those published by the Na-
tional Research Council and U.S. News and World Report. And as has already
been mentioned, fiscal concerns and uncertain employment prospects after
graduation are driving some departments to cut the number of students
they admit.21
Together, increasing demand and a stagnant supply of spaces have raised
the competitiveness of admissions to many doctoral programs. Access has
become what economists Robert Frank and Philip Cook call a “winner-
take-all” market—a type in which “barely perceptible quality margins
spell the difference between success and failure.”22 This type of market frus-
trates both those trying to break into it and those who want to understand
it, because the margins between acceptance, rejection, and the wait-list are
difficult to perceive. Further, what counts as quality is a moving target.
Necessary credentials for admission to graduate programs are just as sub-
ject to inflation as those required for jobs, and faculty judge applicants
not according to a fixed standard but relative to others in the applicant
pool that year.23 Therefore, like undergraduate admissions and other forms
of faculty decision making, doctoral admissions is a process cloaked in
secrecy.
Introduction 7
Defining Merit
People are also anxious about how admission is carried out because it is
thought to provide a barometer for how selective educational institutions
are fulfilling the ideal of allocating opportunities equitably and on the basis
of “merit.” In the United Kingdom, these twin standards constitute what
is known as fair access to higher education. Merit is always a conditional,
not an absolute, assessment. No one inherently merits admission. They do
so because they are judged to possess attributes that decision makers have
deemed legitimate grounds for drawing the line between the many who
would like to enroll and the few who should be given the opportunity. In
the United States, potential for strong academic performance is one such
attribute, but as this book will show, it is hardly the only one.
As in other academic competitions, the conditions under which someone
is judged to merit admission to graduate school are bound up with ideals
of individual or organizational quality. What does and should count as merit
is therefore deeply contested. Is there a single, proper standard we should
be working to define and defend? Is it legitimate if different academic dis-
ciplines use different standards? Why should we elevate individual academic
performance over qualities that may contribute to the common good? Is
merit open to reinterpretation as times and conditions change?24 Personal
opinions and conventional wisdom about these and other questions abound,
but the current research record offers little in the way of clear answers.
Graduate education may play an increasingly important role in shaping
professional opportunities, but we suffer from a relatively one-dimensional
research literature about graduate admissions, especially compared to the
well-developed literature on selective undergraduate admissions.
By examining graduate admissions in practice, from the perspective of
those who make the decisions, I see merit differently than most previous
scholars of graduate admissions. The vast majority of previous research has
tried to statistically model whether applicant characteristics, especially
scores on the GRE and TOEFL, predict various indicators of student suc-
cess.25 Some scholars have been motivated by concern about decision
makers’ use of “explicit cutoffs or tacit minima” when weighing standard-
ized test scores.26 They want to understand just how risky that practice may
be in excluding students with lower mean test scores, who nonetheless
might be academically successful. Others want to determine whether test
scores can be counted upon to deliver satisfactory returns on the investment
that admission represents.27
Whether focused on a single field or a range of disciplines, results of
predictive validity analyses have been mixed. Maria Pennock-Roman
described the assumptions of this body of research and a fundamental

problem with it:
There is one unidimensional [variable] Y, such as college grade-point average,

that measures “success” . . . A predictor X exists, which can be a linear
combination of variables that has demonstrable validity for estimating Y in
advance . . . Since the relationship between predictors and Y is far from per-
fect, some selection decisions will turn out to be correct pre-classifications of
candidates, and others will turn out to be errors.28
The difficulty of reliably predicting long-term outcomes from any infor-

mation in an application is understandable, for “success” and “failure” are
complex concepts with multiple dimensions and debatable definitions.
Further, most existing research has limited generalizability due to their
samples, the restricted range of observed GRE scores in most studies, and
significant changes to the GRE in 2011.29 The most recent study, published
by psychologist Nathan Kuncel and colleagues in 2007, found correlations
between GRE scores and first-year graduate school grades at levels that
testing proponents could hold up as statistically significant, but that skeptics
could dismiss as practically insignificant.30 Neither a student’s application
nor a model developed from information in the application will work very
well as a crystal ball in predicting the probability of a given student’s suc-
cess. Even the Educational Testing Service, which administers the GRE,
recommends that the test be used as just one factor among many in a holistic
review process.31
My observations of admissions committees support previous research
in demonstrating that for better or worse, a few key criteria, including
GRE scores, go a long way in shaping a student’s odds of making the
short list. Very high GRE scores and attending a prestigious college or uni-
versity were clearly among the revealed preferences of faculty in the pro-
grams that I studied. However, my research also suggests that previous
research has misrepresented merit as overly narrow, monolithic, and stable
across disciplines.
As I show in the chapters to follow, what faculty construct as merit in
highly selective graduate programs is complex and dynamic. Faculty use aca-
demic achievements to narrow the pool, so those criteria pattern the
outcomes, but they bring a host of other factors into the conversation to
make their final choices. They do this in part because many students meet the
bar of conventional academic achievement,32 and in part because they see
admissions as an opportunity to enact a variety of values and create the
futures of their departments and disciplines. What counts as a quality ap-
plicant varies by reviewer, committee, department, discipline, and university,
Introduction 9
and involves layers of inferences made from seemingly minor details in the
application. Although statistical methods would be unlikely to pick up on
these details because they are idiosyncratic to individual cases, they are cru-
cial to evaluative cultures of the disciplines and specific academic programs.
Whether in research or in practice, conflating the quality of an applicant
with a narrow set of academic achievements thus misses some of the most
interesting parts of the story about graduate admissions. However, this
definition of merit has other important limitations that also bear men-
tioning. For example, assuming the most accomplished applicants are the
best candidates reduces doctoral education from a developmental process to a
scholarly finishing school, and implies that mentoring relationships and
learning environments matter little to students’ success.33 Focusing atten-
tion solely on student qualities also misses the broader context of who is
defining what is desirable in applicants, how and why they determine this,
and what the consequences are of those choices. Finally, the tendency to focus
on the validity of common admissions criteria ignores important reviewer
effects that also affect the fairness of the admissions process, such as sus-
ceptibility to fatigue and to cognitive and implicit biases.
Reconsidering merit may seem like a radical proposal. By glimpsing the
deliberations of the committees in this book, however, it will become clear
that, already, faculty operate on a more expansive notion of merit than that
of simple academic achievement and academic potential. I found that ad-
missions may start with the official goal of identifying applicants who are
likely to succeed, but organizational interests such as prestige, diversity,
collegiality, efficiency, and fiscal responsibility also drive the process and
endow it with legitimacy in the eyes of important stakeholders.34 Revealed
preferences therefore vary across time and place in response to changes in
applicant pools, the political environment, the mission of the program, and
who is making the decisions. Program and disciplinary priorities, the bal-
ance of student characteristics in an emerging cohort, as well as other pref-
erences that are idiosyncratic to specific committees and reviewers, all frame
judgments of who the “best” or “most qualified” student prospects are.35
For example, under a purely student-centered view of merit and academic
view of quality, one might admit the students who rate highest on criteria
that best predict first-year grades in graduate school. Under this standard,
however, the proportion of students from China in many American doctoral
programs would skyrocket to levels that, to some, would appear unaccept-
able. Cohorts that contain very large or small numbers of any one popula-
tion are often seen as undesirable by faculty and students alike because
they send the message that the program has skewed interests. Diverse
student bodies, on the other hand, are thought to reflect balanced interests
and the richness of our society. A political scientist nicely summarized the
tension between assessing quality in terms of student characteristics versus
cohort characteristics:
I think from practically everybody’s viewpoint getting talented, motivated

people is the top priority . . . But we want to have some balance. I would say
{Pause} my guess is—no, no it was explicit. We had a whole bunch of top Chi-
nese. And {Pause} we decided we don’t want to admit a class consisting of
one-third Chinese and so we didn’t. And it was clearly not because of bias
against Chinese, it was just—we wanted balance.
In practice, merit in graduate admissions is not an absolute assessment of

achievements to date and perceived potential for good grades or a great
dissertation. It is an assessment of admissibility relative to a specific appli-
cant pool, by a set of specific decision makers with specific personal pref-
erences. These preferences include potential and achievements, but an
applicant might also be judged preferable if her admission will appease a
difficult colleague or if it improves the balance of students across depart-
mental concentrations. A solid student from Malaysia or Mongolia, coun-
tries that produce few applicants to U.S. doctoral programs, might be judged
more admissible than a very strong one from China, India, or Korea, which
produce many. A student who grew up in foster care and overcame extraor-
dinary personal challenges might be judged more admissible than a stu-
dent with a similar academic record who grew up in a well-known college
town. There is not a single hierarchy of admissions priorities as can be
implied by tables of coefficients in quantitative studies. Rather, because fac-
ulty use admissions to pursue a variety of interests, multiple hierarchies of
priorities (which sociologists call a heterarchy) simultaneously and inter-
actively shape an applicant’s odds of being admitted.36
What is more, because quality takes many forms in graduate admissions,
no single applicant could possibly personify all that the institution and its
various stakeholders value. Rather, collective cultural priorities are more
likely to be reflected in groups of students than in individual applicants. In
this context, the best that decision makers can hope for is to cobble together
a cohort that, together, represents what is important to them.
For these reasons, I argue that we need to rethink how we talk about
merit in graduate admissions. Discussions about merit can’t be one-sided.
How we understand and recognize merit makes sense only in light of the
larger organizational challenges, goals, and missions that faculty face.
Therefore, we can’t talk about student achievement and potential unless
we also talk about the organizational context that determines how achieve-
ment and potential will be defined. Further, we shouldn’t treat “merit” as
Introduction 11
if it is merely the sum total of an applicant’s “deservingness” based on what

the applicant has done already or how easily he or she will thrive in our
graduate programs as they are currently designed. What it means to war-
rant access to graduate education is more complicated than that, and more
programs would do well to embrace it as such. Professors can use admis-
sions and other student review situations as an opportunity to think criti-
cally about their own professional practice and how their departments and
graduate programs might better realize their educational mission for a
changing labor market and population. In short, those of us with a stake
in graduate education need to broaden the conversation about merit to en-
courage collective responsibility for student learning.37
A conversation about what we value in admissions, and why, thus pro-
vides a natural entry point into questions at the crux of the current debate
over graduate education’s future, or as Leonard Cassuto calls it with re-
gard to the humanities, “the graduate school mess.” This conversation
beckons faculty to align admissions work with program mission and, in so
doing, to consider the professional system and social contexts of which
graduate education is part. For example, Harvard Law professor Lani
Guinier has urged admissions policymakers and decision makers to ask
themselves whether privileging test scores or the first-year grades with which
they are modestly associated will help higher education fulfill its democratic
mission. Her idea of democratic merit advises admissions be conducted with
an eye to selecting students who demonstrate capacity for leadership in a
racially and ethnically diverse democracy.38
To summarize, merit and quality are subjectively assessed and socially
constructed. Although students’ GRE scores and college reputations unde-
niably shape the profile of short list, and therefore of admitted cohorts,
what counts as merit is complex and dynamic, and varies by context. Where
faculty draw the line between admitted and rejected students, it turns
out, is as much a reflection of who is doing the evaluating as who is being
evaluated. Although this insight is new for analysis of graduate admis-
sions, it has propelled research on undergraduate admissions research
since the 1970s, when historian Howard Wechsler declared, “The es-
sence of selective admissions is the subjective judgment of the admissions
officer.”39
Untangling a Paradox of Admissions
I designed this research to untangle an apparent paradox in the research

literature on higher education. On the one hand, diversity40 is a well-
institutionalized value in higher education today, and recruiting applicants
from underrepresented groups is common practice in areas from undergrad-

uate admissions to faculty hiring.41 However, two of the three strongest
predictors of admission to graduate programs, GRE scores and attending
a selective undergraduate institution, privilege populations that already
enjoy an enrollment advantage.42 Male, white, and Asian American students
remain overrepresented in the most selective colleges and universities and,
on average, earn higher scores on the GRE.43
Therein is the paradox. If diversity is valued and concerns about ine-
quality are widely known, why do faculty continue to rely upon criteria
that undermine equity and diversity? One can imagine a range of possible
explanations. Perhaps professors actively resist admissions reform they way
they tend to resist change in general. It could be that they are deeply in-
vested in the entrenched standard and, thus, unwilling to rethink their reli-
ance on specific criteria or vision of ideal applicants. Maybe this paradox
is simply a product of myopia to the implications of current practice. Or
maybe the reasons are more sinister. Are there informal efforts to limit di-
versity beneath public images of inclusiveness? Are they overtly racist or
sexist behind closed doors?
Recent experiments have found evidence that we should not discount
these last possibilities—that faculty judgment in selection situations is
marked by informal discrimination and unconscious (that is, implicit) bi-
ases. A randomized double-blind study by Corinne Moss-Racusin and col-
leagues focused on hiring for a laboratory manager position, found that
faculty rated applications headlined by male names as significantly more
competent and hireable than identical applications headlined by a female
name.44 And in a field experiment with a large sample of 6,500 faculty,
Katherine Milkman, Modupe Akinola, and Dolly Chugh found that
participants ignored email inquiries from prospective students with female,
Indian, Chinese, black, and Latino-sounding names at higher rates than
they ignored those with traditionally Caucasian male names. These findings
held across institutional types and disciplines, but were particularly acute
in higher-paying disciplines and private universities.45 If faculty are dis-
criminating against women and people of color in the informal interac-
tions that precede application and admission, it is important to take a
closer look at how they are interpreting and evaluating the applications
they do receive.
My research into the admissions process did not find evidence of
overt discrimination,46 but it did find that a “colorblind” approach to
admissions—the dominant model in all ten of these programs—also cre-
ates conditions that sustain inequalities. We have reached a point where
policy need not formally exclude or segregate on the basis of race or
Introduction 13
gender, because inequalities can also become locked in (or institutional-

ized) as organizations operate according to shared understandings and in-
formal rules that may look neutral but have a disproportionate, or dispa-
rate, impact on some groups.47 For example, I found that through their use
of shared, discipline-based assumptions to define which applicants were
better, more competent, or deserving, faculty members often accepted in-
equitable admissions outcomes as logical or necessary. What disciplinary
outsiders might have challenged as discriminatory, unjust, or simply wrong
could be deemed perfectly legitimate from an insider’s perspective.
By getting inside the faculty perspective, this book thus uncovers the
common mental pathways scholars use to legitimize a system whose rules
look neutral and by some standards fair, but that nonetheless is marked by
what Charles Tilly called durable inequalities.48 In a society where overtly
racist and sexist behaviors are socially unacceptable and where diversity is
something to celebrate, the institutionalization perspective makes it clear
that durable inequalities are neither inevitable nor natural, but instead are
the result of a process we have created.49 This perspective is also useful in
identifying common perceptions that are out of step with current research,
and in bringing to the surface assumptions that are so deeply held as to be
taken for granted.
I therefore portray the current system of admissions from professors’
own point of view while making clear that the system has cracks—ones
through which students from already underrepresented groups continue
disproportionately to fall. Working with faculty throughout an entire ad-
missions cycle, I gained real sympathy for the magnitude and difficulty of
reviewing files and selecting applicants in these programs. I came to see
that there are unintended consequences for equity from the organizational
apparatus that programs establish to deal with a pile of 800 applications
in total, or 250 from China alone. I observed how ambiguities inherent in
the review process prime faculty to defer to stereotypes, such as when they
judge applicants from China. These findings help reveal why it is so diffi-
cult and complex to make diversity, as one participant put it, “more than a
platitude.”
However, I also found racism and sexism subtly institutionalized in mis-
guided perceptions about what common admissions criteria signal, in
deference to “fit” with the status quo as a core determinant of admissibility,
and in reluctance to take on students from underrepresented backgrounds
whose profiles suggest they may benefit from more intensive mentoring.
Orienting toward traditional ideas of prestige also set up these graduate
programs to reproduce inequalities as they recreate themselves—as William
James noted more than a hundred years ago.
Notes on the Research Design
In this section I provide a general overview of my research design; readers

interested in greater detail may be interested in the methodological ap-
pendix. I conducted 86 interviews with 62 faculty and 6 graduate stu-
dents, the vast majority of whom were sitting on admissions committees
at the time. The heart of my study, though, was the time I spent ob-
serving admissions committee meetings and recruitment events in six of
the ten programs. My perspective in these meetings as an outsider-turned-
insider enabled me to capture routine details of the review process that
committee members may take for granted and to compare the principles
and preferences that faculty espoused with those that they put into
practice.
Due to my agreements with the Institutional Review Boards (IRBs) and
participants, I refrain from naming or describing the universities where I
collected data. What is important for readers to know, and which I can
share, is that two universities were public and one was private, that they
were in two different regions of the country, and that all three are well-
known research institutions. Also of note: Each university had a graduate
school that coordinated the admissions process and offered resources to
help increase the enrollment of women and students of color, such as
diversity-focused fellowships and trainings for faculty engaged in admis-
sions work.
My sample of highly ranked doctoral programs in pure disciplines is not
intended to be representative of American doctoral education or the full
range of fields of study, but it provides insight into the intellectual core
of the academy. It covers the humanities, social sciences, and natural sci-
ences and, within each, intentionally includes fields known for being rela-
tively centralized, hierarchical, and paradigmatic (such as economics,
physics, and philosophy) and others with less intellectual consensus (such
as political science, biology, and linguistics). This variety allows for com-
parisons on multiple dimensions.
Focusing on programs ranked in the top fifteen for their discipline also
has benefits. Most importantly, in highly selective programs like these, the
many demands and sociocultural dynamics of selection come into sharper
focus. Dynamics of elite organizations are also important to understand be-
cause, as sociologists have noted, their practices and priorities often set a
standard that others adopt to improve their standing. A better understanding
of elite organizations—and of efforts within those organizations to resist
prevailing trends—offers a glimpse into the direction that the system, as a
whole, may be headed.50
Introduction 15
I have masked and/or changed information about applicants and faculty

that might be personally identifiable, starting with their names. In referring
to colleges and universities, I tried to balance ensuring anonymity with
conveying a real-world sense of the institutional strata in which these pro-
grams are located. Therefore, when quoting participants who named
specific universities, including their own, I replaced those names with the
results of random draws from fifteen universities in the same tier of program
rankings for the speaker’s discipline. This approach means that the actual
universities in which data were collected could be named due to chance,
but it ensures that readers should be no more able to recognize the data
collection sites than other, similarly ranked institutions.
People commonly ask me how I gained access to the programs, and al-
though I will never be certain about the answer, my typical response is that
it was likely a combination of factors. As a white female student from a re-
spected university, the faculty with whom I interacted may have seen me as
a member of their community and as relatively nonthreatening. I also attri-
bute my unusual access to making clear in early communications what the
study’s confidentiality protections were, the genuine desire many faculty
have to improve admissions, and a dose of divine intervention. There was
also a respondent who agreed to participate out of “karmic obligation to
the many who had participated in [his] own research over the years.”
Most participants were current members of the admissions committees,
and their demographic characteristics are summarized in Table 2. Just
18 percent of the sample were women and only 3 percent were U.S.-born
scholars of color—a decidedly skewed composition that can be thought of
as a limitation and a strength of the book. On the one hand, the sample is
broadly representative of tenured faculty in elite universities, making my
research findings a more trustworthy picture of admissions in highly ranked
graduate programs. On the other hand, my data lacks the voices of women
and scholars of color. The imbalanced sample makes plain the need for re-
search on inequity in graduate education and, more importantly, for re-
cruiting and retaining more diverse cohorts of doctoral students.
Reading This Book
I wrote this book with three audiences in mind: faculty across the disci-
plines who are engaged in graduate admissions work, scholars of higher
education and sociology, and administrators with whom faculty coordinate
to facilitate admissions. Prospective graduate school applicants will also no
doubt be curious to learn how faculty evaluate files and deliberate behind
Table 2 Sample Demographic Characteristics by Program and Subject Area
% Scholars % Domestic # Graduate students

Programs N % Female % International of Color Scholars of Color on Committee
Humanities
Classics 5 40 40 0 0 0
Linguistics 7 29 29 14 0 2
Philosophy (University 1) 6 33 33 0 0 0
Philosophy (University 2) 7 29 14 14 14 0
Subtotal 25 33 26 7 4 2
Social Sciences
Economics 6 0 33 0 0 0
Political Science 8 25 38 13 0 3
Sociology 10 20 33 34 10 0
Subtotal 24 15 35 16 4 3
Natural Sciences
Astrophysics 6 0 33 33 0 1
Biology 6 17 33 17 0 0
Physics 7 0 71 14 0 0
Subtotal 19 6 46 21 0 1
Total/Average 68 1 36 15 3 6
Introduction 17
closed doors. They may be encouraged or dismayed, for example, to learn

that credentials, connections, and effort can propel an application to the
short list, but beyond that, outcomes are almost impossible to predict and
subject to myriad factors that are outside the applicant’s control.
My hope in writing this text was to encourage reflection and dialogue
among those with a stake in graduate education, especially about aspects
of admissions that persist mainly because they are the way things have
always been done. The data do not generalize to admissions everywhere,
but readers who have participated in the process are likely to see some of
their own assumptions and tensions reflected back to them in partici-
pants’ narratives and deliberations.51 Reading how others struggle with
admissions—the tough calls they make, the questionable assumptions they
hold, the displays of inertia or courage—can validate one’s own struggles.
It can also provide positive and negative examples from which to learn. And
as cultural sociologists have demonstrated, “thick description” of cases
and episodes can uncover social mechanisms and concepts that are present
or may apply outside of the samples from which they were derived.52 Con-
cepts emerging from this study include deliberative bureaucracy, disciplinary
logics, and counterscripts.
To build upon the existing sociocultural literature on academic evalua-
tion, I set out to analyze three major issues: the decision-making process in
graduate admissions, the meanings faculty attributed to common evalua-
tion criteria, and disciplinary variation in faculty approaches to admissions.
Those themes are the anchors of Chapters 1 through 3, respectively, and
are helpful in documenting central elements of graduate admissions prac-
tice. However, because I took an inductive approach to analysis and re-
mained open to learning what was important to faculty participants, other
important findings emerged from the data, including several that relate to
the social psychology of faculty identity and judgment. For example, I had
expected that faculty would prefer applicants who shared their elite academic
pedigrees, but I did not expect to see some other dimensions of preference
for students like themselves (such as experiences overcoming poverty and
presenting oneself as cool or hip) (Chapter 4). I had not anticipated that
faculty would circle around to intelligence over and over again as one of
their central concerns (Chapter 5). Finally, committees very rarely men-
tioned the race or ethnicity of domestic students, but they were vocal with
their assumptions about Asian international students, especially those
from China (Chapter 6).
Consistent with constructivist qualitative research, each chapter begins
by presenting faculty participants’ perspectives without supportive or crit-
ical commentary so that readers can immerse themselves in the ways
participants think and deliberate. Each chapter takes up the consequences

of current practice, and examines the extent to which the prevailing mindsets
of faculty participants are consistent with current research. Some chapters
also include results of my searches for disconfirming evidence or alternative
explanations. I will admit: knowing the risk of confirmation bias—the ten-
dency for people to listen mainly to ideas that support their preconceptions—
there are perspectives I hesitated to put in print out of concern that it
would lead readers to become entrenched in ruts they are already in. I
think, for example, about some faculty narratives around intelligence and
belonging within academe. But I have included it all, both to provide the
most honest portrayal of admissions’ good, bad, and ugly, and in trust that
readers will engage participants’ comments and interactions with the same
critical thinking they bring to their own scholarship.
Conclusion
This is, in part, a story about the impact of a system motivated by good
intentions. A common thread in the findings is that faculty enter the ad-
missions process intending to hold firm on their ideals, but that they com-
promise again and again to get the job done. Faculty experience admissions
work as politically, cognitively, and procedurally difficult because it posi-
tions them between impulses, principles, and pragmatism. At the level of
process, they are caught between attractions to a collegial ideal of delib-
erative democracy and the efficiencies of bureaucratic decision making. At
the level of evaluative criteria, they feel that conventional achievements and
pedigree are critical, even as many feel obligated to and see opportunities
afforded by holistic review and a more inclusive notion of excellence. They
struggle with the prospect of rejecting African American, Native American,
and Latino students whose applications receive full committee review, but
they worry about considering diversity as one of their initial criteria. In the
end they exclude many who could be successful and admit some about
whom they feel ambivalent.
More fundamentally, faculty feel caught between satisfying their own
consciences, respecting their colleagues’ values and priorities, and the aims
of the program and discipline whose futures they are trying to shape. De-
termining who should be admitted often becomes an elaborate, ad hoc
compromise rather than an application of specific values and priorities.
In that compromise, good intentions and principles often fall prey to
pragmatic interests, and faculty frequently default to the safety of self-
Introduction 19
reproduction. In this political pressure cooker, it is no wonder that change

comes slowly, if it comes at all.
I wrote this book because faculty often draw the fine line between
admitted and rejected students without a sense of how their program’s
approach compares to others’, without consciousness of the many tacit values
that drive the process, and without clarity on viable alternatives to the status
quo. My hope is to encourage greater awareness on all of these dimensions
by documenting how they play out in departments representing a range of
disciplines. I hope this book puts graduate admissions work into per-
spective, encourages mindfulness about the premises and consequences
of gatekeeping at this level, and builds decision makers’ capacity to bring
about change where it is needed.
chapter one
Decision Making as Deliberative

Bureaucracy
This is hard work. We are all intelligent, competent people trying
to do our best.
— Chair of philosophy admissions committee
H ow should admissions decisions be made? How are they made

in practice? Judgments of admissibility may seem to be the heart of
admissions work, and indeed most sociological research on this topic em-
phasizes how reliance on certain selection criteria reproduces inequities. My
observations suggest, however, that one explanation for overreliance on
those criteria involves the review process used in these programs to manage
the scores or hundreds of applications received. Before delving in subse-
quent chapters into the nature of faculty judgment in graduate admissions,
I therefore start by outlining key dynamics of decision making.
In general, faculty believe that decisions should be the product of an open
debate among equals in which all members have an equal opportunity to
be heard. Following political philosophers Amy Gutmann and Dennis
Thompson, we could call the model of decision making to which faculty
aspired a deliberative democracy. Such a model makes explicit and acces-
sible the reasons for decisions,1 which is thought to legitimize the outcomes
and to encourage mutual respect and collegiality. Under deliberative democ-
racy, governance itself can become an educative process in which individ-
uals learn, through the give-and-take of ideas, to look beyond their own
perspectives.2 Yet despite faculty allegiance to ideals of deliberative democ-
racy, in practice they settled for a model that I characterize as deliberative
bureaucracy.
Decision Making as Deliberative Bureaucracy 21
Deliberative bureaucracy is an organizational adaptation to rising num-

bers of applications and associated demands on faculty time and attention,
through which they strive to maximize efficiency while upholding the
deliberative, democratic norms that lend legitimacy to their work. Norms of
collegiality, shared governance, peer review, and consensus seeking have
long been reflected in the processes surrounding faculty hiring, tenure, and
promotion.3 However, as the volume of graduate applications has risen,
faculty have come to feel that there is simply not enough time to involve
the entire faculty in open discussions of every file.
Unlike deliberative democracy, bureaucratic decision making delegates
individuals to serve as group representatives, and because they are not per-
sonally invested in outcomes, they can prioritize efficiency and technical
expertise.4 Max Weber famously argued for the technical superiority of bu-
reaucratic administration relative to collegiate administration. He argued
that the need to find compromises among competing interests in the col-
legiate model delays decisions and makes the process less reliable and pre-
cise.5 By contrast, bureaucracy’s procedural rules and standard operating
procedures downplay personal interests such that “official business is dis-
charged precisely and efficiently with as much speed as possible.”6 Under a
logic of formal rationality, bureaucracy also tries to minimize ambiguity
by applying quantifiable standards, which are thought to suppress indi-
vidual values and ensure more consistent outcomes.7
And indeed, I found that bureaucratic practices such as delegating file
review and quantifying judgment increased efficiency by reducing the time
that the admissions process required. Many were uncomfortable with
these apparent shortcuts, however, and tried to preserve a spirit of delibera-
tion by discussing a smaller set of borderline files. Yet even in those conver-
sations, the tendency was to focus on procedural matters rather than on
the potentially controversial reasons for their judgments of applicants. By
doing so, they minimized conflict while protecting democratic values such
as collegiality and consensus seeking.
The efficiencies introduced by deliberative bureaucracy came with
costs, however. Compromising on holistic review produced decontextu-
alized interpretations of information in the application, making it more
difficult for reviewers to identify, value, and recruit promising students with
unconventional profiles. Bureaucratic processes also obscured the consid-
erations on which applicant ratings and admissions decisions depended. To
insiders, sacrificing a discussion of the reason for one’s rating or decision
may have facilitated a more orderly, legitimate process because disagreements
were fewer. However, it also masked the use of unseemly considerations
and allowed misperceptions about common criteria to go unchecked. It

too often made the basis for admission and rejection opaque rather than
transparent.
Each chapter in this book is aimed at explaining why faculty in highly
ranked doctoral programs rely on admissions criteria that undermine their
equity and diversity aims. This chapter offers a distinctly organizational ex-
planation: faculty members’ preference for efficient, collegial decision making
made it easier for them to fall back on bureaucratic processes than engage in
the more difficult conversations that may be needed to encourage greater
equity and diversity. In the first section of this chapter, I provide an intro-
duction to doctoral admissions for readers who may be unfamiliar with its
procedures, timeline, or the way that evaluation work is delegated. I then sum-
marize two other characteristics of deliberative bureaucracy—quantifying
judgment and efforts to uphold collegiality and deliberation—before
assessing the costs of this model. Finally I present a provocative episode
from linguistics, a program whose deliberations were marked by both vig-
orous debate and clear collegiality.
An Overview of Admissions Procedures
Before discussing the details of deliberative bureaucracy, some readers may

appreciate a bird’s-eye view of the admissions process. Admissions has an
annual cycle in most universities, beginning with faculty, graduate programs,
schools/colleges, and universities persuading students to submit applica-
tions. There would be no need for decision making without a surplus of
applicants, and faculty develop the applicant pool through the image they
present, their reputations in the field, organized outreach and recruitment
efforts, and early interactions with prospective applicants. After applica-
tions are received, reviewed, rated, and debated, committees make their ad-
mission determinations. In the programs I studied,8 faculty made these
decisions with an eye to building a cohort, but elsewhere in higher educa-
tion the goal is also one of hiring apprentices to support faculty research
and/or teach undergraduate courses. At the point of admission, decision
making shifts back into the hands of admitted students. Faculty move into
a posture of recruitment once again, sometimes using financial aid awards
as a way to signal interest in a student and often inviting admitted students
for a campus visit. One by one, students file their intent to enroll—or their
plans to enroll elsewhere. This book focuses on faculty gatekeeping as a
window into graduate programs as organizations, but the lengthy dance
between selective institutions and prospective students ends not with insti-
tutions choosing or rejecting students, but with students choosing or rejecting

institutions.
Table 3 provides some basic information about admissions in the pro-
grams represented in this research, such as the number of applications re-
ceived, which varied by an order of magnitude, from 80 in classics to 800 in
economics. The admission rates ranged from 5 percent in one philosophy
program to about 20 percent in physics, with a median admissions rate of
9.5 percent. Rather than involving all faculty in selecting doctoral students
or delegating the admissions process entirely to administrative professionals,
as is common in many masters programs, departments entrusted the work
of processing these applications and selecting students to a faculty com-
mittee, sometimes aided by an administrative assistant and/or a few out-
side readers or graduate student members. Admissions chair was in most
cases a rotating position, but two programs strove for continuity of leader-
ship, and there, the same individuals served as chair for twenty-five and nine
years, respectively. One jokingly called himself the “admissions czar” in de-
scribing the power he had wielded over the years to shape the program.
Indeed, at the graduate level, the decision to admit or reject is negotiated
not in a central campus office, but rather by individuals and committees
representing academic programs. One strand of higher education’s history
concerns the transfer of authority over graduate education from university
presidents to departments and programs. Early in the twentieth cen-
tury, scholars’ role shifted from polymath to specialist, disciplines consoli-
dated around common research interests, and departments and programs
became independent—if not autonomous—administrative units.9 Each
of these changes supported department authority over graduate admis-
sions. Recently, leading research universities have reasserted a coordinating
role in graduate education over enrollment management activities such as
admissions, financial aid, and diversity recruitment work. Typically, their
oversight serves to rubber-stamp academic departments’ work and decisions,
build capacity for challenges that are common across disciplines, and make
explicit the connection between doctoral education and sponsored research.
As sociologist Burton Clark noted, disciplines and universities converge in
department activities, and because those activities are so vital to universities’
broader institutional mission, departments are relatively resistant to external
pressures.10
Some admissions committees’ autonomy is limited by institutional policy
or formal law, such as the bans on race-conscious admissions that are
currently in place in eight states. And in programs where admission is as
much about hiring research assistants for specific grants as it is about se-
lecting students for a cohort, the committee may consult with other faculty.
Table 3 Comparison of Admissions Process across Programs
Number who
Number of Applied / Were Rounds Admissions
Committee Committee admitted/ of Key Evaluation Chair’s Gender & Interview (of
Department Size Meetings Enrolled Review Criteria Nativity whom & how) Notes
Biology 5 ~8 150–200/ 2–3 Research Male, domestic Short list, via 3 admissions processes used:
20–35/14 experience video conference standard, via interdisciplinary
program, & in rare cases direct
admit. Numbers reflect
standard admissions process.
Initial review of applications by a
subset of the committee; full
committee reads files on short
list.
2/3 of admitted are international
students.
Astrophysics 5 4 150/20/7 3 Physics GRE, Male, Short list, via Initial review of applications by
research international video conference full committee.
experience, All-male committee.
coursework/ 1 graduate student on committee.
grades Extensive consultation among
committee members via email.
Physics 10 1 450/90/26 1–4 Physics GRE, Male, Non-native Rolling review; initial review of
research international English speakers applications by committee
experience, on the short chair & 2 members.
coursework list, via video Number of rounds of review
conference depends on consistency of
initial ratings.
Committee members required to
provide narrative with their
ratings.
Subsets of committee may meet
to discuss a few specific cases.
Chair has strong managerial role.
Classics 5 1 80/9/4 2 Language Male, Very short list, Department administrator
training, grades, international, during recruit- conducts initial review of
Verbal GRE ment weekend applications, then full
committee reads files on short
list.
Committee includes department
chair.
Philosophy 5 3 200/7 + wait 2 Writing sample, Female, domestic No interviews Initial review of each application
1 list/7 letters of by two committee members;
recommenda- then, full committee reads files
tion, GRE on short list.
Wait list used extensively to ensure
the desired size of cohort.
Chair has strong facilitator role.
Committee appoints diversity
officer.
5-point rating scale used.
(continued)
Table 3 (Continued)
Number who
Number of Applied / Were Rounds Admissions
Committee Committee admitted/ of Key Evaluation Chair’s Gender & Interview (of
Department Size Meetings Enrolled Review Criteria Nativity whom & how) Notes
Philosophy 5 2 225/20/5 1–2 Writing sample, Male, domestic No interviews Initial review by admissions
2 letters of chair; any applicant accepted
recommendation or put on wait list will have
application read by at least 4
people.
4-point rating scale used.
Linguistics 6 5 105/8/4 4 Grades/ Male, Very short list, Initial review of each application
coursework, international via recruitment by 2 committee members; then,
college weekend full committee reads applica-
reputation, tions on short list.
alignment of 3 meetings of committee only;
research 2 meetings with full faculty
interests with deliberation of short list (~30) &
department very short list (~12).
vision Last round of review occurs after
the recruitment weekend.
2 graduate students on committee.
Committee includes department
chair.
Reviewers each give each
applicant a yes/no vote, then
number of yes/no votes used
as proxy for energy behind the
applicant.
Sociology 5 2 225/22/10 1–3 Committee Male, domestic No interviews Number of rounds of review
members free to determined by initial ratings.
use their own Initial review of each complete
criteria application by two committee
members.
Minimal deliberation about files.
Strong reliance on average
ratings.
Chair has a strong manager role.
Political 8 2 350/25/16 2–4 Committee Male, Rarely, usually 1 meeting to set process; 1 to
Science discouraged international intl. students via discuss files. Initial review by 2
from heavy video conference committee members; then full
reliance on committee reads all applica-
GRE tions on short list (~80).
3 graduate students on
committee.
Economics 6 2 800/65/25 3 or Quant GRE, Male, domestic No interviews Additional consultation occurs
more math outside of committee meetings.
coursework 10-point rating scale.
Program chair adjudicates final
decisions.
Admissions chair cuts 50% of
pool before committee review &
has a strong manager role.
Nevertheless, in each of the ten programs I studied, a committee mediates

the individual and collective wills of a department’s faculty. Analytically, I
found that focusing on academic departments and programs allowed me
to analyze up to the levels of the university and disciplines, whose futures
are shaped by admissions, and down to the levels of the committee and
individuals who make the admissions decisions. These committees represent
a vital locus of power in admissions, but their work has rarely been the
focus of empirical analysis.11
My participants were well-aware that forming admissions committees is
a somewhat political task; they noted that members are often the first to
see the pool of prospective advisees and that their decisions affect the
entire department. Most committees included faculty at all three ranks—
assistant, associate, and full—and the astrophysics, linguistics, and political
science programs included graduate students. Only two of the ten chairs
mentioned diversity as a factor in their committee formation decisions,
which may help explain why white males predominated in nearly all of the
programs’ committees.12 As a relatively labor-intensive service appointment,
membership rotated year to year in most programs, although those who
gained a reputation for thorough, thoughtful, timely review found them-
selves tapped more frequently.
Leaders also made appointments to admissions committees to downplay
internal conflicts, protect specific program interests, or buffer the process
from program faculty with outlying perspectives or difficult personalities.
In a humanities program with a new strategic plan, the admissions and pro-
gram chairs appointed committee members whom they knew to be supportive
of the program’s new direction. Two of three committee chairs in the social
sciences were careful to ensure the committee represented the full suite of
program concentrations. Such strategic committee appointments are con-
sistent with findings of previous research. James Wilson’s classic analysis
of bureaucracy argued that organizations use bureaucratic practice to bal-
ance multiple interests and manage untrustworthy subordinates,13 while so-
ciologists John Meyer and Brian Rowan found that delegating work can
serve to institutionalize specific interests and manage uncertainty.14
To enhance efficiency, the admissions chair or an administrative assistant
in most programs conducted the initial screening of the applicant pool. In
a humanities program with 200 applicants, for example, the chair described
her objective for the first round of review as “select[ing] out the top quarter”
before the committee “takes a serious, collective look at fifty [applications]
in the second round.” The chair of a social sciences program encouraged
self-selection by publicizing the range of GRE scores among recently ad-
mitted students, and then relied predominantly on GRE and TOEFL scores
to cut the pool—still hundreds strong—in half. After this first screening,
in all but the two programs with the smallest pools each applicant’s file
would be assigned to two or three committee readers again, to reduce the
time commitment and encourage a close reading. Only if there was a large
discrepancy in reviewer ratings would other committee members review
the application or weigh in on whether it should be advanced to full com-
mittee review.
Interestingly, in the course of delegating admissions work, effort and ex-
pertise were often decoupled from decision-making authority. In all ten
programs, a member of the department’s administrative staff typically or-
ganized files for review, scheduled meetings, and provided institutional
memory. Often, he or she was familiar with all of the files, attended com-
mittee meetings, and was the only person in the room to have had personal
contact with an applicant. This individual was a professional who com-
mitted large numbers of hours to admissions work, yet was the only member
of the committee without a vote. By contrast, when committees included
graduate student members, they had full voting rights, regardless of prior
experience with application file review or decision making.
Perhaps in part because I was a graduate student myself at the time, I
was intrigued to observe differences in how graduate student committee
members handled their role and how often it reflected conventional status
hierarchies. Whereas a white male graduate student in political science con-
tributed with the outspoken self-assurance of tenured faculty on the com-
mittee, a white female graduate student spoke only when spoken to, and a
male student from an underrepresented ethnic background did not speak
at all. In linguistics, an international graduate student member worried
about senior faculty seeing and disagreeing with his ratings, whereas a white
male graduate student on the same committee actively challenged senior
professors’ interpretations of files. In this way, committees operated as mi-
crocosms, small worlds that reflected common department and academic
politics.
Sometimes, leaders involved outside reviewers with specialized research
expertise or knowledge of a country’s educational system. Eight of the ten
admissions chairs consulted with program faculty or doctoral students from
China, for example, in the hope that an additional set of eyes would clarify
ambiguities and ensure a fairer review of those applicants. As Chapter 6
will detail, efforts to improve the quality of review for international ap-
plicants were well warranted. Yet in several departments, the role of “China
expert” fell to one or two individuals year after year and appeared to
be neither voluntary nor accompanied by compensation, recognition of
service, or authority to make admissions recommendations. Similarly,
programs that required writing samples assigned them to be reviewed by

colleagues with subject area expertise, and sometimes this required tapping
faculty colleagues outside the committee. Especially in the humanities, the
writing sample was very important to overall judgments of applications,
but committee leaders recognized that there are idiosyncrasies in, for example,
how a Kantian ethicist versus a philosopher of language would judge what
counts as quality writing in philosophy. One faculty member relayed the
process to me:
We send it out to the experts and then we get the reports. But sometimes there
is disagreement with the experts too. The experts who read the paper haven’t
seen the whole pool, so there has to be some balance too . . . Unless the ex-
perts totally, you know, dismissed it and couldn’t find any value in the paper.
Committees were thus often entrusted to connect outsiders’ judgments of

select cases with their own knowledge of the entire pool. Finally, in eight
programs, the admissions committee also consulted with the department
or program chair,15 either in recognition of the department chair’s authority
or as a source of accountability for the representation of key student con-
stituencies (such as women, international students, and students of color)
among the admitted students.
In two programs, the chair’s approach to delegating work backfired, and
faculty outside the committee actively challenged its decisions. In one pro-
gram, faculty from a particular concentration were upset that students in
the concentration seemed underrepresented in the incoming cohort. In the
other case, a senior scholar complained to the department chair that he
had not “received any good students in several years” and suspected ad-
missions committee members had been “cherry-picking the best students.”
After agreeing to admit an additional student for the next year who would
work with this professor, the admissions chair invited him to participate
in the admissions committee the following year to see for himself whether
the concerns about cherry-picking were warranted. Delegation thus can
save time, but by excluding some stakeholders from the process, it can also
create tension between appointed decision makers and those who feel neg-
atively affected by decisions in which they were not involved.16
Quantifying Judgment
Another bureaucratic practice involved putting numbers to judgments in

order to simplify comparisons among applicants. Quantification in admis-
sions is a foundation of American meritocracy. The movement to incorpo-
rate numbers into admissions began with the optimistic, democratic inten-
tion that standardized aptitude tests might ensure that talented college
applicants from less privileged backgrounds would be judged by the same
standard as their wealthy peers. Henry Chauncey, who founded the Educa-
tional Testing Service, reacted in amazement to the congruence between
his first aptitude test and student grades. Historian Nicholas Lemann wrote,
Chauncey was bowled over. It was magical . . . Testing touched upon the
deepest mythic themes: the ability to see the invisible (what was inside people’s
heads), the oracular ability to predict the future (what someone’s grades
would be in courses he hadn’t even chosen yet).17
In admissions and elsewhere in education and social life, quantifying quality

is a key indicator of deeper trends toward bureaucratization and positivism.
When we want quick and convenient knowledge about student achieve-
ment, college quality, or teaching effectiveness, we turn to numbers.18
Quantification is central to graduate admissions review, in spite of wide-
spread doubts about the reliability and validity of standard measures and
average ratings as proxies for complex constructs like achievement and ad-
missibility. The way that numbers seem to quickly cut through the ambigui-
ties involved in comparing students who are different on many dimensions
has made them a vital decision-making tool, particularly as applicant pools
have grown. Faculty in the programs I worked with quantified their judg-
ments in several ways. They interpreted small increments in test scores as
significant differences, for example, and condensed holistic evaluations of
applicant files into numerical ratings that guided subsequent decision making.
Test score increments. Early in the review process, many faculty fixated on
small increments in test scores or grade point averages, treating them as
the equivalent of significant variance in intelligence, preparation, or gen-
eral admissibility.19 Describing the motivation for this practice to me, a so-
ciologist explained, “If you’ve got a stack from here to the ceiling, you’ve
got to try in a small amount of time to get them down to a small list. The
grades, the GREs, the letters do a big work {pause} in trying to decide who
is on the short list.” A biologist made a similar point on the clarity that GRE
scores provide. Individually composed aspects of the application like the
writing sample or personal statement are not only more time consuming
to review, he argued, but ultimately are incommensurable.20
It’s just too easy given quantitative data like that, it’s just too easy to just rank,
you know, to use that to differentiate . . . Not only is it obviously a time sav-
ings but . . . trying to judge and tease apart these factors and distinguish
applications—sometimes it’s not too bad but a lot of times it is, and so to be
able to just say, “Okay, oh, this person clearly had a better GRE set of scores
than this person. We should rank them in.” . . . It’s something that common.
You know all applicants took the same test, presumably under the same con-
ditions, {Laughs} you know, they were watched over. But anyway it’s one thing
that we can standardize in the sea of variability in these applications.
As the applicant pool increases, so does the burden of holistic review, and
with it, the difficulty of sorting out who should have a place on the short
list. The GRE’s standardization thus becomes central to justifications about
its utility and appropriateness.
Applicant ratings. A second form of quantification involves rating indi-

vidual applicants. Faculty in nine of the ten programs translated their ho-
listic evaluations of individuals on the short list into overall ratings, usu-
ally on a 1–3 or 1–5 scale.21 As one participant described it, ratings offer
“the most efficient way to focus in on the people that really need to be
thought about”—that is, the applicants whose admissibility was debatable.
Two recurrent debates in the committee meetings I observed involved who
should receive the hallowed “1” rating on the one hand, and what that
rating actually meant, on the other. One of the philosophy committees
waxed eloquent on these matters, as I discuss in Chapter 3, leading off the
conversation with one member ruminating, “What is a 1?”
In the end, the meaning of a 1 rating was perhaps best revealed by the
qualities of applicants for whom faculty reserved it—those whom readers
were most enthusiastic about, impressed by, or otherwise eager to admit.
Faculty recognized that, no matter how many seemed worthy of admission,
giving more applicants a rating of 1 only delayed the inevitable task of re-
jecting most in their pool. Allocating the 1’s grew even more difficult in
later rounds of evaluation as the share of promising prospects increased.
An economist explained,
It’s really hard. Once you get down to the final forty, you could conceivably
admit all of these people . . . They probably would come here and do quite well.
It’s just you can’t admit everyone who is actually above a certain bar. You’ve
got to keep raising the bar just to accommodate the numbers that you want.
Above a general threshold of admissibility, faculty were forced to make dif-

ficult choices.
A number whom I spoke with justified their preference for making those
hard choices sooner rather than later by asserting that awarding too many
1’s in the first round of evaluation created more work for one’s colleagues.
After observing that a junior member of the astrophysics committee re-
ceived some teasing from his colleagues for being a “softie,” I asked an-
other member of committee in a follow-up interview to explain the teasing:
If you put forward thirty out of the sixty that you want everybody else to
read—you’re not doing your job to whittle it down to the best of the best.
You’re just making more work for the four other people on the committee who
have to read your thirty—instead of them all having to read your twenty. I
think that’s really the only reason why being harsh would be valued. You’re
trying to pick out the best people. You want to cut the fat away.
Another member of this committee affirmed this:
The list needs to be cut down. We have 180 applications and we had to cut it
down to eighteen admissions or something like that. We had to make cuts. You
know so—you actually need to distinguish them between the merely good and
the very good, and being a harsh critic is helpful for making that distinction.
In some sense, the committee can’t function if we’re all softies. There needs to
be some who are more harsh and demanding. Otherwise it’s impossible to cut
it down.
The perceived need for some reviewers to be “harsh and demanding” was
not limited to the hard sciences. In one of the philosophy departments, for
example, the admissions chair apologized profusely for the number of 1’s
she had awarded, sparking a longer conversation about what constitutes
the “boundary between a 1 and 2” and how each of them had approached
the ratings. The chair exclaimed, “Sorry about all of my 1’s! My 1 might
be another person’s 2 . . . We are all intelligent, competent people trying to
do our best.” A senior colleague replied, “1’s are the crème de la crème. I was
very disciplined. I made a list of all the 1’s, and then I was ruthless to get
to ten 1’s.”
Still, rejecting a large majority of applicants can be a tough pill for new
reviewers to swallow. Many of those who are ultimately rejected could
probably succeed in these programs, and achievements of even rejected ap-
plicants may surpass reviewers’ own accomplishments at a similar stage of
their career. Everyone recognized the difficulty of cutting ostensibly quali-
fied students, and seasoned veterans of admissions review often adopted a
tough attitude to match the toughness of the task. Being “harsh,” “de-
manding,” “disciplined,” or even “ruthless” with one’s ratings was thought
to ultimately serve the committee’s best interests by moving the process
along more efficiently than a “soft,” “generous,” or “lenient” approach. A
sociologist clearly explained the importance of individual ratings, which
many faculty slipped into calling “rankings,” and their influence on admis-
sions outcomes:
There was very much a sense that we all went off in our little corners and read
and submitted our rankings. That’s where the decision really got made.
Individual ratings were key ingredients of collective decisions, both in de-

termining an individual’s place in the rank-ordered list of applicants, and
as a starting point for conversation about his or her particular strengths
and weaknesses.
Rank-ordering applicants. The quantification of judgment continued into

the next phase of the review, as committees averaged individual reviewers’
ratings to develop a rank-ordered list of applicants that would guide subse-
quent discussion and decisions. Here, committees associated each applicant
with two numbers: the average rating of the application (for instance, 2
out a possible 5) and their rank on the overall list (such as a rank of 11 out
of 20 total applicants on a short list). The list, aptly regarded by one philoso-
pher as the “collective assessment” of the committee, thus linked individual
judgment with collective selection.
Several participants described the simple computation of the rank-ordered
list as an efficient substitute for the hard work of comparatively evaluating
the applicants.22 With just a few keystrokes using spreadsheet software, in-
dividual judgments of hundreds of unique applicants could be transformed
without discussion into an apparently clean distribution of their admissi-
bility. Several marveled to me about this substantial feat, with one humani-
ties professor describing the computer’s work with a flourish of his arm
that resembled waving a magic wand.
Letting the computer generate the committee’s collective assessment of
each applicant was functionally important in that it saved time over
consensus-oriented discussions, but its social importance lay in the reason
it saved time: it clarified where agreement already existed and inhibited de-
bate over disagreements. “Consensus at the top tells you that the process is
working,” one social sciences professor summarized. To save time, commit-
tees felt they did not need to discuss areas of agreement, which the spread-
sheet of individual and average rankings allowed them to assess at a glance.
A senior sociology professor elaborated:
Years ago they didn’t score these things. It’s become more bureaucratic. I think
they want to move things along, so once you put a score on it, you just tabu-
late and rank-order . . . When I was at Cornell, we didn’t use this kind of
system. We spent the whole morning deciding the admit list, and in this one it
took us less than an hour after we had already rank-ordered. So we had done
our homework before the meeting.
Once we see the list, then we could discuss, so it becomes very clear. There’s
not much debate . . . I think it’s a good one, because otherwise you can debate
for a long time. They know how to debate and they know how to engage all
these rhetorics . . . If we can keep debating about who is a better student
all morning, we still can’t convince each other . . .
In the end, it’s personal taste. Everyone has good scores on different aspects.
You know, “I like this person’s experience, that person’s background, this
person has better quantitative skills.”
Similarly, one philosopher who reflected on the process noted, “There was
always a consensus that way or another . . . but we didn’t have to agree on
why.” In essence, less time was required for discussion because less nego-
tiation was required. Given the small scale on which ratings were assigned,
average ratings disguised disagreements about individuals’ relative ratings
as minor differences of opinion about appropriate standards, criteria, and
applicant strengths and weaknesses.23
Some committees relied on the list more than did others. Some used it as
a jumping-off point for identifying and discussing borderline cases. In as-
trophysics, one member explained,
The first two-thirds of the list went immediately. There were some people it
was clear we were going to make offers to and some people it was clear we
weren’t going to make offers to, and I think two-thirds of the people fell into
that category. Those two-thirds took ten minutes and the rest of the time was,
“Let’s worry about the boundary cases.”
Other committees used the list to determine sets of applicants who should
be interviewed or whose files the committee should review more closely.
Still others simply admitted as many from the top of the list as their projected
enrollments and yield allowed. A sociologist explained to me that, with the
exception of one case,
there wasn’t a whole lot of discussion at the meetings about moving people
around . . . We kind of went with the rankings. We felt like we could discuss
all these people and certain cases but, you know, we had confidence that ev-
eryone had read everything pretty carefully. And we had thoughtful rankings.
I thought it went really well.
In all cases, allowing each member to have different individual ratings of

applicants and using the average of those individual ratings meant that fac-
ulty could continue to disagree about student quality while still getting the
job done. However, this practice had significant consequences for the trans-
parency of admissions decisions. In highly competitive programs like these, a
single reader’s classification of an application as a 1 versus 2, or as a 2 versus
3, could spell the difference between admission and rejection, yet with only
“the numbers” before them, not even those who assigned the ratings would
necessarily be able to pin down the factors that produced this outcome.24
Faculty committees conducted admissions as a “numbers game” in other

ways. A few readers, for example, took individual rating a step further by
imposing numerical identities onto applicants based on the ratings (“She’s
definitely a 1”). And all programs gambled on their yield rate in determining
how many admissions offers to make. Nevertheless, the three most common
ways faculty quantified judgment were by interpreting small increments in
test scores as substantive differences in admissibility, applying numerical
ratings to applicants, and constructing rank-ordered lists of applicants using
the averages of reader ratings.
“We Definitely Try to Make People Happy”
Admissions may officially be a matter of choosing students, but the selec-

tion process itself is an institutionalized compromise that balances and
reflects multiple, sometimes competing, faculty values.25 And to many pro-
fessors, the easier it is to find this compromise, the better. Therefore, amid
awareness of the tendency toward bureaucratization, the faculty I observed
made concerted efforts to build two professional norms, deliberation and
collegiality, into the admissions process. Perhaps the clearest evidence of
this was in widespread effort to downplay fundamental disagreements
about what the appropriate evaluation criteria should be. About the tacit
agreement to disagree on this matter, a sociologist remarked:
Invariably, we wind up having conversations about norms and priorities, but
there’s no effort to put people on the same page. I would resist that as an indi-
vidual faculty member . . . People weigh pieces of the application packet differ-
ently and I respect that. I would never try to create that kind of consensus.
Another sociology professor related these differences of opinion to the di-

verse viewpoints that he believed were central to research and faculty
culture:
I absolutely believe that the best work I’ve ever done is a pathetic, groping
approximation to some very complicated truth out there. And the more
methods, the more angles of vision I can bring to bear on that complicated
truth, I feel, the better, the closer I get to it, without ever getting there.
I think we’re in the same business when we’re doing admissions. The
complicated truth is this person’s ultimate ability as a sociologist. And I’ve got
different metrics. I’ve got different items of information that are supposed to
be giving me a handle on that. And when I’ve got a faculty member that says,
“Oh, GREs will predict that” . . . I think, “You’ve got to be kidding, right? We
know there’s almost no predictive power there.” But it is a piece of information.
I agree.
So, I’d rather have members of the committee not saying, “It’s GREs.” If
everybody went, “It’s GREs,” we’d have an easy process . . . and a deeply flawed
process. To me it would be very, very far away from that complicated truth.
So I want people coming to it, hopefully, from different angles.
Whether it was to avoid conflict over their colleagues’ divergent approaches

to evaluating applicants, or out of respect for their “different angles,” the
importance of collegiality to perceptions of a legitimate selection process
is clear.
Faculty also displayed a commitment to collegiality in several other ways
through the review process. For example, rather than discuss the thorny
details of many cases, committees redirected most of their deliberations to
less controversial issues, such as the rules of the process, what the numer-
ical ratings mean, and where to “draw the line” for admitted students on
the rank-ordered list. They debated the merits of delegation and quantifi-
cation, philosophically and in practice, reminding one another of the value
of deliberation and making appeals to shared values such as fairness and
peer review. Discussing this level of detail and philosophy about procedure
permitted the appearance of a deliberative process without broaching dis-
agreements that stemmed from differences of priorities.
Moreover, it was natural to discuss procedure because so many faculty
were ambivalent about the admissions process. In both interviews and com-
mittee meetings, participants freely acknowledged imperfections in the
system, such as the false precision introduced when quantifying student
quality through a rank-ordered list. A political scientist compared the chal-
lenge of admissions to that of selecting faculty for awards:
Partly for reasons of expediency, we need some kind of scoring system. It’s easy.
I’ve been on lots of committees on campus this semester deciding award win-
ners. When the pool is twenty-five and you have time to think about your top
fifteen who should get this teaching award, the numbers and the myth of pre-
cision [are] readily acknowledged because you have time to talk about all of
this. But if you have a hundred good people, then . . . [we] say that the confi-
dence interval around these numbers [is] thirty spots plus or minus, but we
still have to go by them.
Members of one philosophy program spoke at length about this problem

as part of a longer conversation about their discomfort with quantifying
judgment and the implications of the subjective nature of the ratings
themselves.26
Along with false precision, another frequently debated set of procedural
questions was whether, how, and where to impose numerical thresholds
for GRE percentiles, TOEFL scores, and average applicant ratings. These
debates provided clear examples of admissions decision making as impro-

visation, with faculty constructing the process as they went along. A lin-
guist summarized the process and his committee’s discussion:
They were ranked, and some of the candidates were already on top of the rank-
ings, so that simplified a lot of the conversations. More disagreement was on
the edges. We cannot admit everybody. You need to put a line somewhere, and
it’s where do you put the line? And then, the people close to the line? That’s
where it starts. And so that’s where the disagreement emerges.
The hard truth for decision makers in programs such as these was that the
admissions process requires many more rejections than admissions. Drawing
a line to separate the two groups, even with a wait-listed set to buffer them,
felt uncomfortable both because it made the scope of the rejections so
plain to see and because it provoked the sort of disagreements that faculty
try to avoid.
Compromise: When Collegiality Trumps Taste
Substituting discussion of uncontroversial procedural matters for questions

of who and what should be valued worked for another reason: Not only did
faculty feel ambivalent about how work was delegated and numbers were
used, but they felt that most admissions decisions were not worthy of ar-
gument, especially when compared to faculty-related decisions about
hiring, tenure, and promotion. In most cases they saw their collegiality as
a bigger priority than their individual preferences. A natural scientist
explained, “I think everyone on the committee wants to have an opinion.
That’s the nature of scientists: they always want to make an opinion. But
I don’t know if a lot of people feel very strongly.” A first-time admissions
committee member in the physics department expressed surprise at the rel-
ative ease with which decisions were made:
I think on the whole there were relatively few, quote, disagreements. Nobody
was—that I could see—no one was particularly adamant that this person must
be admitted. I didn’t see any of that. I was expecting a little bit of that where
students might have contacted directly a professor to say, “I would very much
like to work in your group and your group exclusively” . . . But I didn’t see
any of that. I didn’t see anybody pushing internal candidates. I thought it was
all very congenial and, “Let’s talk this over. What did I miss?”
Indeed, across the six committees I observed, committee members limited

hard advocacy to one or two cases per committee member.
Deferring to a colleague’s judgment, not challenging it, was the norm.

Faculty rationed their confrontations, treating agreement as more impor-
tant than the marginal differences in the quality among those on the short
list. Instead, and somewhat ironically, although admissions committees ag-
onized about the integrity of the process, it was only through deliberation
about its imperfections that they were able to move forward together.
Costs of Deliberative Bureaucracy
The goal of deliberative bureaucracy was to make the process of arriving

at a collective judgment more efficient while preserving core norms such as
deliberation and collegiality. The process clearly accomplishes these aims,
but it also comes with unintended consequences.
Compromising recruitment networks. Bureaucratic practices made it more

difficult for these programs to develop and sustain institutional pipelines
and admit students with nontraditional backgrounds. Comparing current
practice to that of thirty years ago, an emeritus physics professor argued
that pipelines “could be implemented today if they really wanted to. But
everybody is sort of backing off and they want to look at everything and
{Pause} go through the more formalized process.” With decision making
both decentralized and formalized, it becomes harder to maintain relation-
ships with feeder programs outside the network of high-prestige universi-
ties that everyone recognizes. Bureaucratic processes can also undermine
outreach and recruitment programs used to identify talented applicants with
nontraditional academic profiles. Kimberly Griffin and Marcela Muñiz
came to a similar conclusion; these researchers found that the loose cou-
pling between the administrators responsible for recruitment and the fac-
ulty tasked with admissions decision making was a major barrier to the
professional efficacy of graduate diversity officers.27
Decontextualizing applicant characteristics. The approach to review that

I observed during the initial screening, which emphasized the value of stan-
dardized categories and classifications over contextualized interpretations
and individual uniqueness, fostered what Carol Heimer calls a case-based
approach to evaluation.28 This type of review was more efficient, but it also
made it more difficult to identify and admit students with unconventional
profiles. One humanities professor articulated: “Students are first sort of
presented to us and to other search committees—to our competitors—in a
spreadsheet kind of format. There are some that light up. Basically, these
things are scores or numerical numbers like GPA, undergraduate institu-
tions, test scores. There are some that will stand out as being {Pause} that
look promising even without really looking at the application.”
Identifying applicants at the earliest stages of the process by standardized
categories such as test scores, institutional affiliations, or diversity contribu-
tions, rather than a review of their life stories or unique characteristics, had
two important consequences. First, it made the few criteria they did consider
more determinative of ratings and, ultimately, of students’ opportunities. It
also compelled an even greater need for committees to repersonalize ap-
plicants if reviewers submitted widely disparate ratings. They resolved
those differences in most cases not by debating the criteria used to award
a rating, but rather by developing narratives about the person behind the
numbers, achievements, and standard categories. Here the process re-
sembled what Heimer called biographical analysis, and what Mitchell
Stevens called “evaluative storytelling” in his analysis of undergraduate
admissions.29
Obscuring the basis for ratings and decisions. In addition to making pipe-
lines difficult to sustain and decontextualizing information about applicants,
perhaps the most fundamental cost of deliberative bureaucracy was that it
obscured the basis for ratings and, by extension, decisions. What really mat-
tered to faculty when they rated, admitted, and rejected applicants faded
into the background, and core questions of what counts as a “quality” ap-
plicant and which applicants met that definition received only marginal
attention. This tendency is closely related to a common critique of quanti-
fication, that in the course of facilitating decision making, it “simply evades
the deep and important issues.”30
By the standards of deliberative bureaucracy, utilizing numbers as proxies
for complex, debatable judgments had real advantages. Numbers concealed
underlying disagreements that faculty would rather not broach and buff-
ered them from charges of unfairness. I was struck to observe that in only
two of the ten programs were faculty even asked to compose notes that
explained their applicant ratings. Moreover, the very nature of deliberation
about borderline cases was so complex and holistic—rich with detail, im-
pressions, and simultaneous weighing of pros and cons—that it was rarely
possible to determine why, exactly, a given student was admitted, rejected, or
wait-listed. As the chair of the political science department aptly put it,
“Everything matters, and nothing matters the most.” Vague criteria, unjus-
tified ratings, and ad hoc policy all added to the veil of secrecy that pro-
tects these controversial decisions.31
Yet by the standards of deliberative democracy, transparency and reason

giving are central to a decision’s legitimacy. The lack of transparency in-
herent in holistic review has compromised public perceptions of admissions’
fairness. Of particular concern to many are the specific ways that race and
other diversity considerations are factored into decisions. Some assume that
decisions made in secret must be largely a product of committee members’
personal tastes or subconscious biases. Others assume that race plays a piv-
otal role in outcomes, instead of serving as a “factor of a factor of a factor
of a factor,” as Justice Ginsburg noted in her 2013 dissenting opinion on
Fisher v. University of Texas, in which the U.S. Supreme Court evaluated
the constitutionality of the admissions policy at the University of Texas.32
Openness about the basis for ratings and decisions may open the door for
disagreements, but it also forces accountability for evaluation criteria in use
and enables committee members to learn from each other.
Friendly Debate and Drawing Group Boundaries
That reason giving and disagreement need not compromise collegiality—and

indeed can be harnessed to improve admissions work—was exemplified in
three committees: linguistics, political science, and astrophysics. In these
programs, deliberations were characterized by vigorous, friendly debate in
which faculty corrected each other’s misperceptions and proposed alterna-
tive interpretations. They made explicit their tacit assumptions and, in se-
lect cases, challenged one another on those assumptions. An example from
a meeting of the linguistics admissions committee exemplifies how faculty
draw out the assumptions driving their colleagues’ judgments, especially
in the presence of implicit bias.
Near the end of an almost three-hour meeting, left with only a handful
of borderline cases they had previously tabled, the linguistics committee was
losing what little formality had characterized their interactions. Starting the
discussion about Maria, a woman who had attended a small religious col-
lege, a committee member admitted, “I didn’t know anything about the col-
lege. I had to look it up.” Those with some familiarity spent a little time
sharing their impressions and knowledge with the rest of the group.
“Right-wing religious fundamentalists,” one said. “You know, they re-
fused all federal monies so that they can resist the socialists.”
“Supported by the Koch brothers,” another added after a long laugh from
the rest of the committee.
“In all seriousness, it’s actually supposed to be pretty good in the humani-
ties,” one quipped. They started to discuss other specifics of Maria’s case,
noting among other details that she had been homeschooled and that she
had scored at the 99th percentile for the Verbal section of the GRE and the
82nd percentile for the Quantitative section. Although the committee was
near the end of the short list that they were slated to discuss, her case was
the first in which they discussed GRE scores in more than passing detail.
Her educational background clearly had induced skepticism, and they sub-
jected her file to a more stringent review. Those seated on one side of the
table wanted to move her application forward into the next round of eval-
uation, while those on the other side were unconvinced.
“I feel like a jackass for saying this,” a male committee member said,
breaking a long pause, “but she doesn’t seem interested in research to me.”
Others marshaled evidence from her file to the contrary, and the committee
chair smiled and confessed, “I would like to beat that college out of her.”
“I think it already has been,” a senior member of the faculty suggested,
reading an excerpt from Maria’s personal statement that signaled indepen-
dence and critical thinking.
“You don’t think she’s a nutcase?” the department chair asked, to laughter
from most members of the committee. With the three most powerful people
around the table having joked about this case, no one tried to top them.
They agreed to move her forward to the next round but a few weeks later,
when making their final decisions, chose not to extend her an admissions
offer.
Driving home that evening after the meeting, I thought about how the
committee had discussed this case. I wondered what their meetings must
usually be like if this was how they spoke to each other with a researcher
in the room? It was a question I asked myself more than a few times during
the course of my fieldwork, in response to seemingly unfiltered discussions
of candidates with uncommon profiles. It had been an amusing conversa-
tion, but whether Maria had received a fair hearing was debatable, and the
tenor of their discussion was consistent with results of an experimental
study that found disclosing a religious background in one’s application to
psychology doctoral programs reduced the chances of admission.33 On the
one hand, it was frustrating to hear faculty describe any applicant as a pos-
sible “nutcase.” On the other hand, rude comments like this affirmed for
me that I really was seeing behind the curtain. Further, their exchange had
perfectly exemplified how individual and collective judgments could be re-
fined when a committee takes time to work through the details of a case,
especially an unconventional one that might otherwise fall through the cracks
of implicit, collective stereotypes.
On another level, this episode in linguistics showed that interactions of
committees characterized by vigorous but friendly debate were often less
formal and used their deliberations to draw group boundaries. That is, they
used admissions deliberation not only to choose students, but also to send
messages to one another about who they were as a community and who
they were not. It was hardly the only time I observed committees ease ne-
gotiations over sensitive cases by introducing the entertainment and shock
value of humorous or rude comments. Such comments compromised norms
of professionalism, but served an important social function of cutting
through tension, encouraging camaraderie, and clarifying shared values.
Sometimes one committee member would tease another, but more often
jabs were outwardly focused—joking at the applicant’s or another disci-
pline’s expense.
The astrophysics committee was another whose meetings were charac-
terized by friendly debate and humor. In one case, they had reached a point
of deadlock over an applicant, and a junior professor noted that one of the
letters of recommendation in his file seemed “weird and brief.” Another shot
back, “That’s what happens when engineers write,” to cathartic laughter
all around. Whether it was astrophysicists distinguishing themselves from
engineers or members of a large research university distinguishing them-
selves from those in religious colleges, faculty used evaluative deliberations
to affirm their identities and organizational boundaries by admitting those
who were believed to fit and rejecting those who did not.
As my fieldwork progressed, I couldn’t help but note how unique the in-
teractions were in linguistics and astrophysics—how comfortable they
seemed to be with the each other and with the work of ruling out qualified
students. To understand this from a department member’s perspective, I
asked each committee’s admissions chair to reflect upon it in follow-up in-
terviews. The linguistics chair affirmed the value of their deliberation-rich
process,
I think the experience of doing this together is good for us even if we weren’t
admitting anybody. I mean just to read these applications and talk about them,
you find out, you know, about how your colleagues think or find out more
about how they think about the graduate students . . . Their perspective is often
useful and sometimes they know things about such-and-such an institution that
the rest of us wouldn’t know about.
The linguistics and astrophysics committees not only framed critique

and disagreement as natural to social and academic life, they also valued
disagreement as an opportunity to build knowledge and understanding.
And although the political science committee did not seem unanimous
about their comfort with the presence of dissent around the table, its chair
encouraged the members to remember that “disagreement is not a sign
that something is awry.” This perspective stands in contrast to those com-

mittees who perceived disagreement as a precursor to conflict and conflict
as a problem to be avoided. Better, those participants might say, to conceal
disagreements behind numbers and celebrate the average as a compromise
achieved without controversy.
Finally, subtle details hinted at the sort of relationships present in com-
mittees that are characterized by friendly debate. Astrophysics, linguistics,
and political science were the only three committees to serve food during
their meetings, for example. The political science committee had pizza de-
livered to what they called their “marathon meeting.” There were custom
deli sandwiches ordered for one of the linguistics committee’s meetings, and
the administrative assistant serving on the committee in astrophysics cir-
culated a tin of homemade cookies in one of their meetings. I noted also in
two of these three programs that small talk preceding the meeting included
recounting time spent together outside usual work hours and settings. In
astrophysics, two male professors laughed about the play date they had ar-
ranged for their children, and in linguistics, half of the committee bantered
about the mountains they had seen from the air while traveling together
back from a major conference. Obviously, the formation of trusting rela-
tionships is a complicated process involving much more than a sandwich
or some informal time together. Yet it’s also possible that when a group of
professionals will be engaging in potentially contentious work, taking time
for food or fellowship may encourage social conditions, not to mention
blood sugar and energy levels, in which differences of opinion are less likely
to be interpreted as hostility or disrespect.
Conclusion
In some doctoral programs, the sheer volume of applications can make

democratic deliberation about every applicant a near impossibility, espe-
cially if there are profound disagreements about how scarce spaces should
be allocated. The model of decision making I observed, deliberative bureau-
cracy, is an organizational adaptation to these challenges whereby faculty
reduce the potential for divisiveness and increase the potential of getting
the job done with a reasonable time commitment. Three behaviors—
delegating work, quantifying judgment, and shifting the focus of discus-
sion to process matters—serve the interests of bureaucratic efficiency and
safeguard professional norms like collegiality and deliberation.
Deliberative bureaucracy also helps explain why faculty overrely upon
GRE scores. Committees looked to quantitative metrics for efficiency’s sake,
often decontextualizing test scores from their students’ backgrounds. Yet

failing in many cases to discuss the basis for applicant ratings or to read
scores in context allowed narrow framing and misperceptions of what test
scores mean to continue unchecked. Deliberative bureaucracy may serve
faculty interests, but it is a costly model for applicants whose GRE scores
are noisy signals of potential.
Most research on elite preferences in educational settings emphasizes how
common selection criteria can reinforce existing patterns of privilege, and
this is a major theme that runs through subsequent chapters of this book.
However, deliberative bureaucracy demonstrates that the preference for
efficient, collegial decision making also plays an important role in main-
taining inequality by helping explain unexamined use of those criteria in
reference to most applicants.
One of the keys to understanding organizations is understanding how
they coordinate their values and beliefs with their practices and personnel.34
Faculty valued consensus achieved through deliberation, but felt daunted
by the prospect of finding consensus amid the range of disagreements they
held—about acceptable criteria and how applications should be interpreted,
to name just two. Finding consensus required a level of conflict that most
programs preferred to avoid. Moreover, they recognized that any process
they adopted would be seemingly unfair because it would deny enrollment
opportunities to scores or even hundreds of qualified candidates. Absent
the feasibility of consensus or fairness as grounds for legitimate admissions,
they built a case for the legitimacy of their practice on other organizational
priorities—namely, efficiency and collegiality—while permitting a modicum
of disagreement on less controversial matters of process.
Discussing the roots of democratic decision making, Pericles wrote, “In-
stead of looking on discussion as a stumbling-block in the way of action,
we think it an indispensable preliminary to any wise action at all.”35 Using
numbers as proxies for collective judgment and refocusing discussion away
from controversial matters may get the job done, but deliberative bureau-
cracy is ultimately a missed opportunity for organizational learning. The
discussions that lead to selecting students are about more than selection;
they are an annual opportunity for leaders to reassess their department and
discipline’s mission, values, and what they have been and are becoming.36
Through gatekeeping deliberations, faculty can come to understand and ap-
preciate others’ perspectives and more thoughtfully align their organiza-
tional identities and goals as scholars with the people they are initiating to
help fulfill them.
chapter two
Meanings of Merit and Diversity
V ivek, the admissions chair, started the conversation about Denpa.

“He grew up in a yurt in the Himalayas. He was raised by his mom
and grandma after his father died at an early age, and the next neighbors
were two mountains over. He then found his way to a major U.S. research
university and has since started the only organization for the discipline in
the Himalayan region.”
On the other side of the table, William, a recently tenured professor who
seemed to play the role of resident skeptic, asked, “But do we think he can
succeed?” There was a long, loaded silence. None of the five around the table
dared to register a prediction. In an earlier meeting, the committee had
decided not to move Denpa forward in the admission process because his
GRE scores came in after the deadline and because when they did, the
Subject score was at the 10th percentile. They were having this conversa-
tion only because a senior professor in the department, Harold, had peti-
tioned the committee to reconsider his application. Harold had received a
personal email from a longtime friend who was serving as one of Denpa’s
references, discussing how impressed he was with the student’s research ex-
perience and motivation. Harold asked Vivek via email to reconsider their
decision and to offer him admission, conditional on “passing certain reme-
dial courses.” Vivek deferred to the committee in making this call, and after
receiving some pressure from others in the department, they invited Harold
to their meeting to discuss the case.
Meanings of Merit and Diversity 47
Breaking the silence, Vivek acknowledged, “He’s the most amazing case
we’ve ever seen.”
Finally Harold, himself, spoke up. “He would bring some personality to
the department. I commit to look after him and fund him through the pre-
lims. He presents himself as quite intelligent.”
“Excellent idea to give him a chance,” the graduate student member of the
committee affirmed, and with that, Denpa’s case was closed. He was ulti-
mately admitted. He enrolled. He passed the “remedial” coursework, the
core courses, and his qualifying exam. And at the time this book went to
press, Denpa had submitted a paper for publication, was working on his
thesis, and was active in scientific outreach in both the United States and his
home country. Harold remained his advisor, and he collaborated with
William.
Debatable assumptions underlie judgments of quality, which underlie as-
sessments of merit and admissibility. I argued in Chapter 1 that running
admissions as a deliberative bureaucracy effectively prevents difficult con-
versations about those assumptions from dominating or extending the
review process. Yet regardless of whether the politics of quality remain
tacit—due to consensus or avoidance—or are made explicit because indi-
vidual judgments of applicants do not align, they are always present in ad-
missions. They are at the heart of committee and broader debates about
how opportunities should be distributed. Through cases like Denpa’s—in
which faculty debate judgments of unconventional, borderline applicants—
the politics of evaluating student quality and the multiple meanings of merit
are revealed.
I found that the standard used to assess merit during the initial screening
of applications varied substantially from the standard of merit used to judge
candidates who made the short list. The distinct notions of what made a
desirable applicant at each stage of review had distinct implications for the
equity and diversity of who might eventually enroll. Given the distributions
of GRE scores by race, ethnicity, and gender, for example, the extremely
high bar of conventional achievement with which faculty in these programs
initially assessed the pool adversely affected opportunities for Latino, black,
Native American, and female applicants.1 I found that meanings of merit
included diversity only when faculty worked to distinguish among those
on a short list of well-qualified applicants, at which point the ideal of merit
was rooted in envisioning the future of the discipline. Even then, however,
faculty conceived of diversity very broadly. Black, Latino, Native American,
or Southeast Asian applicants might stand out on the short list because they
rarely appeared there, but race/ethnicity was just one of many dimensions
of diversity they weighed in making judgments about quality. And as Denpa’s
case demonstrates, regardless of how faculty assessed individual student

quality, department politics also contributed to the politics of quality and
the operational meanings of merit.
Thus, if a cost of the secrecy and lack of transparency in admissions is
uncertainty about how decision makers are incorporating student identi-
ties into their assessments, an important point in these programs’ admis-
sions story is this: Even in closed-door conversations, race, gender, and other
student identities came up very rarely. When considered at all, race was a
“factor of a factor of a factor of a factor,” as Justice Ginsburg concluded.2
Indeed, it was not opponents of affirmative action but rather its proponents
who would be likely to take issue with the standards of merit by which
graduate decisions were made.
How gatekeepers in elite organizations relate merit and diversity is a
question of deep importance. The Association of American Colleges and
Universities proposes that joining principles of merit and diversity in an in-
tegrated notion of inclusive excellence is “critical to the wellbeing of demo-
cratic culture.”3 Lani Guinier’s advocacy of democratic merit is consistent
with this view, emphasizing that a tendency to narrowly judge excellence
on the basis of individual achievement or aptitude flies in the face of higher
education’s responsibility to cultivate leaders with democratic capacities
such as collaboration, problem solving, and creativity.4 At the institutional
level, higher education researchers have argued that inclusive excellence in
admissions enables postsecondary institutions to “leverage diversity for stu-
dent learning and institutional excellence.”5
Theories of Organizational Excellence
Sociologist Michèle Lamont found that faculty on interdisciplinary review

panels define excellence and rationalize their judgments through evaluative
scripts. The idea traces to Erving Goffman’s dramaturgical theory, which
compares social life to the theater and public behavior to the performance
of familiar roles and scripts. Following prevailing social scripts ensures that
others will find our behavior (the performance) realistic and acceptable,
which led Goffman to conclude that social behavior consists of managing
others’ impressions as much as it consists of managing our own principles
or commitments.6
Applying this idea to the context of evaluation, Lamont found that fac-
ulty use scripts to link formal evaluative criteria with the meanings those
criteria hold. Usually these meanings are rooted in reviewers’ shared iden-
tities and work, and they therefore serve as decision pathways or stories
that reviewers tell to justify their judgments.7 Following these pathways is

a bit like peeling back the layers of an onion. A surface-level criterion has
meaning, but there are reasons—often implicit—that the meaning is itself
important (that is, the meaning carries its own meaning). For example, a
common evaluative script I observed was that some faculty perceive grades
to be a function of academic ability or effort, perceive ability and effort to be
signals of future success, and perceive likelihood of future success to be a
justification for the investment that admission represents. Evaluative scripts
thus allow even controversial outcomes to be regarded as legitimate, because
criteria are understood as proxies for shared, self-evident values. Definitions
of excellence thus come to constitute boundaries that define groups of
scholars and become the terms by which they define others.
Shared goals may also factor into the ideals of excellence that guide de-
cision making, and this is particularly the case in elite organizations because
they attract many applicants who meet basic qualifications. Organizational
theorist Robert Birnbaum concluded that a key factor in elite academic se-
lection is the degree to which a particular applicant “represents the optimi-
zation of one or another institutional goal.”8 Knowing the goals of highly
ranked graduate programs can therefore offer insight into the desired
qualities of applicants and what will likely count as merit. For example,
graduate education and academia have a basic goal of advancing knowl-
edge, and the importance to an academic program of pushing knowledge
in particular directions may help explain the importance in admissions of
perceived “fit” and “match” with departmental intellectual foci or disci-
plinary norms, often over and above conventional academic achievements.9
In broad-based departments or interdisciplinary fields, where fresh perspec-
tive is desirable, committees may be more inclined to select students with
diverse knowledge and perspectives. By contrast, niche programs or disci-
plines with a strong orthodoxy may seek conformity over novelty.10
Status is another important organizational goal in doctoral programs like
the ones I studied, and in higher education more generally. To justify and
maintain their high ranking, leaders order program activities and priorities
for maximum consistency with the norms in their field.11 In admissions, that
work of alignment includes adopting selection criteria that carry academic
cachet, such as high GRE scores and degrees from elite colleges and uni-
versities.12 Whether they name status as a consideration or simply reveal it
as such through their behavior, faculty are attuned to organizational status
competition in defining what counts as merit.
Within this broad framework of priorities, however, research about de-
cision making and the selection of social and academic elites portrays it
as highly negotiated processes that depends on a variety of structural and
cultural conditions, as well as the characteristics of the applicant pool and

quirks of reviewers themselves. As such, we can predict that some outcomes
of selection processes in departments like these will seem far from predict-
able. We need to keep these factors in mind in examining how faculty de-
fine and relate merit and diversity in selecting doctoral students.
The First Cut: Merit as Conventional Achievement
In almost every program, review began with assessments of applicants’ ac-

ademic achievements, especially GRE scores and prior grades. As we saw
in Chapter 1, faculty saw these measures as useful metrics for the first cut
in part because they could be used to quickly compare applicants. The gen-
eral rule was the higher the grades, the better, but the meaning of a partic-
ular GPA could be further calibrated by considering the reputation of the
institution where those grades were earned. Similarly, most insisted upon
interpreting GRE scores in the context of national origin, although they
were reticent to do so on the basis of race, gender, or socioeconomic status.
Programs varied by discipline in the specific grades and test scores they
weighted most heavily at this point in the review process. Physicists and
astrophysicists reported that they looked first and foremost to scores on
the Physics Subject test of the GRE. In response to my usual first interview
question, “What are you looking for in prospective students?,” an astro-
physicist admitted, “I would say—and you will see it in our discussions—
it’s very unlikely that we would consider anyone who has a low Subject
GRE.” One of his colleagues concurred, “If you don’t score high, you’re
probably not going to make the cut.” Those in the humanities emphasized
the GRE’s analytical writing score in their initial screens, and in economics,
high quantitative GRE scores were crucial. The admissions chair in the eco-
nomics program disclosed his approach to making the first cut of their 800
applications: “Personal statements have almost no role. I don’t read them. I
look at the transcript and glance at the GRE. If the quantitative score is
not perfect, don’t bother applying.”13 Broadly, these findings about the vital
role of quantitative metrics support Robert Klitgaard’s assessment of ad-
missions in Harvard University’s Graduate School thirty years ago, that test
scores and grades form the “backbone of the evaluative process,” with test
scores especially attractive due to their “magic simplicity.”14
Yet faculty felt that the rise over time in GRE scores and, in particular,
GPAs had decreased their usefulness in culling the pool. Facing a ceiling
effect, in which distinguishing among candidates was impossible because
so many applicants’ scores and grades were concentrated at the top of the
scales, reviewers felt they needed additional information to compare ap-

plicants. As an astrophysicist put it, “Grade point, most people said it doesn’t
really affect them very much because basically everybody in the pool—
everybody in the final pool—has such high GPAs that it’s not meaningful.”
In response, many reviewers considered grades in light of the reputation of
the institution in which the grades were earned and the rigor of the stu-
dent’s curriculum. A sociologist, for example, compared his department’s
admissions process to its process for faculty hiring. I quote him at length:
Robert: What’s great about hiring professors is we have direct evidence of

exactly what they did . . . It’s not easy, but it is information rich. Whereas I
would say graduate admissions is information poor. So then one tends, or
we tend, a lot to rely on signals that are low quality, like, one of the
frequently used ones is the quality or prestige of the undergraduate school.
Lousy signal, I think.
JP: Why is it valued so much do you think?
Robert: Because you have so little else to go on. You have grades, which I
think are a good signal. But the people we admit are always going to be
right around 4.0. Then you have the ones at the margins coming with a
3.9 or 4.1. So grades are increasingly a lousy signal, especially at these
elite places that just hand out the A’s. So you don’t even have that
anymore. I think only if you don’t get a straight A average or close to it
that it’s something you attend to.
Then you have test scores, which are lousy for all the reasons we
know . . . But what else do you have? So you have the tests, and yeah, we
definitely sort of have an expectation of high scores on the test even
though no one likes to use them. But increasingly, you have plenty of
people who are really high on the test scores and really high on grades.
Tons of those people. So now what do you use?
I: It sounds like you’re looking for variance.
R: Right, right. So you use the prestige of the school.
Reviewers could never be sure how much grade inflation had distorted a
given student’s GPA, and they doubted whether prestige of an undergrad-
uate institution actually predicted student success. Still, they continued to
read grades in light of college prestige because it effectively broadened the
range of grades and test scores. On one end would be a low GPA from an
unknown college or one whose training in the field was regarded as weak.15
On the other end of the spectrum was a high GPA from an Ivy League or
other institution with a reputation for strong undergraduate preparation
in the discipline. A “good” grade from a “good” college “carried more
weight” or could be awarded “higher marks.”
Rankings were only one way faculty gauged institutional quality, how-
ever. Seeing one’s alma mater on the application would stir participants on
an emotional level, as could peer institutions of one’s alma mater. This

first- and secondhand knowledge was no more reliable than rankings as
a measure of current institutional quality or applicant promise, but it pow-
erfully influenced their willingness to trust. Unsurprisingly, the Ivy League
constituted a meaningful institutional trust network, but so did flagship
public universities in the natural sciences and elite liberal arts colleges
like Reed, Williams, and Wesleyan in the humanities. These variations
by discipline and individual experience are consistent with organiza-
tional research that finds that “legitimacy ultimately exists in the eye of the
beholder.”16
I probed faculty in the interviews about why they continued using this
initial filter, knowing that it disproportionately excluded students who were
already underrepresented and knowing there was more to the numbers than
meets the eye. By triangulating their answers with observations from the
meetings, I began to piece together important evaluative scripts in use—the
meanings behind the metrics.
Risk aversion and predicting student success. One evaluative script empha-
sized risk aversion. Faculty believed that they had an institutional impera-
tive to avoid risky candidates. And because slots in their program were so
coveted, they had the luxury to reject those candidates whose test scores
and grades failed to reach a very high standard.
Worried about the possible financial and status consequences of student
attrition from their programs, faculty read applicant qualifications with
their own formal course requirements in mind. They sought students
whom they thought would work hard, but who would not struggle too
much with the curriculum. Often, this curriculum demanded prerequisite
skills not communicated to students and for which they offered few to no
opportunities to learn once on campus. Reflecting on their program’s ten-
dency to prefer students whom they felt “pretty confident can get through,”
a classicist explained,
Graduate admissions is one of the things I think you have to be very humble
about. And there’s always a tension here because we’re always under pressure
to have good numbers for completing a program, completing it in a reason-
able amount of time, and so on. The effect of that is to make you risk averse
because it’s not that hard just to go for the students you’re pretty confident
can get through.
Others discussed risk aversion in terms of the investments of time and en-
ergy that students require of faculty. During an interview with a physicist, I
commented, “You’ve mentioned a couple of times: ‘Can they be successful?’
and ‘Do I think they’ll be successful?’ It sounds like that’s a really key
question you’re asking.” He responded,
Yes, because it’s a big investment for the faculty member who takes on a stu-
dent. And so if you work with the student so closely and then he walks away
or doesn’t make it, then it’s a waste of his time. And in a way, I mean, it’s our
mission to teach, but I’d rather spend my time teaching somebody who actu-
ally can continue my mission and then teach other students than somebody
who realizes, “It’s just too difficult. I can’t do it.”
The tendency to avoid risk was ubiquitous, but it was especially potent in
programs with small cohorts and those that offered full financial support
to their students. Under these conditions, each student has a relatively large
effect on completion statistics and represents a significant financial invest-
ment. Perhaps it should come as no surprise that faculty repeatedly compared
admissions to “risk taking” and “gambling.”17
Participants in this study approached admissions as risk taking in a va-
riety of ways, but for most, the risk was entirely uncalculated. The vast
majority were unaware of research about what affects graduate student
outcomes, and only one program had analyzed their own applicant data.
Nevertheless, I heard a common refrain that a weaker record of upper-level
coursework on student transcripts could be assumed to represent weak
preparation and a good chance that the applicant “may not know what
they’re getting into.” Several participants admitted their tendency to “feel
spooked” when an applicant’s profile called to mind a past student who
had struggled or ultimately not completed. Still others, consistent with re-
search on decision making in elite organizations, preferred to eliminate per-
ceived risks rather than estimate or calculate the level of risk.18
Another approach to risk was born of the tendency toward credential
inflation at the undergraduate level. As more and more students report
higher grades and GRE scores, not to mention other skills, research expe-
riences, and honors, some faculty interpreted the absence of such distinc-
tion as a sign of risk. Finally, four admissions chairs—whose programs were
ranked in the top five for their disciplines—voluntarily discussed the “luxury
of risk aversion” that their large, highly qualified pools permitted. Some
preferred conventional overachievers in the first round of admissions simply
because they could.
Intelligence and belonging in elite intellectual communities. An evaluative

script of intelligence and belonging also helps explain the enduring influ-
ence of GRE scores and student grades early in the review process. Many
associated “the numbers” with intelligence, and intelligence with belonging
in an academic community near the top of their fields.19 I asked each par-
ticipant in individual interviews what they thought GRE scores signaled,
and the first response from more than 50 percent of them included
comments like “innate ability,” “sheer intellectual horsepower,” and “native
intelligence.” These comments were reflected in admissions meetings, where
more than half of the comments about intelligence coincided with com-
ments about high grades and/or Graduate Record Exam (GRE) scores. In
linguistics, one professor commented, “Those are astronomical scores!”
with another responding, “And check out the stellar grades. There’s no ques-
tion she’s smart.” Following up with the department chair, he reiterated,
“Someone who does that well on the GRE is unlikely to be lame-brained.
They are likely to be smart.”
To the extent that faculty connect GRE scores to intelligence, these sup-
posedly objective, standardized criteria may take on a personal quality, for
intelligence is central to academic culture and faculty self-concept.20 I ex-
amine the nuances of how faculty conceive of intelligence in Chapter 5, but
for the sake of understanding conventional achievement as a standard of
merit early in the review process, the key point is that programs’ legitimacy
as elite intellectual communities in some ways hinges on the perceived in-
telligence of those whom they admit and hire. As the linguistics committee
chair put it, they used admissions to “reflect the view the department has
of itself” because the department’s character is “so determined by graduate
admissions.”
That faculty associated quantitative metrics with intelligence and be-
longing helps explain why faculty relied upon them so much when
making the first cut.21 Yet, a sense of the applicant’s belonging mattered
deeply in later phases of review as well. Explaining his committee’s deci-
sion not to admit a borderline applicant, a biologist commented, “[He
was] from a different planet and we were confident that this person was
not going to be one of us. He’s not going to be a full member of the sci-
entific community.” Committees often framed belonging in terms of “fit,”
which was a seemingly unassailable criterion and, in many cases, a bottom-
line consideration.
Convenience. Reading to this point, one might assume simple elitism and
outdated views of intelligence explained why faculty relied so much upon
GRE scores, grades, and institutional prestige. And, indeed, one sociologist
in the sample quipped,
This is an elite university and a lot of the people at the university are elitists.
{He laughs.} Simply said. So they make a lot of inferences about the quality of
someone’s work and their ability based on where they come from.
However, the tendency toward bureaucratic decision making suggests there

may be more to the story. Faculty didn’t simply rely upon grades, test scores,
and institutional prestige because they believed they were the best indica-
tors of achievement. They also relied upon these metrics because they were
convenient.
Time demands and incomplete information were the most frequently
cited answers to my interview question about what makes admissions eval-
uations difficult, and quantifying quality and judgment helped faculty
cope with these difficulties. Some committees went so far as to sort long
spreadsheet lists of applicants by scores and grades. In the physics program,
which drew hundreds of applicants, the chair shared with me his instruc-
tions to the committee for the first round of review: “In reports to me, just
summarize the test scores and GPA because I’m fully capable of reading a
spreadsheet, but I prefer not to have to read the entire file.” Especially in
large programs, some expressed regret that every file did not receive a truly
holistic review in the initial review, including this sociologist:
We receive so many applications, and we are always in a crunch with time—

always. And I have impressions that some of my faculty—senior members—
were simply looking at the GRE. They have a threshold such as, “If it’s not
over 700, I won’t read anything.” And that cuts usually two-thirds of the
applicants.
Drawing large numbers of applicants, faculty sacrificed the multidimen-

sional ideal of quality they believed in to process applications quickly and
efficiently. Letters of recommendation required reviewers to “read between
the lines” and “sift through the superlatives.” Many worried that personal
statements were “subject to gaming,” and not only were writing samples ex-
tremely time-consuming to review, but it was impossible to know how much
support students had received in preparing them. Given these ambiguities,
numbers like the GRE and grade point average seemed to carry an apparent—
if illusory—clarity, simplicity, and precision. A sociologist noted that unlike
work experience, research experience, or awards, which might not be avail-
able for comparing all applicants, “The GRE is something they all have in
common. The fact that it is common to everybody is really useful.” Valuing
scores and grades, then, was not only a matter of perceived belonging and
risk aversion, but also of ambiguity aversion. In the absence of clarity, and
given their shared values of risk aversion and intelligence, faculty settled for
the convenience of the entrenched standard when making the first cut.
The cultural meanings of conventional achievement—risk aversion, in-
telligence, belonging, and convenience—help explain why faculty often
maintain a standard of evaluation about which they feel some ambivalence
and doubt. In addition to the problem that this standard has only marginal
relevance for scholars’ long-term outcomes however, rigidly pursuing this
vision of academic excellence disproportionately prevents women and do-
mestic students of color from advancing beyond the first cut. As long as
systemic disparities in GRE scores and elite college enrollment persist, it
will be more difficult for programs to diversify if their initial standards of
quality rely heavily on these qualities.
Two pathways toward more equitable enrollments are available to fac-
ulty. One involves preserving the current admissions regime, waiting for
more equitable patterns in student scores and institutional affiliations, and
accepting that until that occurs, admission decisions may reproduce within
the academy the racial and gender inequalities present elsewhere in society.
Another pathway involves faculty stepping back from screening customs
and assumptions about student quality that are entirely taken for granted.
They could explicitly treat diversity as a dimension of quality from the be-
ginning of the review process, for example, or reevaluate how they think
about and weigh the GRE.22
The Short List: Merit as the Future

of the Discipline
Faculty revealed their preferences both in the initial criteria they used to
narrow the applicant pool and in the hairs they split to make final selec-
tions. Through holistic and individualized review, faculty adopted a much
broader view of merit in judging the short list. They opened their eyes to
the very minutiae in qualitative components of the application that initially
they wanted to avoid. Evidence of having mastered course material or a
bank of test items was insufficient to impress reviewers at this point, be-
cause the list was replete with such conventional achievers. Rather, they
looked for applicants who had a different relationship with knowledge.
Seeking potential for independent scholarship and innovation born of fresh
perspective, faculty combed through details to make a fundamental infer-
ence: Who might advance the future of the discipline?
Faculty from disciplines that were far afield intellectually shared the sen-
timent that such prediction was very difficult work. An astrophysicist
explained:
You’re really basing it on their potential to be great scientists. But it is much
harder—much harder. What you look at for a particular person varies from
person to person.
And an economist elaborated,
That’s so hard because at one level what you’re looking for is so easy. You’re
basically looking for people who are going to be first-rate researchers and
leaders in the field . . . who are going to be great economists. But figuring out
in a twenty-one-year-old what traits are predictive of becoming great intellec-
tual leaders is incredibly hard. You have to look for proxies at some level.
One of the proxies you look for is, is the person sufficiently quantitative and
mathematically sophisticated? . . . You want to find someone who is creative
and asks great questions. That is so hard to tell. You might look at their thesis
or maybe they worked as a research assistant or their recommendations. It’s
very hard to assess if someone is—{Pause} it takes a certain type of person to
be a researcher.
In classics, fluency in ancient languages mattered at this point as much as

mathematical training did in economics and research experience did in astro-
physics, but in each of these fields, professors struggled to identify talent.
An associate professor of classics described it as “guesswork”:
From the department’s perspective, we want to have somebody who is going

to finish the program and thrive—thrive here and go on and be a successful
faculty member somewhere. So we want to be careful about who we pick . . .
They’re really pretty young, just finishing college and a lot of students really
still are quite unformed in a lot of ways, still taking a lot of other classes on
top of their classics courses. What will that person be like in one or two years?
You know, where will they go? That’s where the guesswork comes in. I think
the more experienced—the more you teach, the more you know students, the
more you choose them and see how they turn out, the better feeling you have
for it. But there’s something. It’s not quantity. A score doesn’t tell you, answer
that question.
Across the disciplines, what faculty sought were displays of potential to

become what the Carnegie Initiative on the Doctorate called a steward of
the discipline, one who “considers the applications, uses, and purposes of
the field and favors wise and responsible application . . . [and] how to
foster renewal and creativity.”23 Reviewers assessed a range of personal
characteristics that might affect whether and how their skills and abilities
would be realized, including affective dispositions and noncognitive quali-
ties that they associated with exemplary scholarship. A physical scientist
explained (emphasis mine): “I look for people who have had research ex-
perience and show a great deal of enthusiasm for it. I look for creativity. I
look for energy. I look for possibilities, potential for innovation . . . a pas-
sion for research.” And in a social sciences program, a senior scholar born
outside the United States listed similar qualities:
If the person is intellectually creative, I think he or she will grow to be a really,

good scholar. To me that is the most important trait I look for . . . He or she
is really thinking. It’s not like—I know this language. I have that skill. I have
learned these courses, all this typical knowledge. I’m curious about all of
the new, interesting phenomena in society or this or that. And I want to explore.
I want to find what’s going on . . . That would make a good scholar.
Such qualities connote a relationship with knowledge that cannot be mea-

sured quantitatively but may be inferred from elements of the application
that many claimed were too onerous and ambiguous to emphasize in the
first round.
Professors were especially eager to glean these qualities from evidence of
previous research experience. Given the importance of conducting original
research to success in doctoral education, their hope was to evaluate re-
search potential as a function of prior accomplishment, as if they were
hiring a faculty member. A biologist commented,
I wanted to see some independent research because I think if you haven’t done
independent research then you don’t really know whether you have a passion
for it. And you kind of don’t know what you’re getting yourself into. This is
what it is all about.
Respondents were divided, however, over whether the most reliable testi-
mony of a student’s research experience and potential came from letters of
recommendation or from students themselves. In four of the programs, re-
viewers wanted students to personally communicate their experience with
research (in the personal statement and/or interview), and in astrophysics
a majority of the committee discussed the importance of “sell[ing] your
ideas.” The admissions chair explained,
Here, we judge them about how well they can explain things. So the selling
component of being an astronomer and selling your ideas is very important.
In theory it’s actually even more important because we discuss ideas all the
time to get new ideas. It’s all about the discourse and why are you doing what
you’re doing. Challenging the fundamental, physical concepts behind it—a
student has to be well versed as to how to defend their own ideas and how to
express them.
Committees also took seriously the alignment of student research inter-

ests with faculty expertise. Some described this alignment in terms of the
potential for mentoring, others in terms of the applicant’s capacity to sup-
port faculty research. In spite of being listed as an author on seven publi-
cations, for example, an applicant in the physical sciences was not admitted
under a rationale that included worries about the depth of her involvement
and weak alignment between her interests and the department’s strengths.
The importance of intellectual fit applied across the disciplines. “We don’t
want clones,” one humanities professor laughed. “But on the other hand, we
do want people who really do match reasonably well to what we’re good at.”
Defining diversity. During later rounds of review, faculty invariably pre-

ferred students who would add to the program’s or discipline’s diversity. If
an applicant with merit is broadly defined as one who deserves admission,
then diversity was not in tension with merit in the later phases of the review
process. Instead, diversity was simply one component of merit.
Why is diversity important to graduate admissions committees today?
Research suggests a variety of reasons that postsecondary institutions seek
diverse student enrollments and strive to be known as diverse. For one, the
educational benefits of diversity is currently the only constitutionally ac-
cepted rationale for affirmative action; therefore, institutions that utilize
race-conscious admissions often lean on the diversity rationale. Others have
found diversity a politically palatable rationale for efforts toward racial or
gender equity, because diversity can be defined with a sufficiently long
“laundry list” of identities to make anyone feel included.24
Diversity also has a clear business case in academe, given its documented
relationships with established metrics of scholarly excellence. Political sci-
entist Scott Page found that a randomly selected group of amateurs out-
performed a homogeneous group of experts in tasks that involve creativity
and problem solving, for example. And through an analysis of the citation
patterns of 1.5 million scholarly papers, economist Richard Freeman found
that papers with author lists that appear ethnically diverse are cited more
frequently. Freeman surmises that ethnically diverse authorship produces
greater diversity of ideas. In graduate education, studies also find that di-
versity contributes to innovation in research and to preparing professionals
who will work effectively in increasingly multiracial communities.25 Both
because diversity aids universities in fulfilling their educational and public
missions, and because the image of insufficient diversity may compromise
an institution’s reputation, many universities and their departments strive
to present themselves as diverse and are explicit about diversity as a goal
that they should use admissions and hiring to pursue.26
The faculty I interviewed and the committees I observed may not have
agreed about exactly which characteristics should “count” as diversity or
why it should be an informal criterion in this phase of the review process.
They did agree to disagree on the matter, though, and in keeping things
vague, made it possible to make the argument that by some standards they
were diverse, in spite of deep gender and racial inequities. Still, the default
mindset about diversity started with racial, gender, and, for a few people,
class stratification. Committees in all three universities adopted the lan-

guage of “diversity candidates” to refer to those who would be eligible
for university fellowships targeting underrepresented groups. Philoso-
phers and physicists expressed the most concern about gender parity. Some
tried to explain long-standing inequality by painting philosophy as “a
macho profession,” evocatively describing current levels of representation
as “abysmal,” “atrocious,” and “awful,” and so on.27 Physicists, how-
ever, felt that progress on gender equity has been made in recent decades.
One elaborated:
I think the drive for broadening gender representation has been quite suc-
cessful, and not without its problems but we have {Pause} we have very strong
representation of women. You can see that in the profession now. {Laughs}
You look at university presidents for example. I think the highest paid presi-
dent in the United States is a woman PhD physicist, Shirley Jackson . . . Anyway,
we now see in our research groups a very strong representation of women and
they’re doing well. So whatever risk we took there, we haven’t had to worry
there. With African Americans, it’s often a different story.
When I asked faculty to describe how their department defined diversity,

they most frequently started with race. Only a few mentioned socioeco-
nomic status, and then only in concert with race. A white female graduate
student on the political science committee was one such individual. Sighing,
she said, “I guess I’m like with {Pause} diversity meaning like class, race,
and gender. It’s a plus. I’m going with like Supreme Court language here
because I just talked about this with my students.” In sociology, a male fac-
ulty member of color expressed that, throughout higher education,
socioeconomic diversity is something I think we don’t talk enough about. It’s

not mutually exclusive with ethnic and racial diversity, but there have been
enough privileged minority kids out there that schools like Harvard and
Princeton can feel like they have a fairly diverse student body if all these kids
are, like, children of doctors and lawyers and second-generation Harvard
students.
In each committee I observed, at least one applicant was counted as a “di-

versity candidate” on the basis of class, socioeconomic status, or family fi-
nancial hardship.
Diversity may be thoroughly institutionalized as a value, but some forms
of diversity are easier to achieve than others, and this appears to have ram-
ifications for how faculty frame their department’s diversity efforts and
goals. Faculty in the disciplines with the deepest racial inequities were
also the ones to most vigorously defend the value of national origin and
disciplinary concentration as meaningful types of diversity. In economics,
several participants held up international students as evidence of achieving

diversity, in spite of persistent gender and racial stratification.28 A professor
of economics who had been born outside of the United States described
his department’s international character in relation to racial and ethnic
diversity:
It is incredibly diverse. There are people from every continent . . . we have

South American and Asian students, European students and Eastern European
students, U.S. students and Canadian students. We get Mexican students every
year. We’re spanning the globe with our students. We’re amazingly diverse in
people’s backgrounds. Everyone, of course, has had a lot of higher education.
But people do have diverse backgrounds and perspectives. And in that sense,
there is a lot of diversity. But in other dimensions of diversity, particularly the
ones you would traditionally measure in U.S. institutions . . . like how many
African Americans do we have? How many Hispanic Americans do we have?
How many women do we have? The situation is dismal. They’re a distress-
ingly small number at every level.
Similarly, a philosopher defended his department’s strong focus on inter-

national students for the diversity of viewpoint it introduces. He argued
that “new kinds of diversity matter” in an increasingly global graduate edu-
cation market, with its changing knowledge production system. In Chapter 6,
I will discuss an exception to this general rule: Applicants from China
occupy a tenuous position due to their strong representation in the pool
and the influence of model minority stereotypes, which carry both positive
and negative connotations.
Whether any personal quality that may introduce a unique perspective
should be considered diversity, or whether diversity efforts should focus on
interrupting historical patterns of inequality, are major questions in many
academic departments today. The tendency of the last few decades to ap-
proach the problem of inequity using a strategy and rhetoric of diversity
may actually help explain why progress has been slow. Research by Liliana
Garces found that statewide affirmative action bans have significantly re-
duced graduate school enrollments among students of color in all fields
studied except business, suggesting that, when permissible, race-conscious
admissions make a difference, and that a general interest in “diversity”
without explicit attention to race may be insufficient to realize equity
aims.29
Evaluative scripts of diversity. In the presence of divergent ideas about how

diversity should optimally be defined, faculty expressed clear patterns in
what diversity means to them. Their answers fall into two themes: obliga-
tion and opportunity.
Obligation. Many faculty whom I spoke with thought about at-

tending to diversity as a professional obligation of admissions work. They felt
a sense of duty, and in a few cases a personal commitment, to improving
the representation of women and racial/ethnic minorities in their programs
and disciplines. According to professors in both philosophy programs, the
attention by disciplinary professional associations to gender inequality in
their fields has heightened the sense of obligation to admit more women.
And in another sign that insufficient diversity may compromise organiza-
tional legitimacy, an awareness among philosophers that other predomi-
nantly male fields have made better progress in attracting and retaining
women has only sharpened their sense of responsibility. A male philoso-
pher elaborated: “We were sort of going backwards while the fields that
you think of as the most paradigmatically male fields were inching to-
wards something.”
However, obligations to diversity were conditional on another obligation
they perceived: protecting well-established standards of conventional
achievement. I started the data collection process expecting to observe some
debate about when diversity should be factored into evaluations, and was
surprised that it only came up in one of the ten programs. In the others it
seemed to be taken for granted that diversity should be weighed only in
the late stages of decision making. However, in the one philosophy program
where this was not the assumed practice, the question generated the com-
mittee’s first real disagreement. During a preliminary committee meeting
held to set norms for application review, a senior scholar asserted, “We
aren’t supposed to consider diversity in the first round. We’re supposed to
select out the top students and seek in the second round to weight the students
who meet diversity criteria.” Two of the five committee members—one male,
one female—questioned him on this. They argued that at least one member of
the committee should be responsible for “reading for diversity” from the start,
so as to ensure that someone would advocate for unconventional students
with exceptional promise. Their minority position was eventually overruled
on the basis of a belief that if diversity were defined broadly enough in the
second reading, that they could achieve the diversity they sought without using
multiple standards of evaluation in the first reading. They, like so many others,
eventually decided to consider diversity only if academic achievements were
roughly equal. A senior sociologist put it this way:
That’s one of the difficult issues, because I would say this: I try not to pay
too much attention. I try to admit students that are the best in my intellect
with no regard for gender or race. And so I understand, I perfectly understand,
that because the world out there is not equal, this simply reproduces inequali-
ties. I understand that. But at the same time—so I will say this—if there are
two students that are, in my view, equal on intellectual merit, then I will
prefer a minority. And I think it is the same—at least from what I observed—
it probably was the same for my other colleagues on the committee. It’s some-
thing that everybody pays attention to—the minority race, gender—for good
reason.
To “pay attention” to diversity in the committees I observed was not to

seek it out in a consistent way from recruitment to admission, but to put it
on the table among the many factors in play and perhaps use it to give se-
lect candidates a positive nudge. Reviewers felt they could uphold diversity
as a legitimate consideration in the review process only by using it to com-
pare applicants who had already surpassed an extremely high threshold of
academic accomplishments.
Opportunity. Other evaluative scripts for diversity reflect a deeply prag-

matic turn that centered on diversity’s organizational benefits, including
financial and intellectual opportunity and the opportunity to maintain a
competitive advantage with other elite programs.
One reason reviewers felt obligated to consider diversity only after can-
didates had sufficient achievement on conventional metrics was because
they believed other top programs in the discipline applied the same stan-
dard. Attracting accomplished, highly sought-after applicants from under-
represented groups may not be part of the algorithms used by U.S. News
and World Report or the National Research Council, but it has become a
way that programs informally evaluate themselves against one another.
Composing student bodies that reflect both traditional notions of academic
quality and demographic diversity was an informal aim of these highly
ranked programs.
Like other top recruits, high-achieving students of color (and, in some
fields, women) were viewed as organizational assets. When a philosophy
professor learned that an Ivy League–educated applicant was also from an
underrepresented minority group, for example, he commented offhand,
“Sounds like he’ll get in everywhere. Everyone will love him.” Conversations
about yield, in which committees worried together about where admitted
students would choose to enroll, were especially useful in revealing this
mindset about diversity. They shared their goals of enticing underrepresented
students away from peer institutions. Asking one other, “Can we get her?”
“Who are we going to get? It’s a gamble,” or “We’ll lose him to Princeton and
Caltech,” they constructed students as human resources in the struggle for
status among peer programs.
Inquiring into these comments in follow-up interviews, participants
framed African American and Latino student underrepresentation using a
set of popular beliefs about the challenge of faculty diversity—that quali-

fied scholars of color are a small pool and those individuals have many at-
tractive offers. An economist, for example, discussed his program’s struggle
to attract black students:
Gender is an issue in that we get good—we get top-notch women as well top-
notch men. Black—we get fewer blacks. It’s true. But we do try—in the past
we’ve tried to attract them. But then they get the same attractive offers from
Columbia and Yale and Stanford and Berkeley and so forth. So it’s a small group
typically who get a lot of attention.
A sociologist of color expressed an almost identical perspective:
We all kind of admit the same pool of applicants—the top ten departments.
Harvard’s going to admit them, Princeton’s going to admit them, Stanford, Co-
lumbia, Michigan, Wisconsin, and Chicago. And so we’re all fighting for the
same applicants, and there’s a lot to compete with, and you know there are a
couple programs right now that are just kind of doing phenomenally well in
terms of placement and training—just dominating sociology.
In a study of faculty hiring, however, higher education scholar Daryl Smith

debunked the claim that enduring inequality is due more to minority
scholars’ many professional options than to institutional decision making.
In a 1996 study of faculty diversity that focused on the trajectories of na-
tional fellowship winners, data from 298 in-depth interviews revealed that
only 11 percent had been through a hiring experience that was anything
akin to a bidding war, and that especially in the STEM fields, faculty of
color began to consider industry positions only after repeatedly feeling that
their credentials were not taken seriously by the institutions to which they
had applied.30 Moreover, diversity within the faculty, even at top-ranked
programs, has failed to keep pace with the increasing numbers of under-
represented PhD recipients in several fields, which creates conditions for
isolation and other negative experiences that recent PhDs may be eager to
escape.31 These studies suggest that the supply-and-demand and bidding-
war arguments used to explain lack of faculty diversity are “more rhetoric
than reality.”32
At the program level, where gatekeeping decisions are made, the aware-
ness that diversity represents an intellectual opportunity was one of the
three most common evaluative scripts for factoring it into evaluations of
those on the short list. Websites for all three universities’ graduate schools
proclaimed the importance of diversity for promoting excellence and the
mission of graduate education. When I asked an associate professor of clas-
sics how her program conceptualizes diversity, she said:
We welcome it. We want it. It’s so much. I think everybody is committed to it

on principle, and we know from past experience how much it enhances our
classrooms and our life in our department. And the university really supports
it, I think . . . Students who come from other backgrounds will have other, will
just focus on other, um, moments or ideas in the text. Also, students who have
different backgrounds will have studied different things. There’s a kind of de-
sire, there’s {Pause} more of a drive in some of the students who come from
different backgrounds.
Consistent with a desire for fresh perspectives on the discipline, diversity

considerations rarely stopped at whether a student identified with a
specified racial, gender, or other social category. Instead, faculty some-
times comparatively evaluated applicants within these categories against
one another, reading between the lines of the personal statement or let-
ters of recommendation to gauge how students’ identities would inform
their perspectives as students, and how these perspectives affect their
communities. Five of the six committees I observed discussed diversity in
this way.
Discussing a female applicant, for example, an astrophysicist asked his
colleagues around the table, “Is it enough to be a woman in science?” This
group of five men proceeded to catalog the ways that women on their short
list had used the personal statement to disclose gender-related experiences
in science. They offered informal judgments of what they thought about
these narratives, and speculated what it might be like to have these students
in the program, noting one applicant’s admission of self-doubt and anoth-
er’s desire to mentor young women because she had personally lacked role
models. In reference to an applicant who disclosed experiences with “teasing
and bigotry” from male teachers and peers, the graduate student member
of the committee quipped, “I’m less persuaded by that story,” and went on
to defend the teacher and suggest she might have an “axe to grind” if she
enrolled in their department. Ultimately, that interpretation was overruled
when an associate professor of color noticed in her personal statement
that she was now “taking action, organizing a lecture series” on women in
science.
Their conversation demonstrates the tendency for the identities and ex-
periences of “diversity candidates” to be scrutinized more critically than
those of applicants from historically privileged backgrounds. Previous
research on the hiring of academic administrators has come to similar con-
clusions. In “Held to a Higher Standard?,” Roberto Haro found that 80 per-
cent of subjects judging a Latino and 75 percent of those judging a white
female for positions in higher education administration cited the doctoral
institution as very important to their judgments, but only 55 percent of
those evaluating a white male cited it as very important. Another study

found that African American women applying for administrative positions
in universities were subjected to additional “filters” (sets of criteria) that
white men and women were not subjected to, including personal values,
personality, and whether she would be likely to “rock the boat or embar-
rass the employing unit or university.”33 My own research here found that
university websites and program publications communicated unequivocal
support for diverse perspectives as a means of enriching the learning envi-
ronment, but that at the level of file review, faculty inferred that some ap-
plicant perspectives would be more appealing than others.
Faculty in these programs also associated diversity with financial oppor-
tunity due to fellowships that all three universities made available to select
students who contribute to the university’s racial, gender, and socioeco-
nomic diversity. These fellowships were especially powerful incentives for
faculty when they enabled a program to admit more students than they ex-
pected. Some committees even adopted the university’s conceptualization
of diversity to ensure they would maximize the opportunity to add a few
fellows to their programs. When I asked an associate professor of sociology
how his program conceptualizes diversity, he admitted:
I think roughly in whatever way the university will pay for. Our conceptual-
ization is the university’s conceptualization, and that’s putting it a little harsh.
But because the university’s commitment is quite good and there are lots of
incentives, we don’t need to add any interests . . . So we’ll just do whatever.
We define diversity as the university defines it.
Depending on your perspective, his admission could be read as promising

or concerning. On the one hand, it suggested that their program’s commit-
ment to diversity is simply a matter of following the money, implying that
they had yet to adopt it as one of their own priorities. On the other hand,
his comments—and my observations in other programs—imply that insti-
tutional financial incentives can change faculty behavior. Whether such
incentives can help shift the composition and culture of the professoriate
so that, in time, such incentives are no longer needed, is a question that
remains open.
To summarize, when faculty evaluated students who made it to the short
list, they invariably considered both qualitative and quantitative elements
of the application, privileging students who seemed to have the research
experience, disposition, and fresh perspective to become leaders and inno-
vators in the field. At this stage, diversity informed broader ideas about
merit and was justified through evaluative scripts about obligation and
opportunity.
Diversity entered higher education discourse in the 1970s as an alterna-

tive basis for affirmative action programs.34 Presidential executive orders
during the previous decade had mandated affirmative action to remedy the
effects of discrimination, but in Regents of the University of California v.
Bakke, the Supreme Court challenged the constitutionality of the remedy
rationale. They ruled that the educational benefits of diversity, broadly de-
fined, are a compelling state interest that justifies the consideration of race
in narrowly tailored admissions policies. Amicus curiae briefs informing
that critical case in 1978 included one written by four of the Ivy League
universities, which has special relevance for graduate education:
A central function of the teacher is to sow the seeds for the next generation of
intellectual leaders, and this, indeed, is a main reason why many university
instructors find that an ethnically diverse student body helps them fulfill their
teaching roles. In short, we hope that by these efforts, the leadership of the
next generation—majority and minority members alike—will be the better, the
wiser, and the more understanding.35
Debate continues today over gender- and race-conscious admissions, the
appropriate basis for it, and the meanings of diversity. In 2015, eight states
had passed bans on affirmative action, but the U.S. Supreme Court and the
5th Circuit Court of Appeals upheld the University of Texas’s affirmative
action policy and the educational benefits of diversity rationale.36
In this uneven policy environment, the multiple definitions and meanings
that professors attach to diversity hardly come as a surprise. Their ideas
about diversity were sometimes principled and sometimes pragmatic, and
could be mobilized to serve a wide range of faculty interests. Yet even as di-
versity shaped conversations about individual, borderline cases, delaying
its consideration until the review of the short list meant that criteria asso-
ciated with conventional achievement carried the day in shaping access to
these selective doctoral programs.
Disconfirming Evidence: Risk Aversion and Academic

Preparation as Final Considerations
I close this chapter with an extended episode that illustrates how faculty
negotiated ideals of merit and diversity in practice. It also is one in a
handful of cases I observed in which questions of preparation and risk re-
surfaced in the late stages of decision making. Two such cases occurred in
linguistics. Toward the end of their second meeting, Peter, the committee
chair, raised the need to discuss two “diversity fellowship students” who had
attended a local college from which the department was striving to develop
a pipeline. He had previously described this recruitment effort to me in an

interview:
We decided to interpret diversity broadly, not just race and gender and all those
things. We have very few students from local colleges so we contacted a local
college and told them that if they have good undergraduates to apply here be-
cause there’s this image that we are snobby so there is no point to apply here. So
we had several applications from the local college and two made the short list.
And, indeed, last year they did admit one student from the college. This
year, although a majority of the committee had voted them through to the
short list in their first meeting, a majority also raised concerns over whether
one student’s research interests were clear and whether the other posed too
much of an academic risk.
After reviewing some basic information about the first of these applicants,
whom I’ll name Amaya, the department chair, Nancy, commented, “If we’re
serious about diversity, these are the students we need to take seriously.”
Interrupting, Lin asked, “Can I please speak up? We want diversity, but we
want excellence in diversity. And when I read that statement, it was all over
the place.” Another member affirmed this view, comparing her to others
in the pool. “We need evidence of research skill, which we had with both
of the last two applicants we discussed.” A third added, “I tried to talk
with her when she visited campus and couldn’t get a line.”
Just as it seemed energy was building against Amaya’s case, Denise, an
associate professor, raised the diversity-risk relationship. “In most cases, di-
versity will involve some degree of risk on our part. We have to not be so
risk averse that we miss opportunities.”
“That’s true,” added one of the people who had expressed concern just
a moment earlier. Then, in the first and only mention of an applicant’s race
or ethnicity I heard in this program, she continued, “And students from her
ethnicity are very unlikely to apply to graduate school.” She did not for-
mally advocate for the student on this basis, but rather reminded the group
that the opportunity to enroll a student from this group—especially in
linguistics—was rare.
They tabled the decision on Amaya, and moved to discuss the other stu-
dent from the local college, Nikki. Peter began by reading an extended por-
tion of a letter of recommendation, which closed with the words “I hope you
get to meet her.” He followed this reading with, “And we will. She’s coming
to campus this Friday for a visit.” Like Amaya, it became clear as they dug
into Nikki’s application that no one felt strongly about moving her for-
ward in the process.
Peter in particular sounded unsure about her, but Nancy highlighted the
extreme personal struggles the student had faced, concluding, “She might
be a bet, but it could be a good bet.” She also reiterated an earlier point. “If
we are going to increase diversity, these are the students we need to take se-
riously.” This perspective reflected a view I heard frequently: that some
bets are safer than others, and improving diversity requires decision makers
to identify applicants who contribute to diversity and have acceptable risk
profiles.
Tentatively, Lin asked, “What’s the diversity?” Denise and Nancy, who
by now were unified in advocating for Nikki, responded with “family fi-
nancial hardship.” The committee agreed at that point to leave her on the
list, but discussion about her and about building diversity continued.
Nancy said, “It will be good for the whole faculty to take a look at her
file. It seems pretty clear that she’s a risk, but if we’re going to increase di-
versity, we have to take risks.” Concurring, Denise added, “And she seems
like a good bet. Increasing diversity will also require these pipelines.”
Two weeks later, after the next faculty meeting, both Nikki and Amaya
visited campus for the recruitment weekend. However, in the final selection
meeting following their visit, in which the goal was to cut four people, the
committee decided not to admit either of them. In Peter’s description,
the committee came to this decision “in the first five minutes, with very little
discussion.” He elaborated:
There was universal agreement . . . No one thought they were up to the bar.
So that got us down to ten. We liked them both [but] we thought they didn’t
have the intellectual capacity to excel in graduate school.
I probed, “What was it during the campus visit that gave you the sense of
that?” He responded,
In the end they just didn’t have the right preparation. They were just, there
were others from universities with great, active linguistics departments where
they probably had taken two years’ worth of graduate courses already, things
like that. The local students just didn’t have that preparation. So they might
have been perfectly capable, but in comparison . . . We felt horrible. We really
wanted them, but it just wasn’t going to happen . . . I’m still wondering what
necessarily the right thing to do is. Both of them may have had the intellectual
ability, but they didn’t have as much background as the other students. So
maybe if there were a way to guarantee them a sixth year of funding so that
they can take an extra year to catch up, but that’s not something we have.37
I followed up with another member of the committee as well, and his nar-
rative corroborated Peter’s:
They were applicants that we were hoping to be able to take a chance on, that
we invited to campus. Sadly, once they got here we decided it would not be a
good gamble. Um, and in fact, I mean we really were disappointed because
we did personally like the people involved and a lot about them but again you
have to think about what you’re doing to their lives as well. {Pause} Some of
the sort of last stage of discussion was, which of the people . . . are likely to be
a good enough risk?
As Peter acknowledged and anyone who has participated in admissions

knows, “the right thing to do” is never a straightforward determination in
borderline decisions. Tough calls like these emerge only through a rough cal-
culus involving risk assessment, principles, emotions, implicit biases, and ex-
periences. They are, however, almost always made easier by consensus among
decision makers, as uncomfortable as they may be with their conclusion.
Bringing a critical eye to this episode raises questions about the assump-
tions underlying the committee’s decision. What is “the right preparation”?
What would it have looked like for these two applicants to be “up to the
bar”? Did those qualities exist in reviewers’ minds before they encountered
the applicant pool, or did they become salient only by comparing appli-
cants with one another? When gatekeepers conflate something as difficult
to judge as the “intellectual capacity to excel” with “the right preparation,”
what does it do for their reputation for “snobbishness”? A case like this
demonstrates that how admissions decision makers conceptualize, assess,
and act on “risk” can have significant implications for diversity. Risk aver-
sion pervades assessments of which applicants are worthy of the investments
of time, money, and energy that doctoral education requires.
Conclusion
A 6 on the GRE Writing section.

A grade of C– in Organic Chemistry II followed by a change in
academic major.
President of the Black Student Union.
When reviewing admission applications, each of these lines is likely to
elicit reactions and value judgments rooted in the meanings a person infers
from them. Those meanings are fragile however, subject to change if one
were to imagine it as a set of qualities from the same application. The perfect
writing score might elicit a different reaction if exchanged for a quantitative
score, and President of the Black Student Union may mean something dif-
ferent to a reviewer than President of Delta Delta Delta or Phi Beta Kappa.
Admissions work involves countless little value judgments made through
what behavioral economist Daniel Kahneman calls “fast thinking,” the same
mode of thought by which we automatically recognize and react to subtle
facial cues or hit the brakes to avoid a collision.38
I found that in the moments that faculty individually and collectively

judged applicant merit, that they instinctively drew upon evaluative scripts.
They wove together available information with cultural meanings, rooted
in shared values, to construct justifications for their applicant ratings and
admission decisions. In response to seeing GRE scores, institutional affilia-
tions, and students’ social identities, for example, faculty employed scripts
like risk, belonging, and opportunity to justify their evaluations. That scripts
of merit play out so quickly in the mind, sometimes without deliberate
thought, poses a challenge for decision making. Professors I spoke with rec-
ognized that the world is full of complexities that invalidate their instincts
and commonly held assumptions. They would be the first to admit that the
assumptions and stories embedded in prevailing scripts might be incomplete
or even inaccurate, especially as they concern individual applicants; never-
theless, they leaned on those scripts because they help to get potentially divi-
sive work done on the basis of shared cultural meanings and values that are
incontrovertible to the community.
In the early rounds of application review, faculty often misconstrued
merit and diversity as a trade-off due to the evaluative scripts they used
and the timing at which they considered key criteria. The excellent appli-
cant was one with a very high GRE score who had earned very high
grades from a well-respected college or university. Such individuals were
thought to have a low risk of failure and to “belong” in a department like
theirs. Just as importantly, faculty thought of these criteria as convenient
means for sifting through the pool. However, evidence of disparities in
GRE scores and enrollment in highly ranked colleges, and of grade infla-
tion in that sector, demonstrate that rigidly defining merit as “conventional
achievement” when sifting the pool is likely to perpetuate enrollment
inequities.39
The findings in this chapter have several implications for admissions de-
cision makers that I will introduce here, in addition to discussing in greater
detail in the Conclusion chapter. First, the interpretation and use of GRE
scores by faculty decision makers requires attention. Female, black, Latino,
Native American, and low-income students may have lower mean GRE
scores,40 but those scores are poor predictors of long-term outcomes, and
research studies have been mixed about whether scores predict the success
of different groups equally well.41 Therefore, ETS advises that admission
decision makers treat GRE scores not as a singular window into a student’s
future of success or struggle, but rather that they interpret scores (a) in the
context of the student’s profile of identities, native language, and previous
educational opportunities, and (b) as one piece of information in a more
holistic review.42
Professors would do well to assess for themselves—either individually or

as committees—the engrained scripts about merit that shape their judg-
ments, which are crucial to cultures of faculty gatekeeping. Take the risk
aversion script, for example. For programs to be self-critical about their use
of this script might mean taking stock of the skills required for doctoral
students’ introductory coursework. They could also gather data to examine
which doctoral student characteristics are associated with success (as they
choose to define it) in their program or discipline, so as not to operate on
assumptions about risk that are invalid for their context.
Like merit, admissions decision makers will give careful thought to
defining what they mean by diversity and equity, and recent scholarship
suggests that academic organizations will benefit from a critical look at di-
versity, itself, as an aim. For no matter how “elaborate [the] organizational
infrastructure for diversity,” sociologist Ellen Berrey notes, it “beautifully
symbolizes inclusion but only sometimes, and only partially, undoes racial
hierarchies.”43 Higher education scholars Alicia Dowd and Estela Bensimon
have also argued that colleges and universities need to enact equity as a goal
distinct from diversity, and specifically, to build organizational capacity for
talking about race. They note, “The majority of diversity initiatives are di-
rected at improving human relations and tolerance rather than on
achieving equity . . . which calls for directing more resources and support
toward students who have greater needs due to social inequality.”44 I ob-
served that many faculty felt a strong obligation to remediate underrepre-
sentation of African Americans and Latinos in their disciplines and the
professoriate, but that they also felt (a) a strong obligation to uphold the
standard of conventional achievement, (b) unsure about what the current
legal parameters were for considering or talking about race in admis-
sions, and/or (c) discomfort raising race, because their colleagues would
regard it as a controversial issue. In short, my findings corroborate Dowd
and Bensimon’s conclusion about the need to build organizational capacity
in this area.
Committees wanted to construct “balanced” cohorts and entice well-
pedigreed students of color, and they believed that this combination is now
a status signal among top-ranked programs.45 Yet several behaviors sug-
gested that racial equity and diversity were not integral to their vision of
excellence, but at best, secondary to and separate from it. Under the cur-
rent review process, many students whose diversity contributions might
have been considered assets in later rounds of review had already been fil-
tered from the pool. An important conclusion of this chapter is that when
a committee factors diversity into recruitment and admissions may matter
much for the outcomes they achieve.46
In these core disciplines of the academy, and in both phases of review,

faculty acknowledged that doctoral students are critical members of their
intellectual communities—not only because of what they would accomplish
as individuals, but because their presence, perspectives, and effort would
affect faculty productivity and the program’s stature. Selecting students was
therefore not only about predicting who would be successful in the future.
It was about creating their programs’ futures by selecting new members
who would uphold the core, identity, and status of the group. Sociologists call
this organizational boundary work, and as subsequent chapters will examine
in greater depth, faculty in these programs did not privilege applicants
who would push their elite communities in bold new directions. Instead,
they favored students who would “fit” and add a bit of fresh perspective.
An unspoken program identity—that of elite academic community—
clearly shaped admission preferences in these ten programs as well, but what
makes the politics of quality in graduate admissions so intriguing is that
faculty negotiated much more than their program’s priorities in choosing
among students. As I’ll discuss next, they also negotiated disciplinary norms
(Chapter 3), their own personal identities (Chapter 4), and ideas about intel-
ligence (Chapter 5), as well as unexamined assumptions about the differ-
ences between American and international applicants (Chapter 6).
chapter three
Disciplinary Logics
To be a Shakespearean scholar, to absorb oneself in black holes,
or to attempt to measure the effect of schooling on academic
achievement—is not just to take up a technical task, but to place
oneself inside a cultural frame that defines and even determines a
very great part of one’s life.
— Clifford Geertz
D octoral programs have long been regarded as “the academy’s

own means of reproduction.”1 They sustain academic departments and
disciplines by developing subject matter expertise, awarding credentials, and
socializing rising scholars to norms, identities, and affective dispositions
such as a sense of “stewardship” for one’s discipline.2 And although the ca-
reer paths of PhDs now vary widely, departments like the ones in this
study saw themselves as leading producers of knowledge for their respec-
tive academic disciplines. For them, doctoral admissions was about identi-
fying rising stars, and doctoral education was about preparing a corps of
disciplinary experts. Through the characteristics of those whom they ad-
mitted, faculty brought to life imagined academic communities.
As Clifford Geertz noted, to do the work of one’s discipline is more than a
“technical task.” The disciplines play an integral role in all of faculty life and,
as I will argue in this chapter, were a critical source of faculty tastes. Cul-
tural sociologists have portrayed academic disciplines as social worlds
unto themselves,3 and education scholars have examined disciplinary vari-
ation in such aspects of doctoral education as mentoring, persistence to de-
gree, and research socialization.
In my case, I expected that faculty in different fields might use or weight
admissions criteria differently. I was also curious to learn whether commit-
tees in fields with relatively unified intellectual paradigms would come to
consensus about criteria, ratings, and decisions more easily than committees
Disciplinary Logics 75
in fields known for using a greater breadth of theories and methodologies.

Within the social sciences, for example, economics has a reputation for
being more paradigmatic than political science. In the humanities, philos-
ophy has a more centralized intellectual tradition than linguistics,4 a once
centralized field into which other disciplinary perspectives and quantitative
research have made significant inroads in recent decades. To ensure that
my sample offered a broad view of core disciplines, I selected three disci-
plines each within the humanities, social sciences, and natural sciences.
Then, to compare whether disagreements on admissions committees varied
by the extent of disciplinary consensus, I chose fields within each broad sub-
ject area that have more and less consensus about preferred theories and
methods for advancing knowledge.5
The Latin for discipline, disciplinare, means to train or to teach. And,
indeed, the patterns of evaluation in programs from centralized, hierarchical
fields like economics, philosophy, and physics exemplified how disciplinary
socialization can train faculty judgment into specific forms. In these
high-consensus fields, admissions preferences and practices exhibited clear
trends. I also observed that, in such fields, the raw material of disciplines—
the prevailing theories, methods, epistemologies, and practical priorities—
produced shared logics by which committee members could legitimize
preferences and decisions to one another that outsiders might find debatable.
A key finding of this study is that, like deliberative bureaucracy, strong
disciplinary logics facilitate decision making while discouraging the selection
of students with unconventional profiles. Disciplinary logics bring shared
priorities into focus, but those shared priorities tend to privilege applicants
who “fit” the disciplinary culture as currently constituted.
In fields with weaker intellectual consensus, such as political science and
linguistics, disciplines also trained individual judgment, but underlying de-
bates in the field manifested in committee deliberation. One participant
called these incidents “worldview moments.” In these moments, the absence
of a common disciplinary logic that might have helped facilitate consensus
became clear to the committee. Admissions in fields like political science and
linguistics therefore involved more disagreement—about applicants, rat-
ings, and what the appropriate process for making decisions should be. With
less underlying common ground to stand upon, the process resembled in-
terdisciplinary review and involved some unique features such as deference
to seniority and the use of serial impressions to describe candidates.
To illustrate these trends, in this chapter I explore in depth two high-
consensus disciplines, economics and philosophy, and then present an
extended episode from deliberations in political science, a low-consensus
discipline. Through these portraits, we see how admissions evaluations
become patterned by disciplinary content and consensus. In Chapter 4 I

highlight an additional feature of review in low-consensus disciplines: pref-
erences among individual reviewers for applicants like themselves.
Economics
Stavros Ioannides and Klaus Nielsen argued that economics is “now widely
seen as the most advanced of the social sciences with its mathematical for-
malization, public prestige, and Nobel Prize awards.”6 Justin Wolfers, him-
self an economist, recently dubbed it “the queen.” Across macro and micro
levels of analysis, economics possesses a cultural coherence around numer-
ical evidence and algorithmic thinking that is stronger than any other so-
cial science discipline. In both the theoretical (mathematical modeling) and
empirical (statistical analysis) approaches to economics research, scholars
rely as much upon numbers as words to communicate their findings to one
another. One priority in economic scholarship, which is reflected in their
emphasis on mathematical reasoning, is precision. Economists have made
inroads into other disciplines by spotting topically related research whose
methodological or theoretical precision could be improved with economic
ideas, reasoning, or analytic approaches. Some argue that their doing so
has simultaneously weakened the intellectual consensus in target fields
and broadened the intellectual jurisdiction of neoclassical and Keynesian
economics. However, a sharp rise since the early 1990s in “heterodox eco-
nomics,” such as behavioral economics, and in membership in groups like
the International Confederation for Associations for Pluralism in Economics
suggests that the field’s orthodoxy may be weakening and/or that other dis-
ciplines shape economics, just as economics is shaping them.7
The viewpoints of economics faculty whom I interviewed epitomized how
common theories, methods, and practical priorities could combine to create
a common logic by which professors could legitimize admissions evalua-
tions and practices that others might question. Decision-making tactics and
priorities reflected foundations in theories of rational choice and human
capital, methods that are mainly quantitative, and a shared priority of main-
taining prestige in the social sciences. Like other disciplines, how econo-
mists represented their values in admissions revealed preferences that were
thought of as self-evidently preferable rather than the result of common
socialization and professional practice.
Relative risk aversion. In a discipline that prizes its status in the social sci-
ences and a department that prizes its standing in the discipline, the eco-
nomics program I studied designed its selection process, curriculum, and

funding model to minimize threats to their status. This orientation com-
ports with one of the most treasured principles of neoclassical economic
theory: relative risk aversion. Their pool was extremely large (about 800)
and academically qualified, so members of this committee routinely denied
admission to anyone whose long-term success they doubted, perceiving such
failure as a financial or reputational risk to the program. As one committee
member explained, “Because we run the program to be highly ranked . . .
basically we have a good chance of getting people if we accept them. We
have the luxury of getting to look at the pool and say, ‘Let’s just take the
people we think have the best prospects as an economist.’ ” Another faculty
member said,
The people that the top programs admit—we admit basically the same people . . .
[Programs in the next tier] don’t want to admit the same people that we admit
because they’re going to lose them . . . I think we have the privilege of taking less
risk than other universities of lower ranking.
He went on to acknowledge that his program’s prestige-focused, risk-averse

orientation had consequences for judgments of applicants with traditional
versus nontraditional trajectories:
If we have a letter from someone who had a difficult background, then I think
that would count a little. I wouldn’t say that that would be the reason why
you get admitted. Based on the experiences we had in the past, these things
are even less important now than it used to be, when some deviation from the
rule that “We only admit the best” was made and it didn’t work out well. So
based on that experience, I think the committee will be more cautious.
Successful organizations generally tend toward risk aversion when facing

the prospect of loss,8 but in this program the intersection of prestige as a
practical priority with risk aversion as a widely shared tenet of economic
theory justified relative risk aversion as a modus operandi of decision
making. The other programs in the sample, which were also highly ranked
for their fields, also revealed risk-averse approaches to admissions, but none
so overtly as in economics.
A foundation of quantification. Perceptions of “the best prospects as an

economist” cannot be understood apart from the quantitative foundation
of the field, for as one participant put it, “Mathematics is the language in
which to explain your argument or your story.” On my asking what intel-
ligence in economics looks like, an associate professor responded, “Many
people who become successful . . . first of all, they all have some basic level
of mathematical sophistication.” Interestingly, economists not only looked
for quantitative expertise; they relied primarily on quantitative measures

to gauge it.
The trend toward quantification seems only to be strengthening with
methodological developments in economics. As one associate professor ex-
plained, “At some point in the middle of the last century, the nature of
discourse among economics became based around models and using math-
ematics for precision and to precisely state hypotheses.” What counts as
high-quality economic research must be cutting-edge, he continued, which
demands cutting-edge statistics. A junior faculty member also volunteered
this position, “Over the decades the field ages, and there’s more and more
work you have to know in getting to the frontier. The frontier is more tech-
nical and deeper.” As the statistics have advanced, so have mathematical
prerequisites for admissions. Today that preparation consists of “ideally a
double major in math and econ,” as one professor summarized.
Thus, this economics program was not unlike the top 25 in the field for
requiring a perfect or near-perfect quantitative GRE score and for giving
careful attention to the score’s percentile, due to the concentration of scores
on the right tail of the distribution.9 They interpreted the score as signal of
competence with skills, and competence as a signal of how the student
would fare in coursework. A junior professor, new to admissions, laid it
out for me:
They do look at whether the student is going to struggle. Technical expertise
is a fact. The trouble is, if a student doesn’t have enough mathematics, they
don’t see the economics the first year. They just see it as mathematics, and it’s
a real problem. What you want is somebody who has enough to distinguish
the math from the economics. I do believe the admission committee does try to
determine whether a student is going to struggle and is prepared. A master’s
program is not necessary but it helps.
Economists’ priorities of near-perfect GRE scores and advanced academic

preparation in mathematics manifested foundations of their field in rational
choice and quantitative evidence. These criteria also supported the disci-
pline’s interest in maintaining its prestige in the social sciences. Together,
these practical, theoretical, and methodological priorities guided admissions
practice and converged in a logic to which they collectively deferred in jus-
tifying their judgments—to me, to one another, and to themselves.
Philosophy
In contrast to economics, “good philosophy,” according to an associate pro-

fessor of philosophy, consists of “charitable conversation with controversy.”
The heart of philosophical scholarship is analytically dissecting ideas about

fundamental questions, and reconstructing them in strong arguments. Having
survived thousands of years, philosophy is among the oldest of disciplines.
Over its history, the topics philosophers have taken up reflect the diversity of
interests of their societies and cultures. Today, there are six primary strands
of work within the field: ethics, epistemology, metaphysics, philosophy of
language and mind, philosophy of science, and aesthetics. Yet across time,
culture, and content, the core of the field has been defined and preserved as
an ongoing conversation about the nature of knowledge and reality.10
Biglan’s original disciplinary classification system named philosophy as
a soft discipline relative to all other fields,11 but it is among the most para-
digmatic in the humanities. The most well-known series of teachers and
students in early Western philosophy—Socrates, Plato, and Aristotle—
initiated modes of discourse that continue to guide Western philosophers’
communication with one another, and most aspects of disciplinary social-
ization in philosophy departments in American universities reflect these
scholars’ influence. The Socratic method, in which hypotheses are eliminated
discursively by identifying ideas that produce contradictions, remains cen-
tral to philosophical discourse. Aristotelian logic and its syllogisms are an-
other enduring branch and method, and training in the use of syllogisms
to argue one’s point is fundamental to the preparation of philosophers.12 It
may not come as a surprise that philosophers’ top priority in evaluating ap-
plicant files was the writing sample or that they engaged in more prolonged
and intense deliberations than did the committees I observed in other disci-
plines. Those deliberations covered their evaluations of applicants, the ra-
tionales for those evaluations, as well as self-critical analysis of their own
ability to judge and select students fairly.
Interpreting the writing sample. Philosophers felt that the writing sample
evinced the substance, skill, and style of what an individual would bring to
the conversation of philosophy. They revealed the crucial role of the writing
sample in several ways. In our interviews, all members of the philosophy
committees I observed at both institutions volunteered to me that it was
“very important” or “the most important” component of the application.
In meetings, they referred to applicants not by name, but by the topic they
had chosen to explore in the writing sample. Philosophers also devoted
more time to discussing writing samples than any other component of the
application, and it became quickly clear to me that applicants needed to
receive a positive assessment of the writing sample to have a chance at an
admission offer. A senior professor explained his approach to evaluating
the writing sample:
When you have eighty of these things to read and you come in in the evening,
you want to get home and watch TV and go to bed. So, you’re looking for a
reason to stop reading. And so I tell [students] what they have to do in their
writing sample especially—they have to try to counteract that tendency. They
have to give me something to make me keep reading.
What holds a philosopher’s interest may vary, of course, and participants

from both philosophy programs were conscious of stylistic differences
among traditions that shaped their interpretations of writing samples. An
associate professor regarded one student’s writing as “murky” and “com-
plexified,” only to learn from a senior committee member that these quali-
ties “come with the territory” in the latest expressivist writing. Admitting
her distaste, because it was not like “her own sort of philosophy,” she de-
ferred to the judgment and expertise of another committee member. To
guard against uncharitable assessments of writing samples based on style,
one committee structured the review process to ensure that each applicant’s
writing sample was read by individuals with expertise in that subject area,
bringing in readers from outside the committee as needed.
Sensitivity to subjectivity. Philosophers revealed in other ways their culture

of self-scrutiny and their emphasis on interpretation as a way of knowing.
They exhibited high levels of introspection about the evaluative process,
for one, and displayed sensitivity to the implications of a process laden with
subjectivity. They tested the defensibility of their judgments, challenging
each other when the grounds for an outlying rating seemed weak. Yet they
almost always crafted these challenges in an exceedingly respectful manner—
as a matter of inquiry, not attack. An associate professor explained the dif-
ference between persuasion born of rhetoric and the philosophical argu-
mentation they practiced as a committee:
Rhetoric is not a craft. It’s an empirical knack that enables you to please people.
But philosophy is [a craft], right? And I think Plato is right. There is such a
thing as learning to give good arguments for what you think. And it’s different
from the skill—such as it is—of just persuading people.
As in their scholarship, philosophy committees approached admissions

as “charitable conversation with controversy.” They were the only commit-
tees to grapple with certain aspects of subjectivity that nearly all other
programs also experienced but seemed to take for granted: the role of emo-
tional responses and quantification of their judgment.
Enthusiasm and other emotions. Enthusiasm for applicants affected

individual ratings and committee deliberations across the disciplines, but
philosophers were the only ones to explicitly acknowledge it. Even one
committee member’s enthusiasm could determine an applicant’s chances
by shaping whether she advocated for the applicant, and this trend be-
came especially apparent in discussions of borderline applicants. When I
asked one philosophy professor what helped the committee decide among
relatively qualified applicants, he admitted:
Well it’s very hard. Let me try to answer your question in a way that may make
the whole process seem a little less credible—but that’s the way it is. What hap-
pens is that one or more members of an admissions committee will catch
some enthusiasm for some applicants but not others. If you go back over, if
you are forced—practically, they’d have to grab you by the scruff of the neck
and force you to look again at a recommendation that you have consigned to
oblivion. There probably are some of them where if you were to say, “Why
weren’t you enthusiastic about the person?,” I’m not sure I could answer. And
I’m sure every year there are . . . people who could easily have been among
the top ones but nobody picked up on them. They just didn’t light a fire under
anyone.
Certainly, graduate admissions is not the only type of elite selection in which
decisions can come down to feelings and gut responses. In an observational
analysis of hiring for investment firms, sociologist Lauren Rivera used Ran-
dall Collins’s theory of emotional energy to explain how “evaluators used
their own emotional responses to candidates as both metrics of merit and
mechanisms of candidate selection.” So influential is emotional “buzz” gener-
ated on behalf of a candidate that it can trump traditional metrics such as
achievement, performance, or ability.
Resigned to quantification. Although every program in this study played

“the numbers game”—collecting ratings of applicants and aggregating them
to simplify selection—the two in philosophy were the only ones to debate
whether it was justified and how it should be executed. Unlike most other
committees, for example, they were explicit about what their rating scales
meant. One committee used the following scale: 1 = admit; 2 = maybe;
3 = wait list; and 4 = a bit of a risk. The other committee used a five-point
scale, and distinguished 1’s from 2’s by the enthusiasm of their own re-
sponses.13 As in other fields, philosophers relied upon quantification to
transform an applicant pool that was diverse along countless dimensions
into a single distribution of admissibility, yet they were deeply uncomfort-
able with the false precision these ranked lists implied.
Philosophers actively discussed subtleties of evaluation that remained
unspoken in most other disciplines, such as the link between enthusiasm
and evaluations and the imperfections of numerical metrics of applicant
excellence. Whereas economists’ comfort with quantitative metrics allowed

them to perceive applicant quality in black and white, it seemed fitting that
during one philosophy committee meeting, every member of the group
wore gray.
Comparing the Cases
Pairing economics and philosophy clarifies how the prevalence of posi-

tivist or interpretive epistemology within a discipline is not isolated to
scholarly research, but can affect how professors carry out administrative,
evaluative work. Philosophers vary in their analytic foci, arguments, and
conclusions, but are distinguished as a group by their method. Similarly,
economists share a strong belief in the value of quantitative analysis and
are committed to making social science as scientific as possible. Though
of course there is some epistemological diversity within both fields, the
predominant orientations were clearly recognizable in their approaches
to admissions. This conclusion fits well with the findings of previous
social research on the power of epistemology in shaping disciplinary
cultures.14
In addition to epistemology, however, other disciplinary cultural quali-
ties were evident in the patterns of faculty evaluation I observed, including
common theories, vocabularies of practice, and practical priorities.15 Table 4
points out these qualities in economics and philosophy, as well as in
physics—the other high-consensus discipline I studied. Given the important
role that admissions and other forms of faculty gatekeeping play in deter-
mining the next generation of the academy, I contend that we should not
interpret evidence of the disciplines in admissions work as a spillover ef-
fect from the intellectual domain to the administrative, but rather as sup-
port for the view that, for professors, administrative and evaluative work
is inherently intellectual. This view is consistent with the perspective on dis-
ciplinary cultures expressed in the quotation by Geertz at the start of this
chapter—that disciplines establish “cultural frame[s]” that shape all sorts
of activity in the lives of scholars.
Particularly in more paradigmatic fields, to work from the vantage point
of a particular discipline leads one to see the world in a particular way.
Committees representing disciplines with stronger intellectual consensus—
such as economics, physics, and philosophy—understandably displayed
more agreement about what counts as a legitimate admissions preference
and less disagreement in their judgments of individual applicants. In the
social sciences, economists tended to draw upon a shared logic more than
Table 4 Indicators of Disciplinary Logics in High-Consensus Fields
Theory Epistemology & Methodology Practical Priorities

Indicator of Indicator of Related Indicator of
Discipline Disciplinary Logic Related Preferences Disciplinary Logic Preferences Disciplinary Logic Related Preferences
Economics Rational choice & Risk aversion; Positivism; deep “Mathematical Concern for Use of admissions
human capital applicants with methodological sophistication”; prestige in criteria with
theories strongest GRE & frontier applicants with social sciences widespread
academic math degrees & legitimacy (e.g.,
preparation near-perfect college prestige &
Quantitative high GRE scores)
GRE scores
Philosophy Valuing prudence Risk aversion; Interpretivism Emphasis on Desire for Applicants with
deliberation letters, personal “distinguished” pre-professional
about as many statement & philosophers writing samples &
applicants as writing sample pedigrees from
possible elite institutions
Physics Goal of discovery Efforts to identify Highly technical Applicants with International International
“spark” and work significant, collaborations applicants with
“diamonds in the relevant in English strong English
rough”; inter- research skills
views that go experience
beyond
credentials
political scientists, and in the humanities, a common disciplinary logic was

more apparent in philosophy than in linguistics.
Yet some disciplinary qualities are admittedly difficult to disentangle from
scholars’ individual identities and program prestige. There was, for example,
the way in which status concerns were engrained in the economics depart-
ment, and the seriousness and pride with which philosophers discussed the
distinctiveness of their own training and work. It is no coincidence that
economists admitted they must resist the temptation to be “dazzled” by ap-
plicants with elite pedigrees, or that philosophers frequently used the word
“prudent” to describe their ideal standard of judgment and “distinguished”
to describe their ideal applicants. Such default tendencies reflected under-
lying disciplinary values, which may not be fully independent from personal
or program values. Regardless of their origin, revealed preferences were for
both specific qualities of desirable applicants and specific practices for iden-
tifying those applicants.
Struggling for Consensus
In contrast to the high-consensus disciplines, which were natural sites to

observe faculty using a common disciplinary logic to build consensus, lower-
consensus disciplines were ideal for seeing how divergent disciplinary logics
could lead faculty to very different interpretations about the same appli-
cant’s admissibility or criterion’s appropriateness. In previous research on
academic disciplines, some scholars have found that disciplines with
competing schools of thought struggle to achieve collective intellectual
progress. For example, scholars in lower-consensus disciplines have a harder
time building upon each other’s findings, are more likely to have turnover
among department chairs, and are less likely to obtain significant research
grants.16 In this study, committees in lower-consensus fields revealed how
individual tastes can shape collective decisions, and were key sites to ob-
serve three other mechanisms of evaluation: serial impressions, worldview
moments, and homophily. I’ll examine the first two ideas in the remainder
of this chapter before devoting Chapter 4 to the third.17
Serial impressions. In fields with weaker disciplinary logics, faculty ex-

plained their ratings of applicants to other committee members using what
I call serial impressions. Drawing on a seemingly bottomless arsenal of ad-
jectives, some faculty sketched their judgments and rationales using long
lists of casual judgments that provided a glimpse into the personality of the
reviewer as much as the qualities of the applicant. It might sound mundane,
even self-evident, that faculty would use adjectives to describe applicants.

But the adjectives faculty chose were revealing: scholars used clusters of
words that reflected their particular tastes and temperaments. Furthermore,
though they stated their impressions as individual responses to individual
applicants, these impressions were often given as much weight in delibera-
tions as universal criteria such as grade point average, work experience, or
positive references.
In political science, for example, the committee maintained an online,
password-protected spreadsheet for tracking their responses and ratings for
each applicant. There was a row for each applicant with corresponding col-
umns that indicated one’s subfield, scores, and comments from each reader,
readers’ average ratings and standard deviations, and the final decisions
on each person. Throughout the file review period, this spreadsheet was
available online for committee members to enter their ratings and comments.
Then, at the meeting, they projected it on a screen for common reference,
and I was struck to see the comment fields. The notes from two readers
about two of the applicants contained: “mature, sophisticated, honest,
stunning, stellar,” and “worry about fit, lukewarm, passionate, committed,
thoughtful, moving, and able.” A few rows down, notes about an applicant
who was ultimately admitted included, “Killer backstory” and “quirky,
which I like.”
In his introductory comments to the group, the political science commit-
tee’s chair noted that one function of their discussion would be to “help
calibrate one another’s comments.” This calibration proved critical as it be-
came clear that some committee members were more effusive than others
in their written praise, and harsher in their judgments, and that at least one
did not provide any notes at all. Moreover, some committee members were
more likely than others to accept their colleagues’ serial impressions—
devoid as they were of reference to the evidence that prompted them—as
valid or trustworthy depictions of the applicants.
In other low-consensus disciplines, a similar use of serial impressions took
place through discussion, though not always with a shared record of ev-
eryone’s notes. In classics and linguistics, most discussions of applicants
took a predictable structure. Faculty would open the discussion of a new
individual with one or two expressing their overall impressions (such as “He
wasn’t the most exciting”). The impressions would be followed by an at-
tempt to jog others’ memory about who the person was through an institu-
tional affiliation or research interest, and another round of adjectives to
summarize the profile (“naive,” “creepy suck-up,” “amazing,” and so on).
Then they would then move into a discussion of the writing sample, letters
of recommendation, and/or personal statements, which were summarized
by invoking several more vague judgments (such as “not entirely lucid,”

“brilliant,” “misguided,” “pretentious”). Discussion would continue, often
comparing the applicant to others in the pool, until at some point the admis-
sions chair would gauge the energy and emotion around the table to dis-
cern what action to take (“There doesn’t seem to be much enthusiasm
around the table”; “So we’re not very enthusiastic”; “Let’s wait and see if
anyone more interesting comes along”). By reiterating individuals’ impres-
sions of the applicant and the file details that inspired those impressions,
committee members resolved their differences of opinion and developed
working conclusions that guided their decisions. Building and attempting
to justify their serial impressions was a process for translating individual
judgment into collective consensus.
Worldview moments. In programs representing lower-consensus disci-

plines, intellectual differences did not prevent committee members from
getting along well on a day-to-day basis or even from becoming friends. A
quantitative linguist brought coffee for a qualitative colleague on the com-
mittee for their three-hour admissions meeting. An experimental astrophys-
icist and a theoretical astrophysicist spent time together on the weekends
with their families. Within departments, what separated scholars from dif-
ferent intellectual traditions was not so much a “gulf of mutual incompre-
hension”18 as a difference in intellectual worldviews, despite residence in
the same organizational space.19 Yet, because disciplinary and evaluative
cultures are not independent from one another,20 collective evaluation, es-
pecially defending one’s ratings and advocating for particular individuals,
may prime scholars in low-consensus disciplines to recognize and engage
worldview differences that most of the time remain comfortably latent.
I observed that weaker underlying intellectual consensus manifested in
what a political science participant astutely called “worldview moments,”
when scholars realized that the differences in how they evaluated an
applicant stemmed from different assumptions about intellectual, theoret-
ical, and metaphysical matters. They recognized that the absence of con-
sensus delayed their progress on decisions. They also understood that
agreeing to disagree would not move them closer to a resolution. In these
cases, the basis for deliberations often shifted to power relations within the
committee and program. For example, if a reviewer’s advocacy was grounded
in disciplinary assumptions that were different from those held by senior
members of the committee or by the committee’s majority, it was unlikely
that the advocacy would bear fruit. Worldview moments provoked some
of the strongest disagreements I observed and, in some cases, led committees
to settle for decision by majority vote in lieu of consensus. I offer an extended

illustration of one such episode in political science.
Worldview Moments in Political Science
Political science can be considered a low-consensus field, with subfield

concentrations that range from political philosophy to mathematical
modeling. The program I studied, however, has been known for its strength
in statistics, with one participant remarking to me with pride, “We are a hard-
line quantitative program.” Indeed, all doctoral students except those
admitted to study political theory were strongly encouraged to take a se-
quenced curriculum of statistics courses in their first year, and admissions
committees typically looked to quantitative GRE scores to predict students’
likelihood of success in that statistics sequence.
On that year’s committee of eight, a full professor whom I’ll call David
explained how scores were typically interpreted. “Lots of eyebrows are
raised if both [sections’ scores] are around the 50th percentile. But if you’re
above 60 and one [section’s score] is high, I think you’re totally solid and
no one is going to bat an eye.” However, if it had been up to that year’s
admissions chair, Thomas, the department would not even collect appli-
cants’ GRE scores. A political theorist, Thomas took seriously the messages
conveyed in the university’s workshop for faculty involved with graduate
admissions about downplaying the GRE and utilizing holistic, individual-
ized evaluation.
The committee’s disagreement about how GRE scores should be inter-
preted and weighed in the context of real applicants revealed the difficulty
that competing disciplinary norms can pose for collective evaluation. In the
group’s initial meeting, Thomas presented his plan for the admissions pro-
cess and suggested how committee members might weigh various criteria.
He discouraged them from using the GRE, noting findings from “a study at
Yale University” that the only thing they could predict was first-year grades.
He also remarked that the graduate college was collecting the scores only
because some programs still wanted to use them.
At this comment, Roger, a full professor on the committee who conducted
quantitative research, interrupted. “I want to respectfully disagree. Evidence
from the research on the GRE can’t give advice to committees like ours be-
cause those studies use censored samples consisting only of those students
who were admitted.” It was the first time someone other than the chair had
spoken since the official beginning of the meeting. David, who was also a
statistician, affirmed this interpretation of the research, adding, “I myself

am going to use the GREs.”
An uncomfortable silence followed, with tension hanging in the air. Linda,
an associate professor, inquired, “What about people who are in, say, the 40th
percentile?” Thomas noted that traditionally there “had been sort of an
informal 50th percentile threshold,” but an assistant professor, Emily,
chimed in that she agreed with the two senior statisticians. Recognizing her
expertise, Thomas affirmed, “And you’re the methodologist here. However,
I do strongly discourage you from using it as the first or only threshold, as
some sort of a litmus test.” After another period of silence, he moved on to
another topic, the question unresolved.
This discussion exemplified how a worldview moment could induce a
more overtly political mode of decision making than was the norm. When
Roger, a senior scholar with authority in the methodological approach fa-
vored by the majority of the department, interpreted research about the
GRE in a way that (1) called into question the findings’ reliability and va-
lidity for their committee and (2) appeared to reinforce the reliability and
validity of a criterion in which several already believed, it empowered other
quantitatively minded people to speak up too. Interestingly, the committee
discussed few details of the research in question (a study from 1997 by
Robert Sternberg and Wendy Williams) and did not note that its authors
anticipated the very critique Roger raised—that the censored range of ob-
served GRE scores would attenuate the odds of finding significant
relationships.21
In a follow-up interview, one of the two senior members of the committee
described his impressions of Thomas and this debate over how research
should inform their work:
He’s a brilliant guy. Also, his heart is in the right place, so when he hears evi-
dence from Yale, he believes it. This is just what he’s waiting to hear, and so
he believes it and he doesn’t approach it in a critical way. Roger, on the other
hand, has done this stuff. This is his bread and butter. He knows instantly that
this is a flawed study. Why did it even get published? Is it just because a lot of
people want to believe it?
His comments suggested skepticism not only about the research study dis-
cussed in the meeting, but also about the motives of those who marshal its
findings to downplay the importance of GRE scores. For his part, Thomas
recognized that others’ decades of experience with quantitative research
surely informed how they approached admissions and GRE scores. He ac-
knowledged to me privately that they had “a perfectly reasonable argu-
ment” about the Sternberg and Williams study, but that more discussion
should have taken place about the study’s nuances than had actually oc-
curred. Finally, Thomas felt that this was a moment in which worldview
and power dynamics in the group both became salient as some committee
members started taking their cues from Roger and David. Despite his posi-
tion as chair, Thomas was the only faculty member on the committee firmly
committed to deemphasizing GRE scores, and he felt that his minority view-
point had been essentially overruled. He was not the only person to see it
this way. In the course of reviewing this exchange with other members of
the committee, I learned that they too thought this was a situation in which
a minority viewpoint had been overruled through a combination of se-
niority and methodological majoritarianism.
The limits of agreeing to disagree. The issue of GRE score interpretation

did not resurface for the political science committee until the end of a very
long meeting in which their task was to create the final lists of admitted
and wait-listed applicants. Committee members frequently cited GRE scores
in their rationales for high and low ratings, to little challenge. None of the
outcomes seemed to rest directly on the strength of a student’s scores, so it
seemed they could agree to disagree about how those scores should be used.
None, that is, until they reached the very last candidate to be discussed.
Like the other two universities in which my case study departments were
located, this university set an early deadline for nominating admitted stu-
dents for diversity fellowships. With an eye on the urgency of this upcoming
deadline, the committee wanted to make a decision on Christian, an ap-
plicant with an unconventional academic profile. They did not explicitly
discuss what made him suitable for the fellowship, except to acknowledge
that he attended a local college that rarely enrolls students in their program.
Christian had the earnest endorsement of two noncommittee faculty in the
department who had met with him; however, his quantitative GRE scores
were under the 20th percentile. “How to think about his scores” was, as
Thomas later reflected, “the closest we came to conflict” in the course of
working together.
Their worldview differences and views on the GRE’s appropriate use un-
resolved, the committee struggled for more than 20 minutes with whether or
not to admit Christian. Emily was participating in the meeting via video
conference and initiated the conversation, arguing with some passion that
there was no basis for admitting a student with such low GRE scores. She
reminded the group that all students except those in the theory concentra-
tion would be essentially required to take a statistics sequence, and that the
student’s expressed research interests did not clearly situate him as a theo-
rist. “When I see him in the intro stats class,” she worried, “I see him failing.”
Picking up on this implicit risk-aversion narrative, Thomas countered,

“All students are gambles in lots of ways. The point is to get local knowl-
edge. Yes, he is problematic on conventional measures, but those are
unreliable in this circumstance.”
David proposed as a compromise that “putting him on the wait list would
be a gesture.” Linda interrupted him, but David piped up, “Let me finish.
Some indicators are unreliable, but less than 20th percentile is scary. Less
than the 20th percentile is not good.”
After a long silence, Thomas offered, “I would read it slightly differently.
He is less prepared than the rest of the cohort, which is his biggest risk.”
He also mentioned Claude Steele’s research on stereotype threat, which
found through experimental research that subconscious anxieties about the
possibility of conforming to negative group stereotypes can depress the per-
formance of women and students of color on standardized tests.22 Thomas
noted that mentoring would be very important if they chose to admit
Christian. At this, David also registered his concern about the student’s
prospects in the statistics sequence, to which Thomas emphatically re-
sponded, “But remember he doesn’t need to do the quant sequence if he
goes the theory route.” Still, Thomas conceded that the student’s scores
suggested underpreparation and that he didn’t like the idea of “throwing
him into a deep pond” with students whose average preparation seemed to
be much greater. One of the three graduate students on the committee then
spoke up, affirming the possibility that mentoring could help address the
applicant’s level of preparation, but also questioning whether there was an
individual who might clearly serve as his mentor. Thomas noted that a fac-
ulty member with interests similar to Christian’s had already volunteered
to serve in this capacity, which produced another long silence. The ball
was in their court.
Understandably, the committee was tired and a little edgy at this point.
They had been working intently together for almost three hours after the
end of the workday. The sun had gone down, and the only light in the room
for nearly an hour had been the unnatural glare produced by their laptop
screens and the image of their applicant rating spreadsheet projected on a
whiteboard. Empty pizza boxes, paper plates, and red Solo cups littered
their large table. Perhaps fitting for these political scientists, the space had
come to feel more like a war room than a conference room. Clearly it was
time to start wrapping up the meeting.
It was several seconds more before anyone spoke. “So, what do we
think?” asked Emily, who had started the conversation. Thomas proposed
a vote by roll call and, to the group’s surprise, the seven persons present
admitted Christian by a margin of one vote. Further emblematic of the

group’s ambivalence, the meeting ended with one individual muttering, “I’m
really worried about what happens if he sinks.”
The conversation had been laced with several disagreements: about
whether to admit the student, of course, but also what the appropriate
grounds for admitting a student with low GRE scores would or should be,
and about what it would mean for a student to “sink.” There were points
of relative consensus as well. All but one expressed concern over admitting
a student with an outlying test score. To me this suggested that an unstated
aim of their selection process was to construct a class that would perform at
a similarly high level from the outset.
Another point of agreement concerned the importance of faculty men-
toring to this student’s prospects. Reflecting on the committee discussion
about the mentoring this student would require, David explained:
I think it is a moral obligation to tell him not to take the quantitative
sequence . . . I think that if his advisor spends night and day tutoring him, he
might overcome this . . . Anything could happen, but I’d say it’s highly
improbable . . . The summer bridge program offsetting twenty years is kind
of unlikely, but it undoubtedly will help . . . It might raise you to the 50th
percentile conceivably. From the 19th to the 50th percentile is a huge increase,
but the students in this program are in the 98th percentile. It’s going to be a
very tough competition.
Linda explained that she would not have voted for him if one of her
colleagues had not committed to serve as his mentor:
I think in the end I felt okay with it . . . If over five were in favor, I was at like
a 4.5 or slightly under. And I decided not to make sort of any problem with it
just because I was very respectful of the committee and I felt {Pause} the fact
that there was this professor here already who was willing to sort of work
with this student. I mean, if there hadn’t been that, it would have been like,
absolutely not. Because this person is likely to get lost and then somebody is
going to be, in some sense, stuck, on top of everything else we’re doing. Then
it takes our time away from our research, our other students, and all this kind
of stuff. But given that there was somebody who was really saying, “You know
I’m going to work with this person,” in the end I felt okay with the decision.
David and Linda both identified a role for mentoring students whose formal
credentials seem like outliers. Yet whereas David worried that mentoring
would be insufficient to offset Christian’s competitive disadvantage among
his peers and might recreate within their department the same inequalities
observed in society, Linda worried about how that mentoring would burden
faculty. Her comments suggest a posture of ambivalence—one that I heard
widely expressed—toward working with students who have less prepara-

tion than their peers.
With so many well-prepared students, faculty in these programs seemed
to prefer students whose success would be less dependent upon their ef-
forts. After all, these students were not only being admitted to learn; they
would also be the very individuals appointed to teach undergraduates and
support faculty members’ own research.
Reflecting on this episode, a graduate student on the political science
committee described support of diversity to me as “a platitude, a noncon-
troversial stance.” Like any institutionalized value, diversity seems natural
to support in principle, and especially so when no competing interests are
perceived or personal investment is required. In the real world of negotia-
tion, however, this is rarely the case, and many faculty struggle to balance
the desire for diversity, the difficulty of excluding any applicant who makes
it to the final round of the review, and their untested assumptions about
risk. Depending upon a scholar’s own commitments and experiences, ad-
vocating for unconventional students may seem natural or it may seem
controversial. David spoke with me about this struggle in a follow-up in-
terview as well.
In what could be read as a sign of a political disciplinary logic at work,
David compared his colleagues’ support for Barack Obama’s presidential
bid in 2008 to their support for applicants of color with unconventional
academic backgrounds:
We all voted for Obama. I doubt if there’s anyone in the department that did
not vote for Obama. And we want to believe sort of, “You can pick them at
random and they’ll all do well,” and it just isn’t true.
David likened the candidacy and election of an inspirational black presi-

dential candidate, who had less experience as a political officeholder than
his white opponent, to the candidacy and admission of compelling applicants
of color with weaker academic credentials than many of the white stu-
dents they enroll. I sat with his comments for the rest of the day, aware
that they could be interpreted in several ways. A critical perspective might
conclude that “We all voted for Obama” was a disclaimer for me—a po-
litical scientist’s way of saying they were not racist, despite how they grap-
pled with merit and diversity. One could also interpret his comments as an
effort to convey his colleagues’ willingness to select students with profiles
that were statistically improbable for their program. President Obama him-
self has spoken about his “improbable journey” to the White House, albeit
in language that names his achievements as the product of collective struggle,
not noblesse oblige.23 Under either of these interpretations, we know that
national political candidates are highly symbolic figures who represent a

platform of ideas, policies, and possibilities. And in America today, we know
there is strong evidence that race and racism continue to affect opportuni-
ties and outcomes in education and politics, regardless of who holds the
office of President.
The political science committee’s discussion teaches us several things
about merit, diversity, and faculty gatekeeping. It offers a concrete example
of informed disagreement about the proper place of GRE scores in admis-
sions and of deference to curriculum in constraining who is judged to be
qualified. It also demonstrates the controlling influence of seniority and in-
tellectual politics on faculty deliberations in the absence of a shared intel-
lectual worldview, and how, in the end, the outcomes of deliberation may
yet surprise everyone at the table. Together, the story from political science
suggests that whatever change or stagnancy may be perceived with respect
to organizational diversity or equity, that the scale on which changes occur
is small—one reviewer, one applicant, and one evaluation at a time.
Conclusion
A comparison of admissions practices and priorities in distinct disciplines

like economics, philosophy, and political science brings into focus the im-
plications of disciplinary values for the work of evaluation. Comparing two
high-consensus fields, I learned that disciplinary methods, theories, and
practical priorities were evident both in scholars’ individual judgments and in
their collective decision making. Vivid differences in disciplinary cultures of
evaluation became clear, as did similarities between the philosophy com-
mittees in two universities. I observed that ideas about merit are enshrined
and understood within disciplinary assumptions that flow from prevailing
theories, methods, epistemologies, and practical priorities. These commonly
held assumptions converged as disciplinary logics that committee members
used to defend controversial judgments and practices that disciplinary
outsiders might question. For instance, the legitimacy of economists’ own
work often rests on the validity of statistical inferences, and I observed that
shared trust in those inferences made it easier for economists to trust infer-
ences from quantitative metrics like college grades, GRE scores, and college
rankings than it was for scholars in the humanities and other social science
disciplines.
In low-consensus disciplines, scholars are socialized in a variety of ways,
which produces a greater variety of theories, methods, and priorities that
inform faculty work. When the committees in such fields sat down to
deliberate, they exhibited divergent disciplinary logics, which yielded les-

sons of their own about how disciplines shape evaluation. I was intrigued
to find an especially strong role for individual tastes in these fields, and to
see those tastes displayed in the use of serial impressions to describe appli-
cants. The episode in political science also illustrated the broader tendency
for committees to defer to majority-minority and seniority politics when
worldview moments erupt, reminding the committee of how difficult it can
be to achieve consensus. A third trend in lower-consensus disciplines—the
strong tendency of faculty to advocate for students like themselves—is the
topic I turn to next.
chapter four
Mirror, Mirror
Mirror, mirror here I stand. Who is the fairest in the land?
— Wilhelm Grimm, Grimm’s Fairy Tales
T hat people connect through their similarities more than their differ-
ences is a fact of social life.1 We know this tendency affects undergrad-
uate admissions and professional hiring, for example, with decision makers
tending to prefer applicants who remind them of themselves or who fit with
the organization’s status quo. Sociologists call it homophily, and have
explained it in terms of simple attraction, reliance on segregated social
networks, and self-serving ideas about merit.2 Preferences for people like
oneself and for qualities like one’s own have significant implications for
equal opportunity, which is especially troubling because such preferences
can be difficult to avoid. Often they occur subconsciously. When those with
privilege look at others with privilege and instinctively see them as “the
fairest in the land,” it becomes difficult for those from more modest personal
backgrounds to break into the upper echelons of organizations. Sociologists
therefore usually discuss homophily as one in a handful of interpersonal
processes that reproduce broader patterns of inequality.
I did not set out to study the preference for self-similarity, but it was im-
possible to ignore in my data. Nothing short of pervasive, it brought to life
the tendency for organizational selection to operate as cultural matching.3
In some cases, reviewers’ personal identities and experiences shaped how
they emotionally responded to specific applicants’ life stories. More gener-
ally, self-similarity shaped reviewer responses to applicants by providing a
familiar frame of reference. Like previous research on elite selection, I found
widespread homophily based on elite academic pedigree. College reputa-

tion was central to pedigree, but the details varied somewhat by discipline.
Each academic discipline has its own awards, rituals, and experiences that
carry cachet, in spite of meaning little, if anything, to those outside the
discipline.
As important as pedigree was, other experiences and identities also shaped
faculty tastes, especially in the lower-consensus disciplines. I mentioned
earlier that when an applicant’s expressed research interests or approach
to scholarship aligned with a reviewer’s, it could earn the application a
closer—and usually more favorable—reading.4 Another type of cultural
matching I observed involved reviewers who tried to present themselves as
cool, or hip, and who preferred applicants with similar qualities. Interna-
tional faculty often made a special effort to see that all international
students’ files received serious consideration. They pushed, in some cases
successfully, for national origin to be weighted like race in diversity
conversations.
Other reviewers felt compelled to pay forward the opportunities for up-
ward mobility that they felt scholars had once given them. Drawing on their
own life experiences, faculty who had experienced social mobility also
introduced what I call counterscripts—alternative interpretations of admis-
sions criteria or student profiles—into their committees’ deliberations. Un-
like differences of interpretation rooted in intellectual worldviews, which
were unlikely to be quickly settled in the middle of a meeting, faculty more
readily opened their minds to counterscripts, perhaps because they came
directly from a colleague’s personal experience. In cases where socially mo-
bile faculty had a critical mass on committees, counterscripts were more
likely to upset the conventional wisdom. And in a few cases, introducing
counterscripts even resulted in decisions to admit applicants with nontra-
ditional profiles. Depending upon an admissions committee’s composition
and how they deliberate, then, I argue that homophily has the potential to
either reinforce or challenge conventional wisdom.5
Academic pedigree was valued in all programs, but only in lower-
consensus fields did I see clear evidence of homophily on international, cool,
and socially mobility dimensions (see Table 5).6 Across settings, identity-based
preferences affected evaluation in four ways. Sometimes they provided
grounds for a reviewer’s claims about an applicant’s “fit” with the disci-
pline’s focus or department’s expertise. Often self-similarity generated
enthusiasm or sympathy for applicants, and these feelings translated into
higher ratings or advocacy in committee. In some cases, personal identities
gave reviewers the sense that they had a unique ability to judge an appli-
cant’s potential. And finally, homophily could stimulate discussion about
Mirror, Mirror 97
Table 5 Evidence of Homophily across Disciplines
Pedigree Cool Social Mobility
Astrophysics x x x
Biology x x
Classics x x
Linguistics x x
Political Science x x x
Sociology x x
Economics x
Physics x
Philosophy x
which characteristics of applicants committee members should weigh and

which ones they should try their best to ignore.
It bears noting that the demographic and intellectual composition of the
committees I studied necessarily biases my findings about which types of
homophily occurred most frequently. For instance, if I had studied grad-
uate programs in applied disciplines rather than ones traditionally desig-
nated as pure, I might have found a more pervasive role for professional
experience or interpersonal savvy. The sample I have, composed mostly of
white and male faculty in pure disciplines in highly ranked programs at se-
lective universities, may help explain the prevalence of pedigree, and may
explain why I did not observe consistent or strong preferences for women
or for students of color from the United States. I explore this possibility in
more detail in the conclusion to this chapter.
This chapter has four sections. First I present evidence that participants
are cognizant that personal similarities shape professional judgment. Then
I offer profiles of the four types of homophily I most frequently observed.
I next describe efforts by socially mobile faculty to rewrite prevailing scripts
of merit; here, I illustrate how counterscripts can challenge well-established
assumptions. Finally, I present disconfirming evidence that points to in-
stances in which faculty valued cultural differences rather than similari-
ties, either as noblesse oblige or in solidarity.
Awareness of Preferences for Self-Similarity
Homophily often occurs in the moments when we form impressions of

others. Nevertheless, faculty in political science, astrophysics, classics, and
linguistics expressed that they were quite conscious of this subconscious
tendency, especially among their colleagues. One astrophysicist recounted,
“That’s how people subconsciously would be attracted to someone—if they

see similar interests, or if they see similar sets of skills or similar emphasis
or a similar way of thinking.” Later he raised the topic again, and spoke at
length about the disagreement it sparked on the committee:
Another interesting thing was the divergence. Different people had very dif-
ferent ways of assessing someone’s worth. In particular, the people whose back-
ground was in theoretical physics or theoretical astrophysics were more likely
to rate a theoretical astrophysicist higher than an observational, experimental
astrophysicist. Whereas I thought they had similar overall qualifications.
There was someone whose background sounded very impressive, and he
listed the projects that he had done. But ultimately, it came down to the fact
that he was capable of integrating on a computer—differential equations—and
that, ultimately, was the basis of all his projects. As an undergraduate that’s a
great thing to be able to do. It’s a skill you could have, and you could do a lot
of very interesting things with it. But ultimately, that’s what his ability was.
And this was a person who theorists on the committee thought was the greatest
person ever.
We also had this guy who was working on building some instrumentation for
telescopes—who I thought in the terms of the skills he had—the skills he had
were certainly equivalent to the skills this other guy had . . . Some of the theo-
rists in the committee looked down upon this person who was an instrumen-
talist. They were subconsciously thinking that people who worked in their
particular subfield were somehow more talented than this other person.
More often than disclosing their own preferences for students with quali-
ties like theirs, participants would volunteer how their colleagues responded
to self-similarities, although they could also point to specific applicants with
whom they personally felt a connection.
Philosophers not only were aware of and discussed with one another how
experiences, identities, and self-interest formed a basis for preferences; they
accepted this tendency as natural and joked about it. One reflected:
There are people who write original papers and they show that they’re . . .
interested in a topic you’re interested in. That has something to do with what
makes people catch fire about an applicant. If they write a good paper about
a topic I know something about and I’m interested in, I can then appreciate
the virtues of this paper. This person is on to a subject. They obviously show
they’re smart because they’re interested in the same thing that I am {He trails
off, laughing}.
Recognizing how difficult it would be for them not to be inclined toward

work like their own, philosophers downplayed homophily as a threat to
legitimate evaluation. Instead, they emphasized how individual expertise
gave them special capacity to judge work in similar areas, and they struc-
Mirror, Mirror 99
tured the review process to ensure that each applicant’s writing sample
would be read by at least one person with expertise in the subject area. This
practice also addressed their concern that an applicant’s writing sample
might be discounted because a reviewer could not bracket their personal
biases or stylistic preferences.
Participants in other programs discussed the importance of constructing
committees that were diverse along several dimensions, including intellec-
tual focus, social identity, rank, and previous experience with admissions.
Ryan, an astrophysicist, elaborated, “It’s valuable for our committee to have
a diversity of backgrounds . . . I think that helps. People tend to view things
through their own experience.” This approach worked, in part, because ad-
missions committees tended to be relatively small, so adding just one or two
individuals from a particular background could easily broaden a group’s
common sense. On a committee of five, for example, a rating that might
have looked like an outlier from one individual would seem plausible from
two, and would become the majority viewpoint from three individuals.
One person might struggle to build enthusiasm for a borderline appli-
cant, but two or three who shared a particular perspective could shift the
entire tenor or direction of a discussion.
Homophily of the Pedigreed
Perhaps unsurprisingly, graduate admissions committees in these top-ranked

programs preferred students with credentials that connoted academic pres-
tige. And although participants respected a broader institutional set than
the “super-elite (e.g., top four)” universities that Lauren Rivera found pro-
fessional service firms to privilege in their hiring decisions,7 college or
university reputation was the most frequently mentioned marker of pedi-
gree. Participants also spoke to the value of recommendation letters from
superstars in the field and to the honor of Phi Beta Kappa membership.
Faculty members with well-pedigreed backgrounds of their own down-
played how socioeconomic status is related to the attainment of elite ac-
ademic credentials, offering several rationales for reframing prestige as
merit. Some projected their own professional success onto the potential of
such applicants, believing that their own early honors set them on a path to
later accomplishments. Others made cultural associations between presti-
gious undergraduate institutions and the doctoral programs in which they
taught, believing that applicants from top-tier colleges and universities had,
as one humanities professor put it, “been preadapted to a program like
ours.” A third group recalled the difficulty they had in gaining admission
into elite undergraduate institutions (and the intelligence they had perceived
among their peers there) as evidence that today’s graduates from such in-
stitutions must truly be “better.” Brian is an astrophysicist who expressed
this perspective:
I was the best at math and physics in my high school, probably the best in five
or ten years. That may be a big fish in a very small pond. Then you get into a
bigger pond. I got to college and was an undergraduate at MIT. The guy sit-
ting next to me in my intro physics class had been on the American Physics
Olympiad Team and was clearly better than I was. I’m good but there’s al-
ways someone better . . . A letter that says, “This is the best person graduating
from Stanford or Princeton” means a lot more than, “This is the best person
graduating from Ohio State or University of Texas . . .” We certainly know the
caliber of people in there. There was one guy who we admitted that a pro-
fessor from Princeton said—“this is the best undergrad I’ve ever worked with.”
That means a whole lot more coming from MIT than coming from a lesser
school. MIT and Harvard are the places where, frankly, the real genius-type
people wind up.
Finally, a relatively small group offered the human capital argument for
pedigree as merit—that applicants who earned their college degrees from
peer institutions must be superior because they had received high-quality
training. Brian also mentioned, “Coming from the places that are our com-
petition, we know what kind of background they’ll be getting.” Across these
rationales, the common thread was an underlying preference for applicants
whose profiles resembled committee members’ own academically distin-
guished biographies.
Pedigree and Privilege in Classics
Pedigree goes deeper than institutional affiliation, however, and every dis-
cipline has its own set of experiences and qualifications that are regarded
as distinctive. To illustrate how it played out in one field, I discuss pedigree
in classics—a small, tight-knit discipline marked by extended socialization
and study of the languages, literature, and art of ancient civilizations. Be-
fore asking questions about admissions, I began each of my interviews by
asking participants to describe how they had come to be professors in the
field, and in classics the responses had clear patterns. All came up through
the American or British academic establishment, all but one received an
early start learning ancient languages, and several felt that travel and study
in Athens, Rome, or England had an important socializing effect on them.
Sure enough, in their judgments of applicants, these very traits emerged as
Mirror, Mirror 101
salient. Not only did they weigh the undergraduate institutions that stu-
dents attended (very important), the honors or awards they earned (nice to
see), and the eminence of reference providers (expected in this small disci-
pline). They also considered years of Greek or Latin training as a yardstick
of preparation, looked for whether students had traveled to and studied in
the ancient world, and prized applicants who had connections to the British
universities that have been at the forefront of classics scholarship for cen-
turies. Given the discipline’s history and its geographic and educational loci
(including the relatively small set of elite preparatory schools and postsec-
ondary institutions that provide early training in Greek and Latin), these
priorities combined to encourage a system that disproportionately serves
the interests of wealthy students.8
The committee used the language of risk aversion to maintain that stu-
dents needed deep, formal training in both ancient and modern languages
to be successful at the graduate level. An associate professor of classics
explained:
We definitely want to know how many years have they studied the ancient
languages and how many modern languages do they have, because if they
don’t have that, they’re not going to be starting out at a certain level here. They
won’t be able to read any of the scholarship, the foreign language scholarship.
They’ll struggle in their classes.
As with committees in other disciplines (such as political science, eco-

nomics, sociology, physics, astrophysics, and linguistics), they felt that
students without formal training would not start out on equal footing with
their peers, which was necessary to ensure a common foundation for in-
dependent scholarship.9 Their preference was to admit cohorts of stu-
dents who would have minimal variation in entering skills, which they
thought would ensure similarly shaped learning curves at the doctoral
level.
As in other disciplines, faculty in classics liked to see evidence of student
effort, but they worried at possible signs of struggle. Rather than selecting
students with underdeveloped potential whom they could envision fash-
ioning into disciplinary experts, they preferred to enroll students who
already embodied ease with disciplinary discourse and norms. This ease,
Bourdieu argued (and other sociologists have agreed), is a hallmark of
cultural capital.10 The disciplinary familiarity that these students already
possessed enabled them to easily transition into the role of classics scholar
and to immediately support faculty members’ research. What is more, a
financial aid structure in which students are paid to support faculty research
further encouraged selecting on pedigree rather than potential.
Yet the tension between pedigree and potential was not lost on classi-
cists. Paula, an associate professor of classics who grew up with professors
in the family and whose institutional affiliations include the Sorbonne, Ivy
League schools, and public research universities, admitted that her depart-
ment was “torn” in how to think about students from less selective colleges.
She said, “I mean you can see how torn we are because, um, we want those
people. We want them encouraged. We would like to support them. {Pause}
On the other hand, they’re nearly always less of a sure thing.” Differenti-
ating herself from such applicants through language like “those people,”
Paula later made a point to deny “bias” in the department. Yet she also
admitted the benefits she personally experienced from her family and well-
resourced undergraduate institution. Paula expounded:
We don’t have any kind of bias. We’re not, you know, “We only take people . . .”
{Interviewee trails off}. We don’t say that, and we would never say that we
only take people from Ivy League schools or from universities as compared to
colleges. Because that’s not, you just never know who the exciting student is
going to be, and it’s in no way a guarantee of where they went to college. The
only thing that can help, that can be an issue, is students who’ve gone to re-
ally good . . . who have had certain advantages . . . either support from family
or support out of a department, have a kind of a confidence that I just myself
felt was needed to get through graduate school.
There may not have been an explicit bias against less renowned colleges
and universities, but there was a clear and consistent preference for the
sophistication of students who graduated from selective undergraduate
institutions.
Students self-taught in ancient languages and those who had attended
less prestigious institutions—including public flagship universities—often
need a postbaccalaureate or master’s degree to be competitive for doctoral
classics programs. The postbaccalaureate was started to help students from
fields like art history and archaeology who wanted to pursue a doctorate
in classics but lacked the language training. Over time, however, students
with bachelor’s degrees in classics have also begun pursuing these de-
grees to gain a competitive advantage over other applicants vis-à-vis lan-
guage skills. Such programs have become a means for current classics
students to preserve their competitive edge in admissions. Notably, these
programs come with high tuition, fees, and expenses for study abroad,
which may work against their original role of broadening access to the
discipline.
In short, it is rare in these programs to break into classics at the PhD
level without an undergraduate degree in the field from an elite institution.
Amy, a female junior faculty member, described the situation to me:
Mirror, Mirror 103
[There is] a structure of privilege which is morally uncomfortable. I was doing

an external review at an Ivy League university and we had a meeting with some
of their graduate students, not all. And I was really struck that of the eight or
so people who actually talked, probably half were Brits. And it came up in the
conversation that three or four of the others were from one Ivy as undergradu-
ates. Now, I did ask them and they had told me they had admitted somebody
from a rural flagship university a couple of years before. You know if you get
classics department students who have been trained in the British system and
students from elite Ivy League schools, you can probably have very successful
completion numbers. But it doesn’t feel right.
In this year’s pool, two individuals with self-taught Greek and Latin language
skills made it to the short list, but their lack of formal training remained a
major concern for the committee, and the committee opted not to send ei-
ther through to the interview stage. A senior scholar also noted it has been
“very rare not to have a senior thesis” among those recently admitted.
The implicit preference of elites for elites reproduces social hierarchies,
and in few places is that pattern more apparent than the upper echelons
of academia. In the case of classics, participants emphasized criteria in
which the lines between human and cultural capital were no doubt blurry.
Extended travel to the ancient world and formal training in prestigious
institutions had been personally formative for committee members, and
hold symbolic value in the field. From their personal and disciplinary frames
of reference, reliance on these criteria was understandable.
Yet when opportunities that are available to only a few pass as prerequi-
sites, it creates, as Amy described it, a “structure of privilege that is morally
uncomfortable.” Unlike fields such as biology or sociology, which have
multiple access points and in which nontraditional trajectories are acceptable
or even valued, this classics program was more like a typical mathematics
program, requiring an early start and years of experience in well-respected
programs to be perceived as a qualified applicant. Of course, the term
“qualified” is loaded. As recent controversies over hiring faculty and ad-
ministrators of color at the University of Pennsylvania have highlighted, to
be judged “qualified” is not only a judgment of skill and talent. It can also
be a function of whether an employer or decision maker feels comfortable
with the applicant due to common intellectual interests, social circles, or
institutional affiliations.11
Inside and outside of classics, participants were mindful of discipline-
specific prestige hierarchies, in addition to their awareness of a general
prestige hierarchy of colleges and universities. In philosophy, astrophysics,
and biology, for example, professors mentioned specific “underrated” uni-
versities whose reputations for strong training in the field surpassed their
college or university’s relatively weak overall ranking. Discipline-specific

honors and rites of passage also counted toward one’s pedigree. Whether it
was travel to Athens or Rome in classics, a journey to one of the major
telescopes in astronomy, a summer spent working in a prominent lab in
physics, or an award for best undergraduate poster at a national linguistics
meeting, accomplishments and experiences unique to one’s discipline could
distinguish an applicant’s pedigree and mark him or her as “one of us.”
Homophily of the Cool
In addition to the old-school culture of the academy that is reproduced

through homophily among the well-pedigreed, another form of cultural
matching that I observed in four programs encouraged characteristics rec-
ognized as cool or hip. I certainly did not go looking for this theme. I began
to notice it, however, in hearing certain applicants and research interests
described as “cool,” rather than the more common descriptors, such as
“good,” “strong,” “solid” or “exciting.” Taking note of which reviewers
used this language, and about whom, clued me into the presence of a cul-
tural archetype. Through further analysis, I came to see it as related to re-
viewers’ and applicants’ self-presentation, technology use, manners of
speech, and the novelty of their research and personal profiles.
Faculty who presented as cool exhibited carefully managed images, in-
cluding specific fashion styles that set them apart from the mode. One such
style I came to think of as the “ascetic sleek.” One male professor seemed
to emulate Steve Jobs’s look: shaved head, black turtleneck, earring. A fe-
male member of that same committee attended a committee meeting with
a fresh pixie haircut, cherry red lipstick, white scoop-neck T-shirt, black
vest, and leather pants. (By contrast, most of the other committee members
wore jeans.) A second type of cool presentation that I observed in several
committees was the “rumpled preppy.” One young scholar’s apparel for
both the meeting and our interview could have come straight off of a man-
nequin in J. Crew, including trendy tortoise shell glasses and a perfectly
rumpled and untucked Oxford shirt under a camel wool sweater—all
bearing the logos of expensive brands. The vast majority of the committee
members counted as cool wore eyeglass frames in trendy styles.12
Self-presentation was thus central to coolness, but it came out not only
in how these faculty dressed, but also in how they spoke and related to one
another. Cool scholars were savvy, with sophisticated social skills. De-
scribing people and things they liked as “cool,” rather than good, great,
and so on, signaled that they felt positively about an applicant but would
Mirror, Mirror 105
not come across as too earnest or passionate (which, by definition, would

not be cool). A certain detachment from the matters that others emotion-
ally engaged with characterized the cool scholars on these committees.
Cool scholars also used technology in a performative way, such as through
the conspicuous display of devices (especially Apple products) and discus-
sion of popular blogs or their own blogging. In one humanities committee
of five people, the gadgets on the table included Mac laptops, an iPad, and
an iPod, and several iPhones. The strong presence of technology in the ad-
missions process was not lost on the political science committee, which
had the largest proportion of cool members among the committees I ob-
served. In one of their meetings, a senior professor commented, “Technology
marches on. Everyone in here has a laptop, some iPads, and some with lap-
tops and iPads.” During their meeting, they projected a spreadsheet on a
screen and committee members searched the Internet during the meeting
for more information about applicants and letter writers. At one point, three
people on the committee had a side conversation about the blogs they main-
tained or to which they contributed.
Although I initially passed off this self-conscious technology talk as an in-
sight into the ubiquity of technology and its use to display status and cultural
affinity, I began to notice engagement with technology emerge in delibera-
tions about borderline applicants. Against the initial vote of the committee’s
majority, a humanities program admitted a nontraditional student with
“crummy grades” from Princeton, largely due to a cool member of the
committee praising the writing skill the applicant had demonstrated
through a popular blog for the discipline. And in another case in the social
sciences, a preppy cool committee member discussed his rating of a bor-
derline applicant, explaining: “I would be willing to go up. I read this app
right after another app in this concentration and wanted to be stingy with
my 1’s. Plus, she writes articles for Slate. Really cool.” An ascetic cool pro-
fessor added, “She seems like an interesting person.”
As these comments indicate, novelty was central to coolness, and in no
domain was novelty or originality judged more frequently than an applicant’s
expressed research interests and experiences. In linguistics, two reviewers
described an applicant’s intention to study emoticons (symbols, such as
smiling faces, inserted in text to add meaning) as “cool” and “really cool.”
That committee was similarly taken by an applicant interested in con-
ducting experimental research on twins’ language use. Political scientists
deemed applicants with field research or extended study in unusual places
as cool. In astrophysics, an associate professor noted that new instruments
generate money, energy, and fertile ground for intellectual breakthroughs
that “impose a fashion” for subsequent scholars:
Astrophysics has become such a large field that you have to be broad in order
to understand what’s interesting—what areas you may move your research to
[in order to] follow or even impose a fashion . . . For example, if a new instru-
ment is going to be built—you know exactly that it’s going to revolutionize a
certain number of fields. What are the reasons for that? These projects are so
large. It’s not like someone can say, “I’m going to understand this question in
the lab. I’m going to set up my own lab to do this.” This is a billion-dollar
project that usually the whole U.S. or international community will use. Students
tend to understand where the landscape is—where they have to go—where
are the sexy questions are that people want answered.
Indeed, perhaps the only praise of research interests that trumped “cool”
was “sexy”—a word that participants reserved for research that was timely,
cutting-edge, and somewhat provocative.
By contrast, a committee member in classics audibly groaned as he read
from an applicant’s personal statement, then commented, “He’s really in-
terested in something that others have done and done and done.” A preppy
cool committee member defended his low rating of a student who had gen-
erated considerable discussion. “He was strong in a conventional way,” he
said. “He was everything you look for, and nothing you weren’t expecting,”
to which an ascetic cool professor added, with a chuckle, “Like a Ford.”
Neither applicant made their admit lists. To cool faculty, the novelty and
originality of one’s research interests and overall profile could make an in-
dividual more likable by enabling that student to stand out from the scores
or hundreds of strong, solid applicants.
Interestingly, not a single mention of a research interest being cool or sexy
came from participants in the higher-consensus disciplines of physics, phi-
losophy, and economics. This finding may relate to a trend in Lamont’s
study of peer review: that centralized, hierarchical disciplines are more in-
clined to believe that academic excellence inheres in the individuals who are
being evaluated rather than in the eye of the beholder.13 It may be that fac-
ulty in higher-consensus disciplines are more inclined to judge scholarship
on the basis of shared, well-established standards than on individual impres-
sions. Centralized disciplines possess more of an intellectual orthodoxy in
which novelty—one of the central meanings of coolness—may be of less
value than qualities like precision and methodological sophistication.
Social Mobility Homophily
Eight of the ten committees included at least one member who advocated
for applicants from modest personal backgrounds. Some of these individ-
Mirror, Mirror 107
uals had been first-generation college students, and all had attended mod-
erately selective or public universities as undergraduates. In interviews with
me, they expressed mindfulness about how far they had come and how they
felt compelled by applicants whose paths reminded them of their own.
Having experienced social mobility, they saw admissions work as an op-
portunity to reach back and promote opportunity for others. Some from
this group also felt that their experience uniquely enabled them to inter-
pret applications from individuals from less privileged backgrounds, and
they offered alternative ways of thinking about institutional prestige and
academic preparation. Contrary to the risk aversion and prestige orientation
of these programs, those from less pedigreed backgrounds encouraged
their colleagues to take a chance on students who seemed to have potential,
apart from their achievements to date.
Narratives from Linda and Louis in the social sciences and from James
and Ryan in the natural sciences14 highlight the unique perspective shared
by faculty who did not attend elite undergraduate institutions. Louis, a so-
ciologist, went to a commuter-focused, moderately selective university
near his home before earning his PhD at an Ivy League university. The tran-
sition for him to graduate school had not been entirely smooth, but after
finding his footing, he had a positive experience. Discussing his current pro-
gram’s approach to thinking about college prestige, Louis remarked,
It’s not like we’re plucking people from only Princeton, Harvard, Yale, Co-
lumbia. But I think prestige is probably a little more important to my col-
leagues, and maybe they’ve internalized it more than I have . . . Some of my
colleagues, in fact people I know are on the committee, really care a lot more
about prestige and think of this university as a much more special place than
other institutions—probably more than I do.
Though he said the university was populated with many “elitists,” its intel-
lectual enterprise had earned his respect. Moreover, although he felt the
degree of coordination between the graduate school’s administration and
program faculty could be improved, the graduate college was “really put-
ting their money where their mouth is on diversity.”
Linda was a social scientist with an international reputation who had
grown up in an impoverished community and was one of the only people
in her high school class to go to college. She spoke to me at length about
her process for choosing a college as a teenager. To save money and stay
geographically close to her family, Linda had enrolled in a college that was
less selective than what she was academically qualified to attend. Higher
education researchers call this undermatching. I asked her whether she felt
her own experiences had affected how she read applications, and she re-
sponded, “Most definitely.”
I will look at somebody who is first-generation {Pause}—I think I understand

the learning curve that they sort of had to go through a little bit more. So I
don’t think I {Pause} favor them. I mean I’ve certainly had people I ranked a
number 1, the highest score, that had gone to the very best schools all along.
But if there’s somebody like—I believe there was somebody we admitted who
got maybe a 4.0 from Washington State—and had I not had my experience
and realized that if you’re coming from a first-generation family, nobody is
telling you how to apply to college, that you should try to go to the best one.
You may be thinking that, if I go here I can help out my family. I won’t have
huge loans or whatever . . .
I think I’m just much more open to looking a lot at: “They got a 4.0. They
did everything they could there . . . The letters of recommendation are really
terrific.” Whereas somebody who just went to Princeton, you might think,
“Even if they have a little bit lower GPA, they might have a better education.”
Linda felt she interpreted grades and institutional prestige more broadly
than the typical script of associating them with intelligence, likelihood
of success, and quality of training received. Her interpretation, she be-
lieved, represented a more “generous” and “gracious” approach than that of
her colleagues.
Linda also considered “distance traveled,” a noncognitive criterion that
many in undergraduate admissions are beginning to weigh. Based on her
firm belief in the role that effort plays in success as a doctoral student, she
assessed not only how accomplished a person seemed, but how far they had
come and how hard they had worked. When I asked her to share an
example of this, she spoke of her own experience as a first-generation col-
lege student, and how that informed her reading of a rural student’s
application:
So that one applicant, I don’t think it’s somebody we admitted, but he said
from his high school something like 50 percent wind up incarcerated . . . some
remote town in Montana. He had to Google how to apply to college. And so
somebody like that, I’m going to sort of look at the file and try to understand
if you started here and already got there {Using hands to show relative dis-
tance} versus somebody who started here and just got a little bit higher than
you. I’m very likely to look at, sort of, where did you start and where have
you been as a very good indicator of where you’ll be.
She also relayed the story of a second-generation immigrant:

One person said that his parents were immigrants and some [other family] had
been killed in the first war in Iraq. And he came here and he was working in
a restaurant washing dishes to get through high school, and in college he had
a job and stuff like that. It was a sort of story of {Pause} tremendous disad-
vantages and struggle and desire to make it in, I guess, handicaps. Of course,
Mirror, Mirror 109
that’s impressive. It implies the person has strong motivation. And also there’s
a kind of sympathy factor. The spoiled brats who’ve always been rich and their
parents gave them everything, well, they can get in anywhere. We don’t need
to admit them.
Linda thus set her expectations of applicants not only in an absolute sense
and by comparing applicants with others in the pool, but relative to the
student’s own previous opportunities and what the person had made of the
chances they had been given.
James and Ryan, two junior scholars in a natural sciences department,
both of whom had attended midwestern land-grant universities, adopted a
similar approach when evaluating institutional affiliations. In their com-
mittee’s initial meeting to develop a short list, they discussed Wilson, a stu-
dent with a near-perfect GPA in the discipline and considerable under-
graduate research experience.
“He went to the University of Nebraska,” James commented. “He’s a lot
like me.”
“Defend your alma mater!” Matthew exuberantly responded, with a
swoop of his arm.
After a moment’s pause that appeared to me to be a pointed nonverbal
response, James calmly continued. “I trust Janine, his letter writer. If she
says he’s one of the best in forty years . . .” He trailed off, and there was
another pause in the conversation.
“I like him,” offered Ryan. “I think he’d be a good fit.”
“Let’s move him up and then do death match against the others,” Mat-
thew urged.
For context, Matthew earned his PhD from Columbia University, and
“death match” was a term he used multiple times to refer to comparing a
small group of applicants with one another in order to identify weak-
nesses.15 In a follow-up interview, James spoke at length with me about
how he felt Wilson and his alma mater had been evaluated.
There was the student from Nebraska who ended up getting in our top
thirty. There was an interview, and I couldn’t make the interview. Apparently the
interview didn’t go very well, and so he didn’t make the top twenty-two. But that
guy, personally, had a research record that looked very much like mine. He had
a very high GPA. If you want to stand out going to a big, state school, you have
to have a really high GPA. Someone from Columbia doesn’t have to have a 3.9
but if you’re from Nebraska you do. So anyway, we’re advocating for other
people. I try to generally take a second look at people from big, state universi-
ties. I know there are a lot of fantastic students there who for very good reasons
decided not to spend $50,000 a year on their undergraduate education.
James continued, discussing how both his own judgment and that of his
colleagues were situated in personal experience:
The people on the committee who have an Ivy League background thought
about things a little differently. That’s why my colleague was trying to criti-
cize someone who went to Nebraska for going to Nebraska . . . I think people
understand what their [own] experience was, and maybe don’t think the others
on the committee may value the experience that they personally went through.
Recognizing that some of the distinctions that their colleagues valued were
also functions of class or wealth, socially mobile faculty made a special ef-
fort to disentangle potential from achievements. Looking back on their
own paths, some expressed skepticism about how well an elite pedigree
would predict long-term scholarly success. They rated unconventional ap-
plicants more highly than their colleagues did, and advocated for uncon-
ventional students in ways that both improved such students’ admissions
chances and, in a few cases, helped their colleagues think more broadly.
Counterscripts and Critical Mass
Rosabeth Moss-Kanter’s breakthrough research on group dynamics coined

the idea of critical mass, “that groups with varying proportions of people
of different social types differ qualitatively in dynamics and process.”16
And, indeed, in contrast to typical committees in which expectations of
elite academic pedigrees were normative, deliberations in more diverse
committees—those with a critical mass of people with a different identity
than the norm—engaged in more debate about the conventional wisdom
around merit. Linda, James, and Ryan’s committees did so, as well as one
of the two in philosophy. In these conversations, members corrected misper-
ceptions and pushed one another on their rationales for particular applicant
ratings. Members of these committees also went beyond correcting and
challenging each other, proposing altogether different interpretations of
common criteria. For how these interpretations deviated from standard
evaluative scripts of merit, I call these perspectives counterscripts.
A common script associated elite pedigrees with belonging in elite com-
munities like theirs. Maintaining this preference helped uphold their organ-
izational identity, and was thought to protect them against financial and
reputational risks of admitting students with lower odds of success. Linda,
Louis, James, and Ryan each pointed out that their current programs and
universities were resource-rich and therefore had the most to give—not just
the most to lose. Contrary to those who saw their programs’ status as be-
Mirror, Mirror 111
stowing “the luxury of risk aversion,” these professors associated status

with an obligation and opportunity to invest in underdeveloped potential.
They advocated for students whose credentials may have seemed less im-
pressive but who seemed to possess “spark” or potential to be “a diamond
in the rough.” In committee, they named the faculty who once took a chance
on them and in one case expressed aloud that they wanted to pay it for-
ward, as the phrase goes. Breaking with the usual script, they argued for
opportunities to be expanded rather than for their own status to be protected.
Their higher ratings, advocacy in deliberation, and challenges to their col-
leagues’ rationales shifted deliberations of borderline applicants and, in a
few cases, the outcomes of deliberation as well.
One person introducing a different interpretation would rarely be suf-
ficient to shift the group’s common sense. Usually it took a coalition of
sorts, a combination of advocacy from someone who shared an aspect of
the student’s profile and support or advocacy from committee members
with more privileged backgrounds. Often that support from more privi-
leged members was no more than a token gesture toward a few applicants
rather than a reorientation toward the pool or of the value of elite creden-
tials. A senior professor of philosophy illustrated:
As things go on, the privileged get more privileges. One thing I like to do when
I can is to include in any class a kind of long shot. That is, a person from a
smaller college or something—someone who has written a writing sample that
really grabs you. Even though you don’t know the recommenders, so there-
fore you don’t really trust the recommendations. You don’t know what it
means that they got a high GPA at this school because it’s not much of a school.
But you know—it looks like this is the sort of person that needs to be plucked
out of mediocrity and given a chance. And I like to do that.
Such noblesse oblige seemed to give those faculty the sense that they had
fulfilled their responsibility to encourage opportunity and mobility in a
system fraught with obstacles. Yet it shifted opportunities for only a very
small minority of applicants, and is unlikely to shift the overall profile of
doctoral students away from the power of pedigree and wealth.
In a few cases a critical mass emerged through ally behavior—individuals
working across different social identities toward a common interest. In phi-
losophy, for example, there was the case of Gerald, who supported Liana’s
and Olivia’s argument that they should not prioritize high quantitative
GRE scores because it would systematically undermine women’s access. In
contrast to noblesse oblige, allies joined with individuals from historically
marginalized groups in questioning the usual review process or rethinking
common criteria with an eye to opening opportunities for categories of
students, not individual cases. Whereas noblesse oblige would selectively

extend support and opportunity in spite of perceived risks, allies poked
holes in the common narratives about “risk.” From an organizational learning
perspective, ally behavior carries the potential to stimulate longer term cul-
tural change by encouraging critical thinking about aspects of admissions
that often are taken for granted. And although examples of advocating
across different identities were rare enough to be considered disconfirming
evidence, they suggest a culture of evaluation that is more complex than
simple theories of homophily suggest.
Conclusion
Especially in lower-consensus disciplines, reviewer tastes were inextricable

from their own identities and experiences, especially their educational, so-
cial, and national origins. James promoted students from flagship public
universities because he had attended one himself, for instance, and had come
to believe they provided undergraduates a training that his colleagues un-
derrated. In classics, Paula liked to see evidence of extensive travel and study
in Greece, Rome, and British universities because those experiences
had been formative to her socialization. Jimmy encouraged his colleagues
in biology to closely consider every East Asian applicant because he had
immigrated to the United States to attend graduate school and has since
become a citizen. The examples are numerous.
Yet what appeared on the surface as a single trend toward homophily
consisted of specific interactions that collectively offer a richer picture of
cultural matching than “birds of a feather sticking together.”17 Sometimes
similarity with applicants provided grounds for claims of fit, which was a
critical consideration in late stages of admissions review. Connecting per-
sonally with an applicant’s experience or trajectory could also generate an
emotional response that affected ratings and spurred reviewers to speak up
in committee deliberations. I also observed scholars treating their iden-
tities and life experiences as knowledge that positioned them to better
interpret the stories and possible trajectories of applicants with similar
experiences—better, that is, than their colleagues who lack such perspec-
tive. Similarity with applicants was thus not only a source of attraction. It
framed participants’ subjective ability to judge and their desire to speak up.
Process-level patterns like these generally align with the organizational,
affective, and cognitive mechanisms that Lauren Rivera identified in her
study of hiring in professional investment firms.18 However, I draw out dy-
namics unique to more diverse selection committees and examine an eval-
Mirror, Mirror 113
uation context in which pedigree is not rooted in a single institutional pres-

tige hierarchy. Whereas Rivera’s research found that elite firms privileged
degrees from a small handful of universities, I found that faculty doctoral
programs were also mindful of discipline-specific status hierarchies.19 The
Ivy League retains a cultural mystique, but faculty in these doctoral pro-
grams were less compelled by an Ivy League institutional affiliation if the
department from which they graduated was different than their own field
or did not have a strong reputation of its own. Likewise, they might be im-
pressed by affiliations with a land grant, state flagship, or other moderately
selective university when it was home to a respected program in the disci-
pline. To identify future stewards of the disciplines, the pedigree that counted
most was pedigree in the discipline.
Interestingly, gender, race, and ethnicity may have been the primary types
of inequality about which faculty expressed concern to me in interviews,
but the committees I observed rarely raised race, ethnicity, or gender in an
explicit way. It could be that distinguishing racial or gender-based ho-
mophily from the review process itself was almost impossible because
straight white men were so well represented on these committees. Indeed,
because white males established most U.S. universities, preferences for cul-
tural qualities associated with whiteness and traditional masculinity are
subtly interwoven into the fabric of values, rules, rewards, and selection
processes in elite higher education. It can thus be hard to see gender and ra-
cial homophily, and social reproduction can occur along these dimensions
without explicit preference or advocacy.
Sociologist Shelley Correll’s research provides excellent evidence of how
judgments of professionally relevant qualities associated with traditional
masculinity reinforce gender inequity. She found with collaborators Stephen
Benard and In Paik that employers judged working fathers as more “com-
petent” and “professionally committed” than working mothers. In another
experiment on the “motherhood penalty,” they controlled for an employee’s
competency and commitment to see whether this removed the preference for
males. Interestingly, they found that male and female research participants
alike rated fathers more highly, because they were judged more “likable”
than professionally successful mothers.20
There are similar dynamics in graduate admissions. Having developed a
suite of professionally relevant rationales for the value of elite college at-
tendance (such as human capital, cultivation of academic ease and confi-
dence), faculty can uncontroversially make explicit their preference for elite
college attendance, in spite of its tendency to rule out many who might suc-
ceed. The focus on personal qualities, which reviewers may subconsciously
associate with social identities, allows more forms of homophily to persist
than those that a committee explicitly discusses. Without a lively debate

among diverse reviewers, particularly one in which individuals debate the
rationales for their rating, committees may not even realize how much their
own identities are shaping the selection process.
Each of the last several chapters has emphasized a social context in which
faculty judgment is situated—one’s program or department, discipline, com-
mittee, and individual identity. Depending upon the situation and group,
different identities can become salient, for identity is multidimensional and
dynamic, not uniform or static.21 In admissions, faculty members’ identi-
ties may be triggered through the committee’s composition, through en-
gagement with applications, and through discussions of those applications
with colleagues (whose own identities have also likely been triggered through
evaluation). The mix of experiences, identities, and preferences that results
reflects a fragmented organizational culture—one composed of multiple
priorities and of fluid perspectives and values. Organizational theorist
Joanne Martin articulated this viewpoint:
Organizations, environments, and group boundaries are constantly changing.

Individuals have fragmented, fluctuating self-concepts. One moment a person
thinks of himself or herself as belonging to one subculture, and a minute later
another subcultural membership becomes salient. People fluctuate in this way
because they are faced with inescapable contradictions, as well as things they
do not understand.
Increasing diversity in the upper reaches of America’s labor market, Martin

also argues, is only likely to increase the complexity of individual identities
and, by extension, how fragmented our organizational cultures are.22
It is hard to predict what the effects of such complexity might be for fac-
ulty evaluation. On the one hand, it could weaken the influence of specific
identity-based preferences. On the other hand, complexity could effectively
increase preference for students like oneself by creating ambiguous condi-
tions under which decision makers look for something familiar. Fragmented
organizational culture might open space for counterscripts that pose a chal-
lenge to conventional wisdom, but these alternative interpretations could
just as easily be folded into the tendency to extend noblesse oblige without
rethinking the system. My data do not suggest any clear answers, but there
are hints to suggest that the cultural profile of the ideal student is no longer
singular, but rather is diversifying. I observed two types of coolness—the
rumpled preppy and the ascetic sleek—as well as the influence of both tra-
ditional and disciplinary-based prestige hierarchies. There is also a growing
value placed on a student’s “distance traveled” and some disdain for ap-
plicants whose applications imply that they have led a charmed life.
Mirror, Mirror 115
One thing is clear. The more that departments hold up a mirror and no-
tice in the reflection not only their collective fairness and beauty, but also
the flaws that stem from who is absent from the image that the mirror pres-
ents, the closer they may come, in time, to embodying the image of diver-
sity they aspire to project.
chapter five
The Search for Intelligent Life

Definitions and understandings of intelligence, like all meanings,
are sensitive to the contexts in which they are developed.
— Jeannie Oakes et al.
W illiam James once hypothesized that self-worth is fragile when

grounded in qualities that are sensitive to others’ judgments. Social
psychologist Jennifer Crocker and her colleagues have amassed an impres-
sive body of evidence for this idea through research on prospective grad-
uate students. They found that graduate school applicants’ self-esteem—
which in most cases is a stable quality—fluctuates dramatically between
the days when they receive acceptance letters and days when they receive
rejection letters. For participants who rate academic competence as central
to their self-worth, the boost to self-esteem that accompanies admission
offers is greater, and the loss of self-esteem from news of rejection is stronger.1
Findings like these, as with the research on human tendencies toward ho-
mophily, support a central tenet of social psychology, that people define
themselves in relation to others.
Interestingly, the areas in which we may feel judged by others, such as
our gender, class, appearance, or intelligence, tend to be the same ones in
which we judge ourselves and others. Our inward judgments create a foun-
dation for identity and self-esteem (self-respect or self-worth). Applied out-
wardly, we use social judgments to discern others’ worthiness to receive
social goods—anything from an admissions offer, to a vote in an election,
to a compliment. Often we subconsciously reserve the harshest judgments
about qualities that are most important to our own self-concept.
In academia, an enterprise oriented around knowledge production, in-
telligence is an important domain of judgment and an informal grounds
The Search for Intelligent Life 117
for distributing opportunities and rewards. Central to academic culture and

the self-concepts of academics, faculty expect intelligence of themselves and
of one another, and they look for it in those whom they train and hire.2 It
may come as no surprise that a senior professor would sum up all of grad-
uate admissions as “finding smart people to replenish the gene pool.”
What “smart” means, however, is of course deeply contested. The status
and opportunities that accompany judgments of intelligence raise the stakes
associated with our definitions, as do historic tendencies to emphasize mea-
sures of intelligence with racial, socioeconomic, and other inequalities.
Disputes also emerge from the wide gap between psychometricians’ formal
measures and the implicit ideas (lay theories) that people use to judge one
another on an everyday basis.3 A related source of disagreement involves
the degree to which intelligence can be separated from cultural context. As
with all meanings, the everyday meaning of intelligence varies widely, de-
pending upon where you are and whom you ask. In a fascinating study of
American kindergarten classrooms, for example, Beth Hatt found that
teachers regarded as “smart” those children who started the school year
with the knowledge that kindergarten aims to teach. This knowledge in-
cluded reading skills, of course, but also the skills of sitting quietly and in-
teracting with docility. A study in Taiwan found that the flexibility to
apply the right ability at the right time was central to judgments of intel-
ligence, whereas Japanese college students and their mothers rated quick
thinking, decisive judgment, and a strong memory as important.4
Few researchers have systematically examined the lay theories of intel-
ligence within specific disciplines or disciplinary cultures. The insights that
have emerged from previous studies are intriguing, however. Lamont’s study
of interdisciplinary peer review found that literary scholars were especially
prone to lapse into weakly substantiated claims of smartness as they eval-
uated others. Data presented in Becher and Trowler’s famous exploration
of the disciplines, Academic Tribes and Territories, suggests that hard sci-
entists compare scientists’ relative intelligence on the basis of subject matter
complexity. One chemist cited a “hierarchy of arrogance” in the sciences
rooted in intellectual activity and qualities of mind: “Physics represents the
hardest, most abstract reasoning—people know they’re smart. Chemists feel
defensive in relation to physicists, but superior in relation to biologists.”5
I did not intend, when I started this research, to examine how faculty
conceive of intelligence across the disciplines or how faculty ideas about
intelligence shape admissions. Yet when four of the first five people I inter-
viewed cited intelligence or brilliance among the characteristics they were
looking for in prospective students, it became clear that this was something
important to them. To understand how they thought about admissions, I
needed to attend to intelligence as well. Disciplinary foundations to ideas
about intelligence and brilliance became quickly apparent as well, because

I conducted early interviews with admissions chairs from a variety of fields.
I learned as the fieldwork continued that intelligence is one of the most
fundamental values driving faculty efforts to create the futures of their re-
spective disciplines. In admissions meetings, I observed that like assessments
of a student’s “fit,” assessments of this nebulous quality almost always went
unchallenged. When an applicant was understood to be “brilliant,” a host
of other flaws could be forgiven or ignored. When an applicant was believed
to lack intelligence, either in degree or in qualities that serve the discipline,
admission offers were unlikely at best. At issue in my analysis is not whether
judgments of intelligence should matter for admission Rather, I focus on
other questions: How do professors conceive of and attempt to assess in-
telligence in doctoral admissions? How do their ideas about intelligence
vary by discipline?
I discussed in Chapter 2 the differences between the criteria that faculty
used in early and later phases of review and how, in each phase, they sought
to infer applicants’ intelligence. Here, I distinguish between two common
ways that faculty understood and assessed intelligence: as a general impres-
sion and as a disciplinary cultural practice. Committee members frequently
expressed summary judgments about how smart they believed an applicant
to be, linking those perceptions to predictions of a student’s likelihood for
success. These overall impressions treated intelligence as a fixed attribute
that some people were assumed to have more of than others. Yet when
prodded in interviews to elaborate on what intelligence means for scholars
in their discipline, faculty transitioned away from impressions and broad
judgments into a view of intelligence as a disciplinary cultural practice. By
assessing applicants’ writing samples, personal statements, or interview re-
sponses, faculty tried to glean qualities of thought and mind, especially as
evinced in research experience and other forms of engagement with the
discipline. Assessments of intelligence thus rested upon a combination of
general impressions and locally defined behaviors—something that an
individual is and something that an individual does.
General Impressions of Intelligence
In offhand comments, reviewers often described applicant intelligence as a

noun or adjective, implying that it is something that one has or is. I men-
tioned in Chapter 2 that at least half of those in the sample instinctively
associated GRE scores with intelligence, although they frequently equiv-
ocated on this. I asked each participant, “What, if anything, does the
GRE signal?” and some caught themselves, mid-answer, overcoming

their instincts with more thoughtful responses. In sociology, for example,
an associate professor responded to my question with the following:
I suppose it signals a certain kind of {Pause} I was about to say native intelli-
gence, but I know that’s not true, because people can dramatically improve
their scores by taking classes—these Kaplan, Princeton Review classes. So
{Pause} I have a hard time knowing what it signals.6
Another sociologist proposed,

GREs tell me something, probably, about—I don’t know—crude about native
intelligence or general intellectual horsepower or something like that. I think
that’s a relevant piece but it doesn’t tell me about the person’s creativity or
sociological imagination or something like that. I tend to look for things in
letters, personal statements, much more than past achievement. And even that
obviously is a real thin predictor.
In committee meetings, discussions of each applicant began and ended with

faculty enunciating their overall impressions of students, and often, that in-
cluded at least one comment that I came to think of as an example of “smart
talk.” Such talk involved language like “intellectually capable” and “plenty
of intellectual horsepower” when judgments were positive, and com-
ments such as “lame-brained” and “I question she has what it takes” when
negative. Broadly, committees in the humanities disciplines displayed more
lenient interpretations of lower scores and grades, but high numbers still
often impressed them. Low scores might be the result of “a bad day” or
“few opportunities for test preparation,” they claimed, but they also said,
“You don’t earn perfect GRE scores by accident.”
Is it appropriate to infer intelligence from scores on the GRE or other
college entrance exams? This question is deeply fraught territory because
of the origins of college entrance and intelligence tests in the eugenics move-
ment and continuing gaps in scores that fall along lines of social identity.7
And like the broader question about what intelligence is, answers about
what can be inferred from GRE scores depend whom you ask. Even the
Educational Testing Service sends somewhat mixed messages. On page 4
of their document “GRE Guide to the Use of Scores,” the ETS insists that
the test measures skills—specifically “verbal reasoning, quantitative rea-
soning, critical thinking and analytical writing skills.”8 They emphasize this
point by placing it inside a large, gray text box within the running text. On
the cover of that same document, however, in fine, italicized print in the
upper right-hand corner, are the words “Assess Ability.”
Rebecca Zwick, author of Fair Game: The Use of Standardized Tests in
Higher Education, acknowledges that if the SAT is the “grandchild” of
intelligence tests, then the GRE and other graduate admissions exams are
“certainly cousins.” Designed to assess skills that are useful apart from dis-
ciplinary content, Zwick argues that the GRE General Test, along with the
SAT, LSAT, and GMAT, tends toward measuring aptitude, whereas the ACT
and MCAT assess specific subject matter learning.9 However, ACT and SAT
scores are highly correlated, and both are promoted as predictors of college
grades. The “IQ test versus achievement test debate is a tempest in a teapot,”
she writes, a minor distinction that has been blown out of proportion.
Zwick quotes sociologist Christopher Jencks:
Many people—including federal judges—think that both intelligence and ap-

titude are innate traits . . . Yet almost all psychologists now agree that while
an individual’s scores on an intelligence or aptitude test depend partly on his
or her genetic makeup, it also reflects a multitude of environmental influences.
Zwick concludes that neither intelligence nor standardized admissions tests

measure “native intelligence” or ability.10
Notably, across the committees I observed, only in the case of one ap-
plicant to one department was a general impression of intelligence—along
with its perceived indicators—enough to generate consensus about ad-
mission. This individual had perfect scores on all three sections of the GRE
and a 4.0 GPA from Harvard University. Unbeknownst to him, he became
an inside joke for the committee when a senior—and typically dour—
professor declared him a “freaking genius.” For the remainder of their
meeting, he became the standard against which many others were jokingly
compared.
Qualities of Thought and Mind
Offhand comments like labeling an applicant a “freaking genius” can speak

volumes about engrained assumptions, and they highlight the tendency to
reduce complicated profiles to stereotypes. During interviews, I therefore
probed respondents to describe more fully what intelligence means for
scholars in their discipline and how they recognize it when they see it. In
these comments, participants indicated that intelligence is not only some-
thing that students have, but something that students do—something that
one embodies and practices. Judging these dimensions of intelligence was
thought to be best accomplished by evaluating students’ experience with
research. As summarized in Table 6, they spoke about intellectual dispo-
sitions as well as qualities of thought such as novelty, rigor, and insight.
A common theme in the humanities was to value qualities of mind that
Table 6 Qualities of Thought and Mind Cited in Relation to Intelligence
Intellectual Intellectual
Novelty Rigor Insight Dispositions Virtues
Creative Analytic Deep Spark Humility

Curious Systematic Reflective Passion Honesty
Imaginative Theoretical Perceptive Intensity Courage
Original Independent Intuitive Grit
Open-minded Abstract “Big picture
perspective”
Critical “Can see
questions”
Precise
Useful
were fundamentally moral in nature, which I characterize as intellectual

virtues.
Intellectual dispositions. Notable among the qualities of mind faculty that

care about are intellectual dispositions, in which reviewers attributed per-
sonal characteristics to the ways in which applicants think, research, or oth-
erwise approach their subject matter. Paradoxically, such dispositions
capture what some would call noncognitive qualities of students’ cognitive
engagement. For example, when speaking of applicants whose approach
to learning or research displayed “spark,” faculty used words like “lively,”
“energetic,” “enthusiastic,” and “excited.” I observed that perceptions of
spark could be especially important as a counterbalance to skepticism about
atypical academic trajectories or subpar test scores or grades. In one phi-
losophy committee, which held high standards for prior disciplinary social-
ization, the committee chair successfully advocated for an applicant who
had attended a small college in the Deep South and lacked formal training
in philosophy. Self-taught, this applicant nonetheless demonstrated excep-
tional insight and energy in her writing statement, the chair argued, and
was potentially a “diamond in the rough” who could inject fresh perspective
into a corner of the discipline that sorely needed it. In the biology program,
which was slightly less selective than most others in the study, faculty
worked hard to recruit students on the basis of spark. The chair emphasized
their interest in students whom other programs might ignore but who would
still likely be successful.
Spark was closely related to affective intellectual dispositions such as
“intellectual intensity” and “passion,” which were frequently mentioned in
answers to my question about what intelligence means in one’s discipline.
Other respondents made direct links with love, such as “You’re supposed
to love truth” and that one should be “in love with research.”
The analogy with love is instructive in that it draws out the distinctly
personal, relational elements of intellectual life that scholars experience but
rarely discuss. In her book Professing to Learn, higher education scholar
Anna Neumann used vivid faculty narratives to illustrate affective and
aesthetic aspects of scholarship that contribute to ongoing scholarly learning.
She concludes that the “construction of subject matter knowledge is hardly
free of emotion and is intimately connected with themselves.”11 Neumann’s
participants relayed stories of rare, but powerful, “experiences of passionate
thought” wherein the beauties inherent in their work eclipsed the usual
struggles that accompanied research, their understanding intensified,
and they felt a sense of exhilaration or intimacy with the subject matter.
As fleeting as these experiences may be, their inspiration fuels everyday
academic life, she argues, and they matter deeply to what it means to be a
scholar.12 It may be that faculty were seeking in prospective students glim-
mers of the same passionate spirit that animates their own work and
careers.
In the unique relationship that scholars have with their work, however,
participants shared with me that intelligent scholars are not only “gripped
by whatever they’re working on.” They also stick with the research process
when the going gets tough. Respondents associated “intellectual grit” and
“tenacity” with intelligence, especially when considering the long haul of
the PhD and today’s competitive research climate. Determination and re-
solve were difficult qualities to infer from the records of these mostly young
students. For those applicants with more experience, however, “persistence
in a research area” could signal persistence through the dissertation. For
others, comments in letters of recommendation that an applicant was
“driven,” “hardworking,” “steady,” or “reliable” in research offered hints
that a student might have the intellectual determination that academic life
demands. To extrapolate from such evidence and become persuaded that
such hints were more signal than noise, reviewers might also look to pat-
terns of grades and courses within a student’s transcript.
Intellectual virtues. In the humanities, participants actively sought intellec-

tual virtue in applicants. Moral qualities such as “intellectual honesty,”
“intellectual courage,” and “intellectual humility” came up regularly in
relation to how applicants seemed to manage their subjectivity and ap-
proach the work of interpretation. Faculty believed intellectual courage
could be inferred from applicants’ willingness and ability to undertake
contentious, complicated, or otherwise difficult projects. They also com-
mented on the value of intellectual honesty and humility. Students could

display these qualities in writing samples through their respect for opposing
viewpoints and disconfirming evidence, and in sensitivity to how their own
background potentially biased their interpretations. For the clearest de-
scription of these qualities, I quote at length Jack, a senior professor of
classics:
Jack: One thing you’re very much looking for is a kind of intellectual
honesty that’s a real feature of quality-of-mind.
JP: What does that mean?
Jack: Above all, that means looking at reasons not to agree with what you’re
saying and treating them fairly. That means making nuanced arguments
instead of overstated ones . . . {Pause} It also involves a certain humility
about—you know, if you are arguing about what a certain expression
simply means at a semantic level. You know if important scholars have
taken views with which you disagree, you are looking for people who
have the courage to disagree but also are aware that . . . there are people
with whom one does not disagree, likely.
JP: That’s very helpful. I also heard intellectual courage mentioned a couple
of times.
Jack: Right. They go—they’re not at all the same. But, do they typically go
together? Yeah, I think so. A lack of profound intellectual honesty is often
a form of cowardice because the kind of intellectual honesty I’m thinking
about requires accepting that you can be wrong—that others’ views are
often entitled to quite a lot of respect. And to be able to be aware of that
and still say what you think is at least one important form of intellectual
courage.
In many cases, faculty folded specific qualities of thought and mind into
broader judgments of intelligence. In a handful of cases in the humanities,
however, they became bottom-line considerations in wrestling with bor-
derline candidates. For example, in the classics program, the committee
debated two applicants from the same, moderately selective public univer-
sity in New England, and ultimately made their decision on the basis of an
applicant’s failure to embody intellectual humility. Jack introduced the first
candidate to the committee by reminding them, “He got his degree in 2010
and has been a legal assistant since then. Has also been a volunteer editor
for a classics website.”
Bill noted, “His letters praise ‘diligence,’ and that he is ‘on the way to
originality.’ ”
“He’s bright and lively,” Linda, the department chair, followed. “Has been
to Athens. There are errors in his Greek, but he seems to have intellectual
courage. He’s curious and is an original character. He has studied with my
old student.”
Jack reminded them that the applicant grew up in rural New England,
and commented, “The place is ideal in some ways for pursuit of classics.
Very pastoral.”
Moving on to the next applicant, Linda expressed amazement that the
student had checked the “some high school or less” box to describe his par-
ents’ education level. Jack said, “He seems bright enough. A little naive. Paper
seemed misguided but intelligent. It’s a good treatment of a new problem.”
Another noted that he “qualified as a diversity candidate,” and Sandy
admitted, “I’ve rated him lower than you all have. The letters stressed
drive and energy more than talent.”
“On general principle, it would be nice to have someone from that part
of New England, from a modest state university,” Linda added. “It’s im-
pressive that they can produce two reasonable applicants.”
Yet with an entering cohort of only four students, the committee agreed
that they should not admit both—that they needed to choose just one. As
they dug into the details of the applications, a committee member discov-
ered a line in one of the letters, describing the second applicant as insensi-
tive to what fellow students thought of his “academic zeal.” Worried that
this alleged lack of intellectual humility might encourage competitiveness
in their small graduate student community, they opted to admit the first
applicant. Qualities of mind, which in the humanities included intellectual
virtues, could thus build up or break down a committee’s enthusiasm and
trust in an applicant.
Research Experience and Disciplinary Engagement
Implicit judgments of intelligence often emerged through inferences made

from GRE scores, but it was through letters of recommendation and ap-
plicants’ records of research engagement that participants gleaned discipline-
specific evidence of intelligence. Consider the comments from Vincent, a
biologist:
JP: What do you think intelligence means here in biology?

Vincent: With how fast science is moving, you have to be able to keep
thinking about the next thing. And so students who can read primary
literature and kind of take several papers and distill them and put together
an idea for what the next experiment should be is really important . . . I’m
looking for students that can critically think about the scientific literature
and students who also have good hands in the lab who can carry out
those experiments they think of.
Intelligence may be fundamental, but it is only valuable to the extent that

it can be demonstrated and thus recognized. In the following narrative from
William, an economist, I highlight qualities of thought in italic, disciplinary
engagement in bold, and overall impressions in bold italic.
JP: What does intelligence mean in economics?
William: That’s a hard question. There are obviously a lot of dimensions of
intelligence . . . First of all, they all have some basic level of mathematical
sophistication, but that can vary a lot. There are people successful in
economics who were mathematically brilliant. They were International
Math Olympians or they were Putnam exam winners and particularly in
economic theory . . . Some of them probably could have been successful
mathematicians. I’m certainly not in that category {Laughing} even though
I was a math major. But then other people are intelligent in the sense that
they think really creatively. And other people are intelligent in the sense
that they’re really good with statistical patterns and data. You’re thinking
about how to use data to get at a particular question . . . You’re looking
for the combination of skills that will serve you well to do research in
the field and it’s some sort of combination. In every field you want people
who think super clearly . . . who are just incredibly clear, logical, precise
thinkers. Although to be honest, there are people who are successful
researchers who think clearly, but they have a hard time articulating
clearly, but they’re so brilliant.
In the same way that successful economists bring a unique blend of tools,
knowledge, and perspective to their scholarly work, the intelligent applicant
was understood to be one whose combination of preparation, achievements,
skills, and qualities of thought signaled capability for producing excep-
tional research. For those students who demonstrated substantive research
experience in their application materials, the committee could more easily
recognize those qualities and, thus, associate them with intelligence.
As William noted, “in every field you want people who think super
clearly.” It was only in humanities, however, that applicants submitted re-
search that faculty could use to personally judge lucidity of thought.13 Fac-
ulty in the humanities scrutinized writing samples with a fine-toothed comb
for evidence of this quality and for evidence of fresh perspective on an issue
of importance to the field. Some social scientists I spoke with lamented the
lack of a writing sample from their applicants, while also acknowledging
that its absence came with positive trade-offs such as reducing the time re-
quired for each file’s review, eliminating ambiguity about how independent
the writing process had been, and obviating the need to debate whether co-
authored work merited the same consideration as a single-authored paper.
Moreover, in the social and natural sciences, an applicant could often
demonstrate clear thinking through facility with quantitative reasoning as

well as the written word.
The importance of research experience. Patricia Gumport, vice provost for

graduate education at Stanford University, wrote that organized research
is the foundation of American graduate education, and selectivity “the key
ingredient reinforcing the pivotal role of the graduate school as a site for
professional socialization and disciplinary reproduction.”14 Especially in doc-
toral programs that strive to develop and produce future “stewards of the
discipline,” it is therefore only logical that evidence of research experience
would serve as a foundation for admissions selections.15 And indeed, I
found that when judging among students on the short list, experience with
and potential to advance the discipline were among the most important con-
siderations that committee members brought to bear on their decisions. A
professor of biology explained:
Probably the most important criteria or variable that I look at is whether the
student has been engaged in research. Because that’s really what we’re trying
to figure out: whether the student can successfully finish a dissertation research
project . . . It’s not to say that I would automatically rule out a student who
had no research experience. But I think my point of view is probably fairly
common amongst most of the faculty.
The physics admissions chair used even stronger language. “If the person
hasn’t taken advantage of [research] opportunities, it’s effectively the kiss
of death, at least from my point of view. I think that probably speaks for
most of the people I’ve talked with about this process.”
With widening opportunities for undergraduate research, faculty increas-
ingly expect that at the point of admission students will already have had
some research experience, but they are not always sure how to accurately
draw meaning from the experience that students report. There were two
common refrains I heard on this topic, and both related to the broader chal-
lenge of incomplete information that plagues admissions decision making.
One interpretive challenge with respect to student research experience is
the question of how well a student’s past experience signals current skills
or future performance. A number of professors expressed to me that they
would ideally like to replicate in admissions their process for hiring faculty
or postdoctoral fellows, in which reviewers judge the quality of completed
manuscripts as a signal of future contributions. Unfortunately, forecasting
a student’s contributions from the admission application and prior research
experience was much more difficult. Substantial learning has yet to take
place, for one, and the nature of skills and dispositions a student may have
gained through undergraduate research experience are difficult to discern

with any accuracy. Faculty in the humanities may have some direct evidence
in the writing sample, but elsewhere, reviewers were forced to make infer-
ences from applicants’ curricula vitae, interviews, and the testimonies of-
fered in letters of recommendation and statement of purpose.
Another common challenge of interpreting student research experience
concerned uncertainty about the relative quality of experiences themselves.
As knowledge spreads among college students that research experience is
desirable for graduate school admission, more students are pursuing these
opportunities and highlighting them in their applications. In some fields,
especially those in STEM, undergraduate research experience is well-
institutionalized in the structure of access to advanced education and
becoming a scholar. In response, faculty feel compelled to “sort out the
quality of undergraduate research experiences.” The gold standard, which
several reviewers offhandedly called “real research,” was understood to
consist of following a study from conception to either publication or pre-
sentation at a national association. Exhibiting some degree of independence
was also critical to “real research.” By contrast, some students’ research
experience was assumed to involve little more than “pressing a button” to
analyze a faculty advisor’s data.
An effect of trying to sort out the relative quality of student research
experiences has been to raise faculty standards about what is necessary to
qualify a student for the research training that doctoral education offers.
The depth and breadth of research experiences available today do indeed
vary widely, with opportunities of course more widely available for students
who attend research universities. Reviewers from several of the programs
admitted their concerns about the rising expectations for research experi-
ence. One, who asked that I not name her discipline, argued that desiring
a degree of independence is understandable, but that admitting increasingly
pre-professional researchers threatens to undermine the developmental, ed-
ucational mission of doctoral education.
Among these ten programs, rising expectations of research experience
were especially salient in philosophy and the natural sciences. About one-
third of the philosophers in each of the two programs I studied worried
that doctoral education was becoming overly professionalized. Whereas
successful applicants might once have been able to submit a slightly revised
term paper, today’s admissions committees expected something much more
sophisticated, original, and polished. Writing samples should demonstrate
the author’s courage and capacity to—in philosophers’ jargon for how
the discipline advances—make a “move” that shifts discourse in the field.
However, several acknowledged that this expectation presumes a level of
socialization that privileged applicants who had received considerable

training and editorial support. In the natural science programs I studied, a
strong base of research experience was valuable because they did not have
structures in place for graduate students to learn the basics; most students
engage as a full collaborator on faculty-supervised research projects from
their first day onward.
A concrete example of a committee discussing a specific applicant’s re-
search record may be useful here as an illustration. Lara had applied to as-
trophysics, and her curriculum vitae listed multiple research publications
accumulated in a full-time research job at a major public university after
completing her bachelor’s degree. This record sounded impressive to me at
first, but the committee was strangely unimpressed. So quiet were they about
this unusually productive student, I wondered if there was something they
were leaving unspoken in my presence. Then they began discussing the very
technical, task-oriented nature of involvement in research, according to the
letters of recommendation. At this, the reason for their ambivalence became
clear. After scrutinizing her background, an associate professor asked, “Is
it our goal to bring in someone who is very productive and can be produc-
tive for me? Or to select those who can be leaders in the field? The applica-
tion provides evidence of coding skill, not science.”
This shifted their discussion to the similarities and differences between
admitting graduate students and hiring postdoctoral fellows, and how these
activities seemed to be increasingly similar in their aims. The committee
chair eventually brought their conversation back around to Lara, asserting,
“What is missing is evidence of scientific intuition, scientific leadership.”
Another associate professor added, “I find no evidence of creativity or
any synonyms for it in the letters.” Ironically, how the committee read her
wealth of technical research experience became her Achilles’ heel. Because
she had years of experience, the committee felt that her potential for scien-
tific leadership, intuition, or creativity should be plain to see. The committee
did not offer her an interview.
Conclusion
Scholars have argued that the informal mental models about intelligence
that each of us hold are at least as important as formal measures of intel-
ligence in how social goods and opportunities are distributed. By their very
nature implicit, these informal views often go unexamined. I found that
there were two dimensions to the implicit model common to many faculty
members in this study. They described intelligence both as something that
a person has or is in a general sense (“they’re so brilliant”), and as some-

thing that one embodies and does (dispositions and “skills that will serve
you well to do research in the field”).
It is possible that a lack of clarity about disciplinary boundaries or quality
standards may compel scholars in low-consensus disciplines, generally, to
cloak the unknown thing they are looking for in vague words whose value
to the whole community is incontrovertible.16 But comparing across fields,
professors expressed specific ideas about intelligence that reflected their
varying disciplinary cultures. In the humanities, faculty assessed how stu-
dents seemed to manage subjectivity, inferring from this such virtues as intel-
lectual honesty and intellectual courage. In one case, doubts about a student’s
intellectual humility—which they inferred through a single line in a letter
of recommendation—even became the deciding factor in that student’s re-
jection. By contrast, it was a strongly quantitative political science pro-
gram that admitted a student dubbed a “freaking genius” on the basis of
his perfect test scores and perfect grades from a famous university. A
study recently published in Science demonstrated that disciplines vary in
gender and racial inequality based on whether scholars in that field believe
innate brilliance is necessary for success. Embedded beliefs about intelli-
gence within the disciplines are thus an important part of the academic op-
portunity structure, and arguably a topic worthy of further research.17
When describing to me what intelligence means within their disciplines,
many participants used active, detailed language to describe specific
research-related skills and qualities of thinking that facilitate scholarly
work. To divine this aspect of intelligence, the applicant’s record of disci-
plinary engagement was key. Committees might get excited about what they
perceived as “raw intellectual horsepower,” but they also wanted to see intel-
lectual dispositions conducive to creative, insightful, rigorous research.
Their ideal applicant would have a clear record of engagement in the discipline
with which to assess these intangible qualities. Their goal was to admit stu-
dents who were not only capable of passing their courses, but whom they
could envision becoming respected scholars. As one economist put it, ideal
doctoral students are “the smartest, most creative people you can find.”
In envisioning applicants’ possible futures as scholars, I was struck by
my participants’ almost single-minded concern with potential for research.
Only in the natural sciences did they even mention an applicant’s potential
for teaching, and no one speculated on applicants’ potential for service or
advising. This emphasis on research may simply have been a matter of the
social context. In elite doctoral programs located in research universities,
teaching is fundamental, but research and writing are sacred. Regardless
of how many people read or cite them, one’s publications create a physical,
indelible legacy of scholarship. More pragmatically, research is the activity

rewarded in research universities’ systems of evaluation for hiring, tenure,
and promotion. It follows that research would also play a formative role
in evaluations for graduate admissions. Over and above their possible fu-
tures as instructors, faculty sought applicants with potential to make inde-
pendent, creative contributions to their collective intellectual record.
Although faculty could speak eloquently about intelligence as a dis-
ciplinary cultural practice, it was the other dimension of intelligence—
something innate, decontextualized, and perceived through general
impressions and GRE scores—that dominated the “smart talk” that I ob-
served in committee deliberations. Ideas like “native intelligence” and “raw
intellectual horsepower” were also the most frequent type of answer to my
interview question about what GRE scores signal. Such thinking is remi-
niscent of Harvard law professor Lani Guinier’s critique of admission test
scores’ misuse:
It is as if this test functions as a thermometer. And you give each person the
test as if you were taking their smartness temperature. And that, unfortunately,
is not how the test functions. Even the test makers do not claim it is a ther-
mometer of smartness. All they claim is that it correlates with first year college
grades.18
ETS has taken steps to revise the GRE away from the format of an IQ test.
Its most recent overhaul, which they have promoted as a more “friendly . . .
flexible test-taking experience,” reduced the amount of rote vocabulary re-
quired for high scores on the Verbal Reasoning section of the exam, among
other changes.19 Their revisions have the potential to reduce group dispari-
ties and better distinguish among test takers on the upper end of the score
range, but improvements to the instrument will not change how faculty in-
terpret the scores.
From professors’ hedging about what GRE scores mean, and from the
nuanced views of discipline-specific intelligence that they articulated, it is
clear that many of them possess a more sophisticated mental model than
the one implied in their discussions with one another. Most know that GRE
scores are only modestly correlated with measures of intelligence, and that
admission tests are not designed to measure it at all. Yet in contrast to some
thoughtful and measured discussions I observed about other topics, smart
talk in committee meetings was off the cuff, to the point of appearing perfor-
mative. Why? I have spent some time trying to make sense of this, because
faculty falling back, in practice, on a more simplistic view of intelligence than
the one they actually hold could be seen as a significant flaw for the integ-
rity and legitimacy of admissions evaluations.
Reflecting on my own assumptions as a researcher in light of Goffman’s

dramaturgical theory was instructive.20 Compared to faculty perspectives
obtained through recorded interviews, which might be tainted by partici-
pants’ desire to express socially desirable views, I viewed the admissions
meetings as an opportunity to see behind the curtain and engage with
“backstage” faculty behavior. But committee members themselves may have
conceived of their group meetings differently. For them, the meetings
were in some ways the front stage of admissions work. The meetings were a
place to perform evaluation using roles and scripts that they thought their
colleagues would interpret as socially appropriate. Awarding ratings to
individual files was a more private, backstage type of work. Further, it is
impossible for me to know what email communication transpired in the
nine programs in which I was not on the committee email list, or what ad-
ditional deliberations about students may have occurred in hallway conver-
sations or private office meetings. To the extent committee members thought
of their meetings as a site for front stage behavior, their glib associations of
GRE scores with intelligence (as described here and in Chapter 2) can be un-
derstood as performance of an evaluative script and scholarly role that
they expected their colleagues would find appropriate.
And indeed, whether it was due to either mutual agreement about its ap-
propriateness or faculty conflict aversion, I observed that it was rare for
people to speak up and correct their colleagues for uncritically using GRE
scores as proxy for either the intellect that graduate school requires or a
student’s likelihood of success. Current research indicates higher scores are
positively associated with first-year grades, but the relationship is more
complicated when analyzed in terms of real grades earned. A compendium
of research published in 2014 by ETS reported that less than half (43 per-
cent) of the students who earned combined GRE scores in the top quartile
also earned first-year grades in the top quartile. GRE scores explain only
about 10 percent of variance in first-year grades, and they have only
modest relationships with formal measures of intelligence and longer-term
outcomes.21
A more sophisticated treatment among decision makers is therefore crit-
ical, both to improve the integrity of admissions review and because mis-
construal of the GRE as a “thermometer of smartness” may feed racial and
gender stereotypes that can contribute to test score disparities. A large body
of research has found that women and non-Asian students of color may
underperform on standardized tests due to underlying fears of earning
scores that will conform to negative stereotypes about their identity group’s
intellectual abilities. For students subject to this risk, which is known as
stereotype threat, the test score is a noisier signal of their skills. And in a
painfully ironic twist, the threat of underperformance is especially strong

among students who feel most closely identified with academic success. The
prompts included in test instructions appear to shape susceptibility to ste-
reotype threat. One of the first experiments on this phenomenon, which was
published by Claude Steele and Joshua Aronson, administered a test to stu-
dents that included items from the GRE. They found that mentioning race
or ethnicity in the instructions induced stereotype threat in African Amer-
ican students.22 In contrast, a recent study eliminated performance gaps
simply by highlighting a test’s fairness within the instructions.23
There is also some evidence of stereotype threat’s flip side, stereotype lift,
in which a privileged group’s knowledge of common stereotypes about in-
telligence elevates their self-worth, thereby boosting their test performance.24
Taken together, this research helps explain why test score disparities may
persist despite structural changes to the instrument and narrowing gaps on
other academic indicators, such as grades. Scores capture qualities of test
takers, to be sure, but also qualities of the environment in which the test is
taken, including “threat[s] in the air” about how intellect will be judged.
On average, test scores from historically privileged groups may therefore
be persistently inflated, and those from historically marginalized groups
somewhat depressed. To strengthen review and reduce inequities, the search
for intelligent life needs to extend beyond these familiar forms of achieve-
ment. Reviewers need to attune their gaze to see intellectual strengths
they may have previously failed to notice and potential in the discipline that
can be nurtured through continued learning.25 The thorough, open-
minded, self-critical approach to evaluating writing samples that I observed
in the philosophy programs encouraged me that such review is possible.
Across programs, many took for granted that an implicit task of admis-
sions is to identify and select the brightest students in the pool. How can
something as difficult to pin down as intelligence play so fundamental a
role? An important part of the answer, I think, is that intelligence is ho-
mologous to the broader idea of merit. In both cases, faculty committees
look for and weigh markers of a quality they struggle to define, whose
meaning escapes consensus, and about which they have little research
knowledge. Yet because merit and intelligence are central to excellence
and to scholars’ own sense of academic self-worth and belonging, the very
act of striving for these interests brings legitimacy to the selection process,
even if it is done in flawed ways. This striving means committees make many
inferences and guesses based on limited evidence and the comforting, if spu-
rious, precision of quantitative metrics. As Chapter 6 will emphasize, those
inferences are especially tricky to make when evaluating across cultural con-
texts and when evaluating international applicants.
chapter six
International Students and Ambiguities

of Holistic Review
The way international applications work is that there is a cloud of
random applications, but good applications come in pipelines.
— Biology admissions chair
G rowing demand for American higher education from international

students is a trend that spans fields of study, institutional types, and
undergraduate, master’s, and doctoral programs alike.1 Between 1900
and 2000, the number of college-aged students enrolled in postsecondary
education exploded from 500,000 to 100 million, a stunning 200-fold in-
crease. Decolonization and the expansion of democracy and human rights
help explain widening participation in higher education, along with eco-
nomic, scientific, and technological development. In countries around the
world, sociologists Evan Schofer and John Meyer argue, these broad trends
have converged to institutionalize a new model of society in which elite
classes established around “schooled knowledge” have replaced traditional
landowning, business, political, and military elites. In the United States,
adults with graduate degrees now comprise almost 50 percent of those in
the top 5 percent of the wealth distribution, up from 28 percent in 1989.2
Colleges and universities are organizational lynchpins to these trends, with
the establishment of new universities around the world cultivating demand
for well-trained faculty and U.S. doctoral programs providing that training
ground for students from across the globe.
International students have driven rising applications, enrollment, and
degrees awarded in U.S. graduate education. Overall first-time graduate en-
rollment has grown by an average of about 3 percent each year since 2001,
largely due to international students, with Chinese students leading the
Table 7 Percent Change in International Offers of Admission by Country or

Region of Origin
2010–2011 2011–2012 2012–2013 2013–2014
International Total +9 +9 +9 +9
Country of Origin
China +21 +20 +5 0
India +2 0 +27 +25
South Korea −2 0 −10 −9
Taiwan — −4 −3 −6
Canada — +9 −1 +4
Mexico — +6 0 −1
Brazil — +6 +46 +98
Region of Origin
Africa — +10 +7 +3
Europe — +2 0 +2
Middle East +16 +17 +12 +9
Note: Prior to 2011, data were not collected for Taiwan, Canada, Mexico, Brazil, Africa,
and Europe.
Source: Council of Graduate Schools, 2013.
way.3 Of the 220,000 international students enrolled in U.S. graduate pro-

grams in 2013, 34 percent were from China alone. In the last four decades,
the percentage of doctorates awarded in the United States to international
students has more than doubled.4 Table 7 reports the percentage change in
offers of admission to international students from U.S. universities, in-
cluding the years of data collection on this project.
International Students and Organizational Interests
Enrollment management is big business in higher education today, and for

good reason. Colleges and universities use the composition of their student
bodies to serve a variety of organizational interests and to signal their pri-
orities to stakeholders. Yet pursuing one set of interests may require
concessions on another, as illustrated in John Cheslock and Rick Kroc’s
analysis of the trade-offs for higher education institutions among fair
access, containing costs, and strengthening academic profile.5 In this vein,
whether one sees rising international student enrollment as a positive trend
for U.S. graduate programs may depend upon whether or not one’s metric
for quality is equitable access, financial security, or strength of academic
profile.
International Students and Ambiguities of Holistic Review 135
International graduate students clearly benefit the United States as fu-

ture “contributors to the U.S. economy as professors, researchers, and en-
trepreneurs.” One recent calculation estimates that international student
enrollment has brought $24.7 billion to the U.S. economy and turned U.S.
higher education into a top service-sector export.6 Colleges and universi-
ties often embrace the demand from international students, as well, both
for financial and intellectual reasons. Public institutions in particular rely
on tuition revenue from non-resident students to offset declining state ap-
propriations, with few states capping tuition increases for out-of-state stu-
dents and international students ineligible for most forms of federal and
state financial aid. International students have also helped sustain faculty
research in STEM disciplines amid the declining share of domestic students
pursuing such majors. International students therefore fill important gaps
in graduate education today.7
Faculty in highly ranked graduate programs may also see international
student enrollment as a strategy for negotiating two pressures—increasing
diversity and maintaining a strong academic profile. As I reported in
Chapter 2, programs that struggled to recruit U.S.-born Latino, black, and
Native American students with extremely high GRE scores would use
well-credentialed international students in one of two ways. Some pro-
grams, such as economics and physics, elevated national origin to the level
of race/ethnicity, gender, and socioeconomic status as a type of valued
diversity. Doing so was consistent with those fields’ disciplinary cultures,
for the major research initiatives in high-energy physics today are deeply
international endeavors and some of the most prestigious economists and
economics programs are in Europe, not the United States. A strategy used
in some other programs was to claim a strong commitment to racial/
ethnic diversity, but downplay whether they were attaining that diversity
through enrollment of domestic versus international students. In both
scenarios, the program reframed how international students “counted” in
programs’ efforts to present a diverse and conventionally high-achieving
image.
Especially in STEM fields, a commonly expressed concern is that inter-
national students may be crowding out domestic students. In 2012, 54.7 per-
cent of international students enrolled in graduate programs in STEM
fields compared to just 17.3 percent of U.S. students. Yet by controlling for
self-selection into graduate study in various fields, research by higher edu-
cation scholar Liang Zhang found more support in non-STEM fields for
the displacement of domestic students. Some graduate programs try to avoid
crowding out domestic students by giving special attention to international
students whose applications indicate independent sources of funding for
their education. Like institutional diversity fellowships for domestic stu-

dents, independent funds for foreign nationals expand program capacity,
allowing them to enroll more students than they anticipated and to avoid
scaling back their support of U.S. students. Yet with whites composing a
declining share of the national population, some scholars note that achieving
diversity in U.S. postsecondary education by increasing international
student enrollment does little to shift the needle on educational inequality
in the United States. Higher education and law scholar Michael Olivas, for
example, has argued that the trend to establish overseas branch campuses
diverts attention from the needs of persistently underserved communities
here in the United States.8
For graduate programs like those I studied, all of these countervailing
forces converge to make rising demand from international students a bit of
a conundrum for enrollment management. These programs are committed
to admitting students with full funding and to maintaining extremely
strong academic profiles. Many could fill their entire cohorts with conven-
tional academic achievers from outside the United States. However, their
values are to some degree symbolized by the backgrounds of their students,
and central though it may be, conventional achievement is not their only
value. They strive to create cohorts of students that are “balanced” on a
variety of dimensions, and many feel obligated to take steps toward re-
ducing racial inequalities. Every program is different, but under these con-
ditions, I found that there are limits to the number of international students
that most faculty feel they can legitimately admit.
Ambiguities of Reviewing International Student Applications
The swelling volume of applications from international students poses an

additional challenge: How should they all be reviewed? Applications from
foreign students for U.S. graduate programs have increased for nine straight
years (by 10 percent in 2014), and overall applications to graduate pro-
grams in research universities have increased at an average annual rate of
5.3 percent for the last decade. For the typical admissions committee, this
steady growth amounted to a daunting 72 percent more applications to re-
view in 2011 compared to 2001.9 The greater the volume of applications,
the greater the strain it places on the system of holistic review. Evaluation
can easily become routinized rather than reflective.
Internationalization also complicates how faculty interpret applicants’
files because students come from widely varying cultural and national
contexts whose distinctions are mostly unknown to faculty.10 Should the
metrics used to assess international applicants be the same or different

from those used to assess American students? If different standards are war-
ranted, should Canadian, Cambodian, and Chinese applicants’ files also be
read differently? Should faculty expect higher quantitative GRE scores
from countries with deep cultures of test preparation? If so, how much
higher should those scores be, and do the usual concerns about gender dis-
parities apply? How well do indicators of English skills in an application
correspond to practical fluency? Should a letter of recommendation that
comes across as odd be understood as the product of different cultural
norms or a red flag about the applicant? These are just a smattering of the
many questions that faculty worry about in trying to construct an equi-
table process of review.
In this chapter I explore how faculty wrestle with a few of these ambi-
guities.11 First I discuss the difficulties that faculty most frequently reported
when tasked with reviewing applications from international students. Then,
in light of previous research that suggests that ambiguities create conditions
in which stereotypical views thrive,12 and having shown that some common
assumptions about Asian students correspond to the “model minority”
paradigm, I offer a brief review of current research about test score va-
lidity and cheating among international students. It would be impossible
to remove ambiguity from the admissions process. Nevertheless, having the
best available information about circumstances in which it tends to crop
up can keep us from succumbing to the stereotypes and cognitive biases
that ambiguity tends to encourage.
A note about labels: Students from China accounted for 29 percent of
all international applications to U.S. universities in 2011–2012, the second
year of data collection in this project,13 so it should come as no surprise
that when faculty spoke about the difficulty of evaluating international ap-
plicants, they frequently referred to “Chinese applicants,” often tying them
to Taiwanese, Korean, and Japanese applicants under the label “East Asian.”
Unless mentioned otherwise, faculty narratives focus on applicants from
this group.
Quantitative GRE Scores and Admissions Ethics
What should reviewers make of international students’ quantitative GRE

results? As I have discussed, faculty relied upon very high GRE scores as
one of a few criteria that could reduce the pool to a size for which close
reading of the application felt manageable. Using GRE scores to winnow
the pool was convenient, and they treated it as a standardized signal of one’s
intelligence, belonging, and likelihood of success. However, they did not al-
ways see scores this way for international students, whose formal training
in basic skills and test preparation might be different than American stu-
dents’. “The educational infrastructure leading up to this test is not the same
across countries,” one physical scientist put it, “so we set the bar at different
places depending on the country and gender.” A social scientist born in
Italy commented more specifically on the relative quality of test prepara-
tion that he believed to characterize different regions:
European students tend to be pretty bad in performing in the standardized
tests . . . They don’t get exposed to this kind of test in their regular curriculum.
For example, I’d never done a test in my life until I had to face the GRE . . .
[So] we tend to underweigh the scores of students from South Asia for the very
simple reason that we know that some of them actually go to school to do the
GRE for months. There’s not a lot of signal in the scores from South Asian
students, especially students from China.
His perspective was hardly unique. A philosopher also shared, “There cer-
tainly is a kind of stereotypical {Pauses, apparently catching himself} Chi-
nese student who will have astronomical test scores.”
With a large pool of East and South Asian applicants who have high
quantitative reasoning scores, many participants came to set higher GRE
score expectations of Chinese, Korean, and Indian applicants than of Amer-
ican and European applicants. “If a kid from the PRC has not essentially . . .
perfect scores on GRE exams,” one linguist admitted, “they’re regarded as
probably brain dead.” In a few fields, however, faculty read Chinese students’
higher mean scores as a sign that the test was less useful for comparing
them, to the point that some dismissed the scores entirely. Describing ap-
plicants from China, Japan, and Vietnam, a midcareer scientist who led his
program’s admissions effort noted:
The scores on the standardized tests are just out-of-sight, just off the charts.
So you can basically throw that out as a discriminator. They’re all doing 90th
percentile and above. The domestic students were all over the place so there
was actually some spread, some dispersion . . . so you could use that more as
one of the quantifiers.
Another, who had been born outside of the United States himself, said,
To evaluate them [Chinese students] relative to Americans and Europeans is
not so easy. The Americans will not push to take the prep on the GREs and
sometimes do very poorly on them . . . If you had to admit them strictly on
the basis of the GRE, we’d basically have . . . all the kids would be Chinese.
There’s no doubt about it.
His comments reflect a broader pattern of faculty attributing high scores

from Asian students to a deep and powerful test-taking culture. This line
of thinking branded students from China, in particular, as “trained to re-
ally achieve very high scores” and supported the notion that China’s “test-
preparation industry is second to none in the world,” and a “well-developed
machine.”
Although standardized testing has a relatively short history in the United
States, in China it dates back to the seventh century A.D. At that time, em-
perors created the Civil Service Examination to identify top performers
and place them into high-ranking positions that bestowed “financial re-
wards, prestige, power, [and] fame.” These highly consequential exams
birthed an intense focus on rote memorization and test-taking skills, sup-
ported by a network of test-preparation schools and “test-coaching” books.
As an informal system of shadow education, China’s test preparatory
infrastructure has evolved into a multimillion-dollar industry and a crit-
ical factor in students’ process of applying to U.S. higher education
programs.14
Thus, even as many professors had come to expect high scores of appli-
cants from China, they also distrusted that those scores signaled the same
things they stood for among American applicants. And without the weight
of their associated cultural meanings, many faculty felt they must be taken
with a grain of salt. Along with widespread questions about the amount of
test preparation that went into producing student scores, participants openly
expressed concern that GRE scores of students from China lacked credi-
bility because “people game these things” and there may be cultural differ-
ences in what counts as fair test-taking, gaming the system, or outright
cheating.
The suspicion that cheating may be quietly prevalent throughout China
was a major concern for some reviewers. Lowering his spectacled eyes, one
professor I interviewed asked me, “You know about the cheating, don’t
you?” I did in fact know a little about the recent cheating scandals, but be-
cause I also knew that his own research included some Chinese history, I
applied the standard interviewing technique of feigning ignorance to keep
him talking. In return for this bit of deception, I was treated to a brief his-
tory that dipped into the Civil Service Exam and gao kao (the National Col-
lege Entrance Examination, or NCEE) as foundations of a culture in which
students will push the boundaries of well-being and ethics to earn high
scores.
During English-examination season, he said, it is not uncommon to find
advertisements for “hired guns” (test takers) who charge students and their
families exorbitant prices to take the TOEFL exam or, in some cases, to pro-
vide answers in advance.15 And in 2001 the Educational Testing Service,
which administers both the TOEFL and the GRE, filed a lawsuit against
China’s largest and most profitable test preparation company for illegally
publishing ETS materials. Official correspondence from ETS notified grad-
uate school deans across the United States of the situation.16 The effect of
these scandals and others, he argued, has been to call into question all test
scores from Chinese applicants. In a system as competitive as this, when
“you’re not 100 percent sure whether they are the person who took it,”
doubts about the trustworthiness of a test score could be used to discredit
the applicant in general.
Others distrusted the signaling value of the GRE for Asian applicants be-
cause it does not capture the work that they worry Asian students will
struggle with in graduate school: open-ended, creative scholarship. This pat-
tern was exemplified in physics and astrophysics, two programs in which
most or all faculty expected strong performance on the GRE, especially the
Physics Subject test. William explained:
If we let in all of the top GRE scores automatically, then we’d have all Chinese
students, which aren’t bad. But historically what’s been found is that students
from China tend to not succeed as well as their GRE scores would indicate
since they’ve learned over the years to learn towards the test. This is a general-
ization, but they generally don’t perform as well as graduate students when it’s
more open-ended because they’ve learned—they’ve been trying to work towards
a particular task their whole time in undergraduate. It’s something that we look
at, but we do a lot of talking about how it is an imperfect measure.
Another physicist discussed his perception of Chinese and Indian students’

scores in relation to the habits of mind that graduate school requires:
It is incredibly rare for any American undergrad to score more than seventies
or eighties in terms of percentile. Like 99th-percentile scores, invariably, are
from China or India or something like that. But often the students aren’t as
successful, because what that proves is that they’re really, really good at mem-
orizing formulas and solving problems based on them. And the way physics is
taught in—as I understand it from talking to people who went through that
system in India and China—is, it is very rote . . . You will solve problems until
you can solve them in your sleep . . . But if that’s what you’ve been trained to
do and you haven’t been trained to think about things that aren’t reducible to
simple formulas, you’ll have a lot of trouble in grad school.
A high GRE score might be earned through single-minded focus on rote

knowledge and by gaining familiarity with the type of questions the test
asks, but it offers no insight into the dispositions and skills that faculty as-
sociated with independent scholarship, nor the English-language skills they

associated with real fluency.
In sum, the difficulty of disentangling test scores from the social context
in which those scores were produced, combined with suspicions of fraud,
made it difficult for faculty to trust Asian students’ Quantitative GRE scores.
One thing was clear: their scores must be interpreted differently than
domestic students. Where many expressed worries that test preparation ren-
dered Chinese students’ scores suspect, not once did a respondent express
similar concerns about formal test preparation courses convoluting the
meaning of a wealthy American student’s score. And although faculty em-
phasized the importance of interpreting GRE scores within a national
context, most were reticent to interpret an African American’s score dif-
ferently than a European American’s score, or a man’s differently than a
woman’s.
The Search for Credible English Skills
Faculty also considered English-language fluency central to international

students’ academic preparation and potential, and many found it difficult
to gauge the credibility of Chinese students’ English skills using information
contained within the application—namely, TOEFL and GRE scores and their
personal statements. There was broad disciplinary variation in the reasons
faculty gave for their concerns about English fluency. Physical scientists
worried about student ability to teach undergraduate courses, whereas those
in the social sciences and humanities wanted to avoid editing students’
written work with a fine-toothed comb. A linguist recalled:
A large proportion of our applicants are from other countries . . . We had a

student like that who was really absolutely brilliant but who got here and
wasn’t successful because their English was not good enough. So with that
kind of student, then, my primary concern is to make sure that their English is
good enough, so that when they come to write their dissertation, they don’t
have to have somebody go over every sentence with them.
Remembering another student, a second linguist ruefully summarized, “He

couldn’t do anything because he couldn’t speak English. It was awful. And
you know, we only have four or five places, so it’s awful to waste one on
somebody who washed out in a year.” A senior professor commented, “I
get really tired of editing . . . so some people do get knocked because of their
language.” One person succinctly captured the ambivalence with which I
heard many faculty describe Asian students. They tend to be “really bright
and interested,” she said, “but their English is just not good enough to
keep up.” As with their evaluations of domestic students, admissions com-
mittees hoped for students who would work hard, but not struggle too
much, because struggling slows students down and drains faculty time
and resources.
And like their interpretations of scores from the quantitative section of
the GRE, some faculty admitted suspicions that Asian students’ English
skills were weaker than implied in the application because applicants had
gamed or cheated the system, or had forged personal statements. One
frankly put it, “I think there’s something notorious out there. The TOEFL
scores somehow can get jacked up.” Like many of the assumptions I picked
up on in our interviews, I probed respondents about what led them to be-
lieve this. Typically, they cited vague memories of publicity from the cheating
scandals or memories of isolated instances in which students arrived on
campus with subpar English skills. Several also discussed rumors of stu-
dents contributing illegally to banks of GRE items after completing the
exam or of hiring someone to stand in for them on the TOEFL or phone
interviews. Recalling his experience at another university, a sociologist
recounted:
There were people applying from China who were getting 100 percent perfect
scores on the verbal part. These are people for whom English is not only not
their native language, but they don’t live in an English-speaking country. And
we found out later there apparently was all kinds of cheating going on.
Perfect TOEFL or GRE Verbal scores are simply not expected of Chinese
students because of the context in which those students are growing up.
Although the economics and physics committees liked to see extremely high
or perfect Quantitative scores, none expected perfect Verbal or Writing GRE
scores—even of American students. Therefore, when such scores showed
up in the application of a U.S.-born student, it was a remarkable surprise,
and when they appeared in a non-native English speaker’s file, it was con-
sidered suspect. And, indeed, uncertainty about whether their GRE Verbal
or TOEFL scores might be ethically tainted often cast doubt upon Chinese
applicants’ fluency, generally.
Those faculty who discussed efforts to gauge fluency from the applica-
tion alone described it as an almost impossible task. In the sociology pro-
gram, several faculty stated the perception that the best evidence for
applicants from China was having studied in the United States, for having
American transcripts and taken the GRE in the United States provided
“a U.S. basis to judge them. They’ve done work in English.” Some programs
conducted video interviews with prospective international applicants to
Table 8 Patterns of Interviewing across Programs
Interviewed All Students Interviewed Only International No Interviews

on the Short List Students on the Short List Conducted
Physics Political Science Economics

Astrophysics Biology Sociology
Linguistics Philosophy Classics
“deal with the paranoia” of admitting students with poor English fluency
or inadequate research experience (see Table 8).
Portraits of Admissions Interviews
In two programs I had the chance to observe several of these interviews

and the faculty debriefings that followed. For the first set I attended, three
male professors of astrophysics huddled around a laptop in a nondescript
conference room. They joked with one another that the time difference—
which had students in China interviewing in the middle of the night—was
a good test of interviewees’ readiness for the rigors of graduate school. After
connecting successfully, Victor, an associate professor born outside the
United States, informally took the lead, and others only occasionally piped
in. He congratulated the applicant—I will call her Chun—on making the
short list, and gave her basic information about the number of applications
they had received, the number they were interviewing, and what the selec-
tion process would be going forward. I noted that right from the begin-
ning, the interview’s tone was comfortable and conversational, like most
interactions I had observed among this committee. The easygoing tone that
Victor and the others set seemed to put Chun at ease.
Transitioning into the main topics of the interview, Victor said that the
committee was “impressed” that she had worked with researchers in two
major areas of the field. She responded that she was most “excited” about
one of them, and spoke with some awe about just how much it was pos-
sible to know with recent advancements in methods and technology. What
she may or may not have known was that the area she cited as a primary
interest was far from the intellectual subject matter that two of the three
committee members present examined. They asked her to talk about her
research interests, and after briefly describing a line of inquiry, she added,
a bit sheepishly, “I am open to learning anything in the first year.” They as-
sured her that they were not “expecting her to define a thesis,” but by my
read, they were looking for more detail than what she was giving.
I sat a few feet behind the committee members during the interview, and
each time I looked up from the field notes I was taking, I could see she was
wearing a wide grin. It appeared that she did not stop smiling from the mo-
ment the interview started until the time they disconnected from the video
conference. I was reminded in the moment of William Tierney and Estela
Bensimon’s idea of “smile work,” an impression management strategy that
women sometimes employ to “fit into departments with a tradition of male
dominance.” Knowingly or unknowingly, women may present themselves
as “pleasing and agreeable” in order to avoid being stereotyped as overly
aggressive.17 Her smiles were not lost on the committee. After finishing the
interview and disconnecting, they leaned back to debrief. They discussed
her “very nice personality,” with one remarking that she was “glowing.” One
admitted he hoped that “she would go into greater detail” about her re-
search, and in the only comment about language, another commented that
she was “not the most articulate person we’ve interviewed.”
“Good but not great . . . Maybe not as serious about us as some,” the
committee chair concluded, to nods from the others. My read of the de-
briefing was that they wanted to like Chun because she seemed so very lik-
able, but that they did not see her as sufficiently professional, at least when
compared with others whom they had interviewed. Interestingly, none of
their comments related to her English skills, which they previously shared
with me was a primary motivation for the video interviews, or other issues
specific to her status as an international applicant.
Finishing up their conversation in the minutes before the next interview,
two of the committee members commented on gender dynamics they had
observed in interviews over the years. Victor said he felt women tended to
give concise responses and “speak only when they know what they are
talking about” whereas men will “just keep talking to convince you that
they know what they’re talking about.” It was not clear whether they
made these comments because I was in the room or because questions about
the depth of Chun’s responses hung in the air after her interview, but it
was clear that the pattern they noted with women had been true of Chun.
Similarly, the pattern they attributed to men was evident in the next inter-
view, with Bradley, a male in his thirties.
The two other committee members joined the huddle to participate in
Bradley’s interview. They may have judged Chun “good, but not great” after
her interview, but that was the general impression of Bradley held by two
of the five before his interview began, largely because of an unconventional
trajectory into the discipline (he had worked for several years in the arts)
and his undergraduate education at a moderately selective public univer-
sity. He had just barely made the short list, and they wanted to video in-
terview him to gather more information about his background. The two
skeptics wanted to be present to take a closer look, and the three who were
already impressed wanted to attend out of curiosity—and to make sure the
skeptics did not grill Bradley too harshly. By the end of the interview,
however, he had completely won them all over with the story behind his
unusual background (a triple major, including astrophysics and art), his
obvious ease discussing his research in detail, and by the incisive questions
he asked the committee. For example, whereas Chun’s first question for
the committee was “What is it like to live in your town?,” Bradley’s first
question for the committee was, “It seems from the website like you distin-
guish your department in specific ways. Could you talk about your current
priorities as a department?” He even went so far as to directly inquire
about the committee’s thoughts on a personal and legally protected char-
acteristic that he disclosed in the personal statement, and which, indeed, one
member of the committee had raised as a red flag in their last meeting. “We
think it is worth something,” one of them assured him. Victor added, “It’s
worth a lot.”
In the closing back-and-forth comments between Bradley and the com-
mittee, the chair’s tone and comments seemed to presume admission and
enrollment, and in debriefing his interview after it finished, the committee
displayed none of the ambivalence or equivocation that I heard after Chun’s
interview. Instead, discussion repeatedly included such words as “superb.” It
was clear that they wanted badly to attract him to their program, and indeed,
at the end of the process he was one of the first to be added to their “admit”
list. In the end, however, Bradley chose to enroll elsewhere.
I do not mean to imply that Chun’s and Bradley’s interviews represent
how the committee judged women versus men or Chinese versus American
students. This committee was thoughtful in their deliberations, and the con-
clusion to my field notes from that day includes the line, “If I could admit
two of the four people interviewed today, I would select the two men.”
However, these interviews did draw out just how much more evaluation
may take place in interviews than is formally communicated to students.
As a professor in another department admitted to me, “That’s what you’re
going to tell them, that you’re going to screen them for language. But frankly,
we think a little broader.”
As with GRE scores, suspicions of fraud surfaced in respondents’ stories
to me about their interviews with students from China, South Korea, and
Japan. Some faculty admitted that they would continue to be skeptical of
international students’ English skills until they arrived on campus. A member
of the committee described above recalled the case of a recent South Korean
applicant:
They interviewed him on the telephone. His English was fine, and he was very,
you know, his credentials were good. We admitted him. But he couldn’t talk
English. I mean he couldn’t understand what you said to him. He couldn’t
speak English. The only thing we could think was that he’d gotten a ringer to
do the interview.
Only one person openly acknowledged that the widespread skepticism

about Asian applicants’ English skills “seems in some ways like racial pro-
filing.” Yet a posture of distrust was pervasive, and consistent with the
contagiousness of stereotypes found in social psychological research. Stereo-
types spread through rumors, storytelling, and innuendo as well as through
firsthand experience.18
Cultural Distinction Work
Interviews were also important to many faculty in managing a third ambi-

guity: the perception that students from China are so similar to one an-
other as to be essentially fungible. Given that the basic task of admissions
evaluation is making relevant distinctions among applicants, faculty ap-
proached the short list with a desire to identify doctoral applicants’ unique
qualities and contributions. Some admitted to me in interviews, however,
that they found this more difficult to do with Chinese applicants, especially
using only the application materials. The chair of the physics program ex-
plained this in some detail:
There will be awful, many, Chinese students that look very, very similar. They
have similar scores from similar universities, similar everything. They all get
stellar recommendations because that’s the style of the Chinese recommenders.
And how exactly do we tell the difference? Well, we interview. And that gives
you much, much better sense, certainly. Each of these people that were just
numbers and looked exactly the same, suddenly, they have different person-
alities and different interests and they become alive. So that’s great. It really
works very well and not restricting really strictly just to language screening.
Difficulties in seeing differences among students from China began with

seeing their names on the applications. “How do you compare six students
from China, who all have the same last name?” one humanities scholar
wondered aloud to me. Although her comment overstated the pattern, it is
true that the United States has a much wider distribution of surnames than
China. The hundred most prevalent surnames in China account for 87 per-
cent of its huge population, whereas only 17 percent in the United States
have one of the hundred most common surnames. Li, Wang, and Zhang
alone comprise 22.4 percent of the Chinese population, and the Chinese
phrase “three Zhang, four Li” translates in English to “just anybody.”19
Their expectation that Chinese students would have high quantitative

GRE scores and weak English skills also reflected the troubling tendency
to think of students from China not as individuals, but a profile of group
averages.
To grasp individuality, several committees involved additional faculty
members in the review process. A Chinese-born professor in one program,
not a member of the committee, cross-checked ratings of Chinese applicants
when there were specific ambiguities in an applicant’s file or when readers
gave widely divergent ratings. Two departments each informally dubbed
one of their faculty members who had served on the admissions committee
for several years their “China specialist” or “China expert.” This role was
thought to reflect significant social capital, with one described as “know[ing]
pretty much all the top astrophysicists in China. He knows what all the
good universities are.” In one program, this expertise was instrumental in
conducting what amounted to a separate evaluative process of the hundred
or so applications they received each year from Chinese students.
Colleagues with special expertise would also aid committee members in
sorting out letters of recommendation—a part of the application that holds
promise for making distinctions but which can be an interpretive minefield
without an understanding of how recommendation letters may vary across
cultures. Faculty were divided on the utility and trustworthiness of letters
of recommendation to inform judgments of international students. Para-
doxically, letters were both the most frequently cited way that participants
said they came to trust international applicants and the most frequently
cited source of concern about applicant credibility. A close analysis of fac-
ulty comments on this issue revealed that this paradox can be explained in
terms of how well an international student’s letters of recommendations
reflect American cultural norms.
When a letter of recommendation conformed to American norms, it could
“speak to the record” of an international student or “compensate” for un-
certainty about academic preparation. When letters failed to fit the Amer-
ican style, however, faculty declared them “pretty useless” and “basically
worthless.” One senior professor summarized, “Now, 15 to 20 percent of
our letters are from Asia. If you’ve ever read them, you can’t compare them
with Americans. You basically have to ignore the letters.” Another scholar
summarized a series of cultural trends that I heard participants from var-
ious disciplines express:
There’s also a difference in culture in what the letters of recommendation look
like. American letters of recommendation tend to be pretty descriptive. They
tend to be pretty positive. Often European letters of recommendation are more
reserved. Ones from China in particular are often extremely perfunctory. They
tend to be one paragraph. “This is a good student. He is honest.”
The most common concern faculty named about reading letters was how
to interpret a less-than-exuberant tone. Could they safely use tone as a clue to
the letter writer’s feelings about the applicant? Or was it a function of cul-
tural norms or the letter writer’s personality? After tone, faculty struggled
with how to make sense of content, especially in letters from scholars in
China. Letters from China often emphasized applicant morality, when what
faculty primarily wanted were hints about fit and the quality of their research
experience. Also, what would the applicant be like as a junior colleague for
five or more years? How did she or he compare to other students?20 The
chair of the economics committee described their general understanding
of the situation: “Internationally, they [the committee] know the programs
they trust, and there are some institutions producing very good econo-
mists, but their faculty don’t know how to craft the kind of recommenda-
tion that at U.S. reader will find persuasive.” Recognizing this tendency, he
has made it a point to explain these norms to scholars he meets at interna-
tional meetings and collaborations.
Like differences in the cultures of test preparation, cultural differences
in letter-writing norms complicated faculty members’ efforts to get to know
individual applicants and their skills. Ultimately, that ambiguity prevented
most Chinese applicants from receiving the same boost that a well-received
letter could give students with nontraditional backgrounds.
Ambiguities Create Conditions in Which Stereotypes Thrive
Challenges of ambiguity in decision making have drawn scholars’ atten-

tion for over fifty years, and recent experiments have found that decision
makers are more likely to interpret ambiguous information using expec-
tancies or stereotypes.21 I found that the layers of ambiguity involved in
evaluating international applicants created conditions under which stereo-
types thrived. Reviewers leaned on what they knew (or thought they knew)
about students from a particular group in order to manage the uncertain-
ties. Most frequently, the ambiguities of evaluating students from China and
other East Asian countries—what to make of their GRE scores, English
skills, and individual differences—primed faculty to adopt elements of the
“model minority” paradigm. For East Asian applicants to distinguish them-
selves therefore required evidence that refuted the stereotypes associated
with this paradigm and suggested alignment with American norms.
The model minority paradigm portrays people of Asian22 descent as con-
forming to a monolithic model of academic and economic success. It re-
duces the complex Asian and Asian American diaspora to a set of simplistic
stereotypes23 and emphasizes Confucian cultural values (assimilation, strong

work ethic, deference to authority) and dissimilarity from American
norms.24 These stereotypes are rarely challenged in mainstream social
discourse. For example, the popular media draw on model minority ste-
reotypes in their typical portrayals of Asians and Asian Americans—as
nerdy, passive, socially awkward, perpetual foreigners with poor English-
language abilities, but who also are courteous, well-educated, overachieving,
hardworking, and talented in math.25 This mix of qualities takes on spe-
cial meaning in the context of recent economic development in many
Asian countries, the history of cheating scandals, Asian Americans’ mean
test scores advantages, and their representation in selective colleges. It
paints Asian immigrants and Asian Americans as a threat to white America’s
“educational, economic, and political opportunities.”26 As such, the model
minority stereotype is fundamentally one of ambivalence. As seen in the
comments of so many faculty in the current study, views of Asians and
Asian Americans often mix respect and resentment with a basic sense of
distrust.
Scholars argue that because it focuses on cultural differences associated
with race, rather than phenotypical characteristics, the model minority par-
adigm is an example of contemporary racism, or neoracism. The paradigm
has been especially toxic to American race relations, pitting Asian Ameri-
cans against other minority groups, and holding up their example as evi-
dence that the United States must have an open, meritocratic opportunity
structure in spite of its racialized history. Higher education scholars Samuel
Museus and Peter Kiang have found that in higher education, the model
minority stereotype manifests in assumptions that Asian American students
are all the same, are not really minorities, are immune from negative expe-
riences with race, and neither seek nor require academic resources and
support.27
In addition to the specific racial stereotyping Asian students face, research
has found that foreign graduate students in general may face stereotyping,
such as the assumption that they are intellectually backward strangers to
the community of science.28 International students from Asia, India, Latin
America, and the Middle East report proportionally more experiences with
discrimination than students from Europe, including negative remarks
about their home countries or cultures; hostility toward nonfluency in En-
glish; social exclusion; conflicts with professors, administrators, and class-
mates; and direct verbal insults or physical attacks.29 For those whose ex-
perience of American higher education requires them to contend with model
minority and foreigner stereotypes, the degree may come with a psycho-
logical price tag as high as the tuition bill.30
Decision makers need to be aware of the existence of these stereotypes

because the nature of admissions review involves conditions under which
stereotypes are easily activated. When trustworthy information is scant, in-
dividuals are being compared, and assessments are routinized rather than
reflective, humans have a tendency to replace ambiguity and complexity
with simpler explanations. In this sample, the composite stereotype that fac-
ulty held of Asian applicants was one of mathematically sophisticated test
takers and status seekers with weak English skills, who are difficult to dis-
tinguish from one another. Thus, it makes sense that the international
students whose applications impressed faculty had characteristics that con-
tradicted this stereotype.
The effort to create broadly diverse departments compelled faculty to
consider uniqueness and freshness of perspective when taking stock of the
short list. Amid many conventional achievers, admissions committees sought
applicants who were both demographically diverse as well as diverse in their
personal backstories, personalities, and nonacademic interests. When fac-
ulty evaluated applicants who seemed demographically and academically
similar, these latter types of personal uniqueness were especially salient.
“That kind of stuff jumps off of the page,” one physicist explained. “It sort
of says, ‘Hi, I’m a little bit different.’ You’re looking for something to dis-
tinguish all these great test scores.”
I nonetheless observed striking patterns across disciplines in the personal
qualities faculty considered desirable among applicants from China and
Korea. One of the important interview questions I posed to each participant
asked them to describe applicants—domestic or international—who fit
the profiles of two “ideal types”: borderline students, those about whose
admissibility the committee disagreed, and easy-to-admit students, of whom
the committee all thought highly. Among those whom participants described
as easy to admit, seven happened to be from China, Korea, Taiwan, or
Japan. Juxtaposing the descriptions of these individuals took me aback.
They were remarkably similar to one another, and clearly contradicted the
model minority stereotype.
Of the seven, Table 9 presents descriptors of “easy admits” from the four
most lengthy profiles that participants provided. Not only mathematically
skilled, three were recognized for artistic and creative pursuits, including
two poets. In contrast to the image of a quiet student struggling with En-
glish, three were described as “communicators.” Not only seeking status or
opportunity for themselves, three of the four were identified for their aca-
demic and social engagement on behalf of others (activist, advocate, tutor).
And breaking the stereotype that students from East Asian countries are
indistinguishable from one another, three of the four were acknowledged
Table 9 Descriptors Used by Faculty in Profiles of “Easy to Admit” Students from China,
South Korea, and Japan
Theme Student 1 Student 2 Student 3 Student 4
Creative Poet Poet; Renais- Photographer

sance type
Communication Communicator Communicator Extremely good
communicator
Outreach Activist Advocate; tutor Tutor
Unique Original; Different; Broad interests;
interesting special eclectic tastes
Engaged Motivated Excited; Enthusiastic;
interested genuinely
interested
Other Talented; Distinguished; Not poor, not
very good genuine super-rich
Note: Descriptors that faculty used in the other three profiles of “easy to admit” students from East
Asian countries were “thoughtful,” “knowledgeable,” “really good,” “very strong,” “imaginative,” and
“independent.”
for their uniqueness. In sum, stereotypes associated with the model minority
paradigm informed faculty judgment, but when it came to reviewing the stack
of outwardly “indistinguishable” Chinese applicants, those with the best
chances of being admitted contradicted those stereotypes.
Conclusion
Under even the most straightforward conditions, admissions evaluation

is work filled with uncertainty, but reviewing international students’ files
introduces specific ambiguities. Some of these may predispose faculty to
default to model minority stereotypes. This possibility was exemplified in
participants’ expectations of Chinese students’ GRE scores and English
skills, and their perception that these students did not differ substantively
from one another. Two forces seem to fuel the doubts cast on Chinese ap-
plicants as a group. Personal memories and departmental tales of admitted
Chinese students who came to campus and confirmed negative stereo-
types loomed large in faculty members’ minds. There was also widespread
skepticism about the credibility of Chinese students’ applications. Some
recalled the cheating scandals as if they were yesterday, and others be-
lieved the rumors that cheating, forgery, and other efforts to game the ad-
missions system define all Chinese students’ efforts to gain admission to
American universities.
A natural question is: To what extent is there evidence for the types of
assumptions that faculty held? On the question of whether TOEFL and
GRE Verbal scores are valid indicators of English proficiency and student
success, more than forty years of research have yielded few clear trends
about the TOEFL’s utility in predicting international students’ academic per-
formance. A recent study, which sampled 1,733 international students en-
rolled in U.S. graduate programs between 2005 and 2009 (22 percent of
whom were Chinese), found that TOEFL scores did not predict graduate
school academic performance. Other studies suggest that the TOEFL is a
more useful metric in the humanities and social sciences than in the natural
sciences and more effective in predicting academic performance of students
with lower English proficiency. An interesting study of international grad-
uate teaching assistants found that those who received negative teaching
evaluations from students were also likely to have lower TOEFL and
Verbal GRE scores. These mixed findings suggest that faculty should not
count on the TOEFL alone to gauge students’ fluency. When doubts are
present, seeking additional information to assess applicant English profi-
ciency appears to be a reasonable step.31
It is true that, on average, international students have higher GRE Quan-
titative scores than U.S. citizens, by 6 points in 2012. Chinese students
have the highest mean GRE Quantitative scores in the world (162.9 points),
more than 13 points higher than the mean for U.S. test takers.32 However, Fu
found that, among international students, GRE Verbal, GRE Quantitative,
and undergraduate GPA together explained only 3.1 percent in the vari-
ance of first-year graduate school grade point average, less than half of
that explained in an identical model with domestic students. She also found
that GRE scores did not on their own significantly predict international stu-
dents’ graduate school GPA,33 suggesting that reviewers had a point in not
taking international students’ scores too seriously. If the test’s weakness for
some U.S.-born African American and Latino students is that scores are
susceptible to noise from stereotype threat, then for many Asian-born stu-
dents, the weakness may be scores’ sensitivity to large amounts of practice
and preparation. Rather than reducing students to their performance on
tests, faculty should weigh TOEFL and GRE scores as part of a holistic set
of considerations about students.
Reliable evidence on the presence, nature, and extent of academic cheating
is difficult to obtain, and to my knowledge there is no academic research
on the topic of cheating in graduate school admissions.34 Incidents of fraud-
ulent TOEFL and GRE test taking in China seem to have reinforced the
fundamental distrust with which many faculty regard applicants from that
country, even though the incidents were thought to involve only a small
proportion of applicants to U.S. graduate programs.35 It is difficult to gain

an understanding of cheating in China because the application process in
China is decentralized through private consultants, and cultural norms vary
in what should be considered to be cheating, gaming, or fair play. A white
paper written by one such private consultant, Zinch China, claimed that
parental pressure, aggressive private college counselors, and high schools’
complicity all carry some responsibility for the pervasiveness of cheating
in China. Admitting that the findings would be impossible to verify, Zinch’s
paper concluded that “the common refrain” from 250 interviews with
students in top Chinese high schools and their parents (who were asked to
ask their friends), as well as several agents, was, “Everybody cheats.”36
Regardless of the reliability of evidence, the clear correspondence between
judgments of students from China and a common racial stereotype is worth
calling to mind when faculty assume that they are immune from implicit
biases about other groups, whether women, LGBTQ (lesbian, gay, bisexual,
transgender, and queer) students, those from southern states, or any other
group. Perhaps because model minority stereotypes receive little challenge
in American society—and because Asian success has been positioned as a
threat not only to people of color but also to white America—participants
were more vocal and candid about their distrust and ambivalence about this
group than about any other single demographic group.
Finally, faculty members’ review of applications from China clearly
demonstrates some malleability to ideals about merit, a point that may be
worth remembering when claims are made that a program should not take
into account domestic students’ previous contexts for learning. Rather than
judging students against a uniform ideal of quality or as representatives
of broader categories or identities, the aim of holistic review is to evaluate
students as individuals, in the context of their own opportunities and
potential.
Conclusion
Merit beyond the Mirror
My first thing about graduate admissions is that we do it as con-

scientiously as we can, but it is a crapshoot. It really is. We don’t
know who’s going to blossom and who isn’t. We have not found
reliable predictors.
— Senior professor of classics
I t is no wonder that issues surrounding admissions are among the

country’s most controversial topics. It’s not only educational creden-
tials, but where they are earned, that increasingly shapes professional op-
portunities, both in academia and in the broader labor market. Two recent
studies have found that faculty hiring occurs within largely closed net-
works and that most of the faculty who are hired into tenure-track posi-
tions possess PhDs from a small set of institutions.1 And in industry, firms
in some sectors only seriously recruit from a very small set of “super-elite”
undergraduate institutions, creating what sociologist Lauren Rivera called
a “golden pipeline” into society’s most lucrative positions.2 In a system
like this, the transparency of selection mechanisms into selective colleges
and doctoral programs is critical for equity in the system. At both the un-
dergraduate and the graduate level, however, the basis for any applicant’s
acceptance or rejection is usually opaque to applicants and admissions
personnel alike. Uncertainty about what admission requires, combined
with the sense that it will play a determining role in their life outcomes,
raises the stakes and anxieties for applicants and their families. For those
tasked with making admissions decisions, the process may be known, but
the basis for individual outcomes may be just as difficult to articulate.
Conclusion 155
Summary
We know this much about how faculty evaluate prospective graduate stu-
dents: Test scores, college grade point average, and college reputation play
a formative role in the initial review of applications, which explains their
relationship with the probability of admission, generally. From the current
study, I learned that faculty conceive of “merit” at this point in review prin-
cipally as a matter of conventional achievement, although scholars across
the disciplines may vary in the sections of the GRE they weigh and in con-
sidering overall GPA versus grades earned within one’s major. They do not
interpret every student’s scores the same way. Due to China’s deep-rooted
culture of test preparation and history of admissions fraud, for example,
many faculty believe that students should have very high GRE and TOEFL
scores but that those scores cannot be trusted to reliably predict student
skills. Intriguingly, although it is standard practice to contextualize test scores
by national origin, and grades by institutional prestige, most faculty do
not contextualize test scores in light of their distributions by race, gender, and
socioeconomic status. Just two respondents, an economist and a philoso-
pher, actively encouraged consideration of diversity and/or students’ social
identities in the initial round of review.
Why the double standard? The most common explanation I heard was
that such an interpretation would introduce diversity into the conversation
too soon. As a sociologist in the study put it: “First you have to be above a
bar, then we can ask the diversity question.” Setting extremely high bars on
GRE scores and college prestige, however, disproportionately excludes the
very populations whom university websites and mission statements claim
they wish to attract—and who are already underrepresented in many fields
at the levels of doctoral education and the professoriate. For example,
16 percent of Asian American high school graduates enrolled in highly se-
lective colleges and universities in 2004 compared to 2 percent of African
American high school graduates. In the physical sciences, 82 percent of
Asian and white students earned a 700 or above on the Quantitative sec-
tion of the GRE, compared to 5.2 percent of underrepresented minority stu-
dents.3 The informal admission standards that elite doctoral programs
have established in many fields therefore make it extremely difficult for
people of color to gain access.
As a matter of procedure, committees commonly maximize efficiency and
minimize conflict by discussing only the cases in which initial ratings di-
verge from each other. As long as committee members rate an applicant
similarly, the average rating can serve as a proxy for “the collective assess-
ment of the committee.” Through this process, they can quickly eliminate
a large share of the pool using common, academic criteria. But deliberations
about borderline applicants are handled quite differently; these decisions often
come down to hair splitting or making subtle distinctions between appli-
cants based on novel criteria that were not considered relevant to the assess-
ments of other candidates.
Academic preparation is just the tip of the iceberg when it comes to
“merit,” because faculty judge many more students smart and competent
than their doctoral programs have capacity to accommodate. For the small
number whose applications survive to receive full review, judgments are
thus holistic, complex, and unpredictable. When comparing generally qual-
ified students, seemingly small matters may matter very much to an appli-
cant’s chances because of the layers of inference involved. In the words of
one tenured philosopher, “We are in the business of making fine distinc-
tions.” A single, ambiguous line in a letter of recommendation; the appeal
of a writing sample’s introduction; the poor reputation of a letter writer
for speaking too highly of too many students; an applicant’s weekend hobby
or hometown—reviewers may read meanings and value judgments into
each of these, in ways that can spell the difference between a candidate’s
admission or rejection. With few cases discussed and evaluations subject to
a myriad of considerations, pinning down what counts as merit late in the
process is more difficult than it is earlier in review, when a few key criteria
go a long way in shaping ratings. “Everything matters, and nothing matters
most,” as one faculty member commented.
In this type of situation, any basis for comparing applicants can have the
effect of a “preference.” The individualistic analyses that are inherent to ho-
listic review can elevate opportunities for one student in spite of reducing
chances of another with different, albeit also desirable, qualities. When a
small program like classics admitted one of two applicants from rural New
England because a committee member envisioned it as a “pastoral” place
for early socialization in classics, it impinged on opportunities for many
applicants whose geographic origins were not perceived as salient. In po-
litical science, when a student with mediocre grades was admitted in part
because a committee member thought it was “cool” that she wrote for an
online magazine, it came at the cost of another borderline applicant who
had written a book. Committees in several disciplines admitted applicants
from China who disclosed creative hobbies over dozens of their fellow Chi-
nese nationals who did not. What counts as merit when comparing bor-
derline cases sometimes comes down to details that appear idiosyncratic
and far afield from conventional achievement.4
Many of the grounds for judgment late in the review process are spe-
cific to one or two applicants rather than applied to the entire pool, but
Conclusion 157
those evaluations are all part of an effort to shape the future of their pro-
grams and disciplines. Leaving behind the focus on conventional achieve-
ment in most cases, deliberations revolved around how applicants “fit” with
the program in the present and their idealized academic communities of the
future. They sought students who offered fresh perspectives and strong
grounding in discipline-specific research dispositions and skills. Diversity,
broadly defined, was integral to the fresh perspectives they sought, and a
majority strived to construct “balanced” cohorts of students. Rarely, how-
ever, did they discuss race, gender, or socioeconomic status in explicit or
substantive ways. In contrast to the state of affairs that some worry about,
in which race considerations drive admissions in ways that compromise
the fairness of the entire process and its outcomes, I was more taken
aback by the almost complete silence on these topics in the meetings that
I observed.
Explaining the Gap between Principles and Practice
This observation returns us to the questions that motivated my research.

With problems of inequality widely known, why do so many faculty rely
upon selection criteria that obstruct access for women and underrepresented
minorities? If they value diversity, why are they loathe to make racial diver-
sity part of the conversation? I investigated the culture of faculty decision
making in elite doctoral programs as a root cause. From there, the answers
depend in part on the data and in part on the lenses through which we read
it. I will synthesize my findings from three theoretical perspectives before
turning to their implications for admissions practice.
Multiple Interests, Multiple Contexts
Assessments of the sort that happen in admissions committees are by na-

ture an elaborate compromise, according to Luc Boltanski and Laurent
Thévenot’s theory of situated judgment.5 Graduate programs choose their
next cohorts of students with many social goods in mind—student success,
prestige, and diversity, to name a few—and each of these will be more or
less salient for a particular discipline, program, committee, or reviewer.
Practically speaking, admissions decision making consists of negotiating
multiple hierarchies of priorities (a heterarchy) that emerge from disci-
plinary logics, program values, committee dynamics, and personal identi-
ties. This perspective would argue that because faculty are trying to satisfy
perceived demands of multiple evaluative contexts that it is rare for any

one interest or criterion to consistently hold up as decisive or determina-
tive across reviewers and rounds of review. Only those interests constructed
as core across contexts are likely to withstand the layers of compromise. A
situated judgment angle on my data would conclude that diversity may be
one in a constellation of interests that faculty would like to maximize, but
especially with respect to race, it has not yet attained the status of a major
priority when faculty are thinking about their own values, and those of
their programs, committees, and disciplines. Here, it helps to contrast the
role of diversity in admissions with that of prestige, an even more well-
institutionalized interest.
Prestige. The powerful influence of pedigree and prestige can be understood

from two related angles. Pierre Bourdieu’s perspective argues that the power
of pedigree reflects a broader tendency for elite educational institutions to
organize their activities and define quality in ways that preserve their status
in the field.6 Because institutional reputation (“peer assessment”) and en-
rolled doctoral students’ mean GRE scores and grade point averages con-
tribute to popular university-ranking systems, prestige-oriented graduate
programs use admissions to boost their academic profile.7 Often this en-
rollment management decision comes at significant cost to other espoused
priorities, such as broadening access for underrepresented groups.
However, status competition does not explain the many individual pref-
erences that also shape admissions judgments. These more complicated
patterns signal locally defined—even self-defined—ideas about quality and
the tendency for faculty to judge prospective students in the same domains
they judge themselves.8 Michèle Lamont explains how drawing such
identity-based boundaries can contribute to social reproduction:
Exclusion is often the unintended consequence or latent effect of the defini-

tion by the upper middle class of its values and indirectly of its group identity
and its nature as a community . . . Only when boundaries are widely agreed
upon (i.e., only when people agree that some traits are better than others) can
symbolic boundaries take on a widely constraining (or structural) character
and pattern social interaction in an important way.9
I observed faculty using informal conversation to express shared tastes,

identities, and goals, thereby constructing symbolic boundaries that guided
their assessments of fit and belonging. The student from a politically con-
servative, religious college was debated as a possible “nutcase.” The student
with a strong but conventional file was mockingly compared to a Ford. In
physics, a lack of research experience was called the “kiss of death.”
Conclusion 159
Prestige thus drives graduate admissions in programs like these because

it helps them maintain their status and because it is central to their iden-
tity. More than admitting a group whose average characteristics preserve
the program’s position in nationally defined status orders, it helps graduate
programs create the sort of community they feel they are and aspire to
be. For such organizations, which already see themselves as successful, the
status quo therefore represents a powerful default. This is particularly
the case, as I will discuss, if changing the basis for admission may require
some soul searching about flaws in their collective identity or goals, and if
admission already requires more time and effort than is desirable.
Diversity. In contrast to prestige, diversity is relatively new among admis-

sions priorities. How has it emerged as an interest at all? According to the
organizational theory of institutionalism, organizations survive by adapting
their practices and priorities to changes in the institutional and political en-
vironment. Shifting values in society can bring about new policy in organi-
zations like colleges and universities by motivating changes that might
otherwise have been avoided.10 This institutionalist angle helps explain
why educational equity came to be recognized as a social imperative during
the Civil Rights Movement, and why diversity has come to be viewed as a
practical interest for higher education today.11
A very brief history can situate the current state of diversity as an in-
terest for higher education stakeholders. Civil Rights protests changed the
political environment in the 1950s and 1960s to support demands for
improved access to higher education and employment for African Americans,
Latinos, Native Americans, and, in spaces where they were still excluded,
white women. The Civil Rights Act of 1964 and executive orders for
affirmative action under Presidents Lyndon Johnson and Richard Nixon
made these shifting values visible, and they presented them to the public as
means of redressing the effects of long-standing policies of segregation,
discrimination, and exclusion. These macro-level influences, coupled with
acute pressures that campus-based student movements placed on univer-
sity administrators, compelled the adoption and diffusion of race-conscious
admissions policies by selective colleges and universities across the country.
The difference that scholars and institutional researchers found diversity
to make for student learning and development provided an empirical foun-
dation for another affirmative action rationale—the educational benefits
of diversity. Proponents sought to broaden the appeal of affirmative action
by emphasizing its benefits for all students’ development, namely white stu-
dents, not only for those who bore the “present effects of past injustice.”
Thus, whereas universities implemented affirmative action due to changes
in the political environment, university efforts since the 1970s have been
to protect the legitimacy of race-conscious admissions by elevating the ra-
tionale with the widest public support.
Institutions have also focused on diversity’s educational benefits because
the courts have established narrow parameters for the constitutionality of
race-conscious admissions. In 1978 in Regents of the University of California
v. Bakke the U.S. Supreme Court ruled racial quotas unconstitutional, and
in 2003 in Gratz v. Bollinger it also struck down the awarding of “auto-
matic, predetermined point allocations” to underrepresented students.12
Justice Lewis Powell’s controlling opinion in Bakke ruled consideration of
race permissible as a “plus factor” in admissions and financial aid decisions,
but this consideration requires narrow tailoring to a compelling state
interest.13 What counts as a compelling state interest? In Bakke, whose
precedent has been upheld in such recent decisions as Fisher v. University
of Texas, the Court rejected three of the four interests in affirmative action
asserted by the University of California–Davis Medical School. However,
Powell’s opinion affirmed the educational benefits of diversity, broadly de-
fined, citing promotion of diversity in higher education as a “compelling
governmental interest.” Powell also upheld educational institutions’ discre-
tion to determine the selection procedures that best suit their needs, asso-
ciating it with academic freedom and the First Amendment.14 In the years
since Bakke, voters have banned affirmative action through ballot initia-
tives in Arizona, California, Michigan, Nebraska, Oklahoma, and Wash-
ington, and the legislature and governor passed anti-affirmative policies in
New Hampshire and Florida. These bans are responsible for declining
racial/ethnic diversity in selective undergraduate institutions and in many
graduate fields of study, amid increasing racial and ethnic diversity in the
population.15 For academic institutions in the other forty-two states, how-
ever, principles from Powell’s opinion are still the law of the land, and are
important for those engaged in admissions to know. A brief from the Civil
Rights Project at UCLA summarizes them:
1. Reserving seats or proportions of seats specifically for underrepre-

sented students is not permissible.
2. Reviewers should use a common process of review for all applicants.
3. Race should be one of several individual characteristics assessed as a
plus factor in the effort to promote diversity.
4. Every applicant should be evaluated as an individual, and should not
be assumed to represent a broader identity category.
5. Programs should not single out specific racial/ethnic groups, but
consider the contribution that all groups make to diversity.16
Conclusion 161
Bakke focused on admission to medical school and Grutter on law school,

but the Supreme Court has not specifically examined graduate school
admissions among the arts and sciences. However, economists, psycholo-
gists, sociologists, as well as higher education and legal scholars have all
advised that selection processes and their rationales be tied to educational
mission.17 In addition to ethical and economic rationales, the distinct
mission of doctoral education may therefore elicit interests in equity and
diversity that are distinct from those in undergraduate and professional ed-
ucation. As discussed earlier, research by teams possessing diverse social
identities is associated with core characteristics of scholarly quality. Het-
erogeneous research teams have demonstrated advantages in creativity
and problem solving, and publications that result from ethnically and
gender diverse research collaborations are cited more often.18 Relatively
few people I interviewed associated increasing social diversity with intel-
lectual excellence or the health of their fields. Important new analyses also
assert that, if reducing inequities in higher education is the goal, then “di-
versity” itself merits critical evaluation as a rhetorical strategy for organi-
zational behavior.19
From the perspective of situated judgment, what counts in practice as
merit is as an institutionalized compromise across multiple interests
associated with the multiple social contexts that decision makers represent.
Professors strive to simultaneously maximize individual, committee, program,
and disciplinary interests, and to uphold their personal values, interpersonal
relations, and institutional and wider policy. Prestige and diversity may both
be organizational interests, but only if something is conceived as a core in-
terest across those contexts will it be likely to survive as a priority in the
negotiation process. We can explain continuing inequality in spite of diver-
sity’s institutionalization through (1) the entrenched value that status and
prestige hold for organizations and participants like the ones in this study,
(2) the more tenuous place that diversity holds (legally, politically, and dis-
cursively) as a value in their disciplines, programs, committees, and for
many of the individual reviewers, and (3) evaluative scripts that associate
very high GRE scores and attending elite institutions with intelligence and
belonging (two additional shared values) in prestigious communities like
theirs, and that construe lower GRE scores or less selective college enroll-
ment as a trade-off with excellence. However, same deliberative processes
by which faculty collectively define merit can be used to diversify what
counts as merit. Decision makers may encourage alternative interpretations
of common criteria, for example, or use admissions and hiring to encourage
new perspectives or collective goals.
Preferences Imply Aversions
Decision theory offers a second perspective from which to interpret this

study’s findings. Although research and public debate about admissions
often revolve around the justice and necessity of specific evaluation criteria,
a contribution of this research is to show that decision-making processes
are also implicated in ongoing reliance on pedigree and GRE scores.
Specifically, participants’ preference for a process that reduces uncertainties
and preserves collegiality revealed their shared aversions to ambiguity,
risk, and conflict.
Ambiguity aversion. The complexity of comparing hundreds of applicants

who come from different national, institutional, and personal backgrounds
makes faculty ambiguity averse. Faced with a “cloud of random applicants,”
professors hunger for clarity and conviction about which ones are likely to
thrive as scholars and bring vitality to their departments and disciplines.
From this standpoint, reliance upon test scores is not exclusively about what
they signal, but instead about how they make it easy to compare applicants.
Participants associated GRE scores with convenience because their apparent
standardization, precision, and clarity seem to cut through all of the apples-
to-oranges comparison that evaluating students’ letters of recommenda-
tion, essays, and writing samples requires.
Professors in the humanities were both more thoughtful about and less
averse to the subjectivity inherent in judging and comparing their appli-
cants. Yet they too ultimately quantified their judgments to bring closure
to deliberations, and they were especially likely to do so for the sake of expe-
diency. Though fallible, “the numbers” provided something on which hu-
manities faculty felt they could stand. By contrast, in the physical sciences,
greater trust in numbers engendered trust in the validity of distributions of
test scores and average ratings. In both cases, quantifying differences in
perceived admissibility was a hallmark of deliberative bureaucracy, one that
increased efficiency by masking perceived ambiguities. For busy profes-
sors, the GRE’s appeal was in no small part the possibility that it would
simplify their work.
Risk aversion. There is also abundant evidence that faculty in these presti-
gious programs engaged with admissions work from a posture of risk aver-
sion. This tendency, paired with the specific assumptions about risk that
they frequently made help clarify why they favored criteria that undermine
increased racial diversity. Many respondents associated less selective under-
graduate institutions and lower test scores and grades with less prepara-
Conclusion 163
tion and intelligence, and they associated less preparation and intelligence
with a higher risk of academic struggle or failure. This risk aversion script
pervaded ratings and deliberations at all stages of review. Risk aversion also
provided a bottom-line basis for rejecting well-pedigreed individuals who
had personality red flags, research interests that aligned only marginally
with faculty expertise, or letters of recommendation that raised eyebrows.
Organizational theorist James March has found that a propensity to risk
aversion is common in high-status organizations’ decision-making pro-
cesses.20 In this case, faculty rationalized risk aversion as an organizational
luxury, a matter of financial prudence, and a foundation of responsible de-
cision making. They knew they could afford to be risk averse because their
highly ranked programs attracted so many qualified candidates. Further,
they convinced themselves they should be risk averse in light of the finan-
cial investment involved and what admitted students represent—namely, the
program, discipline, and university’s quality, status, and future.21
High-status organizations are also more prone to fundamental attribu-
tion error—the tendency to attribute the outcomes of their members to
intention and skill rather than to the conditions in the environment. Under
these conditions, high-status organizations often simply exclude prospects
perceived as risky rather than measuring the risk and making a decision on
that basis.22 This point highlights a critical detail about the risk aversion I
observed—that it was almost always grounded in perceptions of risk rather
than generalizable evidence or calculations of it.
Availability bias offers a third way of understanding risk aversion in
terms of cognitive biases. Trusting the reliability and validity of the limited
information that is currently available to decision makers can lead them to
overestimate the likelihood of rare but memorable events. Availability bias
facilitates group-based stereotyping, and I observed it to be a process by
which racism was subtly institutionalized in the admissions process. For
example, a number of faculty regarded African American, Latino, or Na-
tive American applicants from less selective colleges or non-native English
speakers from East Asian countries as admissions risks because their pro-
grams did not enroll many of these students and recollections of one or two
students who had enrolled and struggled were hard to shake from their
memories. Several participants admitted that they felt “spooked”—three
using this particular word—when reading applications from students whose
profiles reminded them of individuals who did not graduate.
Availability bias is especially dangerous when combined with fundamental
attribution error. The combination of these two biases locks in faulty as-
sumptions about who is successful and why, and then uses those faulty
assumptions as a basis for distributing future opportunities. First, attribution
errors blind academic departments to how the climate and culture that
they create may contribute to students’ successes and struggles. Then,
availability bias encourages extrapolation of a few students’ struggles
(which resulted in part from organizational conditions the department
could change) to a larger category of students as a basis for evaluating their
future potential, and thus their deservingness for opportunities. Taking steps
to actively check the natural tendency toward these biases can prevent the
experiences of a few individuals from becoming the basis for judging whole
categories of future applicants as risks. It also highlights the importance of
encouraging faculty mindfulness about their own role in shaping student
success.23
Conflict aversion. Quantification also suppresses conflict among faculty.

Conflict aversion is the flip side of the deeply rooted professional norm of
faculty collegiality. Admissions committee members felt that quantifying
their judgment—whether through test scores, individual ratings, or average
ratings—provided more uniform interpretation than the many understand-
ings that the committee might generate about individuals and their relative
admissibility. With a few keystrokes, numbers could be sorted into a single
ranked list, obviating the need for difficult negotiations about who should
be deemed more or less admissible. Most departments viewed some delib-
eration as prerequisite to legitimate admissions decisions. However, com-
mittee members preferred not to argue with each other, especially about
matters that were rooted in deeply held epistemological differences, such
as the validity of statistical inferences about GRE scores. As such, two key
elements to the deliberative bureaucracy decision making model that I ob-
served included careful committee appointments and substituting poten-
tially controversial discussion of applicants with relatively uncontroversial
discussion of admissions procedure. The formal task of admissions may be
to identify applicants with the most impressive achievements and greatest
potential, but the process of collective selection was one of political com-
promise and avoiding uncomfortable conversations.
Resistance to Change, Ambivalence about Diversity
By combining elements from the first two theoretical perspectives—

inconsistencies across context in how faculty weigh diversity and the im-
portant role of aversions in shaping faculty decision making—we can
also interpret deference to entrenched evaluation criteria as a matter of
resistance to change. An entire scholarly literature has developed around
Conclusion 165
resistance and ambivalence to organizational change, with Louis Menand

going so far as to argue that twenty-first-century American professors are
trying their best to maintain a nineteenth-century system. They want to
make a better world through their research and teaching, but often es-
chew making change in their own ranks to help bring it about. Among the
admissions committee chairs I interviewed, outright resistance to change did
not seem to be the problem. Indeed, three of them welcomed me to observe
their admissions committees to learn how their systems could be improved.
However, their colleagues were in many cases more ambivalent—about
changing admissions, generally, and about rethinking their applicant pools
and diversity, specifically.
Why is it that faculty might stand by equity and diversity in principle,
but feel ambivalent about practical efforts to achieve them? For one thing,
many thought of these as obligations dictated by social norms and for their
pragmatic benefits, not as conditions that demonstrably improved their
work or their community. Faculty typically build their careers on the visions
of academic excellence they inherit from their academic forebears, and for
most, diversity is not yet integral to that vision.
Thus, in political science, linguistics, astrophysics, and philosophy, fac-
ulty supported their colleagues’ advocacy for a small number of borderline
applicants with nontraditional profiles, but they were reticent to take on
such students as their own advisees. They expressed concerns about the ad-
ditional mentoring investment that might be required. The self-trained
philosopher from the Deep South, the astrophysicist born in the Himalayas,
and the political scientist with unusually low GRE scores each received ad-
missions offers, for example, but only after extended debate and assurance
that someone else would take responsibility for their supervision.24 Most
faculty could be persuaded to admit specific individuals in the name of di-
versity, but I observed little evidence that diversity was either understood
to be a collective responsibility or enacted as a shared value in the day-
to-day of department life.
Postracial discourse in American society may also contribute to ambiva-
lence about reform. For one, claims that racism is a problem of America’s
past undermine attention to continued structural inequalities and everyday
biases.25 Constraints on affirmative action reflect postracial sensibilities, and
I learned that regardless of whether or not a program was subject to specific
legal constrains, that bans, rumors of bans, and worries about lawsuits all
made faculty reticent to discuss race at all with respect to recruitment and
admissions.
However, like the other factors I have discussed, uncertainty about options
for equity-based reform must be read as an explanation, not an excuse.
Research by Roger Worthington and colleagues found that “colorblind-

ness” and personal privilege predispose those within higher education to
an overly rosy view of the status quo on their campuses.26 Most faculty in the
ivory tower—especially in elite institutions—are distanced and well sheltered
from the struggles with exclusion, marginalization, and discrimination
that have historically compelled equity-minded reform efforts in higher
education. I found that, with the exception of some sociologists and a handful
of others, most participants in this study did not see equity or diversity as
their issues, or the students admitted under these banners as their students.
I did not collect data to measure implicit bias against women and students
of color, but we must not cast aside the evidence of this pattern uncovered
in prior research. Faculty judgments of many applicants from China were
consistent with model minority stereotypes, providing a window into the
likely presence of other identity-based biases of which faculty may have
been unaware or more careful to conceal.
Finally, organizational change tests collegiality, especially when it occurs
through democratic processes. It can be costly in time and effort, two scarce
resources for faculty operating in a reward system that privileges research
productivity over student development or improving departmental struc-
tures. Were faculty to receive rewards or incentives for time spent strength-
ening admissions, not to mention student mentoring, maybe the gap between
expressed and enacted commitments would not be quite so wide.
With myriad social forces acting upon faculty as they select and discuss
applicants, we need to think systemically when we think about improving
admissions. Faculty members’ ambivalence about admissions reform has
several dimensions, and must, itself, be understood as one among several
explanations for continued inequality. Shared aversions to ambiguity, risk,
and conflict shape the work of decision making in powerful ways, and pres-
tige often trumps other organizational interests when faculty seek shared
values to which they can collectively defer. In addition, the perceived costs
of admitting more diverse cohorts, a lack of awareness about compelling
alternatives to affirmative action, satisfaction with the way things are, priv-
ilege and personal distancing from equity-related issues, and a changing
sociopolitical context each stand as barriers to change.
From this angle, the slow pace of change is hardly a surprise, because
change is needed on many fronts. There are no silver bullets, easy answers,
or quick fixes to systemic problems. Faculty may be so reluctant to
wrestle with the complexity of diversity and inequality that they avoid
altogether the conversations required to get the process started. Comments
from Jonathan, a white male associate professor, summarize these points
perfectly:
Conclusion 167
It’s uncomfortable to be reminded that I’m part of the problem. What I don’t
want are things that will feel unnatural, take 300 percent more time, or that my
colleagues will fight. I want to keep my sanity as a professional, and I don’t want
admissions work to consume me. What I want are straightforward ideas that
help me be part of the solution.
Implications for Practice
With this idea and Jonathan’s quote in mind, I turn now to discuss the im-
plications of this research for graduate admissions practice. “Admissions
decisions are actionable choices,” as higher education scholars Rachelle
Winkle-Wagner and Angela Locks wrote, and although professors may in-
herit their colleagues’ approach to admissions, they also have autonomy in
most universities to change the way things are done. My research was not
designed to encourage policy prescription, but integrating its findings with
those of other recent research studies is useful for sketching a framework for
structural changes needed in many fields and programs. Within this frame-
work, graduate programs can craft and evaluate practices that suit their
own unique needs.
Revisit Admissions Routines and Make Them Explicit

All but two of the graduate programs in my sample would have benefited
from taking a fresh look at their routines for recruitment, admissions, and
awarding financial aid. For example, developing a thorough list of the many
steps in admissions that involve subjective judgments may aid faculty in
checking themselves for implicit biases. Which email inquiries a professor
chooses to follow up on with a phone conversation, which colleges a grad-
uate program chooses to visit for recruitment, who should be awarded
financial aid packages as a recruitment tool—each of these procedural
decisions involves a degree of subjectivity.
Also, considering the important role of initial reviewer ratings, many
graduate programs can improve admissions generally by improving the re-
view of individual files. Like selection criteria, approaches to evaluating
files come with benefits and drawbacks, and each program must decide for
itself what the optimal review process is, in light of what they are trying to
accomplish. I will describe here some benefits and drawbacks of two models
for evaluating files.
Open, holistic review incorporates into evaluation all available informa-
tion about students. One strength of a holistic approach is that students
Table 10 Example Rubric
Contributions to {Insert addtitional
Academic Potential for Dept., Cohort, evaluation
Preparation Scholarship Discipline standards} Basis for Ratings
Weak-Acceptable- Weak-Acceptable- Weak-Acceptable- Weak-Acceptable-
Rating Strong Strong Strong Strong
Applicant 1
Applicant 2
Applicant 3
. . . and so forth
Conclusion 169
are easily assessed in the context of their individual trajectories and oppor-
tunities. Open-ended review can lack transparency, however, putting the
applicant in the position of playing a game whose rules are unknowable.
This opacity maximizes institutional discretion, but it can also allow un-
seemly preferences and biases to enter the review process.
More structured approaches to review define in advance the criteria on
which every application will be assessed. The National Science Foundation’s
ADVANCE program has developed rubrics and other evaluation tools that
codify and define selection criteria and ask reviewers to provide a rationale
for their ratings. Table 10 presents a very basic skeleton of a rubric that
departments could customize. Rubric-based evaluation raises transparency
and does not preclude reviewers from contextualizing applicant charac-
teristics or attending to underlying rationales. Rubrics can also be de-
signed to note desirable qualities that come along so rarely that it makes
little sense to include them in evaluations of everyone.27 Using a rubric in-
creases equity, by comparing all applicants using the same criteria, and ef-
ficiency, by focusing reviewer attention on key information. Rubrics do
raise the stakes associated with each criterion, so they need to be chosen
with care.28
Strengthen Recruitment and Align It with Admissions

The need to strengthen outreach and recruitment early in the admissions
cycle was an almost universal pattern across the ten programs. Two pro-
grams had experimental outreach efforts in place to build the diversity of
their applicant pools, but weak coordination with admissions committees
undercut those initiatives’ efficacy. The problem in the other eight programs,
however, was that early recruitment strategies were virtually nonexistent,
consisting of little more than email responses to inquiries from prospective
applicants. Departments looking to build a more diverse applicant pool
have several options, including developing relationships with sister depart-
ments in institutions whose undergraduate alumni of color frequently go
on to earn doctoral degrees. A recent study examined the baccalaureate ori-
gins of African American, Latino, and Asian/Pacific Islander doctoral degree
recipients, and assembled lists of the colleges and universities that produce
the most doctorates of color for each racial/ethnic group. Table 11 lists these
institutions, many of which are minority-serving institutions (MSIs).29
Graduate school administrators and diversity officers could assist academic
programs in developing discipline-specific lists of this sort, which would
enable faculty to recruit where students are, rather than waiting for stu-
dents to approach them.
Table 11 Top 30 Institutions Producing African American, Latina/o, and Asian / Pacific
Islander PhDs
Top 10 African Top 10 Asian / Pacific

American Top 10 Latina/o Islander
1 Howard University University of Puerto University of

Rico–Piedras California–Berkeley
2 Spelman College University of Puerto University of
Rico–Mayaguez California–Los Angeles
3 Florida A&M University of Massachusetts Institute
University California–Los Angeles of Technology
4 Hampton University University of Harvard University
Texas–Austin
5 Southern University Florida International University of
A&M College University California–San Diego
6 Jackson State University of Texas–El Cornell University
University Paso
7 Morehouse College Harvard University Stanford University
8 University of University of Florida University
Michigan–Ann of Hawaii–Manoa
Arbor
9 North Carolina A&T University of New University of
University Mexico Michigan–Ann Arbor
10 University of University of Arizona University of
California–Berkeley California–Davis
Sources: Lundy-Wagner, Vultaggio, and Gasman 2013, 158, citing Survey of Earned Doctorates 2009.
Informal recruitment relationships can become institutionalized over

time, such as the Fisk-Vanderbilt Masters-to-PhD Bridge Program. This re-
spected multidisciplinary program identifies promising undergraduates from
Fisk University who would like additional preparation or research experi-
ence before beginning a PhD. Students enroll in a master’s degree program
at Vanderbilt, which subsequently offers them “fast-track admission” to
participating doctoral programs there.30 Bridge programs are growing in
popularity as a means of broadening access to doctoral education, espe-
cially in STEM fields, and the National Science Foundation regularly fields
calls for proposals for institutions that would like to develop them.
Undergraduate research programs also develop students who are inter-
ested in and qualified for doctoral study. Faculty-supervised research expe-
rience has long been promoted as a nonremedial strategy to encourage the
success of undergraduates from underrepresented backgrounds. Among
the studies that find positive outcomes of undergraduate research participa-
tion, scholars at UCLA’s Higher Education Research Institute recently con-
Conclusion 171
cluded through a rigorous quantitative methodology that it increases black,

Latino, and Native American students’ participation in graduate education.31
Other research, using qualitative methods, indicates that student research
promotes continuation to graduate education through mutually reinforcing
processes: (1) attracting students by solidifying aspirations and providing
early socialization, and (2) making applicants, especially with less conven-
tional profiles, more attractive to faculty in the admissions process. Through
extended engagement with research, students cultivate forms of cultural
capital deemed valuable in doctoral education and the academy, including
fluency with disciplinary jargon, ease in discussing technical details and the
significance of one’s research, and national presentations or publications.32
I found in this study that professors make critical inferences about prospec-
tive students from their research experiences, such as how they might con-
tribute to the future of the discipline. As with many admissions criteria that
involve some subjectivity, decision makers may want to set norms before
they review files about how they will evaluate research experience, rather
than allowing themselves to be impressed by students with the most prom-
inent research experiences.
Finally, I want to offer a few comments about a common recruitment
dilemma. Graduate programs striving to increase their enrollment of a par-
ticular group often find themselves stuck in a negative feedback loop: For
admitted students from the underrepresented group debating whether or
not to enroll, an absence of individuals who share their identity can raise
red flags about what the quality of their experience will be. This uncertainty
may contribute to their decision to matriculate elsewhere, which perpetu-
ates the group’s underrepresentation. Breaking this cycle requires strong
leadership and evidence that diversity is, for the program, more than a plat-
itude. Those in recruitment roles should be ready to initiate honest conver-
sations about how and why diversity matters, what it means for their schol-
arly work, what students from underrepresented backgrounds will gain
from and offer to the program, and what the climate is like for students
from that group in the program, campus, and broader community. Espe-
cially for people from privileged backgrounds, it can be hard to speak
frankly about these issues because they are always works in progress. How-
ever, developing the proficiency to do so in recruitment, admissions, men-
toring, and instructional contexts is part of the change process itself.
Examine Assumptions about Merit

Perhaps the most fundamental implication of this study is the need to re-
frame relationships among fundamental principles like excellence, merit,
and diversity. Diversity was one dimension of merit when faculty assessed
applicants on the short list, but many spoke about “diversity with excel-
lence” as well—as if these are independent entities accompanying one an-
other. The Association of American Colleges and Universities recommends
educational institutions work toward a culture of inclusive excellence,
which
integrate[s] diversity, equity, and educational quality efforts into their missions
and institutional operations . . . The action of making excellence inclusive re-
quires that we uncover inequities in student success, identify effective educa-
tional practices, and build such practices organically for sustained institutional
change.33
In admissions, inclusive excellence means broadening recruitment efforts

beyond the relatively small network of colleges and universities through
which opportunities tend to flow, reassessing assumptions about what it
means to be qualified (and what the grounds are for those assessments),
and tracking the equity implications of current selection practices and any
proposed reforms.34
Making excellence inclusive also means recognizing that common per-
formance metrics do not tell the full story about underrepresented students’
potential.35 Research is needed to refine measures and assess their predictive
validity, but there is a growing movement to formally assess students’ non-
cognitive strengths as part of a holistic definition of what it means to be
qualified for doctoral education.36 For example, the Fisk-Vanderbilt
Master’s-to-PhD Bridge Program added an interview to their selection pro-
cess that assesses students’ grit and resilience. It asks them to reflect upon
their interest in science, how they persevered through challenges, and the
resources and relationships to which they turn in struggles. Their outcomes
have been impressive. Since 2004, 81 percent of its entrants have continued
on to doctoral studies, which flips a national trend in which 80 percent of
students of color with STEM bachelor’s degrees do not continue to grad-
uate school.37 Raising the profile of promising alternatives to current edu-
cational and selection models has value when a dearth of alternatives is
often presented in defense of the status quo.
Stop Misusing Standardized Test Scores

For some faculty, meeting a standard of fairness in admissions will require
that they learn more about the proper interpretation and limits of standard-
ized test scores (GRE, TOEFL, GMAT, others) and reassess their use in the
admissions process. Documentation from ETS, for instance, indicates that
Conclusion 173
the GRE measures skills. It advises that raw scores and percentiles should
not be used as sole or primary admissions considerations, and that score
thresholds should not be applied.38 However, decision makers in these pro-
grams routinely used scores as a primary consideration in the first round
of review. About half thought of scores as signaling intelligence, and it was
common practice to assess applicants against formal or informal score or
percentile thresholds. Simply put, they misused the GRE.
The appeal of simple metrics is undeniable. Indeed, anything that makes
this complicated review process more convenient is hard to ignore. How-
ever, the patterns I observed both run contrary to the psychometric
properties of the test and put already underrepresented populations at a
disadvantage. By failing to read scores in context, reviewers made overly
simplistic comparisons about students’ relative potential. Many had a gen-
eral awareness that GRE scores have patterns associated with race, gender,
and first-year graduate school GPA. Fewer knew that research studies have
come to very different conclusions about the strength of the correlation
between test scores and longer-term academic outcomes, or that stereotype
threat helps explain group-level disparities.39 Professors thus need better
knowledge of ETS guidelines, of research about the full scope of factors
that test performance can reflect, as well as its validity for different groups
of students and for short- versus long-term outcomes. With this knowledge,
admissions leaders can come to more thoughtful decisions about whether to
require the test and, if they choose to do so, how to appropriately use and
interpret its results.40
Awareness of the many problems with standardized test scores—that they
are more complicated than they appear, that reliance on them can under-
mine equity efforts, that they do not reliably predict students’ long-term
success—has led some higher education institutions to reduce their reliance
on these measures or eliminate their requirement entirely. As of summer
2014, the National Center for Fair and Open Testing reported that 103 se-
lective colleges and universities had implemented test-optional admissions
processes. At the graduate level, doctoral programs in at least thirty-two
different fields have eliminated their GRE requirement, including ones at
Columbia University, University of Pennsylvania, Johns Hopkins Univer-
sity, and the University of Michigan.41 Research is needed about the effects
of going test optional, both at the undergraduate and graduate levels. In
one study, results indicated that making standardized test scores optional
resulted in an increase in the enrollment of women, students of color, and
international students, without changing the institution’s mean GPA or gradu-
ation rate. But in another study, researchers found that going test-optional did
not on its own increase diversity, but instead increased selectivity because only
the applicants with higher scores submitted theirs.42 If increasing the enroll-
ment of students of color is specifically a goal, making GRE scores optional or
declining to review them are options that should be on the table. However,
they will be more successful as part of a coordinated set of efforts.43
Seek Small Wins in Admissions as Part of

a Multifaceted Change Process
Changing culture and confronting inequality can seem daunting prospects,
and that may be part of the problem. According to organizational theorist
Karl Weick’s idea of “small wins,” how we define long-term challenges af-
fects the chances of making progress on them, because when social prob-
lems are conceived at large scales, people become immobilized from taking
action and they perform more poorly when they try. “When the magnitude
of problems is scaled upward in the interest of mobilizing action,” he writes,
“the quality of thought and action declines, because processes such as
frustration, arousal, and helplessness are activated.” Weick therefore rec-
ommends that organizations approach major challenges by “identify[ing] a
series of controllable opportunities of modest size that produce visible re-
sults,” particularly in how they interpret and classify problems, which
shapes the scope of possible solutions.44 Among the examples of small wins
that Weick cites is Alcoholics Anonymous conceptualizing sobriety as a goal
that is achieved one day—even one hour—at a time.
Improving admissions is a small win relative to systemic problems of in-
equality in academia and society. Yet to bring the scale down to something
even more manageable, admissions itself holds possibilities for small wins
because the work is replete with classification. The endgame, of course, is
to classify a subset of the applicant pool as admitted, but along the way
decision makers employ an elaborate mental classification system when they
infer complex qualities like intelligence and potential from the application.
Counterscripts—alternative interpretations of common criteria—serve as
a small win by systematically shifting how decision makers mentally clas-
sify students. By focusing on engrained assumptions, small wins can build
consensus and create sustainable change.45 By shifting discourse they help
shift conventional wisdom, motivating change as they signal that such change
is already under way.46
The field of philosophy offers an excellent case of how small wins in ad-
missions can be part of a broader strategy for equity-minded change, in
their case, with respect to gender. In one of the two the committees I ob-
served, female members and male allies actively challenged their colleagues
on misinterpretation of GRE scores, and they offered counterscripts about
Conclusion 175
how to read undergraduate college prestige. On the latter, they introduced

into the deliberations some of the reasons other than talent (such as finan-
cial barriers, family commitments, or opting out of the prestige chase) that
students attend less-selective institutions. The support of a critical mass for
these counterscripts shifted the dominant narratives about at least three bor-
derline applicants. Instead of playing a determinative role, scores and affili-
ations counted as two among many considerations.
In addition to broadening the meanings attributed to GRE scores and
college prestige, individual faculty, graduate students, and small groups in
both of the philosophy departments I worked with are engaged in other
departmental and national efforts to increase women’s participation. At the
department level, they are hosting coffee hours and dinners for aspiring fe-
male philosophers, and creating structures through which individuals can
safely report experiences of discrimination or harassment. At the national
level, female faculty in these departments are working through formal and
informal collectives and campaigns.47 There are active networks promoting
equity and opportunity for women through the American Philosophical As-
sociation’s Committee on the Status of Women (APACSW) and the Women
in in Philosophy Task Force, for example. One recent APACSW campaign
gathered and posted online information from master’s and PhD programs
about their enrollment, retention, and job placement rates by gender, as well
as comments about each one’s efforts and achievements to increase wom-
en’s participation. Another popular campaign compiled photos of more
than 200 women in philosophy into posters with the line “Philosophy. Got
women?” There are also anonymous, cooperatively published websites
(titled “Feminist Philosophers” and “What is it like to be a woman in phi-
losophy?”) that provide a public forum for speaking out and sharing expe-
riences. As philosophers work to change the culture of philosophy, changing
the assumptions that decision makers bring to admissions evaluations is just
one part of their efforts.
Conclusion
I have argued that judgments of the “best” graduate programs in the country
and the “best” applicants in the pool come about only by negotiating across
priorities associated with specific social contexts. Thus, there is wisdom in
my colleague Paul Courant’s admonition to “beware the tyranny of best
practices.” The best practices for admission to a department’s doctoral
program may be different from those used to admit master’s-level students,
who may have different professional goals. What is best for recruiting
graduate students to New York or Los Angeles may be different from what
it takes to draw students to Madison or Boulder. From this perspective, a
final implication of this study is not the need for a specific set of practices,
but rather for faculty to approach their gatekeeping work with a different
state of mind. Instead of proceeding by default or adopting a specific check-
list of procedures developed for some other program’s needs or goals, ad-
missions decision makers need to approach gatekeeping with mindfulness
of their own situation and needs. When evaluation and selection happen
by default rather than intention, the social consequences of decision
making may be misrecognized as the normal course of events—the way
things should be rather than the way things ended up. And over time those
outcomes can be thought of as natural rather than deliberate, for the more
deeply held a belief is, the more likely it is to keep a person from seeing
things any other way.
Today’s faculty choose students on the basis of an array of perceptions
that only sometimes have a strong evidentiary basis. In a process so com-
petitive that the mere presence of doubt can seal an applicant’s fate, per-
ceptions often carry the weight of truth. They drive decision makers to
act—to rate and advocate, to admit and reject—as if perception were
reality. As the theorem made famous by sociologists William and Dorothy
Thomas puts it: “If men define situations as real, they are real in their
consequences.”
Context and circumstance will always establish broad parameters within
which we make choices. Yet social constructionism teaches us that the so-
cial world is as we perceive it to be because we act upon our perceptions.48
Admissions judgments are socially situated; they result from inference upon
inference, and from filters associated with organizational contexts and
goals. The benefit of research like this study, which puts participant inter-
pretations at the center of analysis, is in uncovering these taken-for-granted
filters, for it is only by stepping outside of and analyzing them that we can
understand their power. Like the prisoners in Plato’s cave, misled to believe
that shadows on the wall were reality, scholars may not even realize that
the way they have been conditioned to see excellence in themselves and
others is not natural, but constructed to serve specific ends. Yet unlike the
prisoners, scholars can shape the social contexts in which their judgment is
situated. In doing so, they refine the lenses through which their own un-
derstanding—and that of their colleagues and students—will be filtered.
Indeed, one of the great values of understanding an organization’s gate-
keeping systems is that it lays bare cultural values that drive policy, which
are often so engrained as to be taken for granted. Too often our preferences
and aversions go unspoken—even unrealized at a conscious level—so that
Conclusion 177
misguided assumptions and implicit biases are never challenged and con-
tinue to shape the outcomes of review. Honest dialogue about the values and
norms that shape interpretations of scholars’ records is therefore needed,
not only in admissions, but across the disciplines and the range of gate-
keeping processes the academy employs. These conversations are relevant
to how we admit students; how we hire postdoctoral researchers and fac-
ulty; how we select individuals for awards, grants, and fellowships; how
we advance students to candidacy; and how we assess faculty scholarship
for tenure and promotion. Each of these evaluation processes is fundamen-
tally an activity of defining and monitoring organizational boundaries, of
determining the grounds for membership, belonging, and recognition.
The faculty I spoke with want their students to succeed, and they like
the idea of increasing diversity of many sorts. They carry out gatekeeping
activities consistent with field-level norms and with those of their own
training. Yet having mostly graduated from selective doctoral programs
themselves, and with limited exposure to other models, their inclination
may be to so strongly associate their own training with “quality” training
that they recreate its selection process and the educational gauntlet in the
interest of preserving what they see to be the purity of the discipline and
excellence of the program.
However, society is changing. The labor market for PhDs is changing.
And higher education is changing, too, both intellectually and demograph-
ically. If doctoral education and our means of identifying talent do not
change with them, it will be doctoral education and the professoriate that
fail by falling behind, not our students. Re-creating academic programs and
disciplines in our own image may be the natural tendency, but in a changing
world—one that is more diverse, more collaborative, more interdisciplinary—
stewardship of our disciplines’ futures means being flexible to recognize
what the world needs from our fields of study and adapting our ideals of
excellence accordingly.
Methodological Appendix
Research design follows from one’s research questions, and the two questions that
guided my design were straightforward: How do faculty members individually judge
and collectively select prospective doctoral students? How do disciplines shape judg-
ments of admissibility? I could have answered these questions through a number of
methods but opted to design an ethnographic comparative case study after weighing
several options. I could have statistically inferred faculty priorities from the char-
acteristics of students that they admitted, but I wanted to explore reasons for the
apparent contradiction between diversity as an institutionalized value and con-
tinuing inequalities in higher education. I also wanted to see how admissions nego-
tiations unfolded over multiple rounds of review—knowing from previous research
that there were likely to be differences between how reviewers approached the pool
as a whole compared to how they approached the short list.1 Third, quantitative
research only picks up patterns present across many people, and many of the rea-
sons that individuals are selected for graduate programs may be idiosyncratic to
committees, programs, disciplines, or specific applicant pools. Thus, the benefits of
greater generalizability that might have come from findings of a quantitative study
would have come at the cost of capturing the very complexities that make it a so-
cially important, and intriguing, phenomenon.
My own conceptualization of faculty evaluation also suggested the need for nat-
uralistic research. I think of evaluation as an activity situated in, and thus shaped
by, social forces associated with multiple organizational contexts, whose priorities
may or may not align. I therefore sought to be closely engaged with the phenom-
enon as it was happening, where it was happening. Observations would also help
triangulate what the evaluative scripts in use were, per methods used in Goffman’s
180 Methodological Appendix
Table 12 Overview of Sample and Data Collection
Discipline Subject Matter Year University Interviews Observations
Astrophysics Natural Science 1 B X X

Biology Natural Science 2 C X
Physics Natural Science 2 C X
Philosophy 1 Humanities 1 A X X
Philosophy 2 Humanities 2 C X X
Linguistics Humanities 2 C X X
Classics Humanities 2 C X X
Economics Social Science 1 A X
Political Social Science 2 C X X
Science
Sociology Social Science 1 A X
initial work on scripts, Garfinkel’s ethnomethodological tradition, and more recent

organizational and institutional scholarship.2 I knew that observational data would
also be useful in drawing out aspects of decision making that are important for out-
comes but that faculty may take for granted as routine. Finally, collecting and ana-
lyzing data at multiple levels of analysis would facilitate theory elaboration, which
is often an implicit aim in inductive, comparative case studies.3
After weighing these considerations, I conducted comparative ethnographic case
studies of the PhD admissions cycle in ten departments that represented disciplines
in the humanities, social sciences, and natural/physical sciences. They were located
in three research universities with very high research activity, and data collection
took place over two admissions cycles (2010–2011 and 2011–2012). See Table 12
for details.
Case Study as a Methodology

Merriam defines case study as “in-depth description and analysis of a bounded
system.”4 This definition has four critical elements: depth, description, analysis, and
bounded system. Case study methodologists strive to illuminate a particular phe-
nomenon by learning as much as possible about the particulars of one or a handful
of cases rather than learning a few things about a very large sample.5 Cases are care-
fully selected for how they are likely to reveal the phenomenon of interest.
Description and analysis are the researcher’s means of revealing the case. “Thick
description” uses details of what is immediately observable to take readers beyond
the immediately observable into the realm of deeper social meanings. In his famous
article on the topic, Clifford Geertz illustrates thick description by considering the
difference between an eye twitching and winking.6 Thin description would note
muscle contraction in one eye, but miss whether it was involuntary (a twitch) or
conspiratorial (a wink) because this distinction may also require detail about the
actors, setting, other communication, and circumstances. Description along these
latter dimensions ensures that the meaning of a socially important moment is
not lost. Small matters—the lighting in a conference room, the order in which
applicants are reviewed, the topics of side conversations preceding the “real”
conversation—may matter much as cultural insights, and through describing
such details I have tried to capture the social and organizational complexity of
faculty evaluation.
Case study methodology does not prescribe specific analytic methods, but weaving
together a case narrative that both accounts for details and sorts out participants’
signification structures is empirical, interpretive work.7 To make sense of the volu-
minous data typically collected for case studies, the researcher’s meaning-making is
a key intellectual activity of analysis.
The final element in Merriam’s definition is the case itself—the object of study—
as a bounded system. The case can be an instance of a population or of a concep-
tual class, but either way it may be thought of as the phenomenon in context.8 I
strove to capture individual and collective faculty judgment in situ, so my case
boundaries were organizational (highly selective PhD-granting programs) and tem-
poral (the admissions cycle from application to yield). Case selection often takes
place by selecting on the dependent variable (identifying a case or cases that maxi-
mize the phenomenon of interest). In a regression this sampling approach is under-
stood to produce selection bias, but comparative case study researchers make no
assumptions about broad generalizability from the case. Instead, their goal and in-
tellectual contribution is deep understanding, including capturing necessary (versus
sufficient) conditions.9 For case study researchers, it is more important to gain a
rich picture of a phenomenon in a meaningful context.
Sampling
Rather than generalizing from their small n’s to broader populations, case study
scholars hope that the theoretical leverage obtained through thick description and
in-depth analysis generates understanding that can illuminate similar cases. Case
study sampling involves selection decisions at two levels: (1) the case, which serves
as the primary unit of analysis, and (2) subjects within the case, whose perspectives
help the researcher craft a holistic depiction.10 At both levels I used criterion sam-
pling, in which the researcher chooses cases that meet a predetermined criterion or
set of criteria.11
Because the majority of doctoral degrees are awarded in research-focused uni-
versities, I focused on programs within three universities that have the Carnegie clas-
sification Research Universities–Very high research activity (RU/VH). Two are
public and one is private, and they are located in two regions of the United States.
Both for participant and program confidentiality and because institutional details
are less salient to my analysis than details of the departments, programs, and disci-
plines, I do not discuss in the results the names or other identifying details of the
universities in which data were collected. Initially I was disappointed by the Insti-
tutional Review Board’s constraint on discussing the details of the universities in
which I collected data. However, as my time in the field went on, I came to see that
the universities varied more in their yield—the ability to draw admitted students—
than in their selection processes, which was my focus.
Table 12 summarizes the sampling design. Within each university, I sampled PhD-
granting programs representing disciplines in humanities, social sciences, and
physical/natural sciences. I expected that such programs would evince disciplinary
influences more clearly than applied or interdisciplinary programs. I also narrowed
my choice of institutions to those in which the fields of interest maintain programs
ranked in the top fifteen in their field.12 In these contexts, many qualified individuals
apply; therefore, the competing demands of selection come into sharper focus.
Although I initially hoped to study the same departments at multiple universities,
there was only one instance in which I could gain access from a second department
that was also ranked in the top fifteen in the discipline. The exception to this is phi-
losophy, in which I collected data at two of the three universities. However, within
the humanities, social sciences, and natural/physical sciences, I sampled at least one
higher-consensus discipline (such as classics, economics, or physics) and one lower-
consensus discipline (such as linguistics, political science, or biology), which has lent
structure and intellectual focus to my cross-case comparison.
My sample in each of the ten programs consisted of the chair of the department
and/or admissions, members on the admissions committee, and one emeritus pro-
fessor. Where this group totaled fewer than six people due to a small committee
size or a small proportion of the committee who consented to be interviewed, I in-
terviewed additional faculty members. The astrophysics, linguistics, and political
science departments each included one to three graduate students on the admissions
committee as well, for a total of six graduate student participants. I wanted to create
a picture of admissions as it is currently practiced in these departments, and there-
fore I did not oversample by race/ethnicity or gender. As a result, my sample re-
flects the same problems of underrepresentation by female and domestic faculty of
color that the U.S. professoriate does.
Recruitment
Sensitive to the ethical and reputational implications of this research for the pro-
grams and individuals who participated, I tried to design and write it in ways that
protected them, and to make it as easy as possible for prospective participants to
feel comfortable in our early interactions. In the initial contact, I framed my study
as a matter of decision making and the challenge of balancing multiple interests.
Although I probed for issues around diversity throughout data collection, I saved
the protocol questions about diversity for the end of interviews, after some rapport
and the beginnings of trust had been established. Also, I outlined in both the initial
email and the informed consent document the measures I took to ensure confiden-
tiality. In summary, although the nature of the research undoubtedly deterred some
from participating, I made a concerted effort to minimize the risks to participants
and maximize their anonymity.
With a target sample of four to six programs per year, I initiated contact via email
with seventeen programs in the first year of data collection and eighteen programs
in the second year. My first contact with each was an email message in September
to the department chair or program director in qualifying programs at prospective
universities. I requested a thirty-minute informational interview to discuss their pro-
gram’s approach to admissions. The department chair referred me in most cases to

the admissions committee chair for this interview. Ten and nine programs agreed
to informational interviews in the first and second years, respectively. I continued
the recruitment process during those interviews by explaining the research study
design, but I encouraged the chair to consult others in the department before making
a decision. In a few cases, general consent to conduct the departmental case study
was immediately provided or declined at the informational interview, but more often
this occurred via subsequent email communication. We agreed in all cases that in-
dividuals would each have the opportunity to consent or decline participation, and
that if any faculty member on the admissions committee preferred for me not to
observe, that I would not observe that department’s committee.
Once general consent to conduct the case studies was provided, I requested an
official approval letter to file with the IRB, printed on department letterhead and
signed by the department chair. I also requested from the admissions chair a list of
their committee members for the year and the names of one or two emeritus fac-
ulty whom I might recruit for an oral history of the department and its admissions
practices. Then, using an email template similar to that used with department chairs,
I emailed each committee member an invitation to be interviewed for the study.
As described above, in departments with small committees or where fewer than
six committee members and emeriti agreed to be interviewed, I used snowball sam-
pling within each department to develop a list of individuals with few and many
years of admissions experience whom I could also recruit to participate.
Data Collection
To achieve deep and holistic understanding of a case, leaders in case study research
advise using all three of the common qualitative data collection techniques—
interviews, observation, and artifact analysis.13 In this study, data collection consisted
of observations, interviews, and collection and analysis of department websites
and participants’ curricula vitae. In the first year, I collected data from one pro-
gram at one university and three more at another university. Then, in the second
year, I collected data from six programs at one university.
Observation. Observation of admissions committee meetings in six of the ten pro-

grams was a crucial complement to the interview data by laying bare the priorities
and preferences that shaped admissions outcomes in practice. I feel privileged to
have had access to this invaluable source of knowledge, which also highlighted
similarities and differences across disciplines in how faculty make admissions
decisions.
I conducted observations in four types of events over the two years of data col-
lection: admissions committee meetings, faculty training workshops at two univer-
sities, Skype interviews of applicants in two programs, and campus visit weekend
events in one program. Every member of the admissions committee in six of the ten
departments kindly consented for me to attend their admissions committee meet-
ings, for a total of ten meetings observed overall. The shortest meeting was eighty
minutes and longest was three hours and fifteen minutes (both in the political science
department), for a total of twenty-one hours of meeting observations. I took struc-

tured field notes longhand to document the following:
1. The frequency with which various criteria and applicant traits were mentioned
2. The patterns of participation and interaction among faculty
3. How faculty discussed issues of merit and diversity
4. Which members of the committee appeared to have power or to be
marginalized
5. How applicants were compared with one another and with external standards
6. Examples of consensus and conflict and how these were reached/handled
7. The physical setting and participants’ nonverbal communication in the setting
The astrophysics committee also added me to their email list, so that I could ob-
serve their email interactions.
In addition, I received permission from administrators in two of the three uni-
versities to attend admission workshops conducted by their graduate schools for
admissions committee chairs and faculty members throughout the university. These
training events were one and two hours long, respectively. The astrophysics and
physics departments also offered to have me sit in the room during select Skype
interviews they planned with prospective students. Over two hours, I sat in on a
total of four such interviews—two with domestic students and two with students
from China—as well as the subsequent debriefing faculty had with one another fol-
lowing the interviews. Finally, the astrophysics department invited me to sit in on
the two-hour introductory session of their campus visit weekend hosted for admitted
students. I took open-ended field notes in these settings.
Insider status presents decided advantages in interview settings, but when it comes
to observation, being an outsider has clear benefits, assuming that access to the
group is permitted and group members do not make significant changes to their
language or behavior. Chief among the benefits is that outsiders are able to notice
what has become routine to or taken for granted.14 For all observations, I sat in the
back of the room or in a corner to minimize my obtrusiveness, and within twenty-
four hours I transcribed the field notes into electronic format and composed memos/
reflections. According to Gold’s classic typology of observers’ stances toward the
observed, I was an “observer as participant.”15 That is, the group knew my activi-
ties and research aims, but did not count me as one of its members or invite me
into the activities I observed. There were moments early in a few meetings where
those around the table commented on my presence or seemed to self-monitor, but
the many times faculty spoke in ways that were socially undesirable led me to be-
lieve that I was seeing more than a performance of their best behavior.
Interviews. I conducted eighty-six interviews ranging from twenty-three to eighty-

five minutes in length with four groups: department and/or admissions committee
chairs, other committee members, additional faculty, and faculty emeriti. As de-
scribed above, recruitment for the study involved thirty- to sixty-minute semistruc-
tured informational interviews with program chairs and/or admissions committee
chairs in prospective programs. These unrecorded interviews in faculty offices in-
quired into details of the department’s admissions process and assessed interest in
further participation in the study. Of the seventeen informational interviews I

conducted, eleven resulted in approval for me to interview other department fac-
ulty; however, I opted out of one because only one other committee member in their
program was willing to be interviewed. The final sample therefore includes ten
departments. Of these, eight of the ten chairs also agreed to an unstructured follow-
up interview after final offers were made. In some of the most candid interviews of
the study, we discussed their reflections on how the admissions process went, ad-
missions challenges, and deliberations about applicants on whom it was difficult to
find consensus.
The initial research design proposed two thirty-minute interviews with each com-
mittee member—one before and one after file review. However, in the first year of
data collection, many participants expressed concern about the scheduling difficul-
ties this would pose, so I proposed as an alternative a single, longer interview. That
approach proved much more appealing and resulted in response rates in excess of
70 percent. In the second year I made the single, longer interview the norm for non-
chairs, and had almost 100 percent participation from all six committees. Consis-
tent with a semistructured approach, these interviews varied somewhat, depending
on when in the admissions process the interviews were conducted.
Standard faculty interviews began with faculty providing a brief narrative of their
route to the professoriate, emphasizing their personal experience with graduate ad-
missions so that I might ascertain any personal factors that might prove salient to
their admissions work in the present. We also discussed important criteria and the
meanings and signals they carry; perceptions and challenges of committee deliber-
ations; profiles of easily admitted, easily rejected, and borderline cases; and how
traits of highly valued faculty candidates compare to those of compelling PhD pro-
gram applicants. These interviews averaged approximately fifty minutes each, and all
but one took place in faculty/student offices. Upon my request in the informed
consent document, all but three consented to have the interviews audiotaped
and transcribed. Consistent with principles of emergent design in qualitative re-
search, I adapted the interview protocol in response to unexpected themes that
emerged in early interviewing. Most importantly, after the first five interviews in
the first year of data collection, it became apparent that a common quality that
participants sought in applicants was intelligence, but that their beliefs about what
this meant varied widely. In response, I added a question to the interview asking
them to reflect on the meaning of intelligence in their discipline and how they try
to recognize it in applicants.
Finally, to contextualize current admissions practice in these programs, I inter-
viewed an emeritus professor in each program to gain an oral history of develop-
ments in the department and discipline, and to probe for connections between the
discipline’s epistemology and selection criteria. These audiotaped interviews aver-
aged just over sixty minutes each, and took place in either their homes or campus
offices, per the participant’s preference.
I developed the interview protocols using principles advocated by Robert Weiss
in Learning from Strangers: The Art and Method of Qualitative Interview Studies
and Herbert and Irene Rubin in Qualitative Interviewing: The Art of Hearing
Data.16 Asking main questions was intended to “elicit the overall experiences and
understandings” of participants, while follow-up questions clarified responses and

probes encouraged the participant to provide detail that could produce a more com-
plete narrative.17 A number of the questions had been used in or revised from a
small-scale study I conducted in the winter and spring of 2009 on merit and diver-
sity in graduate admissions. Using interview strategies employed in studies by Wil-
liam Tierney and Estela Bensimon, Stephen Barley, and Michéle Lamont, a set of
important questions inquires into “ideal types” of applicants.18 First advocated by
Weber as a research technique, understanding ideal (and less than ideal) types of ap-
plicants is useful, not because they represent typical applicants or individuals who are
admitted, but because they clarify symbolic boundaries and serve as models or heu-
ristics to guide our thinking about social phenomena.19 By integrating the observa-
tional and interview data, I have tried to outline common scripts of merit and
other decision pathways.
Audio-recorded interviews were transcribed verbatim. I personally transcribed
25 percent of the interviews to aid in reflecting on the quality of my interviewing
and the effectiveness of the interview protocol. Personal transcription also enabled
me to remain mindful that transcripts do more than put talk in print; they re-present
discourse, including both what is said and not said.20 Transcripts thus noted the
relative length of pauses, verbal pauses, events that interrupt the flow of conversa-
tion, laughter, in addition to words spoken. I augmented select transcripts with re-
searcher memos as a forum for reflexivity, or self-critique about my work and self
as researcher and instrument of research.21
History, place, and thick description. Understanding contemporary social and ed-
ucational phenomena requires us to consider how they are situated spatially and
temporally.22 Present events and prevailing logics constitute “the signature of his-
tory.”23 Details of space and setting similarly shape social phenomena. What may
appear to be background details are more than local color, for the dimensions of time
and space create dependencies and contingencies that help explain how and why a
case plays out as it does under our analysis. Howard Becker describes such detail
as “the environing conditions under which the things we stud[y]—the relationships
we uncover, the general social processes whose discovery we want to brag about—
exist.”24 While rigorously adhering to the standards of confidentiality that human
subjects policy and informed consent agreements demanded, my analysis involved
careful observation, investigation, chronicling, and sifting of details about the set-
tings and histories of the programs, disciplines, and universities to draw out
context-specific variation and provide rich case descriptions.
Data Management
Case studies are notorious for the amount of data they generate, and I knew that
my ability to craft a meaningful narrative would depend partly on its overall orga-
nization and the accessibility of individual bits of data. My first step in managing
the data was creating an inventory, for which I maintained a comprehensive Excel
workbook. To organize the field notes, interview transcripts, documents, and memos,
I utilized NVivo 9.2. NVivo was also a critical tool for me in data analysis.
Data Analysis
The aims of case study—to “uncover patterns, determine meanings, construct con-
clusions, and build theory”—provided a guide to selecting appropriate analytic
methods.25 Techniques from the grounded theory tradition provided a systematic
means of capturing patterns and meanings and of developing conclusions and
theoretical contributions. Grounded theory’s constant comparative method is prev-
alent across various qualitative research traditions, including case studies.26 I drew,
in particular, from Strauss and Corbin’s approach, which involves three stages of
coding (that is, seeking patterns in the data).27 Data collection and analysis occurred
simultaneously to facilitate ongoing reflection on the data and adjustments to my
interviewing and observations. However, analysis continued long after data collec-
tion concluded. I initially analyzed the data one department at a time and used those
analyses to write cases of each department. Then I conducted a cross-case analysis
in which I compiled all of the cases into a new NVivo project and interpreted case
similarities and differences by disciplinary type and paradigm strength using the
same micro-level analytic steps described below. Analysis was iterative, alternating
between aggregated data from specific cases, cases themselves, and the sample as a
whole.
Analysis at the micro level proceeded according to the three stages of coding rec-
ommended in grounded theory: open, axial, and selective. During the initial open
coding phase, I stayed close to my original research questions, looking for themes
in faculty members’ decision making processes, the meanings they associated with
common criteria, and evidence of disciplinary cultures. However, the second round
of open coding was more inductive, and through this I developed a variety of ad-
ditional codes including homophily, bureaucracy, aversions, intelligence, and many
more. For the purposes of an ethnographic case study, I think of axial coding as
data-driven theory generation and of selective coding as an opportunity that case
study presents to intersect theory generation with thick description. It was through
selective coding that I developed a narrative that accounts for the relationship of
the codes to one another and to critical details of the case. By interrelating the data
and themes across levels of analysis and abstraction, my findings include a combi-
nation of description and interpretation at both the individual level and the case
level.
Case Development and Cross-Case Analysis

In the months following each admissions cycle and my initial analysis of data from
programs studied that year, I composed cases of each program. Most were between
five and fifteen pages, single-spaced. Writing these cases was a substantial task due
to the number of departments I had studied and the amount of data I had gener-
ated about each. Cases were understandably shorter in the programs where I con-
ducted only interviews, and longer in the departments where I also observed ad-
missions meetings and other events. My goal in composing each case was to represent
what the results of the study might have been if I had sampled only that case. I devel-
oped and utilized a case template that was organized around answering the research
questions and reconstructing specific episodes in which faculty negotiated ques-

tions of merit, diversity, and equity (either the equity of their choices or the equity
of their process). The goal of cross-case analysis was to answer the research ques-
tions in a rigorous way that would enable me to make broader theoretical
propositions.28
Protecting Participant Confidentiality

Protecting participant confidentiality ranked high on my list of concerns in this
study, as did protecting the anonymity of any student applicants discussed by the
committees. Among the several items in the suite of protections, I masked or gave
pseudonyms to all information that might be personally identifiable. In referring to
specific institutions, I sought to balance the need to ensure anonymity with the need
to convey a real-world sense of the institutional strata in which these programs are
located. Therefore, when the need arose to name specific universities including
the data collection sites, I randomly drew from a list of fifteen institutions in the
same tier of program rankings for that discipline. This means that the actual uni-
versity in which data was collected could have potentially been named due to
chance, but readers should be no more able to recognize that institution as being
the data collection site than any other, similarly ranked institution.
Trustworthiness and Reliability

I engaged in a number of practices to increase the trustworthiness and reliability of
the findings. These include data triangulation across interviews and observations,
thick description, reflexive researcher memos, member checking, and searching for
disconfirming evidence. Member checking is rooted in the idea that the qualitative
researcher’s role is not to impose one’s own voice on those of participants, but rather
to get inside of participants’ perspectives and give voice to their concerns. Practi-
cally, member checking consists of consulting research participants to learn “whether
the data analysis is congruent with the participants’ experiences.”29 My objective
was to discuss whether and how preliminary findings about each department reso-
nated with the respective admissions committee chairs, who had practical, profes-
sional knowledge of the processes I was trying to capture.
I circulated a draft of each program’s case to its respective admissions chair and
offered to exchange feedback in person or via email. Most opted for face-to-face
conversations, and I did not receive any negative responses about the credibility or
clarity of the findings; however, not everyone liked what they read about their col-
leagues and their own practice. One department chair provided me with some
additional organizational history that helped me make sense of a gap in the data.
Another explained how their interpretation of test scores has changed in the transi-
tion to the new version of the GRE. In several cases, the admissions chairs with
whom I met also asked for my recommendations about specific changes they might
make to improve the department’s admissions process.
With respect to falsifiability, Popper advocated an approach to knowledge pro-
duction in which progress is measured not only by the volume of evidence that can
be amassed in support of a theory, but by its ability to withstand a search for

disconfirming evidence.30 Lareau supports this view.31 In the name of presenting ap-
parently clean results, she argues that contemporary qualitative researchers in soci-
ology and education too frequently sacrifice the rigor of searching for and responding
to disconfirming evidence. Therefore, a late stage of data analysis in my study in-
volved an additional round of analysis for disconfirming evidence of the prevailing
themes and propositions. Patterns in disconfirming evidence or alternative interpre-
tations are presented before the conclusion section in most chapters of findings.
Limitations
The principal limitations of this study concern data imbalances across departments
and demographic groups. Female, African American, or Latino faculty are under-
represented in the highly ranked departments I partnered with and on the commit-
tees I studied; their voices are therefore underrepresented in my sample. Although
this means that my results capture prevailing viewpoints in these departments
(and thus, I hope, cultural dynamics that help sustain structural inequality), the
results do downplay the alternative viewpoints that individuals from these back-
grounds may contribute.
Another limitation is that I have interview data from all ten programs, but ob-
servational data from only seven. Seven of the ten permitted observation, and one
of these (physics) does not convene for traditional admissions committee meetings.
Though I observed two Skype interviews of applicants in physics, I am unable to pro-
vide concrete examples from deliberations in four departments: biology, economics,
physics, and sociology.
Also, the committee sizes and response rates for consent to be interviewed varied
widely across departments. In an extreme case, although the admissions chair was
eager for my involvement with his committee, only one other member consented to
be interviewed; therefore, I withdrew this case from the set of cases (for a final total
of ten). On the other end, in astrophysics, each member of the five-person committee
was willing to be interviewed multiple times. In several departments the committee
consisted of fewer than six individuals or not every member of the committee con-
sented to be interviewed; in these departments, I recruited faculty to interview from
outside the committee, using snowball sampling starting with the admissions chair.
Researcher’s Role and Establishing Rapport

Origins of this research. My interest in graduate admissions began in the course of
working from 2003 to 2007 with the McNair Scholars Program at the University
of Northern Colorado. McNair is a federally funded TRiO program that facilitates
research opportunities for low-income, first-generation, and underrepresented mi-
nority students, with an eye toward long-term improvement in faculty diversity. I
taught introductory research design courses, advised them in working with their fac-
ulty mentors, and supported those who elected to continue their education through
the graduate school admissions process. Each summer, program staff facilitated
GRE preparation workshops, and we invited a graduate admissions consultant for
a two-day seminar. Then, each fall, I worked one-on-one with students as they se-
lected prospective graduate programs, initiated contact with prospective advisors,
wrote their personal statements, solicited letters of recommendation, and managed
the myriad of details, deadlines, and emotions that the process requires and creates.
Through this work, I became interested in several aspects of graduate admissions.
As mentioned in the Preface, I struggled to find cross-disciplinary, empirical studies
that could inform my advising work, and thought this absence of research was
curious considering how much has been published on undergraduate admissions.
The similarities between doctoral admissions and faculty hiring processes also
became striking to me, including the role that research potential and establishing in-
stitutional fit play in perceived merit. When I decided to pursue a PhD myself, it
was with an interest in building knowledge about the current structure of access to
graduate education.
Positionality and building rapport. I began the research as an outsider to these de-
partments, disciplines, and the professoriate itself. According to Jorgensen, a major
strength of the ethnographic approach is that long-term involvement allows time for
the building of insider perspective and, with it, trust in the researchers.32 Conversely,
where trust is low, the risks are higher that participants will withhold or distort in-
formation, act out their roles differently than usual, or provide socially desirable
answers to interview questions.33 Thus, it came as no surprise that I encountered a
wide range of comfort and candor levels from participants. Five participants of the
eighty-six who consented to be interviewed projected a clear sense of guardedness
throughout the interviews. These interviews were also relatively short. The other re-
spondents thawed over the first fifteen minutes of the interview or displayed ease
from the beginning and an interest in co-constructing bridges of rapport.
Rapport was demonstrated in ways including but not limited to: socially unde-
sirable and politically incorrect responses (which I probed), an offer to provide me
with an office within the department for my work (which never panned out due
to space limits), an invitation to have our families meet for a playdate (which I
declined), requests for additional conversations about the department’s admissions
reform efforts (which I agreed to), and expressions of emotion such as tears, excla-
mations, and laughter (which I noted with gratitude).
Although it was impossible to know how forthcoming any research participant
was being, the faculty I interviewed seemed to be especially direct with me when
interviews took place after I had already observed an admissions meeting in that
department. Having seen them involved in this difficult work, they spoke with the
honesty that one has when there is nothing left to hide. Interviews with emeritus
faculty were also rich, and often entertaining as they passionately recounted to me
the tides of change and currents of stability in their careers and the lives of their
departments and disciplines. Faculty emeriti, I learned, have not only a wealth of
knowledge about the politics and “dirty laundry” of their departments and disci-
plines. Unlike graduate students and untenured professors, they have little to lose
in disclosing it. I also learned they have a deep, emotional appreciation for their
work and the academic communities who have supported them.
Reflecting on this rapport, I brought a few characteristics to the research rela-

tionships that may have unexpectedly worked in my favor in cultivating trust or
candor. My having been a graduate student may have led professors to view opening
up to me as less threatening than if I were a professor, for example.34 Furthermore,
informational interviews suggested that my affiliation with the University of Mich-
igan brand lent credibility, and that my ability and interest in establishing a yearlong
relationship with the departments rather than swooping in for a one-time interview
may have helped build rapport. Finally, with a sample that was 82 percent male
and 85 percent white, my identity as a white woman undoubtedly affected the
majority of participants’ comfort and candor during interviews. I came to realize
in the course of the first year of data collection that when I presented in stereo-
typically female ways—such as wearing pink or wearing lipstick—male inter-
viewees were often more candid with me. Though it never sat well with my feminist
sensibilities, I kept my one pink shirt in regular rotation for the sake of the research
and bought another mid-year off of the clearance rack. Interview-based research is
hardly immune from the social forces that affect everyday interactions.
Conclusion
Each experience one has for research is an opportunity for continued learning, both
about one’s subject matter and the research process. Although I am unable to share
the original data from this study with others due to my agreements with partici-
pants and the Institutional Review Board, I would be pleased to discuss further de-
tails of the research methodology or findings with readers who are interested.
Notes
Introduction
1. All faculty, applicant, and university names, as well as other personally iden-
tifiable information, have been changed or masked to protect participant
confidentiality.
2. My first thought when I scoped out the room was that I must be in the wrong
place. “Come to the conference room off of the copier room,” my email instruc-
tions from the committee chair had read. The only such room I could find, how-
ever, was about 10 feet by 10 feet, with walls on three sides covered with storage
units and filing cabinets. In the middle was a table for four—but which could
seat six if you were willing to get cozy. I walked the hallways to see if there
might be another copier room for the department. By the time I made my way
back I found three of the committee members seated at the table, and concluded
that this tiny space must be it after all.
3. These numbers refer specifically to admission/rejection in doctoral programs in
research universities with very high research activity. For this and other statistics
about admission rates in master’s and doctoral programs in universities with
additional Carnegie classifications, see Allum, Bell, and Sowell 2012, 5.
4. Other correlates of admission include college curriculum and grade point av-
erage, stated religion, and perceptions of professional competency and interper-
sonal and intrapersonal problems (Campbell 2009; Brear, Dorrian, and Luscri
2008; Attiyeh and Attiyeh 1997; Gartner 1986).
5. For a few of the explanations behind rising graduate degree pursuit, see Collins
2002, Labaree 1997, Pappano 2011, Walker et al. 2008, and Wulff et al. 2004.
194 Notes to Pages 2–6
6. Schmidt 2010; Schofer and Meyer 2005. For detailed trend data about U.S. doc-
torates in the twentieth century, see Thurgood, Golladay, and Hill 2006. Find-
ings from the Council of Graduate Schools’ annual International Graduate
Admissions Survey offer useful information about trends in application, admis-
sion, and enrollment across countries, regions of origin, and fields of study.
For example, Chinese nationals comprised 34 percent of international students
enrolled in American graduate programs in 2014. See J. Allum et al. 2014.
7. Torche 2011.
8. U.S. Department of Education, National Center for Education Statistics 2012.
9. Gonzales, Allum, and Sowell 2013. Note that across racial/ethnic groups,
women comprised the majority attaining doctoral degrees in 2010 if first pro-
fessional doctoral degrees (e.g., MD, JD, DDS) are included in the calculation
(U.S. Department of Education, National Center for Education Statistics 2012).
10. Bell 2010. In 2012, U.S. universities awarded just 102 new PhDs to Native
Americans (Patel 2014).
11. Leslie et al. 2015; Attiyeh and Attiyeh 1997.
12. National Science Foundation 2010.
13. Gonzales et al. 2013.
14. On rising pursuit of the PhD, see Jones 2003, Gibney 2012, and Cyranoski et al.
2011.
15. Nettles and Millett 2006, 101.
16. Schillebeeckx, Maricque, and Lewis 2013;Cassuto 2015. Recent attention in
the popular press to the academic job market in the humanities includes a forum
in the New York Times (“Room for Debate” 2013) and an article by Michael
Bérubé (2013) in the Chronicle of Higher Education.
17. Cassuto 2015; Kezar and Sam 2010.
18. James 1903.
19. To maintain a clear focus, this research excluded from analysis professional doc-
toral degree programs (e.g., MD, JD) because processes of admissions to such
programs are unique in several ways. Professional staff often review applica-
tions, and criteria emphasize experience with and potential for practice. Not-
withstanding the range of careers those with a PhD may pursue, PhD programs
continue to emphasize research training over (or in addition to) training for
professional practice. Different standardized tests are also used for professional
admissions, and scores on those tests are more heavily weighted in admissions
decisions than in most PhD programs.
20. APA-accredited clinical psychology programs—arguably among the most elite
in the discipline—offered admission to only 10.5 percent of the nearly 19,000
applicants in 2002 (Landrum and Clump 2004). Within and across the disci-
plines, admissions selectivity and overall doctoral program enrollment explain
large proportions of reputational scores used to calculate the 1995 National
Research Council rankings (Grunig 1997) and the 2008 U.S. News and World
Report rankings (Sweitzer and Volkwein 2009). The exception to this trend is
in education, where average admissions test scores for graduate programs are
not significantly associated with program reputation (Sweitzer and Volkwein
2009). Nevertheless, Vanderbilt University’s Peabody College of Education, the
Notes to Pages 6–8 195
highest-ranked college of education nationally in 2010, admitted just 6.5 per-

cent of its PhD applicants in 2009 (K. Tanner, personal communication, Au-
gust 24, 2010).
21. Jaschik 2009.
22. Frank and Cook 1995, 3.
23. Collins (2002) examines the implications of credential inflation in the labor
market for the future of higher education. Mitchell Stevens (2007) finds in a
study of undergraduate admissions that quality is also a moving target in the
review process because decision makers often judge applicants not according
to a specific set of standards but relative to the characteristics of others in the
pool.
24. Beneath the debates over these questions are fundamental debates about what
our standards of fairness should be, and whose standards of fairness should
carry the day.
25. For some of the earliest research on graduate admissions, see Cureton, Cureton,
and Bishop 1949; Borg 1963; Newman 1968; Madaus and Walsh 1965; and
Lannholm 1968.
26. Sternberg and Williams 1997, 630.
27. See, for example, Kuncel and Hezlett 2007.
28. Pennock-Roman 1986, 252.
29. Further, the methods most scholars have used do not allow for causal inference
at all; they have been almost exclusively correlational and meta-analytic, with
few controls for confounding variables. Predictive validity research typically
samples admitted students, not the applicant pool, restricting the range of ob-
servations to the right tail and therefore producing attenuated correlation co-
efficients. If samples were to include the outcomes of rejected students who
enroll elsewhere, a broader range of GRE scores would be available for anal-
ysis and GRE scores would be more likely to predict academic outcomes. Fur-
ther, the covariance structure of admissions variables consists of negative
correlations because admissions decision-making is usually a holistic process
in which there are a large number of practical considerations, whose relative
importance varies widely from person to person (Dawes 1971, 180).
30. Kuncel and Hezlett 2007. For two examples of meta-analyses, see Kuncel, He-
zlett, and Ones 2001; and Morrison and Morrison 1995. For examples of
single-discipline predictive validity studies, see Grove, Dutkowsky, and Grodner
2007; Ruscingno 2006; and Wilkerson 2007. Through meta-analysis in several
disciplines, Kuncel, Hezlett, and Ones (2001) concluded that the GRE is a gen-
eralizably valid predictor of passing comprehensive examinations. This finding
holds in a separate study of economics students, although the GRE does not
predict completion of the PhD itself (Grove, Dutkowsky, and Grodner 2007).
In Morrison and Morrison’s 1995 meta-analysis, there is no significant corre-
lation between GRE scores and first-year GPA.
31. See Educational Testing Service, 2014a for ETS guidelines on the appropriate
interpretation of scores.
32. Stevens (2007) found a similar pattern in his ethnographic study of undergrad-
uate admissions in a liberal arts college. As economist Robert Klitgaard put it,
highly ranked universities and graduate programs within them tend to receive
applications from many students who fall in the “right tail of the distribution
of talent” (Klitgaard 1985, 12).
33. A few scholars have noted how environmental characteristics shape the rela-
tionship between academic admissions criteria and student outcomes, such as
Walters, Lee, and Trapani, 2004. Organization-level characteristics, such
as those of the test center (e.g., ethnic match between test proctor and test-
taker), graduate program (e.g., number of same-gender peers in a graduate pro-
gram), or university (e.g., perceptions of climate) may also explain the relationship
between applicant criteria and later outcomes. Regression analyses that fail to
control for organizational-level factors will produce upwardly biased estimates
about the strength of the relationship between an admissions criterion and later
measures of success.
34. For examples of research on higher education that examines the use of hiring
and admission to encourage broader organizational goals, see Birnbaum 1988,
Klitgaard 1985, Killgore 2009, Nivet 2011.
35. Stevens 2007.
36. Lamont 2012.
37. The organizational perspective on merit that I propose draws upon ideas from
the sociology of evaluation and sociocultural research on undergraduate ad-
missions and the professoriate. See, for example, Karabel 2005, Kahlenberg
1996, and Karen 1990. In higher education research, specifically, sociocultural
scholars approach selection and promotion in higher education as contextual-
ized and socially constructed. See Tierney 1988, and Peterson and Spencer 1990,
13. Beginning with Burton Clark (1960) sociocultural paradigms have been
used to study process-based phenomena such as socialization and tenure (Tierney
and Bensimon 1996) and the institutionalization of interdisciplinarity (Lattuca
2001). Stevens 2007, Birnbaum 1988, Twombly 1992, and Lamont 2009 all
bring sociocultural sensibilities to bear on selection in higher education. I also
thank Janet Weiss and Andrew Kinney for their insights into this angle on merit.
38. Cassuto 2015; Guinier 2015.
39. Wechsler 1977, 244. See also Stevens 2007 and Stampnitzky 2006 for institu-
tional accounts about the production of preferences in undergraduate
admissions.
40. A narrow definition of diversity is heterogeneity based on individual character-
istics such ethnicity, race, gender, gender identity, sexual orientation, class, reli-
gion, geography, experience, and viewpoint. Beyond this narrow framing, which
Daryl Smith calls “the laundry list approach,” the pursuit of diversity through
graduate admissions may encompass a wide range of important considerations
about the nature of the knowledge produced, the unique strengths and weak-
nesses that prospective cohort members hold, the representation of historically
marginalized groups in an increasingly multiethnic society, and the place of
justice and opportunity among an organization’s aims (National Center for In-
stitutional Diversity 2011).
41. Attiyeh and Attiyeh 1997; Karabel 2005; Grodsky 2007; Stevens 2007; Lamont
2009. Inequalities in graduate education condition the equity in the labor
market, and indeed, women and people of color are underrepresented in many
corners of academia, especially tenured faculty positions and selective colleges
and universities. For example, women constitute just 21 percent of the popula-
tion of employed philosophers (Crasnow 2007).
42. Attiyeh and Attiyeh 1997; Klitgaard 1985; Grodsky, Warren, and Felts 2008;
Campbell 2009.
43. According to Educational Testing Service (2013a), among U.S. citizens who
took the GRE in 2012, men had higher scores than women on both the verbal
and the quantitative reasoning sections. Including international students, men
had higher scores only on the quantitative section of the test. Among the racial/
ethnic classifications that ETS uses, whites and Asians earned the highest scores
on the verbal and quantitative sections of the test, respectively. There are just
three studies that point to how faculty actually use test scores or other criteria
in graduate admissions decisions. In a 1984 ETS-sponsored study of 333 fac-
ulty members, participants claimed that their most important admissions crite-
rion is grade point average, followed by letters of recommendation, and then
GRE scores. However, as the authors admit, “We are dealing with self-reports
and not observed actions” (Oltman and Hartnett 1984, 7). A decade later,
Landrum, Jeglum, and Cashin (1994) found that GRE scores were the second
most important criterion in psychology admissions outcomes, and Attiyeh and
Attiyeh (1997) found GRE scores the strongest predictor of admission.
44. Moss-Racusin et al. 2012.
45. Milkman, Akinola, and Chugh 2014.
46. That I did not find evidence of discrimination does not mean it does not occur,
but I designed the research as an open-ended inquiry into possible explanations
for inequality rather than as a specific test for prejudicial behavior.
47. See Roithmayr 2014 for a description of the ways in which white privilege is
reproduced through everyday choices, including in college admissions.
48. Tilly 1998.
49. See Bonilla-Silva 2010, 8, on colorblind racism and how contemporary social
scientists may unwittingly reinforce racial hierarchy by describing racial in-
equalities without explaining the processes that create those inequalities.
50. DiMaggio and Powell 1983; Stevens 2007.
51. Findings from research like these cannot be used to directly prescribe policy, but
they can be used to inform a more thoughtful approach to policy formation.
52. Sociologists such as Lyn Spillman call the uncovering of mechanisms and con-
cepts through thick description “theoretical generalization.” See her excellent
2014 paper in Qualitative Sociology for an example.
1. Decision Making as Deliberative Bureaucracy

1. Gutmann and Thompson 2004, 3–7.
2. Ibid. For examples of other settings in which deliberative democracy has rele-
vance, see Fishkin 2009, Hess and Posselt 2002, Ryfe 2002, and Schoem and
Hurtado 2001.
3. Salthouse, McKeachie, and Lin 1978; Twombly 1992; Jackson 2006.
4. Weber 1978b; Wilson 1989.

5. Weber 1978b.
6. Ibid., 974.
7. Morrison 2006, 226.
8. Some of the departments in this study conduct admissions differently for their
MA/MS vs. PhD programs. Given my focus here on doctoral education, I use
the language of “program” rather than “department;” however, some respon-
dents used these words interchangeably.
9. Walker et al. 2008, 22.
10. See Geiger 1986 and Clark 1987 for helpful histories and analyses of academic
departments as organizations, and Gumport 1993 for an excellent treatment
of the central role that sponsored research plays in American graduate
education.
11. Analyses by Twombly (1992) and Salthouse, McKeachie, and Lin (1978) dis-
cuss committee-based selection of students and faculty. Landrum et al. (1994)
claim that “a wide range of studies have examined the decision-making
process within graduate admissions committees” (240), but this claim is quite
a semantic stretch. The studies they cite do not focus on the work of decision
making or consider the process itself, but rather develop prediction models to
approximate the likely outcome of decision-making processes. Landrum et al.
(1994) come closest by asking fifty-five department chairs to describe, in writing,
their department’s decision-making processes and then conducting a content
analysis of the answers.
12. Baseline program demographics surely contributed to the absence of racial/
ethnic diversity on most committees as well.
13. Wilson 1989.
14. Meyer and Rowan 1977; March, Simon, and Guetzkow 1958.
15. In the other two programs, the department chair was a voting member of the
admissions committee.
16. Through follow-up interviews, I gathered that the faculty and administrative
committee members’ biggest concern with this episode was not about the charge
of cherry picking, which they dismissed as unfounded to begin with, but with
Will’s inserting himself into “their” process. They felt his involvement violated the
bureaucratic procedures that lend legitimacy and efficiency to their work. Strong
feelings motivated Will’s challenge, and they also provoked strong feelings among
other participants. Participants saw the challenges and the associated feelings as
corrupting influences on an otherwise orderly and objective process.
17. Lemann 2000, 18.
18. Stevens 2007; Espeland and Stevens 2009; Porter 1996.
19. Educational Testing Service (ETS), which administers the GRE, recommends
against this practice.
20. Espeland and Stevens (1998) define commensuration as “the transformation
of different qualities into a common metric” (314). If two things are incom-
mensurable, it implies they are sufficiently distinctive that they should not be
considered in relation to the other.
21. In linguistics, the only program that does not assign numerical ratings, indi-
vidual faculty members commensurate their judgment into a yes/no vote for
each applicant.
22. Porter 1996, 6.
23. Judgment is quantified in other ways as well, but not as consistently across pro-
grams. For example, in biology the faculty developed a system for rating along
several dimensions the applicants whom they interview. Summary scores from
those evaluations are averaged and tabulated to develop a rank-ordered list that
drives the final round of decision making.
24. Only in political science and physics were faculty formally asked to provide
written comments to justify their ratings.
25. I thank Michael Bastedo for this insight.
26. Indeed, their discussion did not only reveal differences in interpretation about
what a rating of 1, 2, or 3 means. It turned out that one committee member
had used the three-point scale as three categories and others on the committee
had regarded the numbers as ordinals.
27. Griffin and Muñiz 2011.
28. Heimer 2001.
29. Ibid.; Stevens 2007.
30. Porter 1996, 5.
31. Klitgaard 1985, 13.
32. Fisher v. University of Texas, 570 U.S. 11-345 (2013).
33. Gartner 1986.
34. Wilson 1989.
35. Thucydides (trans. Crawley) 1951, 105–106.
36. Birnbaum 1988; Twombly 1992.
2. Meanings of Merit and Diversity

1. By the standards of disparate impact law, “facially neutral” hiring practices
(those that appear to be race-neutral) that have adverse effects on opportuni-
ties for specific populations may be judged discriminatory. Disparate impact
has been ruled constitutional only when “business necessity” motivates the dis-
criminatory action. In Smith v. City of Jackson, police officers over forty years
old brought suit against the city for granting larger raises to junior officers. Be-
cause the city offered the raises to achieve parity with “comparable positions
in the market,” the basis for age discrimination was ruled a matter of business
necessity, and therefore constitutional in spite of the negative, disparate impact
it had on more senior officers (Smith v. City of Jackson [03–1160] 544 U.S.
228, 2005).
2. Fisher v. University of Texas, 570 U.S. 11-345 (2013).
3. Association of American Colleges and Universities 2014.
4. Guinier 2015.
5. Milem, Chang, and Antonio 2005, v.
6. Lamont 2009; Goffman 1959.
7. Lamont 2009. Note that evaluative scripts are distinct from interaction scripts
qua Barley: “outlines of recurrent patterns of interaction that define in observ-
able and behavioral terms, the essence of actors’ roles” (Barley 1986, 83).
8. Birnbaum 1988, 498.
9. Birnbaum 1988; Brink 1999; Klitgaard 1985; Twombly 1992.
10. Chapter 3 examines how these trends play out in specific disciplines.
11. March 1994; Karen 1990.
12. Espeland and Sauder 2007; Bourdieu 1984; Attiyeh and Attiyeh 1997;
Hersch 2014.
13. When I reviewed my findings with the chair of economics (i.e., what qualita-
tive research methodologists call member-checking), he mentioned that he now
pulls applications from black and Latino students to ensure a diverse pool in
later rounds of review. He reported that they have relaxed the GRE score
threshold since the test’s 2011 revision, which he believes “nicely distinguishes
among highly qualified people.”
14. Klitgaard 1985, 32.
15. Their devaluing of unknown institutions is consistent with previous research
on status judgments of organizations, such as that by Bitekine (2011).
16. Zimmerman and Zeitz 2002, 416.
17. Bowen, Rudenstine, and Sosa’s (1992, 11) study of doctoral education in ten
institutions found that graduate programs with small cohorts had higher com-
pletion rates and shorter time to degree that those with larger entering cohorts.
18. March 1994, 38. Just two programs among the ten in this study had been col-
lecting data on applicant characteristics in relation to students’ academic and
professional outcomes. Neither used the data to inform current admissions
practice.
19. ETS does not claim that the GRE measures intelligence. For guidance on inter-
preting scores on tests taken after August 2011, see ETS 2014a.
20. Bourdieu 1988; Lamont 2009.
21. Mitchell Stevens (2007) called this phase “coarse sorting” in his study of un-
dergraduate admissions.
22. Although my study was not designed to assess alternative approaches to using
the GRE, there are a number of possibilities, including interpreting scores in
light of known disparities, waiting to consider GRE scores until the rest of the
application has been reviewed, or cutting the GRE from their considerations
entirely.
23. Walker et al. 2008, 11.
24. For a thoughtful discussion of drawbacks to a “laundry list” approach to di-
versity in higher education, see Smith 2009.
25. For a few of the many studies on the benefits of educational diversity for un-
dergraduates, see, for example, Bowen and Bok 1998; Chang, Astin, and Kim
2004; Gurin et al. 2002; and Jayakumar 2008. Some studies have found that
educational experiences with diversity can strengthen professional preparation
and research innovation; these include Page 2008 and Harvey and Allard 2012.
26. For studies that examine how colleges and universities strive to present them-
selves as diverse, see Osei-Kofi, Torres, and Lui 2013; regarding admissions
viewbooks, see Stevens and Roksa 2011; regarding mission statements, see
Morphew and Hartley 2006. Stevens and Roksa argue for a “metrical” under-
standing of diversity as an indicator of organizational excellence in higher ed-
ucation today.
27. Two respondents in philosophy noted that women have better representation
in ethics and the history of philosophy than they do in other subdisciplines.
28. In an interesting example of how one’s own identity may frame conceptualiza-
tions of diversity, I found that the majority of people who counted national
origin among salient dimensions of diversity had, themselves, been born out-
side of the United States. Similarly, individuals in two programs who had grown
up in Appalachian states defended the importance of regional diversity.
29. Garces 2012.
30. Smith et al. 1996.
31. Beutel and Nelson 2005; Kulis, Chong, and Shaw 1999; Olivas 1994; Smith
et al. 1996; Trower and Chait 2002.
32. Smith 2009, 149.
33. Haro 1995, 196; Danowitz Sagaria 2002, 689.
34. Garces 2014; Marichal 2009; Smith, 2009.
35. Brief for Columbia University, Harvard University, Stanford University, and the
University of Pennsylvania as Amici Curiae at 3, Regents of the University of
California v. Bakke, 438 U.S. 265 (1978) (No. 76–811).
36. Smith 2009; Garces 2014.
37. It was unclear from these comments whether the faculty assumed that the
courses the students had taken signaled their “intellectual capacity,” whether
some committee members were concerned about preparation and others about
capacity, or whether Peter was just uncomfortable describing their rationale for
eliminating these two individuals and so spoke in apparent contradictions.
38. Kahneman 2011.
39. Bastedo and Jaquette 2011; Bielby et al. 2014; Posselt et al. 2012; Rojstaczer
and Healy 2012.
40. For data on group differences in GRE scores and ETS recommendations for
the appropriate use of GRE scores in the review process, see ETS 2013a
and 2014a; Fischer, Schult, and Hell 2013; and Miller 2013.
41. See, for example, Fischer, Schult, and Bell 2013; Young and Kobrin 2001. In the
most recent study of differential predictive validity on college admissions exams
by gender, meta-analysis by Fischer, Schult, and Bell (2013) found that the
problem of performance underprediction is worse for undergraduate admissions
tests than for the GRE. Like all meta-analyses, however, the findings are only
as generalizable as the data from which the studies’ findings are drawn. Often in
the research on graduate admissions, studies disproportionately examine stu-
dent scores in psychology, which may not generalize to other disciplines.
42. ETS, 2014a.
43. Berrey 2015, 3.
44. Dowd and Bensimon 2014, 58.
45. Stevens (2007) noted that diversity has become an informal “index of prestige”
(182) among selective liberal arts colleges.
46. Park (2013) highlighted the power of intentionality for meaningful diversity in
her analysis of the racial reconciliation efforts of a religious student organiza-
tion at a California university affected by Proposition 209, the state’s ban on
affirmative action.
3. Disciplinary Logics
1. Walker et al. 2008, xi.
2. The findings of the Carnegie Initiative on the Doctorate, described in full in
Walker et al., The Formation of Scholars (2008), borrowed imagery and ideas
from the training of clergy to promote an ideal of doctoral education that in-
cludes developing “stewards” of the discipline. See also Abbott 2001, Austin 2002,
Thelin 2004, and Wulff et al. 2004 on doctoral education as credentialing and
socialization for the professoriate.
3. Becher 1981; Clark 1987; Hermanowicz 2005.
4. Johnson (2008) discussed increasing divergence in the approaches to scholar-
ship used by linguists.
5. It would have been convenient to defer to the disciplinary classification schemes
developed through Hagstrom’s (1964) or Biglan’s (1973) analyses of disciplinary
consensus or that of Lodahl and Gordon (1972) based on paradigm develop-
ment. However, their classifications are forty or more years old, and disciplinary
changes during this time period make adoption of previous scholars’ assessments
of paradigm strength a spurious task (Pfeffer 1993). Therefore, I adopted a his-
torical approach to categorizing each discipline as high-, moderate-, or low- con-
sensus relative to others within their respective subject area of humanities, social
sciences, or natural sciences. To do so, I used (a) surveys of the history of the
disciplines from secondary sources, (b) interviews with emeritus members of
each department, and (c) member-checking with admissions committee chairs.
Given what we know about the dynamics of change in disciplines, these clas-
sifications should be interpreted, not as static, enduring, or absolute, but rather
as reflections of assessment in the context of the fields’ historical trajectories.
6. Ioannides and Nielsen 2007, 1; Wolfers 2015.
7. For brief overviews of economics as a discipline, see Becker 1978, Skousen
2009, and Harvey and Garnett 2006.
8. March 1994.
9. Under the GRE’s previous scoring rubric, the mean quantitative reasoning score
of students enrolled in the top six economics programs in 2002 was 785,
and 765 in the programs ranked 15–30 (Stock, Finnegan, and Siegfried 2006).
10. Russell 1972; Durant 2012.
11. Biglan 1973.
12. Durant 2012.
13. In classics, the department chair said that ratings were a “guide to the feelings
that people have about candidates,” but this was not established as the norm
for the entire committee.
14. Knorr-Cetina 1999; Lamont 2009.
15. See, for example, Clark 1987 and Hermanowicz 2005.
16. Pfeffer 1993.

17. These findings could also be due to less consensus motivating more delibera-
tion and, with it, challenges to be explicit about the basis for their ratings.
18. Snow 1959.
19. Indeed, if Abbott’s (2001) argument about the fractal nature of knowledge
within disciplines is true, then even high-consensus fields have considerable in-
tellectual diversity.
20. For discussion and analysis of relationship between disciplinary cultures and
evaluation, see Becher and Trowler 1989, Knorr-Cetina 1999, Lamont 2009,
and Swales 2004.
21. The study in question (Sternberg and Williams 1997) analyzed GRE scores in
relation to a number of academic outcomes and other admissions criteria for
167 current and recently graduated psychology students at Yale University. They
found that the median correlation coefficient of GRE scores in four areas (Verbal,
Quantitative, Analytical, and Subject) in predicting first-year grades was 0.17,
with the Subject test (not the Quantitative reasoning) bearing the strongest
relationship. GRE scores only weakly predicted “other aspects of graduate perfor-
mance: ratings of analytical, creative, practical, research, and teaching abilities
by primary advisers and ratings of dissertation quality by faculty readers.” The
authors noted “substantial” standard deviations and score ranges of 550, 520,
400, and 360 points (Sternberg and Williams 1997, 636–637).
22. See the conclusion section of Chapter 6 for a more complete discussion of ste-
reotype threat research in relation to graduate admissions tests.
23. For one example of Barack Obama’s personal characterization of his “improb-
able journey,” see the transcript to his July 24, 2008 speech in Berlin, Germany,
available at http://edition.cnn.com/2008/POLITICS/07/24/obama.words/.
4. Mirror, Mirror
1. McPherson, Smith-Lovin, and Cook 2001, 416.
2. Kanter 1977; Karabel 2005; Rivera 2012b.
3. DiMaggio 1992.
4. This tendency would be defensible if faculty were selecting students to be ap-
prentices of their own, since it would be hard for a professor to advise a stu-
dent in areas of research far afield from one’s own. However, in the committees
I observed, the committee members’ task was to select a cohort of students for
the department as a whole.
5. I began to notice it in the course of analyzing how disciplinary context shapes
personal preferences, and how personal preferences affect admissions outcomes.
I noted that preferences for self-similarity seemed especially prevalent in low-
consensus disciplines, the very fields in which personal tastes often played an
especially important role in evaluation, generally. Realizing that individual level
preferences for particular types of students could not, in many cases, be divorced
from reviewer backgrounds, the four types of homophily that I observed most
frequently were by and for pedigreed, cool, socially mobile, and international
scholars. This chapter focuses on the first three.
6. This finding could be due to the absence of individuals from these backgrounds
in high-consensus disciplines (i.e., a selection effect) or the absence of a common
disciplinary logic, making personal tastes more salient (i.e., a disciplinary
effect).
7. Rivera 2011b.
8. What I observed is a perfect example of Pierre Bourdieu’s description of elite
reproduction through education. He noted that selection processes in elite
French educational institutions treat elite social connections and cultural knowl-
edge as forms of capital on par with financial resources. Social and cultural
capital operate as arbitrary, and ultimately exclusionary, bases for selection that
reproduce existing structures of privilege.
9. Given the small range in grades typically assigned for graduate courses, small
increments carry greater meaning—both to the faculty to assign the grades and
to the students who receive them.
10. Bourdieu 1984; Khan 2011; Lamont and Lareau 1988.
11. Patton 2013.
12. Professors may not be cool by society’s standards, but signature frames may be
a way some scholars try to express style while owning their intellectual status.
13. Lamont, 2009.
14. In this case, I do not mention disciplines to protect anonymity.
15. “Death match” can be thought of as the opposite of an undergraduate admis-
sions model of “building” (Bastedo 2012), in which reviewers assess and col-
lect applicants’ strengths as grounds for moving an individual forward in the
process.
16. Kanter 1977, 965.
17. McPherson et al. 2001.
18. Rivera 2012b.
19. Rivera 2011b.
20. Correll, Benard, and Paik 2007; Benard and Correll 2010. Benard and Correll
argue that the judgments of likeability have to do as much with the importance
of likeability on the job as the perception that professionally successful women
have betrayed their proper gender role. The perceptions of mothers that Cor-
rell and colleagues have found can skew how opportunities and rewards are
allocated. They also risk pigeonholing women—or other groups about whom
there are strong moral overtones to their social roles—into persistently second-
class status by nudging them to self-select out of careers and fields where their
success may be negatively judged.
21. Sellers et al. 1998.
22. Martin 1992, 9–10. Such complexity might help explain trends that statistical
research has represented as either weak or broad preferences (i.e., what soci-
ologists have dubbed cultural omnivorism).
5. The Search for Intelligent Life

1. Crocker, Sommers, and Luhtanen 2002.
2. On the self as defined in relation to others, see Oyserman, Elmore, and Smith
2012; Tajfel 2010; and Zerubavel 1993. On the centrality of intelligence to
academics’ self-concept, see Bourdieu 1988; Crocker, Sommers, and Luhtanen

2002; and Lamont 2009.
3. Robert Sternberg, one of the most highly published scholars of intelligence in
his generation, argued that “how we should conceive of intelligence” was his
research community’s most fundamental matter of debate (Sternberg 2003, 4).
4. For discussions of socially constructed ideas of intelligence, see Hatt 2012,
Lamont 1992, and Oakes et al. 2008. Yang and Sternberg 1997 conducted the
cited study of ideas about intelligence in Taiwan, and Azuma and Kashiwagi
1987 published the study of Japanese college students’ and their parents’ ideas
of intelligence. Furnham 2001 offers a review of differences in appraisals of
intelligence across gender and national context.
5. Lamont 2009; Becher and Trowler 1989, 102.
6. Research by Rebecca Zwick (2002) finds that simple practice is the primary
mechanism by which standardized test scores increase, and that official test-
prep courses contribute less to score changes than is commonly assumed (as
illustrated through the quote here).
7. It was within a sociohistorical context of racial hierarchy and striving for mer-
itocratic college admissions that, after World War I, eugenicist Carl Campbell
Brigham and colleagues tweaked the Army IQ test into the SAT. So initially
entwined were the two tests’ underlying factors that Brigham developed an
algorithm for translating SAT into IQ scores (Lemann 2000). By the time he
actively promoted the SAT to colleges and universities, however, he had come
to distrust the notion that IQ tests measured an innate trait, and eventually
he renounced his landmark study, A Study of American Intelligence. In 1934
he wrote,
The test movement came to this country some twenty-five or thirty years ago ac-
companied by one of the most glorious fallacies in the history of science, namely,
that the test measured “native intelligence” purely and simply without regard to
training or schooling. I hope nobody believes that now. The test scores very defi-
nitely are a composite including schooling, family background, familiarity with
English, and everything else, relevant and irrelevant. The “native intelligence”
hypothesis is dead (quoted in Lemann 2000, 34, emphasis in original).
8. ETS 2014a.
9. Zwick 2002. The ACT’s core areas are English, Science, Math, and Reading,
for example, and the MCAT focuses on knowledge of Chemistry, Physics, and
Biology.
10. Ibid., 32–33.
11. Neumann 2009, 54–55.
12. Ibid., 54–55.
13. Comments from Bruce Cole, former chair of the National Endowment for the
Humanities, suggest that writing quality and cognitive clarity are inseparable
in the humanities. He said, “Writing is thinking. To write well is to think clearly”
(National Endowment for the Humanities 2002).
14. Gumport 1993, 226.
15. Walker et al. 2008.
16. My thanks to Elizabeth Knoll for this insight.
17. Leslie, Cimpian, Meyer, and Freeland 2015.
18. Mindful of higher education’s democratic mission, Guinier (2003) urges ad-
missions policymakers and decision makers to ask themselves whether test scores,
intelligence, or first-year grades are robust conceptions of excellence for a
democratic society.
19. For details of the revised general test, see ETS 2014b.
20. Goffman, 1959.
21. Wendler and Bridgeman 2014. ETS has released a compendium of research
about the GRE that will be worthwhile for readers interested in learning
more about the details of the test’s validity. Most of the research was con-
ducted before implementation of the revised GRE General Test.
22. Steele and Aronson 1995. For an excellent narrative about the results of ste-
reotype threat research and the development of this line of scholarship, also
see Claude Steele’s accessible 2010 volume, Whistling Vivaldi: How Stereo-
types Affect Us and What We Can Do about It.
23. See Logel et al. 2012 pp. 46–49 for additional recommendations about making
college learning environments stereotype-safe. The 2008 field experiment by
Catherine Good, Joshua Aronson, and Jayne Ann Harder revealed that stereo-
type threat can shape outcomes in a common STEM gatekeeping course, but that
highlighting a test’s fairness could eliminate performance gaps. They manipu-
lated the instructions that students in a college calculus class received on a test
with items similar to those on the GRE. To half of the class, the researchers
described the test as diagnostic of mathematical ability; instructions to the
other half stated, “This mathematics test has not shown any gender differences
in performance or mathematics ability.” The researchers found no significant
gender differences in either group’s final course grades and or in the test scores
of students in the stereotype-inducing condition. However, women’s perfor-
mance in the stereotype-safe condition exceeded that of men’s, implying that
stereotype-neutralizing test instructions make a difference (Good, Aronson,
and Harder 2008).
24. Gregory Walton and Geoffery Cohen’s 2002 meta-analysis estimated that ste-
reotype lift gives white males a bump of about 50 points on the SAT (Walton
and Cohen 2003, 456).
25. Like the call to preserve democratic deliberation, this recommendation may
seem to some a stumbling block on the already rocky path to decision making.
But, hopefully, it will compel those with gatekeeping roles to growing in the
definition of intelligence that literary critic Barbara Christian (1987) articu-
lated: “a tuned sensitivity to that which is alive and therefore cannot be known
until it is known.”
6. International Students and Ambiguities of Holistic Review

1. I am indebted to Kimberly Reyes and Kelly Slay in equal measure for their sup-
port with the data analyses and review of literature that brought about this
chapter.
2. For discussion of the trends that contribute to rising enrollment of international
students, see, for example, Schofer and Meyer 2005 and Schmidt 2010. For the
share of adults on the top end of the wealth distribution with graduate degrees,
see Thompson 2013.
3. Applications from Chinese students dipped 3 percent in 2012 and 1 percent in
2013, after increasing by 21 percent in 2010 and 19 percent in 2011 (Allum
et al. 2014, 7).
4. For the proportion of the adult population, by gender, with graduate degrees,
see Torche 2011. See Gonzales, Allum, and Sowell 2013 for statistics on inter-
national students in U.S. graduate education. Tracing the pathways from
students’ countries of origin to U.S. doctoral programs, as well as growth of
foreign enrollments in U.S. doctoral programs and trends by educational devel-
opment in the country of origin, Blanchard, Bound, and Turner (2009) found that
students from countries with well-developed higher education systems typically
come to the United States for doctoral degrees only when highly ranked programs
admit them, in contrast to countries with fewer doctoral programs, which send
students to lower-ranked U.S. doctoral programs as well. Their findings sup-
port a broader pattern of international students’ strong preference for highly
ranked universities. Fischer 2013 cites a study by IDP Education, a firm that
facilitates study abroad opportunities, that found that no fewer than 85 percent
of international students reported prestige to be an important consideration in
their college choice process, and one-third said it was their top priority.
5. On trade-offs among enrollment management goals, see Cheslock and Kroc
2012. For the use of admissions and student body composition to signal pri-
orities to stakeholders, see, for example Karen 1990 and Karabel 2005.
6. For rates of growth in first-time graduate school enrollment, see Gonzales,
Allum, and Sowell 2013. For national economic benefits of international
student enrollment, see National Foundation for American Policy (NFAP) 2013,
Roach 2013, and Institute of International Education 2013.
7. See David 2012, Jaquette and Curs 2015, Rizzo and Ehrenberg 2004, for the
role of nonresident student tuition in offsetting state appropriations. See Redden
2013 for the role of international students in sustaining STEM faculty research.
See NFAP 2013 for the role of international students in filling gaps.
8. Olivas 2009.
9. Allum et al. 2014; Allum, Bell, and Sowell 2012.
10. Interestingly, they worried less about their ability to judge across American ap-
plicants’ diverse institutional, geographic, and personal contexts.
11. In 1961, economist Daniel Ellsberg classically defined ambiguity as a type of
uncertainty present “in situations where available information is scant or obvi-
ously unreliable or highly conflicting; or where expressed expectations of dif-
ferent individuals vary widely; or where expressed confidence in estimates
tends to be low” (Ellsberg 1961, 660–661).
12. See Epley and Krueger 2005 for an example of research into how ambiguities
create conditions for stereotyping.
13. Gonzales, Remington, and Allum 2013.
14. For the quotation about the Civil Service Exam, see Suen and Yu 2006, 48. For
more about China’s test preparation apparatus, see Golden 2011, Magnier
2006, Özturgut 2011, Xueqin 2011, and Zhao 2001.
15. Larmer and Lin-Liu 2002; Montlake 2006; Mooney 2005.

16. Chang 2001; Xu 2001; Zwick 2002.
17. For more on “smile work,” see Tierney and Bensimon 1996, 83.
18. Epley and Kruger 2005.
19. Fish 2013.
20. An exception to this seems to be French philosophers, who, in the words of
one participant in philosophy, are “very explicit about how they’re ranking
{laughs} the various people.”
21. Epley and Krueger 2005, 419.
22. In this context, “Asian” refers to both Asian immigrant and Asian international
populations.
23. According to the United States Census, more than 30 ethnic groups are repre-
sented in the Asian American category alone (Iwamoto and Liu 2010).
24. Chou 2008; Lee 1994; Osajima 2005.
25. Taylor and Stern 1997; Suzuki 2002; Zhang 2010.
26. Maddux et al. 2008, 86.
27. These stereotypes have implications for higher education research, as well as
for the educational policies and practices that are informed by that research
(Museus 2008; Museus and Kiang 2009; Nadal et al. 2010; Ng, Lee, and Pak
2007; Teranishi et al. 2009). In their critical review of educational literature on
Asian Americans, for example, Ng, Lee, and Pak (2007) argue that even
existing higher education scholarship is tainted by the model minority stereo-
type. The literature not only represents Asian American faculty as hardworking,
nonconfrontational individuals incapable of departmental leadership, it also
overlooks Asian American students’ multiple social identities and their rele-
vance to student support service needs and campus climate concerns (Ng, Lee,
and Pak 2007).
28. Beoku-Betts 2004; Iwamoto and Liu 2010.
29. Lee and Rice 2007; Hanassab 2006.
30. Clearly detrimental, these experiences are expressed in higher reports of dis-
comfort, stress, and homesickness; lower self-esteem; self-segregation from do-
mestic students; and suppression of feelings of depression (Poyrazli and Lopez
2007; Schmitt, Spears, and Branscombe 2003; Trice 2004; Wadsworth, Hecht,
and Jung 2008; Wei et al. 2008).
31. For studies on the TOEFL’s validity, see Fu 2012; Wait and Gressel 2009; Light,
Xu, and Mossop 1987; Johnson 1988; Yule and Hoffman 1990.
32. ETS 2013a.
33. Fu 2012, 105.
34. There is a growing body of creative scholarship on administrative pressure and
K–12 teachers’ efforts to game or cheat high-stakes standardized tests. See de-
veloping work by Hibel and Penn, as well as by Maloney, for example.
35. Chang 2001; Xu 2001.
36. Melcher 2010, 3.
Conclusion
1. See Clauset, Arbesman, and Larremore 2015; Fowler, Grofman, and Masuoka
2007.
2. See Rivera 2011b and 2012a on professional investment firms’ consideration
of elite college status for entry-level hiring. It’s important to note that sorting
out which colleges deserve “elite” status has become a national preoccupation,
and simply gaining admission to the ones so deemed is now a veritable status
symbol for both the young adults who attend them and their families. A growing
share of families organize their children’s time and very lives to put them on a
trajectory toward one of these selective institutions. They do so because, to
many, admission to elite colleges not only symbolizes status achieved. It also
represents a gateway to status and a hope for inoculation from the financial
stresses that plague so many Americans today. Graduate school is among the
subsequent opportunities that may follow from attending an elite college, and
attending graduate school is on the minds of young people from as early as
tenth grade. A recent study I conducted with Matthew Holsapple found that U.S.
tenth-graders who aspired to a graduate degree or who thought a college’s grad-
uate school placement rate was “very important” had higher odds of enrolling
at one of U.S. News and World Report’s top 100 colleges and universities (Hol-
sapple and Posselt 2010).
3. For racial inequalities in selective college enrollment from 1972 to 2004, see
Posselt et al. 2012, p. 1091. For percentage of women, men, underrepresented
minority, and other ethnic groups earning scores of 700 or above, see Miller
and Stassun 2014.
4. For an excellent analysis of trade-offs and opportunity costs in admissions, see
Espenshade and Chung 2005. An important and related point is this: When a
program fixes the number of students they will admit, they create a zero-sum
game. Trade-offs for access and opportunity are inherent to these kinds of com-
petitions, so the question is not whether there will be trade-offs, but what they
will be, how they will be determined, and what their implications are. As we
saw in affirmative action lawsuits by Abigail Fisher and Jennifer Gratz, con-
troversy can erupt when a student denied admission assumes that she would
have been admitted next if some preference or another had not been in play.
Like other external sources of student financial support, institutional diversity
fellowships for graduate students may therefore be most powerful as an incen-
tive for increasing diversity when they increase program capacity, because it
changes the very nature of the admissions competition.
5. Boltanski and Thévenot 2000.
6. Bourdieu 1984; Bourdieu 1988.
7. For details of the methodology behind the National Research Council’s as-
sessment of research doctoral programs, see http://www.nap.edu/catalog
.php?record_id=12676#toc. For details of the U.S. News Best Graduate
School Rankings, see http://www.usnews.com/education/best-graduate-schools
/articles/2014/03/10/how-us-news-calculated-the-2015-best-graduate-schools
-rankings.
8. Research context (French postsecondary vs. U.S. doctoral education) could also

explain the insufficiency of Bourdieu’s theory to capture the trends I observed.
In the United States, cultural boundaries are fuzzier (Lamont 1992), elite tastes
are more omnivorous (Bryson 1996), and institutionally valued cultural objects
are defined by intersections of race, class, and gender rather than class alone
(Yosso 2006).
9. Lamont 1992, 178.
10. Scott 2008.
11. Smith 2009; Stevens and Roksa 2011.
12. Regents of the University of California v. Bakke, 438 U.S. 265 (1978); Gratz v.
Bollinger, 539 U.S. 244 (2003).
13. Narrow tailoring and compelling state interests are the two primary hurdles
required to pass the legal test of strict scrutiny.
14. See Garces 2014 for an excellent history of legal developments shaping affir-
mative action in graduate admissions. Garces notes that introducing the test of
strict scrutiny was a turning point in this history because it “equated efforts to
promote access to education for racial minorities with discriminatory practice
against whites” (460).
15. Ibid.
16. Civil Rights Project 2002.
17. Tienda 2013; Guinier 2003; Klitgaard 1985; Gurin et al. 2002.
18. Two studies have found that diversity on research teams has advantages for
creativity and problem solving (Committee on Equal Opportunities in Science
and Engineering 2013; Page 2008). Two studies that have found diverse teams’
papers are cited more frequently (Campbell et al. 2013; Freeman and Huang
2014). A limitation of the citation analysis studies is that they rely upon the
researchers’ judgments of author gender and ethnicity from names alone, rather
than on authors’ self-reported gender and ethnicity.
19. Ahmed 2012; Berrey 2015; Dowd and Bensimon 2014.
20. March 1994.
21. Hagedorn and Nora 1996, 31.
22. March 1994.
23. Research using experimental, survey, and qualitative methodologies finds sig-
nificant costs to the quality of intergroup relations and to students of color who
experience racial isolation in the classroom. For example, Deo (2011) writes
that graduate students of color are less likely to contribute to discussion in
classes in which they feel tokenized, compromising their learning experience
and the spirit of the broader group’s discussion.
24. Some of this may have to do with a skewed knowledge base. Faculty were more
aware of racial and gender gaps on standardized tests and other educational
outcomes than they were of the research evidence that professors’ own actions
and learning environments contribute to student outcomes or that social di-
versity supports excellence in research.
25. See Bonilla-Silva 2006 and Carter and Tuitt 2013 for excellent discussions of
colorblind racism and postracial discourse in U.S. education and society.
26. Worthington et al. 2008; Garces 2014.
27. Neglecting context is what Kahneman (2011) calls narrow framing, and it
causes reviewers to misread the justifiable reasons some applicants might be
weaker or stronger than expected on a desired criterion.
28. An extreme version of structured review is “policy capturing,” which a small
group of scholars recommended in the 1970s (Dawes 1971). Using the charac-
teristics of recent cohorts to reveal a department’s preferences, policy capturing
would replace human judgment in admissions with regression models that mea-
sure an applicant’s admissibility.
29. Lundy-Wagner, Vultaggio, and Gasman 2013.
30. See Fisk-Vanderbilt (2014) for a full description of the Fisk-Vanderbilt Masters-
to-PhD Bridge Program. A similar bridge program links Spelman and Georgia
Tech. Others using the bridge label offer the master’s degree at a research uni-
versity as a bridge to doctoral program admission, but do not recruit solely
from an MSI.
31. Eagan et al. 2013.
32. Nagda et al. 1998; Posselt and Black 2012. Through the National Science Foun-
dation’s Research Experiences for Undergraduates program and the Department
of Education’s Ronald E. McNair Postbaccalaureate Achievement Program,
the federal government has financially invested in student research as a means
of diversifying graduate education and the professoriate. There are institution-
specific programs, such as the successful Meyerhoff Scholars Program at the
University of Maryland–Baltimore County, as well as multi-university efforts
such as the Summer Research Opportunities Program, administered through a
cooperative effort of Big Ten universities and the University of Chicago. The
mission of programs like these is to develop professional interest in and com-
mitment to research among underrepresented populations, but their partici-
pants may also be attractive recruits for graduate programs.
33. Association of American Colleges and Universities 2014.
34. Hersch 2014. See Rogers and Molina 2006 for research into the qualities of
psychology doctoral programs that have been especially successful in recruiting
and retaining students of color.
35. Rather than assume risk associated with metrics like college grades, TOEFL,
or GRE scores, some leaders may want to start by measuring the probability
of specific outcomes for students in their program relative to specific student
characteristics. To do this well, though, there are a few measurement issues to
take into account. For programs that have historically enrolled few people from
a specific group (e.g., women, international students with lower TOEFL scores,
students from less selective colleges), the validity of analyses may be hampered by
small sample sizes. Also, measures concentrated in a small range, like enrolled
students’ college GPA, may produce downwardly biased estimates of correla-
tions with the outcome (researchers call this attenuation bias). Finally, because
student success is the product of a complex interplay of student inputs and the
learning environments faculty create, such analyses would need to account not
only for characteristics of students, but also for characteristics of the environ-
ment that shape student performance (e.g., perceptions of climate and/or ste-
reotypes, proportion of same-identity individuals in one’s cohort). Hierarchical
linear modeling is well suited to analyses of this sort. See Walters, Lee, and
Trapani 2004 for a fine example.
36. Among the noncognitive characteristics that scholars have proposed for grad-
uate admissions, Sedlacek (2004) has developed measures for the following: Pos-
itive self-concept, Realistic self-appraisal, Successfully handling the system,
Preference for long-term goals, Availability of a strong support person, Leader-
ship experience, Community involvement, Knowledge acquired in a field. The
ETS Personal Potential Index provides opportunity for those writing letters of
recommendation to assess the following qualities of applicants: Knowledge,
Creativity, Communication skills, Teamwork, Resilience, Planning and organi-
zation, Ethics and integrity (Kyllonen 2008).
37. Powell 2013; Miller and Stassun 2014.
38. ETS 2014a.
39. For analyses of white and African American students’ scores in relation to back-
ground characteristics, ethnic match with test proctors, and characteristics of
the GRE test center, see Walters, Lee, and Trapani 2004. An interesting finding
of this study, one that the authors argue contradicts a trend toward stereotype
threat, is that a match between student ethnicity and proctor ethnicity is sig-
nificantly associated with test scores for white students, but not for African
American or Hispanic students.
40. ETS (2013b) promoted the most recent GRE overhaul as a “friendlier, more flex-
ible test-taking experience.” Among other changes, they have reduced the amount
of rote vocabulary knowledge required for high scores on the Verbal section.
41. For the list of colleges and universities that have test-optional policies, see Na-
tional Center for Fair and Open Testing 2014. For a list of PhD programs that
do not require the GRE see http://ainsleydiduca.com/phd-programs-no-gre
-required/.
42. For findings about the effects of going test-optional at the undergraduate level,
see Hiss and Neupane 2004; Belasco, Rosinger, and Hearn 2015.
43. About test-optional undergraduate admissions, Belasco et al. (2015) wrote, “Ar-
guably, institutions that fail to reach a majority of underrepresented students
through recruitment or other outreach initiatives, will find it difficult to im-
prove diversity in meaningful and significant ways, regardless of their admis-
sions criteria” (219).
44. Weick 1984, 40. There are physiological explanations for freezing up in the face
of significant challenges, such as overarousal in the limbic system, with observ-
able physiological effects such as an inability to concentrate.
45. For doctoral programs that choose to require GRE scores, for example, another
small win would be more nuanced classification of “high” and “low” scores. If,
when reviewers sit down with files, they are provided with some discipline- or
institution-specific data about the current distributions of GRE scores by na-
tional origin, race/ethnicity, and gender, they might interpret scores differently.
One of the three graduate schools with which I worked already shares similar
information in their admissions workshop for faculty, providing institutional
data about the wide racial disparities in the number of students who score
above 700 on each section of the test.
46. Weick 1984, 41.

47. Tilly 2005; Rojas 2007.
48. Berger and Luckmann 1966.
Methodological Appendix
1. Stevens 2007.
2. Goffman 1959; Garfinkel 1967; Lamont 2009.
3. Vaughan 1992; Eisenhardt 1989.
4. Merriam 2009, 39.
5. Stake 2005, 448.
6. Geertz 1973.
7. Charmaz, Denzin, and Lincoln 2000; Stake 2005; Yin 1994; Cronbach 1975;
Geertz 1973.
8. Abbott 1992; Miles and Huberman 1994; Merriam 2009; Smith 1978.
9. Dion 1998.
10. Merriam 2009.
11. Patton 2002.
12. National Research Council 2010; “Best Graduate School Rankings” 2010.
13. Merriam 2009; Yin 1994.
14. Merriam 2009.
15. Gold 1958.
16. Weiss 1995; Rubin and Rubin 2009.
17. Rubin and Rubin 2009, 152–153.
18. Tierney and Bensimon 1996; Barley 1990, 1996; Lamont 1992, 2009.
19. Lamont 2009; Barley 1996.
20. Green, Franquiz, and Dixon 1997; Scheurich 1995.
21. Lincoln and Guba 2000.
22. Gieryn 2000; Clark 1987.
23. Gould 1990, 283.
24. Becker 2008, 54, emphasis in original.
25. Patton and Appelbaum 2003, 67.
26. Glaser and Strauss 1967, 22, 25, 101–116.
27. Strauss and Corbin 1998.
28. Rubin and Rubin 2009, 201.
29. Curtin and Fossey 2007, 92.
30. Popper 1959.
31. Lareau 2010.
32. Jorgensen 1989.
33. LeCompte and Schensul 1999.
34. Michéle Lamont, personal communication, 2010.
References
Abbott, A. 1992. What do cases do? Some notes on activity in sociological anal-
ysis. In C. C. Ragin and H. S. Becker, eds., What is a case? Exploring the foun-
dations of social inquiry. New York: Cambridge University Press.
———. 2001. Chaos of disciplines. Chicago: University of Chicago Press.
Ahmed, S. 2012. On being included: Racism and diversity in institutional life.
Durham, NC: Duke University Press.
Allum, J., et al. 2014. Findings from the 2014 CGS International Graduate Admis-
sions Survey: Phase II: Final applications and initial offers of admission. Wash-
ington, DC: Council of Graduate Schools.
Allum, J. R., N. E. Bell, and R. S. Sowell. 2012. Graduate enrollment and degrees:
2001–2011. Washington, DC: Council of Graduate Schools. http://www.cgsnet
.org/graduate-enrollment-and-degrees-2001-2011.
Association of American Colleges and Universities. 2014. Making excellence inclu-
sive. http://www.aacu.org/compass/inclusive_excellence.cfm.
Attiyeh, G., and R. Attiyeh. 1997. Testing for bias in graduate school admissions.
Journal of Human Resources 32 (3): 524–548.
Aud, S., M. A. Fox, and A. Kewal-Ramani. 2010. Status and trends in the educa-
tion of racial and ethnic groups (NCES 2010-015). U.S. Department of Edu-
cation, National Center for Education Statistics. http://nces.ed.gov/pubs2010
/2010015.pdf.
Austin, A. E. 2002. Preparing the next generation of faculty: Graduate school as
socialization to the academic career. Journal of Higher Education 73 (1):
94–122.
Austin, A. E., and M. McDaniels. 2006. Preparing the professoriate of the future:
Graduate student socialization for faculty roles. In J. C. Smart, ed., Higher
216 References
education: Handbook of theory and research, vol. 21, 397–456. Dordrecht:

Springer.
Azuma, H., and K. Kashiwagi. 1987. Descriptors for an intelligent person: A Japa-
nese study. Japanese Psychological Research 29 (1): 17–26.
Barley, S. R. 1986. Technology as an occasion for structuring: Evidence from ob-
servations of CT scanners and the social order of radiology departments. Ad-
ministrative Science Quarterly 31 (1): 78–108.
———. 1990. Images of imaging: Notes on doing longitudinal field work. Organi-
zation Science 1:220–247.
———. 1996. Technicians in the workplace: Ethnographic evidence for bringing
work in to organizational studies. Administrative Science Quarterly 41 (3):
404–441.
Barley, S. R., and P. S. Tolbert. 1997. Institutionalization and structuration: Studying
the links between action and institution. Organization Studies 18 (1): 93–117.
Bastedo, M. N. 2009. Convergent institutional logics in public higher education:
State policymaking and governing board activism. Review of Higher Educa-
tion 32 (2): 209–234.
———. 2012. Cognitive repairs in the admissions office. Unpublished paper pre-
sented at the annual meeting of the Association for the Study of Higher Edu-
cation, Las Vegas.
Bastedo, M. N., and O. Jaquette. 2011. Running in place: Low-income students and
the dynamics of higher education stratification. Educational Evaluation and
Policy Analysis 33 (3): 318–339.
Becher, T. 1981. Towards a definition of disciplinary cultures. Studies in Higher Edu-
cation 6 (2): 109–122.
Becher, T., and P. Trowler. 1989. Academic tribes and territories. Columbus, OH:
Open University Press.
Becker, G. S. 1978. The economic approach to human behavior. Chicago: Univer-
sity of Chicago Press.
Becker, H. S. 2008. Tricks of the trade: How to think about your research while
you’re doing it. Chicago: University of Chicago Press.
Belasco, A. S., K. O. Rosinger, and J. C. Hearn. 2015. The test-optional movement
at America’s selective liberal arts colleges a boon for equity or something else?
Educational Evaluation and Policy Analysis 37 (2): 206–223.
Bell, N. E. 2010. Graduate enrollment and degrees: 1999 to 2009. Washington, DC:
Council of Graduate Schools. http://www.cgsnet.org/ckfinder/userfiles/files/R
_ED2009.pdf.
———. 2012. Findings from the 2012 CGS International Graduate Admissions
Survey, phase I: Applications. Washington, DC: Council of Graduate Schools.
http://www.cgsnet.org/ckfinder/userfiles/files/R_IntlApps12_I.pdf.
Benard, S., and S. J. Correll. 2010. Normative discrimination and the motherhood
penalty. Gender & Society 24 (5): 616–646.
Beoku-Betts, J. A. 2004. African women pursuing graduate studies in the sciences:
Racism, gender bias, and third world marginality. NWSA Journal 16 (1):
116–135.
References 217
Berger, P. L., and T. Luckmann. 1966. The social construction of reality: A treatise

in the sociology of knowledge. New York: Anchor Books.
Berrey, E. 2015. The enigma of diversity: The language of race and the limits of
racial justice. Chicago: University of Chicago Press.
Bérubé, M. 2013. The humanities, unraveled. Chronicle of Higher Education, Feb-
ruary 18. http://chronicle.com/article/Humanities-Unraveled/137291/.
Best graduate schools rankings. 2010. U.S. News & World Report. http://grad
-schools.usnews.rankingsandreviews.com/best-graduate-schools.
Beutel, A. M., and D. J. Nelson. 2005. Gender and race-ethnicity of faculty in top
science and engineering research departments. Journal of Women and Minori-
ties in Science and Engineering 11:389–403.
Bielby, R., J. Posselt, O. Jaquette, and M. Bastedo. 2014. Why are women under-
represented in elite universities? A non-linear decomposition analysis. Research
in Higher Education 55 (8): 735–760.
Biglan, A. 1973. The characteristics of subject matter in different academic areas.
Journal of Applied Psychology 58:195–203.
Birnbaum, R. 1988. Presidential searches and the discovery of organizational goals.
Journal of Higher Education 59 (5): 489–509.
Bitekine, A. 2011. Toward a theory of social judgments in organizations: The case
of legitimacy, reputation, and status. Academy of Management Review 36 (1):
151–179.
Blanchard, E., J. Bound, and S. Turner. 2009. Opening (and closing) doors: Country-
specific shocks in U.S. doctoral education. In R. G. Ehrenberg and C. V. Kuh, eds.,
Doctoral education and the faculty of the future, 224–248. Ithaca, NY: Cornell
University Press.
Bleske-Rechek, A., and K. Browne. 2014. Trends in GRE scores and graduate en-
rollments by gender and ethnicity. Intelligence 46:25–34.
Bobo, L., J. R. Kluegel, and R. A. Smith. 1997. Laissez-faire racism: The crystalli-
zation of a kinder, gentler, antiblack ideology. In S. A. Tuch and J. K. Martin, eds.,
Racial attitudes in the 1990s: Continuity and change, 15–42. Westport, CT:
Praeger Press.
Boltanski, L., and L. Thévenot. 2006. On justification: Economies of worth.
Princeton, NJ: Princeton University Press.
Bonilla-Silva, E. 2006. Racism without racists: Color-blind racism and the persis-
tence of racial inequality in the United States. Lanham, MD: Rowman and
Littlefield.
Borg, W. R. 1963. GRE aptitude scores as predictors of GPA for graduate students
in education. Educational and Psychological Measurement 23 (2): 379–382.
Bourdieu, P. 1977. Outline of a theory of practice. Cambridge: Cambridge Univer-
sity Press.
———. 1984. Distinction: A social critique of the judgment of taste. Cambridge,
MA: Harvard University Press.
———. 1988. Homo academicus. Stanford, CA: Stanford University Press.
Bourdieu, P., and J. C. Passeron. 1977. Reproduction in education, society, and cul-
ture. Thousand Oaks, CA: Sage.
218 References
Bourdieu, P., and L. J. Wacquant. 1992. An invitation to reflexive sociology. Chi-

cago: University of Chicago Press.
Bowen, W., and D. Bok. 1998. The shape of the river. Princeton, NJ: Princeton Uni-
versity Press.
Bowen, W. G., N. L. Rudenstine, and J. A. Sosa. 1992. In pursuit of the PhD.
Braxton, J. M., and L. L. Hargens. 1996. Variations among academic disciplines:
Analytical frameworks and research. In J. C. Smart, ed., Higher education:
Handbook of theory and research, vol. 11. New York: Agathon Press.
Brear, P., J. Dorrian, and G. Luscri. 2008. Preparing our future counselling profes-
sionals: Gatekeeping and the implications for research. Counselling and Psy-
chotherapy Research 8 (2): 93–101.
Brink, W. J. 1999. Selecting graduate students. Journal of Higher Education 70 (5):
517–523.
Bryson, B. 1996. “Anything but heavy metal”: Symbolic exclusion and musical dis-
likes. American Sociological Review 61 (5): 884–899.
Burris, V. 2004. The academic caste system: Prestige hierarchies in PhD exchange
networks. American Sociological Review 69 (2): 239–264.
Campbell, J. 2009. Attitudes and beliefs of counselor educators toward gatekeeping.
Unpublished doctoral dissertation, Old Dominion University.
Campbell, L. G., S. Mehtani, M. E. Dozier, and J. Rinehart. 2013. Gender-
heterogeneous working groups produce higher quality science. PLoS ONE 8
(10): e79147.
Cantwell, B., M. Canche, J. Milem, and F. Sutton. 2010. Do data support the dis-
course? Assessing holistic review as an admissions process to promote diver-
sity at a US medical school. Unpublished paper presented at the annual meeting
of the Association for the Study of Higher Education, Charlotte, NC.
Carter, D. J., and F. Tuitt. 2013. Contesting the myth of a “post-racial” era: The
continued significance of race in U.S. education. New York: Peter Lang Inter-
national Academic.
Cassuto, L. 2015. The graduate school mess. Cambridge, MA: Harvard University
Press.
Chang, L. 2001. Greater China: Cheating allegations anger Chinese; Row over
TOEFL, GRE exams reveals cultural, educational, legal chasm; Students fear
their scores now have no credibility with U.S. schools. Asian Wall Street Journal,
February 21.
Chang, M. J. 2002. Preservation or transformation: Where’s the real educational
discourse on diversity? Review of Higher Education 25 (2): 125–140.
Chang, M. J., A. W. Astin, and D. Kim. 2004. Cross-racial interaction among un-
dergraduates: Some consequences, causes, and patterns. Research in Higher
Education 45 (5): 529–553.
Charmaz, K., N. K. Denzin, and Y. S. Lincoln. 2000. Handbook of qualitative re-
search. Thousand Oaks, CA: Sage.
Cheslock, J. J., and R. Kroc. 2012. Enrollment management. In R. Howard, B. Knight,
and G. McLaughlin, eds., The handbook for institutional researchers, 221–
236. San Francisco: Jossey-Bass.
References 219
Chou, C. C. 2008. Critique on the notion of model minority: An alternative racism

to Asian American? Asian Ethnicity 9 (3): 219–229.
Christian, B. 1987. The race for theory. Cultural Critique 56 (6): 51–63.
Civil Rights Project. 2002. Constitutional requirements for affirmative action in
higher education admissions and financial aid. http://civilrightsproject.ucla.edu
/legal-developments/legal-memos/constitutional-requirements-for-affirmative
- action - in - higher- education - admissions - and - financial - aid / constitutional
-affirmative-action-financial-aid.pdf.
Clark, B. R. 1960. The open door college: A case study. New York: McGraw-Hill.
———. 1987. The academic profession: National, disciplinary, and institutional set-
tings. Berkeley: University of California Press.
Clauset, A., S. Arbesman, and D. B. Larremore. 2015. Systematic inequality and
hierarchy in faculty hiring networks. Science Advances 1 (1): e1400005.
Collins, R. 2002. Credential inflation and the future of universities. In S. Brint, ed.,
The future of the city of intellect: The changing American university, 23–46.
Stanford, CA: Stanford University Press.
Committee on Equal Opportunities in Science and Engineering (CEOSE). 2013.
2011–2012 biennial report to Congress: Broadening participation in America’s
STEM workforce. CEOSE 13–01. Arlington, VA: National Science Foundation.
http://www.nsf.gov/od/iia/activities/ceose/reports/Full_2011–2012_CEOSE
_Report_to_Congress_Final_01–28–2014.pdf.
Correll, S. J., I. S. Benard, and Paik. 2007. Getting a job: Is there a motherhood
penalty? American Journal of Sociology 112 (5): 1297–1339.
Council of Graduate Schools (CGS). 2009. Graduate enrollment and degrees: 1999
to 2009. http://www.cgsnet.org/portals/0/pdf/R_ED2009.pdf.
———. 2013. Press release: First-time enrollment of international graduate students
up 10 percent. Council of Graduate Schools. http://www.cgsnet.org/sites/default
/files/Intl_III_2013_release_final_0.pdf.
Crasnow, S. 2007. Report on the APA Status of Women. Paper presented at the an-
nual meeting of the Central Division American Philosophical Association,
Chicago.
Crocker, J., S. R. Sommers, and R. K. Luhtanen. 2002. Hopes dashed and dreams
fulfilled: Contingencies of self-worth and graduate school admissions. Person-
ality and Social Psychology Bulletin 28 (9): 1275–1286.
Cronbach, L. J. 1975. Beyond the two disciplines of scientific psychology. Amer-
ican Psychologist 12 (11): 116–127.
Cureton, E. E., L. W. Cureton, and R. Bishop. 1949. Predictions of success in
graduate study at the University of Tennessee. American Psychologist
4:361–362.
Curtin, M., and E. Fossey. 2007. Appraising the trustworthiness of qualitative
studies, Australian Occupational Therapy Journal 54 (2): 88–94.
Cyranoski, D., N. Gilbert, H. Ledford, A. Nayar, and M. Yahia. 2011. Education:
The PhD factory. Nature 472:276–279.
Danowitz Sagaria, M. A. 2002. An exploratory model of filtering in administrative
searches: Toward counter-hegemonic discourses. Journal of Higher Education
73 (6): 677–711.
220 References
David, J. 2012. Michigan colleges welcome Chinese students. Detroit Free Press,
March 6.
Dawes, R. M. 1971. A case study of graduate admissions: Application of three prin-
ciples of human decision making. American Psychologist 26 (2): 180–188.
———. 1975. Graduate admission variables and future success. Science 187 (4178):
721–723.
Deo, M. 2011. The promise of Grutter: Diverse interactions at the University of
Michigan Law School. Michigan Journal of Race & Law 17 (1): 63.
DiMaggio, P. 1987. Classification in art. American Sociological Review 52 (4):
440–455.
———. 1992. Nadel’s paradox revisited: Relational and cultural aspects of social
structure. In N. Nohria and R. Eccles, eds., Networks and organizations: Struc-
ture, form and action, 118–142. Boston: Harvard Business School Press.
———. 1997. Culture and cognition. Annual Review of Sociology 23 (1): 263–287.
DiMaggio, P. J., and W. W. Powell. 1983. The iron cage revisited: Institutional iso-
morphism and collective rationality in organizational fields. American Socio-
logical Review 48 (2): 147–160.
———, eds. 1991. New institutionalism in organizational analysis. Chicago: Uni-
versity of Chicago Press.
Dion, D. 1998. Evidence and inference in the comparative case study. Comparative
Politics 30 (2): 127–145.
Dowd, A. C. and E. M. Bensimon. 2014. Engaging the “race question”: Account-
ability and equity in US higher education. New York: Teachers College Press.
Duguid, M. M., D. L. Loyd, and P. S. Tolbert. 2012. The impact of categorical
status, numeric representation, and work group prestige on preference for de-
mographically similar others: A value threat approach. Organization Science
23 (2): 386–401.
Durant, W. 2012. The story of philosophy. New York: Pocket Books.
Eagan, M. K., S. Hurtado, M. J. Chang, G. A. Garcia, F. A. Herrera, and J. C.
Garibay. 2013. Making a difference in science education: The impact of
undergraduate research programs. American Educational Research Journal 50
(4): 683–713.
Educational Testing Service (ETS). 2013a. A snapshot of the individuals who took
the GRE Revised General Test. http://www.ets.org/s/gre/pdf/snapshot_test
_taker_data.pdf.
———. 2013b. Test content and structure. https://www.ets.org/gre/revised_general
/about/content/.
———. 2014a. GRE: Guide to the use of scores. http://www.ets.org/s/gre/pdf/gre
_guide.pdf.
———. 2014b. Test content and structure. https://www.ets.org/gre/revised_general
/about/content/.
Eide, E., D. Brewer, and R. Ehrenberg. 1998. Does it pay to attend an elite private
college? Evidence on the effects of undergraduate college quality on graduate
school attendance. Economics of Education Review 17 (4): 371–376.
Eisenhardt, K. M. 1989. Building theories from case study research. Academy of
Management Review 14 (4): 532–550.
References 221
Ellsberg, D. 1961. Risk, ambiguity, and the Savage axioms. Quarterly Journal of
Economics 75 (4): 643–669.
Epley, N., and J. Kruger. 2005. When what you type isn’t what they read: The per-
severance of stereotypes and expectancies over e-mail. Journal of Experimental
Social Psychology 41 (4): 414–422.
Espeland, W. N., and M. Sauder. 2007. Rankings and reactivity: How public mea-
sures recreate social worlds. American Journal of Sociology 113 (1): 1–40.
Espeland, W. N., and M. L. Stevens. 1998. Commensuration as a social process.
Annual Review of Sociology 24 (1): 313–343.
———. 2009. A sociology of quantification. European Journal of Sociology 49 (3):
401–436.
Espenshade, T. J., and C. Y. Chung. 2005. The opportunity cost of admission pref-
erences at elite universities. Social Science Quarterly 86 (2): 293–305.
Feldman, M. S., and B. T. Pentland. 2003. Reconceptualizing organizational rou-
tines as a source of flexibility and change. Administrative Science Quarterly
48 (1): 94–118.
Fischer, F. T., J. Schult, and B. Hell. 2013. Sex-specific differential prediction of col-
lege admission tests: A meta-analysis. Journal of Educational Psychology 105
(2): 478.
Fischer, K. 2013. American universities yawn at global rankings. Chronicle of
Higher Education. http://chronicle.com/article/American-Universities-Yawn-at
/141947.
Fish, I. S. 2013. Why do so many Chinese people share the same names? Foreign
Policy, April 26. http://blog.foreignpolicy.com/posts/2013/04/26/why_do_so
_many_chinese_people_share_the_same_name.
Fisher v. University of Texas at Austin. 2013. 570 U.S. 11–345.
Fishkin, J. 2009. When the people speak: Deliberative democracy and public con-
sultation. Oxford: Oxford University Press.
Fisk-Vanderbilt. 2014. Fisk-Vanderbilt Masters-to-PhD Bridge Program. http://www
.vanderbilt.edu/gradschool/bridge/descript.htm.
Fowler, J. H., B. Grofman, and N. Masuoka. 2007. Social networks in political sci-
ence: Hiring and placement of PhDs, 1960–2002. PS: Political Science & Poli-
tics 40 (4): 729–739.
Fox, C. R., and A. Tversky. 1995. Ambiguity aversion and comparative ignorance.
Quarterly Journal of Economics 110 (3): 585–603.
Frank, R. H., and P. J. Cook. 1995. The winner-take-all society. New York: Free
Press.
Freeman, R. B., and W. Huang. 2014. Collaborating with people like me: Ethnic
co-authorship within the US. National Bureau of Economic Research. NBER
Working Paper No. 19905. http://www.nber.org/papers/w19905.
Friedland, R., and R. R. Alford. 1991. Bringing society back in: Symbols, practices,
and institutional contradictions. In P. J. DiMaggio and W. W. Powell, eds., The
new institutionalism in organizational analysis, 232–263. Chicago: University
of Chicago Press.
Fu, Y. 2012. The effectiveness of traditional admissions criteria in predicting col-
lege and graduate success for American and international students. Unpublished
222 References
doctoral dissertation, Arizona State University. http://arizona.openrepository

.com/arizona/handle/10150/217056.
Furnham, A. 2001. Self-estimates of intelligence: Culture and gender difference in
self and other estimates of both general (g) and multiple intelligences. Person-
ality and Individual Differences 31 (8): 1381–1405.
Garces, L. M. 2012. Necessary but not sufficient: The impact of Grutter v. Bollinger
on student of color enrollment in graduate and professional schools in Texas.
Journal of Higher Education 83 (4): 497–534.
———. 2014. Aligning diversity, quality, and equity: The implications of legal and
public policy developments for promoting racial diversity in graduate studies.
American Journal of Education 120 (4): 457–480.
Garfinkel, H. 1967. Studies in ethnomethodology. Malden, MA: Prentice Hall.
Gartner, J. D. 1986. Antireligious prejudice in admissions to doctoral programs in
clinical psychology. Professional Psychology: Research and Practice 17 (5):
473–475.
Geertz, C. 1973. Thick description: Toward an interpretive theory of culture.
In C. Geertz, ed., The interpretation of culture: Selected essays, 3–30. New
York: Basic Books.
———. 1982. The way we think now: Toward an ethnography of modern thought.
Bulletin of the American Academy of Arts and Sciences, 14–34.
Geiger, R. L. 1986. To advance knowledge. Scientist 1 (2): 24.
Gibney, E. 2012. So many aspirant doctors, so little time. Times Higher Education,
October 4. http://www.timeshighereducation.co.uk/news/so-many-aspirant
-doctors-so-little-time/421356.article.
Giddens, A. 1984. The constitution of society: Outline of the theory of structura-
tion. Berkeley: University of California Press.
Gieryn, T. F. 2000. A space for place in sociology. Annual Review of Sociology
26:463–496.
Glaser, B. G., and Strauss, A. L. 1967. The discovery of grounded theory: Strategies
for qualitative research. New York: Aldine.
Goffman, E. 1959. The presentation of self in everyday life. New York: Overlook
Press.
———. 1974. Frame analysis: An essay on the organization of experience. New
York: Harper and Row.
Gold, R. 1958. Roles in sociological field observations. Social Forces 36 (3):
217–223.
Golden, D. 2011. China’s test prep juggernaut. Bloomberg Businessweek. http://www
.bloomberg.com/bw/magazine/content/11_20/b4228058558042.htm
Gonzales, L. M., J. R. Allum, and R. S. Sowell. 2013. Graduate enrollment and
degrees: 2002 to 2012. Washington, DC: Council of Graduate Schools.
http://www.cgsnet.org/ckfinder/userfiles/files/GEDReport_2012.pdf.
Gonzales, L. M., J. R. Remington, and J. R. Allum. 2013. Findings from the 2013
international graduate admissions survey. Washington, DC: Council of Grad-
uate Schools. https://www.cgsnet.org/ckfinder/userfiles/files/Intl_II_2013_report
_final.pdf.
References 223
Good, C., J. Aronson, and J. A. Harder. 2008. Problems in the pipeline: Stereotype

threat and women’s achievement in high-level math courses. Journal of Applied
Developmental Psychology 29:17–28.
Gould, S. J. 1990. Wonderful life: The Burgess Shale and the nature of history. New
York: W. W. Norton.
Graduate Employees and Student Organizations. 2005. The (un)changing face of
the Ivy League. New Haven, CT: Yale University.
Gratz v. Bollinger. 2003. 539 U.S. 244.
Green, J., M. Franquiz, and C. Dixon. 1997. The myth of the objective transcript:
Transcribing as a situated act. TESOL Quarterly 31:172–176.
Griffin, K. A., and M. M. Muñiz. 2011. The strategies and struggles of graduate
diversity officers in the recruitment of doctoral students of color. Equity & Ex-
cellence in Education 44 (1): 57–76.
Grodsky, E. 2007. Compensatory scholarship in higher education. American Journal
of Sociology 112 (6): 1662–1712.
Grodsky, E., J. R. Warren, and E. Felts. 2008. Testing and social stratification in
American education. Annual Review of Sociology 34:385–404.
Grove, W. A., D. H. Dutkowsky, and A. Grodner. 2007. Survive then thrive: Deter-
minants of success in the economics Ph.D. program. Economic Inquiry 45 (4):
864–871.
Grunig, S. D. 1997. Research, reputation, and resources: The effect of research ac-
tivity on perceptions of undergraduate education and institutional resource ac-
quisition. Journal of Higher Education 68 (1): 17–52.
Grutter [et al.] v. Bollinger [et al.]. 2003. 539 U.S. 306.
Guinier, L. 2003. Admissions rituals as political acts: Guardians at the gates of our
democratic ideals. Harvard Law Review 117:113.
———. 2015. The tyranny of the meritocracy: Democratizing higher education in
America. New York: Beacon Books.
Gumport, P. J. 1993. Graduate education and organized research in the United States
and views from American campuses. In B. Clark, ed., The research foundations
of graduate education: Germany, Britain, France, United States, Japan, 225–
293. Berkeley: University of California Press.
———. 2002. Academic pathfinders: Knowledge creation and feminist scholarship.
Westport, CT: Greenwood Press.
Gurin, P., E. Dey, S. Hurtado, and G. Gurin. 2002. Diversity and higher education:
Theory and impact on educational outcomes. Harvard Educational Review 72
(3): 330–367.
Gurin, P., J. Lehman, E. Lewis, with E. Dey, G. Gurin, and S. Hurtado. 2004. De-
fending diversity: Affirmative action at the University of Michigan. Ann Arbor:
University of Michigan Press.
Gutmann, A., and D. Thompson. 2004. Why deliberative democracy? Princeton, NJ:
Princeton University Press.
Hagedorn, L. S., and A. Nora. 1996. Rethinking admissions criteria in graduate and
professional programs. New Directions for Institutional Research, no. 92:
31–44.
224 References
Hagstrom, W. 1964. Anomy in scientific communities. Social Problems 12 (2):

186–195.
Hammersley, M., and P. Atkinson. 2003. Ethnography: Principles in practice, 3rd
ed. New York: Routledge.
Hanassab, S. 2006. Diversity, international students, and perceived discrimination:
Implications for educators and counselors. Journal of Studies in International
Education 10 (2): 157–172.
Haro, Roberto. 1995. Held to a higher standard: Latino executive selection in higher
education. In Walter Bennis, ed., The leaning ivory tower: Latino professors in
American universities, 189–207. Albany: SUNY Press.
Harper, S. R. 2008. Realizing the intended outcomes of Brown: High-achieving Af-
rican American male undergraduates and social capital. American Behavioral
Scientist 51 (7): 1029–1052.
Harvey, C., and M. J. Allard. 2012. Understanding and managing diversity: Read-
ings, cases, and exercises. Boston: Pearson Prentice Hall.
Harvey, J. T., and R. F. Garnett. 2006. Future directions for heterodox economics.
Ann Arbor: University of Michigan Press.
Hatt, B. 2012. Smartness as cultural practice in schools. American Educational Re-
search Journal 49 (3): 438–460.
Heimer, C. 2001. Cases and biographies: An essay on routinization and the nature
of comparison. Annual Review of Sociology 27 (1): 47–76.
Hermanowicz, J. C. 2005. Classifying universities and their departments: A social
world perspective. Journal of Higher Education 76 (1): 26–55.
Hersch, J. 2014. Catching up is hard to do: Undergraduate prestige, elite graduate
programs, and the earnings premium. Vanderbilt Law and Economics Re-
search Paper No. 14–23. http://papers.ssrn.com/sol3/papers.cfm?abstract_id
=2473238.
Hess, D., and J. Posselt. 2002. How high school students experience and learn from
the discussion of controversial public issues. Journal of Curriculum and Su-
pervision 17 (4): 283–314.
Hibel, J., and D. M. Penn. 2014. An organizational perspective on the patterns and
predictors of educator cheating on standardized accountability tests. Paper pre-
sented at the 2014 Sociology of Education Association Conference, Pacific
Grove, CA, February.
Hiss, W., and P. Neupane. 2004. 20 years of optional SATs. Bates News. http://www
.bates.edu/news/2004/10/01/sats-at-bates/.
Holsapple, M., and J. R. Posselt. 2010. The best in the country? High achieving
rural students’ enrollment in America’s top-ranked colleges and universities.
Paper presented at the annual meeting of the American Educational Research
Association, Denver, April.
Howard, J. A. 2000. Social psychology of identities. Annual Review of Sociology
26: 367–393.
Hurtado, S. 2001. Linking diversity and educational purpose: How diversity affects
the classroom environment and student development. In Gary Orfield, ed., Di-
versity challenged: Evidence on the impact of affirmative action. Cambridge,
MA: Harvard Education Publishing Group.
References 225
Institute of International Education. 2013. Open Doors Data. http://www.iie.org

/research-and-publications/open-doors/data.
Ioannides, S., and K. Nielsen. 2007. Economics and the social sciences: Synergies
and trade-offs. In S. Ioannides and K. Nielsen, eds., Economics and the social
sciences. Cheltenham, UK: Edward Elgar.
Iwamoto, D. K., and W. M. Liu. 2010. The impact of racial identity, ethnic identity,
Asian values, and race-related stress on Asian Americans and Asian interna-
tional college students’ psychological well-being. Journal of Counseling
Psychology 57 (1): 79.
Jackson, J. F. 2006. Hiring practices of African American males in academic leader-
ship positions at American colleges and universities: An employment trends and
disparate impact analysis. Teachers College Record 108 (2): 316–338.
James, W. 1903 The Ph.D. octopus. Harvard Monthly, March. http://www.des.emory
.edu/mfp/octopus.html.
Jaquette, O., and B. R. Curs. 2015. Creating the Out-of-State University: Do Public
Universities Increase Nonresident Freshman Enrollment in Response to De-
clining State Appropriations?. Research in Higher Education, 1–31. doi:
10.1007/s11162-015-9362-2.
Jaschik, S. 2008. Non-cognitive qualities join the GRE. Inside Higher Ed, May 22.
http://www.insidehighered.com/news/ 2008/05/22/ets.
———. 2009. Ph.D. admissions shrinkage. Inside Higher Ed, March 30. https://www
.insidehighered.com/news/2009/03/30/phd.
Jayakumar, U. M. 2008. Can higher education meet the needs of an increasingly
diverse and global society? Campus diversity and cross-cultural workforce
competencies. Harvard Educational Review 78 (4): 615–651.
Johnson, K. 2008. Quantitative methods in linguistics. Malden, MA: Blackwell.
Johnson, P. 1988. English language proficiency and academic performance of un-
dergraduate international students. TESOL Quarterly 22 (1): 164–168.
Jones, E. 2003. Beyond supply and demand: Assessing the Ph.D. job market. Oc-
cupational Outlook Quarterly. Washington DC: Bureau of Labor Statistics.
Jorgensen, D. L. 1989. Participant observation: A methodology for human studies.
Thousand Oaks, CA: Sage.
Kahlenberg, R. D. 1996. Class-based affirmative action. California Law Review 84
(4): 1037–1099.
Kahneman, D. 2011. Thinking, fast and slow. New York: Macmillan.
Kallingal, A. 1971. The prediction of grades for black and white students at Mich-
igan State University. Journal of Educational Measurement 8 (4): 263–265.
Kant, I. 1951 [1790]. Critique of judgment. Translated by J. H. Bernard. New York:
Hafner Press.
Kanter, R. M. 1977. Some effects of proportions on group life: Skewed sex ratios
and responses to token women. American Journal of Sociology 82 (5):
965–990.
Karabel, J. 2005. The chosen: The hidden history of admission and exclusion at
Harvard, Yale, and Princeton. New York: Houghton Mifflin.
Karen, D. 1990. Toward a political-organizational model of gatekeeping: The case
of elite colleges. Sociology of Education 63: 227–240.
226 References
Kezar, A., and C. Sam. 2010. Special Issue: Understanding the New Majority of
Non-Tenure-Track Faculty in Higher Education—Demographics, Experiences,
and Plans of Action. ASHE Higher Education Report 36(4): 1–133.
Khan, S. R. 2011. Privilege: The making of an adolescent elite at St. Paul’s school.
Killgore, L. 2009. Merit and competition in selective college admissions. Review of
Higher Education 32 (4): 469–488.
Klitgaard, R. 1985. Choosing elites: Selecting the” best and brightest” at top uni-
versities and elsewhere. New York: Basic Books.
Knorr-Cetina, K. 1999. Epistemic cultures: How the sciences make knowledge.
Cambridge, MA: Harvard University Press.
Konrad, A. M., and J. Pfeffer. 1991. Understanding the hiring of women and mi-
norities in educational institutions. Sociology of Education 64 (3): 141–157.
Kuhn, T. S. 1970. The structure of scientific revolutions, 2nd ed. Chicago: Univer-
sity of Chicago Press.
Kulis, S., Y. Chong, and H. Shaw. 1999. Discriminatory organizational contexts and
black scientists on postsecondary faculties. Research in Higher Education 40
(2): 115–148.
Kuncel, N. R., and S. A. Hezlett. 2007. Standardized tests predict graduate students’
performance. Science 315:1080–1081.
Kuncel, N. R., S. A. Hezlett, and D. S. Ones. 2001. A comprehensive meta-analysis
of the predictive validity of the Graduate Record Examinations: Implications
for graduate student selection and performance. Psychological Bulletin 127 (1):
162–181.
Kyllonen, P. C. 2008. The research behind the ETS® Personal Potential Index (PPI).
Princeton, NJ: Educational Testing Service. http://www.ets.org/Media/Products
/PPI/10411_PPI_bkgrd_report_RD4.pdf.
Labaree, D. F. 1997. How to succeed in school without really learning: The creden-
tials race in American education. New Haven, CT: Yale University Press.
Lamont, M. 1992. Money, morals, and manners: The culture of the French and
American upper-middle class. Chicago: University of Chicago Press.
———. 2009. How professors think: Inside the curious world of academic judg-
ment. Cambridge, MA: Harvard University Press.
———. 2012. Toward a comparative sociology of valuation and evaluation. An-
nual Review of Sociology 38 (1): 201.
Lamont, M., and A. Lareau. 1988. Cultural capital: Allusions, gaps and glissandos
in recent theoretical developments. Sociological Theory 6 (2): 153–168.
Lamont, M., and V. Molnar. 2002. The study of boundaries in the social sciences.
Annual Review of Sociology 28 (1): 167–195.
Landrum, R. E., and M. A. Clump. 2004. Departmental search committees and the
evaluation of faculty applicants. Teaching of Psychology 31 (1): 12–17.
Landrum, R. E., E. B. Jeglum, and J. R. Cashin. 1994. The decision-making pro-
cesses of graduate admissions committees in psychology. Journal of Social Be-
havior and Personality 9:239–239.
Lannholm, G. V. 1968. Cooperative studies of predicting graduate school success.
Graduate Record Examinations Special Report Number 68–3.
References 227
Lareau, A. 2010. Families, schools, and inequality: A reflection. Keynote address at

the annual meeting of the Sociology of Education Association Conference. Pa-
cific Grove, CA, February 19.
Larmer, B., and J. Lin-Liu. 2002. Diplomas for dollars: In China, education has al-
ways been sacred; Now, tainted by corruption, it’s become big business, too.
Newsweek, December 2.
Lattuca, L. R. 2001. Creating interdisciplinarity: Interdisciplinary research and
teaching among college and university faculty. Nashville, TN: Vanderbilt Uni-
versity Press.
LeCompte, M. D., and J. J. Schensul. 1999. Designing and conducting ethnographic
research. Lanham, MD: Altamira Press.
Lee, J. J., and C. Rice. 2007. Welcome to America? International student percep-
tions of discrimination. Higher Education 53 (3): 381–409.
Lee, S. J. 1994. Behind the model minority stereotype: Voices of high and low
achieving Asian American students. Anthropology & Education Quarterly 25
(4): 413–429.
Lemann, N. 2000. The big test: The secret history of the American meritocracy. New
York: Farrar, Straus and Giroux.
Leslie, S.-J., A. Cimpian, M. Meyer, and E. Freeland. 2015. Expectations of bril-
liance underlie gender distributions across academic disciplines. Science 347
(6219): 262–265.
Light, R. L., M. Xu, and J. Mossop. 1987. English proficiency and academic per-
formance of international students. TESOL Quarterly 21 (2): 251–261.
Lincoln, Y. S., and E. G. Guba. 2000. Paradigmatic controversies, contradictions,
and emerging confluences. Handbook of Qualitative Research 2:163–188.
Lodahl, J. B., and G. Gordon. 1972. The structure of scientific fields and the func-
tioning of university graduate departments. American Sociological Review 37
(1): 57–72.
Logel, C. R., G. M. Walton, S. J. Spencer, J. Peach, and Z. P. Mark. 2012. Un-
leashing latent ability: Implications of stereotype threat for college admissions.
Educational Psychologist 47 (1): 42–50.
Lounsbury, M. 2002. Institutional transformation and status mobility: The profes-
sionalization of the field of finance. Academy of Management Journal 45 (1):
255–266.
———. 2007. A tale of two cities: Competing logics and practice variation in the
professionalizing of mutual funds. Academy of Management Journal 50 (2): 289.
Lundy-Wagner, V., J. Vultaggio, and M. Gasman. 2013. Preparing underrepresented
students of color for doctoral success: The role of undergraduate institutions.
International Journal of Doctoral Studies 8:151–172.
Madaus, G. F., and J. J. Walsh. 1965. Departmental differentials in the predictive
validity of the Graduate Record Examination aptitude tests. Educational and
Psychological Measurement 25 (4): 1105–1110.
Maddux, W. W., A. D. Galinsky, A. J. Cuddy, and M. Polifroni. 2008. When being
a model minority is good . . . and bad: Realistic threat explains negativity to-
ward Asian Americans. Personality and Social Psychology Bulletin 34 (1):
74–89.
228 References
Magnier, M. 2006. The world: The academic equation in China is making college
a long shot. Los Angeles Times, June 18.
Maloney, P. 2014. Gaming or cheating? How teachers and administrators in low-
income schools adapt to high stakes standardized tests. Paper presented at the
2014 Sociology of Education Association Conference, Pacific Grove, CA,
February.
March, J. G. 1994. A primer on decision making: How decisions happen. New York:
Free Press.
March, J. G., H. A. Simon, and H. Guetzkow. 1958. Organizations. Cambridge,
MA: Blackwell.
Marichal, J. 2009. Frame evolution: A new approach to understanding changes in
diversity reforms at public universities in the United States. Social Science
Journal 46 (1): 171–191.
Marston, A. R. 1971. It is time to reconsider the Graduate Record Examination.
American Psychologist 26 (7): 653–655.
Martin, J. 1992. Cultures in organizations: Three perspectives. Oxford: Oxford Uni-
versity Press.
Mason, M. A., M. Goulden, and K. Frasch. 2009. Why graduate students reject the
fast track. Academe 95 (1): 11–16.
Maton, K. I., J. L. Kohout, M. Wicherski, G. E. Leary, and A. Vinokurov. 2006.
Minority students of color and the psychology graduate pipeline: Disquieting
and encouraging trends, 1989–2003. American Psychologist 61 (2): 117.
Mayhew, M. J., H. E. Grunwald, and E. L. Dey. 2005. Curriculum matters: Cre-
ating a positive climate for diversity from the student perspective. Research in
McAdam, D., S. G. Tarrow, and C. Tilly. 2001. Dynamics of contention. Cambridge:
Cambridge University Press.
McDonough, P. M. 1997. Choosing colleges: How social class and schools struc-
ture opportunity. Albany: SUNY Press.
McPherson, M., L. Smith-Lovin, and J. M. Cook. 2001. Birds of a feather: Ho-
mophily in social networks. Annual Review of Sociology 27:415–444.
Melamed, D., and S. V. Savage. 2013. Status, numbers and influence. Social Forces
91 (3): 1085–1104.
Melcher, T. 2010. Busted: The top 5 ways that Chinese students cheat on their un-
dergraduate applications to American schools (and what schools can do about
it). In The China conundrum: Insights for US schools. White Paper No. 4,
Zinch, Inc. http://www.aplu.org/document.doc?id=4004.
Merriam, S. B. 2009. Qualitative research: A guide to design and implementation.
San Francisco: Jossey-Bass.
Meyer, H. D., and B. Rowan. 2006. The new institutionalism in education. Albany:
SUNY Press.
Meyer, J., and B. Rowan. 1977. Institutionalized organizations: Formal structure
as myth and ceremony. American Journal of Sociology 83 (2): 340–363.
Milem, J. F., M. J. Chang, and A. L. Antonio. 2005. Making diversity work on
campus: A research-based perspective. Washington, DC: Association of Amer-
ican Colleges and Universities.
References 229
Miles, M. B., and A. M. Huberman. 1994. Qualitative data analysis: An expanded

sourcebook. Thousand Oaks, CA: Sage.
Milkman, K. L., M. Akinola, and D. Chugh. 2014. What happens before? A field
experiment exploring how pay and representation differentially shape bias on
the pathway into organizations. Social Science Research Network. http://dx.doi
.org/10.2139/ssrn.2063742.
Miller, C. 2013. Diversity in physics: Impact of using minimum acceptable GRE
scores for graduate admissions. Bulletin of the American Physical Society 58.
http://meetings.aps.org/Meeting/MAR13/Session/F38.3.
Miller, C., and K. Stassun. 2014. A test that fails. Nature 510:303–304. http://www
.nature.com/naturejobs/science/articles/10.1038/nj7504–303a.
Montlake, S. 2006. High-tech cheating in Asia’s high-stakes exams. Christian Sci-
ence Monitor, June 9.
Mooney, P. 2005. Mercenaries of grammar. Chronicle of Higher Education, Feb-
ruary 4.
Morphew, C., and M. Hartley. 2006. Mission statements: A thematic analysis of rhet-
oric across institutional type. Journal of Higher Education 77 (3): 456–471.
Morrison, K. 2006. Marx, Durkheim, Weber: Formations of modern social thought.
Thousand Oaks, CA: Sage.
Morrison, T., and M. Morrison. 1995. A meta-analytic assessment of the predic-
tive validity of the quantitative and verbal components of the GRE with grad-
uate GPAs representing the criterion of graduate success. Educational and
Psychological Measurement 55 (2): 309–316.
Morse, R., and S. Flanigan. 2013. Methodology: Best education schools. U.S.
News & World Report, March 11. http://www.usnews.com/education/best
-graduate-schools/articles/2013/03/11/methodology-best-education-schools
-rankings.
Moss-Racusin, C. A., J. F. Dovidio, V. L. Brescoll, M. J. Graham, and J. Handelsman.
2012. Science faculty’s subtle gender biases favor male students. Proceedings
of the National Academy of Sciences 109 (41): 16474–16479.
Museus, S. D. 2008. The model minority and the inferior minority myths: Under-
standing stereotypes and their implications for student learning. About Campus
13 (3): 2–8.
Museus, S. D., and P. N. Kiang. 2009. Deconstructing the model minority myth and
how it contributes to the invisible minority reality in higher education research.
New Directions for Institutional Research, no. 142: 5–15.
Nadal, K. L., S. T. Pituc, M. P. Johnston, and T. Esparrago. 2010. Overcoming the
model minority myth: Experiences of Filipino American graduate students.
Journal of College Student Development 51 (6): 694–706.
Nagda, B. A., S. R. Gregerman, J. Jonides, W. vonHippel, and J. S. Lerner. 1998.
Undergraduate student-faculty research partnerships affect student retention.
Review of Higher Education 22 (1): 55–72.
National Center for Fair and Open Testing. 2014. Colleges and universities that do
not use SAT/ACT Scores for admitting substantial numbers of students into bach-
elor degree programs. http://www.fairtest.org/schools-do-not-use-sat-or-act
-scores-admitting-substantial-numbers-students-bachelor-degree-programs.
230 References
National Center for Institutional Diversity. 2011. Mission statement. http://www

.ncid.umich.edu/about/mission.shtml.
National Endowment for the Humanities. 2002. David McCullough interview with
Bruce Cole. http://www.neh.gov/about/awards/jefferson-lecture/david-mccull
ough-interview.
National Foundation for American Policy. 2013. The importance of international
students to America. Arlington, VA: National Foundation for American
Policy.
National Research Council. 2010. A data-based assessment of research doctorate
programs in the United States. Washington DC: National Academies Press.
National Science Foundation. 2004. Women, minorities, and persons with disabili-
ties in science and engineering: 2004. NSF 04–317. Arlington, VA: NSF Divi-
sion of Science Resource Statistics.
———. 2010. Doctorate recipients from U.S. universities: 2009. Arlington, VA: NSF
Division of Science Resource Statistics.
———. 2013. Doctorate recipients from U.S. universities: 2012. Arlington, VA: NSF
Division of Science Resource Statistics. Tabulation engine at https://ncses.norc
.org/NSFTabEngine/#TABULATION.
Nee, V. 1998. Sources of the new institutionalism. In M. C. Brinton and V. Nee,
eds., The new institutionalism in sociology, 1–16. Stanford, CA: Stanford Uni-
versity Press.
Nettles, M. T., and C. M. Millett. 2006. Three magic letters: Getting to Ph.D. Bal-
timore: Johns Hopkins University Press.
Neumann, A. 1991. The thinking team: Toward a cognitive model of administrative
teamwork in higher education. Journal of Higher Education 62 (5): 485–513.
———. 2009. Professing to learn: Creating tenured lives and careers in the Amer-
ican research university. Baltimore: Johns Hopkins University Press.
Newman, R. I. 1968. GRE scores as predictors of GPA for psychology graduate
students. Educational and Psychological Measurement 28 (2): 433–436.
Ng, J. C., S. S. Lee, and Y. K. Pak. 2007. Contesting the model minority and per-
petual foreigner stereotypes: A critical review of literature on Asian Americans
in education. Review of Research in Education 31 (1): 95–130.
Nivet, M. A. 2011. Commentary: Diversity 3.0: A necessary systems upgrade. Aca-
demic Medicine 86 (12): 1487.
Oakes, J., A. Wells, M. Jones, and A. Datnow. 1997. Detracking: The social con-
struction of ability, cultural politics, and resistance to reform. Teachers Col-
lege Record 98 (3): 482–510.
———. 2008. The social construction of ability, cultural politics, and resistance
to reform. In C. Grant and T. K. Chapman, eds., History of Multicultural
Education, 2:293–315. New York: Routledge.
Olivas, M. A. 1994. The education of Latino lawyers: An essay on crop cultivation.
Chicano-Latino Law Review 14:117–138.
———. 2009. What the “war on terror” has meant for U.S. colleges and universi-
ties. In R. G. Ehrenberg and C. V. Kuh, eds., Doctoral education and the fac-
ulty of the future, 249–258. Ithaca, NY: Cornell University Press.
References 231
Oltman, P. K., and R. T. Hartnett. 1984. The role of GRE General and Subject test
scores in graduate school admission. GRE Board Research Report GREB.
http://origin-www.ets.org/Media/Research/pdf/RR-84–14-Oltman.pdf.
Orfield, Gary, ed. 2001. Diversity challenged: Legal crisis and new evidence. Cam-
bridge, MA: Harvard Education Publishing Group.
Ortiz, A. 2003. The ethnographic interview. In F. K. Stage and K. Manning, eds.,
Research in the college context: Approaches and methods, 35–48. New York:
Brunner-Routledge.
Osajima, K. 2005. Asian Americans as the model minority: An analysis of the pop-
ular press image in the 1960s and 1980s. In K. Ono, ed., A companion to Asian
American studies, 215–225. Malden, MA: Blackwell.
Osei-Kofi, N., L. E. Torres, and J. Lui. 2013. Practices of whiteness: Racialization
in college admissions viewbooks. Race Ethnicity and Education 16 (3):
386–405.
Oyserman, D., K. Elmore, and G. Smith. 2012. Self, self-concept, and identity.
In M. R. Leary and J. P. Tangney, eds., Handbook of self and identity, 2nd
ed., 69–104. New York: Guilford Press.
Özturgut, O. 2011. Standardized testing in the case of China and the lessons to be
learned for the U.S. Journal of International Education Research 7 (2).
Page, S. E. 2008. The difference: How the power of diversity creates better groups,
firms, schools, and societies. Princeton, NJ: Princeton University Press.
Pallas, A. M. 2001. Preparing education doctoral students for epistemological di-
versity. Educational Researcher 30 (5): 6–11.
Pappano, L. 2011. The master’s as the new bachelor’s. New York Times, July 22.
Park, J. J. 2013. When diversity drops: Race, religion, and affirmative action in
higher education. Newark, NJ: Rutgers University Press.
Patel, V. 2014. Why so few American Indians earn Ph.D.’s, and what can be done
about it. Chronicle of Higher Education, May 27. http://chronicle.com/article
/Why-So-Few-American-Indians/146715/.
Patton, E. and S. H. Appelbaum, S. H. 2003. The case for case studies in manage-
ment research. Management Research News 26(5): 60–71.
Patton, M. Q. 2002. Qualitative research and evaluation methods. Thousand Oaks,
CA: Sage.
Patton, S. 2013. At the Ivies, it’s still white at the top. Chronicle of Higher Educa-
tion, June 9.
Pennock-Roman, M. 1986. Fairness in the use of tests for selective admissions of
Hispanics. In M. A. Olivas (Ed.), Latino College Students (pp. 246–280). NY:
Teachers College Press.
Peterson, M. W., and M. G. Spencer. 1990. Assessing academic climates and cul-
tures. New Directions for Institutional Research, no. 68: 3–18.
Pfeffer, J. 1993. Barriers to the advance of organizational science: Paradigm devel-
opment as a dependent variable. Academy of Management Review 18 (4):
599–620.
Pfeffer, J., and G. R. Salancik. 1978. The external control of organizations: A re-
source dependency perspective. New York: Harper and Row.
232 References
Pfeifer Jr., C. M., and W. E. Sedlacek. 1971. The validity of academic predictors for
black and white students at a predominantly white university. Journal of Edu-
cational Measurement 8 (4): 253–261.
Popper, S. K. 1959. The logic of scientific discovery. Fayetteville, AR: Hutchinson.
Porter, T. 1996. Trust in numbers: Pursuit of objectivity in science and public life.
Portes, A. 2003. Social capital: Its origins and applications in modern sociology.
Annual Review of Sociology 24:1–24.
Posselt, J. R. 2014. Toward inclusive excellence in graduate education: Constructing
merit and diversity in Ph.D. admissions. American Journal of Education 120
(4): 481–514.
Posselt, J. R., and K. Black. 2012. Developing the research identities and aspi-
rations of first generation college students: Evidence from the McNair
Scholars Program. International Journal of Researcher Development 3 (1):
26–48.
Posselt, J. R., O. Jaquette, R. Bielby, and M. N. Bastedo. 2012. Access without eq-
uity: Longitudinal analyses of institutional stratification by race and ethnicity,
1972–2004. American Educational Research Journal 49 (6): 1074–1111.
Powell, K. 2013. Higher education: On the lookout for true grit. Nature, no. 504:
471–473. http://www.nature.com/naturejobs/science/articles/10.1038/nj7480
–471a.
Powell, W. W. 1985. Getting into print: The decision-making process in scholarly
publishing. Chicago: University of Chicago Press.
Powers, D. E., and S. S. Swinton. 1981. Extending the measurement of graduate
admission abilities beyond the verbal and quantitative domains. Applied Psy-
chological Measurement 5 (2): 141–158.
Poyrazli, S., and M. D. Lopez. 2007. An exploratory study of perceived discrimina-
tion and homesickness: A comparison of international students and American
students. Journal of Psychology 141 (3): 263–280.
Rao, H., and S. Giorgi. 2006. Code breaking: How entrepreneurs exploit cultural
logics to generate institutional change. Research in Organizational Behavior:
An Annual Series of Analytical Essays and Critical Reviews 27:269–304.
Rao, H., P. Monin, and R. Durand. 2005. Border crossing: Bricolage and the ero-
sion of categorical boundaries in French gastronomy. American Sociological
Review 70 (6): 968–991.
Redden, E. (2013, July 12). New report shows dependence of U.S. graduate pro-
grams on foreign students. Inside Higher Ed. http://www.insidehighered.com
/news/2013/07/12/new-report-shows-dependence-us-graduate-programs-forei
gn-students.
Regents of the University of California v. Bakke. 1978. 438 U.S. 265.
Rivera, L. A. 2011a. Go with your gut: A theory of emotion and inequality in hiring.
In Academy of Management Proceedings, 1–2. http://proceedings.aom.org
/content/2011/1/1.262.short.
———. 2011b. Ivies, extracurriculars, and exclusion: Elite employers’ use of edu-
cational credentials. Research in Social Stratification and Mobility 29 (1):
71–90.
References 233
———. 2012a. Diversity within reach: Recruitment versus hiring in elite firms. An-
nals of the American Academy of Political and Social Science 639 (1): 71–90.
———. 2012b. Hiring as cultural matching: The case of elite professional service
firms. American Sociological Review 77 (6): 999–1022.
Rizzo, M., and R. G. Ehrenberg. 2004. Resident and nonresident tuition and
enrollment at flagship state universities. In C. M. Hoxby, ed., College choices:
The economics of where to go, when to go, and how to pay for it, 303–354.
Chicago: University of Chicago Press.
Roach, R. 2005. Affirmative action fallout. Black Issues in Higher Education 22
(11): 24–27.
———. 2013. U.S. international education enrollment reaches all-time high. Diverse
Issues in Higher Education, November 11. http://diverseeducation.com/article
/57394/#.
Robins, R. H. 1997. A short history of linguistics, 3rd ed. Harlow, UK: Longman.
Rogers, M. R., and L. E. Molina. 2006. Exemplary efforts in psychology to recruit
and retain graduate students of color. American Psychologist 61 (2): 143.
Roithmayr, D. 2014. Reproducing racism: How everyday choices lock in white ad-
vantage. New York: NYU Press.
Rojas, F. 2007. From black power to black studies: How a radical social movement
became an academic discipline. Baltimore: Johns Hopkins University Press.
Rojstaczer, S., and C. Healy. 2012. Where A is ordinary: The evolution of American
college and university grading, 1940–2009. Teachers College Record 114 (7).
Room for debate: The fate of the humanities. 2013. New York Times, November 4,
http:// www. nytimes . com / roomfordebate / 2013 / 11 / 04 / the - fate - of - the
-humanities.
Rubin, H. J., and I. Rubin. 2009. Qualitative interviewing: The art of hearing data,
2nd ed. Thousand Oaks, CA: Sage.
Ruscingno, G. 2006. Admission variables and academic success in a doctor of physical
therapy program. Unpublished doctoral dissertation, Seton Hall University.
Russell, B. 1972. The history of Western philosophy. New York: Simon and Schuster.
Russell, S., M. Hancock, and J. McCullough. 2007. Benefits of undergraduate re-
search experiences. Science 316:548–549.
Ryfe, D. M. 2002. The practice of deliberative democracy: A study of 16 delibera-
tive organizations. Political Communication 19 (3): 359–377.
Salthouse, T., W. McKeachie, and Y. Lin. 1978. An experimental investigation of
factors affecting university promotion decision: A brief report. Journal of
Sauder, M., F. Lynn, and J. M. Podolny. 2012. Status: Insights from organizational
sociology. Annual Review of Sociology 38:267–283.
Scheurich, J. J. 1995. A postmodernist critique of research interviewing. Interna-
tional Journal of Qualitative Studies in Education 8 (3): 239–252.
Schillebeeckx, M., Maricque, B., & Lewis, C. 2013. The missing piece to changing
the university culture. Nature biotechnology 31(10): 938–941.
Schmidt, F. L., and J. Hunter. 2004. General mental ability in the world of work:
Occupational attainment and job performance. Journal of Personality and So-
cial Psychology 86 (1): 162.
234 References
Schmidt, F. L., R. H. Johnson, and J. F. Gugel. 1978. Utility of policy capturing as

an approach to graduate admissions decision making. Applied Psychological
Measurement 2 (3): 347–359.
Schmidt, P. 2010. Graduate school applications spiked after economy’s plunge.
Chronicle of Higher Education, September 14. http://chronicle.com/article
/Graduate-School-Applications/124387/.
Schmitt, M. T., R. Spears, and N. R. Branscombe. 2003. Constructing a minority
group identity out of shared rejection: The case of international students. Eu-
ropean Journal of Social Psychology 33 (1): 1–12.
Schoem, D. L., and S. Hurtado, eds. 2001. Intergroup dialogue: Deliberative de-
mocracy in school, college, community, and workplace. Ann Arbor: University
of Michigan Press.
Schofer, E., and J. W. Meyer. 2005. The worldwide expansion of higher education
in the twentieth century. American Sociological Review 70 (6): 898–920.
Scott, W. R. 2008. Institutions and organizations: Ideas and interests. Thousand
Oaks, CA: Sage.
Sedlacek, W. E. 2004. Beyond the big test. San Francisco: Jossey-Bass.
Sellers, R. M., M. A. Smith, J. N. Shelton, S. A. J. Rowley, and T. M. Chavous. 1998.
Multidimensional model of racial identity: A reconceptualization of African
American racial identity. Personality and Social Psychology Review 2 (1): 18–39.
Skousen, M. 2009. The making of modern economics, vol. 2. Armonk, NY: M. E.
Sharpe.
Smith v. City of Jackson. 2005. (03-1160) 544 U.S. 228.
Smith, D. G. 2009. Diversity’s promise for higher education: Making it work. Bal-
timore: Johns Hopkins University Press.
Smith, D. G., L. E. Wolf, B. Busenberg, and Associates. 1996. Achieving faculty
diversity: Debunking the myths. Washington, DC: Association of American
Colleges and Universities.
Smith, L. M. 1978. An evolving logic of participant observation, educational ethnog-
raphy, and other case studies. Review of Research in Education 6:316–377.
Snow, C. P. 1959. The two cultures and the scientific revolution. London: Cam-
bridge University Press.
Spence, M. 1973. Job market signaling. Quarterly Journal of Economics 87 (3):
355–374.
Spillman, L. 2014. Mixed methods and the logic of qualitative inference. Qualita-
tive Sociology 37(2): 189–205.
Stake, R. E. 2005. Qualitative case studies. In N. K. Denzin and Y. S. Lincoln, eds.,
The Sage handbook of qualitative research, 3rd ed., 443–466. Thousand Oaks,
CA: Sage.
Stampnitzky, L. 2006. How does “culture” become “capital”? Cultural and institu-
tional struggles over “character and personality” at Harvard. Sociological
Perspectives 49 (4): 461–481.
Stanton-Salazar, R. D. 1997. A social capital framework for understanding the so-
cialization of racial minority children and youths. Harvard Educational Re-
view 67 (1): 1–40.
References 235
Steele, C. M. 2010. Whistling Vivaldi: How stereotypes affect us and what we can
do about it. New York: Norton.
Steele, C. M., and J. Aronson. 1995. Stereotype threat and the intellectual test per-
formance of African Americans. Journal of Personality and Social Psychology
69 (5): 797.
Stephens, N. M., S. A. Fryberg, H. R. Markus, C. S. Johnson, and R. Covarrubias.
2012. Unseen disadvantage: How American universities’ focus on independence
undermines the academic performance of first-generation college students.
Journal of Personality and Social Psychology 102 (6): 1178.
Sternberg, R. J. 2003. Wisdom, intelligence, and creativity synthesized. Cambridge:
Sternberg, R. J., and W. M. Williams. 1997. Does the Graduate Record Examina-
tion predict meaningful success in the graduate training of psychologists? A
case study. American Psychologist 52 (6): 630–641.
Stevens, M. 2007. Creating a class: College admissions and the education of elites.
Cambridge, MA: Harvard University Press.
Stevens, M. L., and J. Roksa. 2011. The diversity imperative in elite admissions.
In L. M. Stulberg and S. L. Weinburg, eds., Diversity in American higher
education: Toward a more comprehensive approach, 63–73. New York:
Routledge.
Stock, W. A., A. T. Finnegan, and J. J. Siegfried. 2006. Attrition in economics Ph.D.
programs. American Economic Review 96 (2): 458–466.
Strauss, A., and J. Corbin. 2008. Basics of qualitative research: Techniques and pro-
cedures for developing grounded theory. Thousand Oaks, CA: Sage.
Studley, R. E. 2004. Inequality, student achievement, and college admissions: A
remedy for underrepresentation. In R. Zwick, ed., Rethinking the SAT: The fu-
ture of standardized testing in college admissions, 321–344. New York:
Routledge.
Suen, H. K., and L. Yu. 2006. Chronic consequences of high stakes testing? Les-
sons from the Chinese civil service exam. Comparative Education Review 50
(1): 46–65.
Suzuki, B. H. 2002. Revisiting the model minority stereotype: Implications for stu-
dent affairs practice and higher education. New Directions for Student Services,
no. 97: 21–32.
Swales, J. 2004. Research genres: Explorations and applications. Stuttgart: Ernst
Klett Sprachen.
Sweitzer, K., and J. F. Volkwein. 2009. Prestige among graduate and professional
schools: Comparing the US News’ graduate school reputation ratings between
disciplines. Research in Higher Education 50 (8): 812–836.
Swidler, A. 1986. Culture in action: Symbols and strategies. American Sociological
Review 51 (2): 273–286.
Swidler, A., and J. Arditi. 1994. The new sociology of knowledge. Annual Review
of Sociology 20 (1): 305–329.
Tajfel, H., ed. 2010. Social identity and intergroup relations, vol. 7. Cambridge:
236 References
Tajfel, H., and J. Turner. 1986. The social identity theory of intergroup behavior.
In J. T. Jest and J. Sidanius, eds., Psychology of intergroup relations, 7–24. Chi-
cago: Nelson Hall.
Taylor, C. R., and B. B. Stern. 1997. Asian-Americans: Television advertising and
the “model minority” stereotype. Journal of Advertising 26 (2): 47–61.
Temp, G. 1971. Validity of the SAT for blacks and whites in thirteen integrated in-
stitutions. Journal of Educational Measurement 8 (4): 245–251.
Teranishi, R. T., L. B. Behringer, E. A. Grey, and T. L. Parker. 2009. Critical race theory
and research on Asian Americans and Pacific Islanders in higher education. New
Directions for Institutional Research, no. 142: 57–68.
Thelin, J. R. 2004. A history of American higher education. Baltimore: Johns Hop-
kins University Press.
Thompson, J. 2013. The changing demography of wealth: Demographics by wealth
and wealth by demographics in the SCF. Washington, DC: Federal Reserve
Board of Governors.
Thorndike, R. L. 1949. Personnel selection: Test and measurement techniques. New
York: Wiley.
Thornton, P. H. 2004. Markets from culture: Institutional logics and organizational
decisions in higher education publishing. Stanford, CA: Stanford Business Books.
Thornton, P. H., and W. Ocasio. 1999. Institutional logics and the historical contin-
gency of power in organizations. American Journal of Sociology 105 (3):
801–843.
———. 2008. Institutional logics. In R. Greenwood, C. Oliver, R. Suddaby, and
K. Sahlin-Anderson, eds., The Sage handbook of organizational institution-
alism, 99–129. Thousand Oaks, CA: Sage.
Thucydides. 1951. The history of the Peloponnesian War. Translated by R. Crawley.
New York: Random House.
Thurgood, L., M. J. Golladay, and S. T. Hill. 2006. U.S. doctorates in the 20th cen-
tury: Special report. NSF 06-319. http://www.nsf.gov/statistics/nsf06319/pdf
/nsf06319.pdf.
Tienda, M. 2013. Diversity ≠ inclusion: Promoting integration in higher education.
Educational Researcher 42 (9): 467–475.
Tierney, W. G. 1988. Organizational culture in higher education: Defining the es-
sentials. Journal of Higher Education 59 (1): 2–21.
Tierney, W. G., and E. M. Bensimon. 1996. Promotion and tenure: Culture and so-
cialization in academe. Albany: SUNY Press.
Tilly, C. 1998. Durable inequality. Berkeley: University of California Press.
———. 2005. Social movements, 1768–2004. Boulder: Paradigm.
Torche, F. 2011. Is a college degree still the great equalizer? Intergenerational mo-
bility across levels of schooling in the United States. American Journal of Soci-
ology 117 (3): 763–807.
Tough, P. 2012. How children succeed: Grit, curiosity, and the hidden power of char-
acter. New York: Houghton Mifflin.
Trice, A. G. 2004. Mixing it up: International graduate students’ social interactions
with American students. Journal of College Student Development 45 (6):
671–687.
References 237
Trower, C. A. 2009. Toward a greater understanding of the tenure track for mi-
norities. Change 41 (5): 38–45.
Trower, C. A., and R. P. Chait. 2002. Faculty diversity: Too little for too long. Har-
vard Magazine, March/April.
Tumminia, D. 1992. Hard work, luck, and uncontrollable variables: A guide to get-
ting into graduate school. Sociological Practice Review 3 (1): 37–41.
Twombly, S. 1992. The process of choosing a dean. Journal of Higher Education
63 (6): 653–683.
Urrieta, L. 2007. Orchestrating habitus and figured worlds: Chicana/o educational
mobility and social class. In J. A. VanGalen and G. W. Noblit, eds., Late to
class: Social class and schooling in the new economy, 113–140. Albany: SUNY
Press.
U.S. Census Bureau. 2008. 2008 American Community Survey 1-year estimates.
Washington, DC: U.S. Census Bureau.
U.S. Department of Education, National Center for Education Statistics. 2012. The
condition of education 2012. NCES 2012-045. Indicator 47. http://nces.ed.gov
/fastfacts/display.asp?id=72.
Vaughan, D. 1992. Theory elaboration: The heuristics of case analysis. In C. C.
Ragin and H. S. Becker, eds., What is a case? Exploring the foundations of so-
cial inquiry, 173–203. Cambridge: Cambridge University Press.
Wadsworth, B. C., M. L. Hecht, and E. Jung. 2008. The role of identity gaps, dis-
crimination, and acculturation in international students’ educational satisfac-
tion in American classrooms. Communication Education 57 (1): 64–87.
Wait, I. W., and J. W. Gressel. 2009. Relationship between TOEFL score and aca-
demic success for international engineering students. Journal of Engineering
Education 98 (4): 389–398.
Walker, G. E., C. M. Golde, L. Jones, A. C. Bueschel, and P. Hutchings. 2008. The
formation of scholars: Rethinking doctoral education for the twenty-first cen-
tury. San Francisco: Jossey-Bass.
Walters, A. M., S. Lee, and C. Trapani. 2004. Stereotype threat, the test-center en-
vironment, and performance on the GRE General Test. Princeton, NJ: Educa-
tional Testing Service.
Walton, G. M., and G. L. Cohen. 2003. Stereotype lift. Journal of Experimental
Social Psychology 39 (5): 456–467.
Weber, M. 1978a. Economy and society, vol. 1. Berkeley: University of California
Press.
———. 1978b. Economy and society, vol. 2. Berkeley: University of California Press.
Wechsler, H. S. 1977. The qualified student. New York: Wiley.
Wei, M., T. Y. Ku, D. W. Russell, B. Mallinckrodt, and K. Y. H. Liao. 2008.
Moderating effects of three coping strategies and self-esteem on perceived dis-
crimination and depressive symptoms: A minority stress model for Asian in-
ternational students. Journal of Counseling Psychology 55 (4): 451.
Weick, K. E. 1979. Cognitive processes in organizations. Research in Organizational
Behavior 1 (1): 41–74.
———. 1984. Small wins: Redefining the scale of social problems. American Psy-
chologist 39 (1): 40.
238 References
———. 1998. Improvisation as a mindset for organizational analysis. Organiza-

tion Science 9 (5): 543–555.
Weidman, J. C., and E. L. Stein. 2003. Socialization of doctoral students to aca-
demic norms. Research in Higher Education 44 (6): 641–656.
Weiss, R. S. 1995. Learning from strangers: The art and method of qualitative in-
terview studies. New York: Free Press.
Weitzman, R. A. 1972. It is time to re-reconsider the GRE: A reply to Marston.
American Psychologist 27 (3): 236–238.
Wendler, C., and B. Bridgeman, eds. 2014. The Research Foundation for the GRE
revised General Test: A compendium of studies. Princeton, NJ: Educational
Testing Service. http://www.ets.org/s/research/pdf/gre_compendium.pdf.
Wendler, C., B. Bridgeman, F. Cline, C. Millett, J. Rock, N. Bell, and P. McAllister.
2010. The path forward: The future of graduate education in the United States.
Princeton, NJ: Educational Testing Service.
Wilkerson, T. 2007. The relationship between admissions credentials and the suc-
cess of students admitted to a physics doctoral program. Unpublished doctoral
dissertation, University of Central Florida.
Williams, J. D. 1969. Judgment analysis for assessing doctoral admission policies.
Journal of Experimental Education 38 (2): 92–96.
Willingham, W. W. 1974. Predicting success in graduate education. Science 183
(4122): 273–278.
Winkle-Wagner, R., and A. M. Locks. 2013. Diversity and inclusion on campus:
Supporting racially and ethnically underrepresented students. New York:
Routledge.
Wilson, J. 1989. Bureaucracy: What government agencies do and why they do it.
New York: Basic Books.
Wolfers, J. 2015. How economists came to dominate the conversation. New York
Times, January 24. http://www.nytimes.com/2015/01/24/upshot/how-econo
mists-came-to-dominate-the-conversation.html.
Worthington, R. L., R. L., Navarro, M. Loewy, and J. Hart. 2008. Color-blind ra-
cial attitudes, social dominance orientation, racial-ethnic group membership
and college students’ perceptions of campus climate. Journal of Diversity in
Higher Education 1(1): 8–19.
Wulff, D. H., A. E. Austin, J. D. Nyquist, and J. Sprague. 2004. Paths to the profes-
soriate: Strategies for enriching the preparation of future faculty. San Fran-
cisco: Jossey-Bass.
Xu, X. 2001. Chinese ETS test results doubted. China Daily, February 21.
Xueqin, J. 2011. Selecting the right Chinese students. Chronicle of Higher Educa-
tion, November 3.
Yang, S. Y., and R. J. Sternberg. 1997. Taiwanese Chinese people’s conceptions of
intelligence. Intelligence 25 (1): 21–36.
Yin, R. K. 1994. Case study research: Design and methods, 4th ed. Thousand Oaks,
CA: Sage.
Yosso, T. J. 2006. Whose culture has capital? A critical race theory discussion of
community cultural wealth. Race, Ethnicity, and Education 8 (1): 69–91.
References 239
Young, J. W., and J. L. Kobrin. 2001. Differential validity, differential prediction,

and college admission testing: A comprehensive review and analysis. College
Board Research Report No. 2001-6. New York: College Entrance Examina-
tion Board.
Yule, G., and P. Hoffman. 1990. Predicting success for international teaching as-
sistants in a US university. TESOL Quarterly 24 (2): 227–243.
Zald, M. N. 1971. Occupations and organizations in American society: The
organization-dominated man? Cambridgeshire, UK: Markham.
Zerubavel, E. 1993. The fine line. New York: Free Press.
Zhang, L. 2009. Do foreign doctorate recipients displace U.S. doctorate recipients
at U.S. universities? In R. G. Ehrenberg and C. V. Kuh, eds., Doctoral educa-
tion and the faculty of the future, 209–223. Ithaca, NY: Cornell University
Press.
Zhang, Q. 2010. Asian Americans beyond the model minority stereotype: The nerdy
and the left out. Journal of International and Intercultural Communication 3
(1): 20–37.
Zhao, Y. 2001. Preparing early for the first of the all-important tests. New York
Times, October 10.
Zimmerman, M. A., and G. J. Zeitz. 2002. Beyond survival: Achieving new ven-
ture growth by building legitimacy. Academy of Management Review 27 (3):
414–431.
Zucker, L. G. 1977. The role of institutionalization in cultural persistence. Amer-
ican Sociological Review 42 (5): 726–743.
Zwick, R. 2002. Fair game? The use of standardized admissions tests in higher edu-
cation. New York: Routledge.
———. 2016. Who gets in? Strategies for fair and effective college admissions. Cam-
bridge, MA: Harvard University Press.
Acknowledgments
The trust that enables prospective research participants to become forthcoming

members of a study is a foundation of social inquiry, and I am grateful to the par-
ticipants in this study for sharing with me their experience of admissions work and
for helping me to see it from their perspectives. I extend special thanks to the de-
partment leaders and admissions committees who allowed me into the protected
space of their admissions meetings. My observations of those meetings have added
more richness to my findings than scores of additional interviews could have
provided.
This project was also made possible through financial support from a few sources.
The Center for Public Policy in Diverse Societies and the National Association of
Graduate Admissions Professionals awarded grants that supported the fieldwork
when the project was just an idea. I was thankful as well to receive a fellowship
from the Rackham Graduate School, which afforded a level of focus that resulted
in deeper analysis and better writing than would otherwise have been possible. I
hope the quality of the outcome is commensurate with my gratitude for each of
these groups’ support.
I was delighted to connect with an academic press that shared my vision of dis-
seminating this research to graduate education stakeholders outside the higher ed-
ucation and sociology circles for whom I usually write. Harvard University Press
has been outstanding to work with from day one, and I am indebted in particular
to Andrew Kinney and Elizabeth Knoll, who both served as editors. Elizabeth ac-
quired the project, encouraged me to add the chapter on evaluation of international
students, and provided sage guidance on each chapter in the manuscript. I looked
forward to her manila envelopes, with official feedback accompanied by postcards
242 Acknowledgments
with witty remarks. After she transitioned out of HUP, Andrew’s editorial skill and
eye for detail also contributed to the strength of the manuscript, and I’m grateful for
his steady support through the publication process. I would be remiss not to acknowl-
edge my appreciation to the two anonymous external reviewers, both of whom of-
fered constructive and thoughtful advice that helped me improve the manuscript.
Chapter 4 develops ideas originally treated in “Disciplinary Logics in Doctoral Ad-
missions,” published in the Journal of Higher Education.
Outstanding research and editorial assistance from University of Michigan doc-
toral students Kimberly Reyes, Kelly Slay, and Aurora Kamimura also supported the
production of the book. I am particularly appreciative of Kim’s dedicated editorial
support with early chapter drafts. I also want to acknowledge my transcriptionist,
Mary Ann Spitale, whose efficient, accurate work increases my research capacity.
I have been graced with several wise mentors whose counsel and support have
been instrumental as I worked on this project. This book would not exist without
Mike Bastedo’s generous mentoring through the research and publication processes.
I am deeply grateful to Deborah Carter for her wisdom, wit, and for all that I learned
from her about qualitative research. Janet Lawrence and Janet Weiss offered astute
interpretations of my early findings, and they helped me think about my data in
new ways. I am also indebted to Annemarie Palincsar, Lisa Lattuca, Diana Hess,
Chris Golde, Liliana Garces, and Tabbye Chavous for professional guidance and
review of specific chapters.
For conversations and insights that strengthened the design, findings, and/or mes-
sage of this work, I extend my sincere thanks to Elizabeth Armstrong, Ann Austin,
Deborah Ball, Estela Bensimon, John Bound, Phil Bowman, Bryan Brayboy, Lenny
Cassuto, Steve DesJardins, Kimberly Griffin, Eric Grodsky, Patti Gumport, Ozan
Jaquette, Michèle Lamont, Anna Neumann, Gary Rhoades, Ed St. John, Mitchell
Stevens, Barrett Taylor, and Rebecca Zwick.
Through the years I have conducted research on graduate admissions—the long
haul of design, recruitment, fieldwork, analysis, writing, and revising—scholarly
friendships with others who study higher education have been a great source of sus-
tenance. I especially want to acknowledge Cassie Barnhardt, Nick Bowman, Julio
Cardona Raya, Gloryvee Fonseca, Leslie Gonzales, Nathan Harris, Matt Holsapple,
Ignacio Hernandez, Susana Hernandez, Marc Johnston, Anat Levtov, Chris Linder,
Angela Locks, Johanna Masse, Matt Mayhew, Carmen McCallum, Genevieve
Negron-Gonzales, Penny Pasque, Rosie Perez, Awilda Rodriguez, Shelley Strickland,
and Kerri Wakefield.
Finally, my family. Blessings to each and every one of you. Special thanks to my
parents, Janet and Gary Schmidt, and to Derek and Daniel Posselt, my dear hus-
band and son. Thank you for being my foundation, and for your strong support of
my work and its demands. Derek and Danny boy: I absolutely love our life together,
and I dedicate this book to you.
Index
Abbott, Andrew, 203n19 Asians, 149, 153, 155, 163, 170, 197, 108,
Academic preparation, 67–70, 78, 156, 162 197, 208n22; Asian international
Access, 2, 6, 7, 11, 15, 20, 67, 102–103, students, 17, 61, 112, 137, 139, 140, 141,
111, 127, 134, 155–159, 170, 210n14 142, 148–150, 152, 163, 208; East
Accountability, 41 Asians, 112, 137, 138, 148, 150–151,
ACT, 205n9 163; South Asians, 138; Southeast Asians,
Administrative staff, 1, 29, 44 47
Affirmative action. See Race-conscious Association of American Colleges and
admissions Universities, 48, 172
African Americans, 3, 4, 159, 189, 200n13, Astrophysics, 2; admissions committee
212n39; baccalaureate origins of, dynamics observed in, 32–33, 35, 41, 43,
169–170; and compositional diversity, 18, 44; and admissions interviews, 143–146;
61, 135; evidence of bias against, 12, 66, assessments of merit in, 46–47, 50, 51,
163; and GRE scores, 47, 132, 141, 152, 56, 58, 147; description of departments
155, 212n39; and stereotype threat, 132, included in study, 16, 28, 180, 182, 184,
152; and undergraduate research 189; discussion of gender, 65; evaluation
programs, 171; views regarding of GRE scores in, 140; evidence of
underrepresentation of, 60, 63–64, 72 homophily in, 97, 98, 99, 100, 105–106;
Ally behavior. See Solidarity and risk aversion, 101, 103–104, 128,
Ambiguity, 55, 114, 125, 133–153, 156, 165
207nn11–12
Ambiguity aversion, 162 Bastedo, Michael, 199n25, 204n15
American Philosophical Association, 175 Becher, Tony, 117
American Psychological Association, 194n20 Belonging, 53–54, 161
Aronson, Joshua, 132, 206nn22–23 Benard, Stephen, 204n20
Asian Americans, 3, 4, 12, 148–149, 155, Bensimon, Estela, 72, 144, 186, 196,
208n23, 208n27; baccalaureate origins 208n17
of, 169–170 Berrey, Ellen, 72
244 Index
Bias, 102, 123, 137; cognitive, 163–164 Classification, 174–175

Biglan discipline classification, 79 Climate, 211n35
Biology, 2, 133, 205n9; admissions Cohort design, 136, 157, 200
committee dynamics in, 31, 54, 58, Cole, Bruce, 205n13
199n23; and admission interviews, 143; Collegiality, 9, 18, 20–22, 36–39, 41, 44–45,
assessments of merit and intelligence, 117, 162, 164, 166
121, 124, 126; description of departments Collins, Randall, 81, 195n23
in study, 14, 16, 24, 180, 182, 189; Colorblind, 12, 166, 210
evidence of homophily in, 97, 103, 112 Colorblind racism, 197n49, 210n25
Birnbaum, Robert, 49 Columbia University, 173
Blacks. See African Americans Commensuration, 31, 198n20
Boltanski, Luc, 157 Committee, formation and composition, 28,
Boundaries, 43, 49, 114, 129, 158, 210n8; 99, 114, 164; based selection, 23, 34–35,
admissions as boundary work, 73 157, 198, 203n4; dynamics, 41–42, 44,
Bourdieu, Pierre, 101, 158, 204n8, 205, 130, 143–146, 198n16; critical mass
210n8 within, 96, 99, 110–112
Bridge programs, 211n30 Competency, 78, 113, 113, 116, 156, 193n4
Brigham, Carl Campbell, 205n7 Competition, 6, 7, 122, 209n4; academic
Brilliance. See Intelligence status competition, 39, 49, 63, 158; for
British universities, 100–103 admission, 35, 91, 102; among doctoral
Bureaucracy, 18, 20–22, 28, 30, 31, 34, 55; programs, 39, 63, 100; among doctoral
bureaucratic efficiency, 44; bureaucratiza- students, 124, 140; competing demands,
tion, 31, 36; deliberative bureaucracy, 17, interests, values, 21, 36, 84, 87, 92, 182;
20, 21, 39–40, 43–45, 47, 75, 162, 164, for doctoral students, 64
187 Compromise, 36–38, 157, 161, 164
Confirmation bias, 18
Carnegie, Classification of Institutions of Conflict aversion, 164
Higher Education, 193n3; Initiative on Consensus, 34–35, 45, 84–94, 174
the Doctorate, 57, 202n2 Convenience, 54–55, 162
Cassuto Leonard, 11 Conventional achievement, 50, 54, 55, 62,
Chauncey, Henry, 31 135–136, 150
Cheating, suspicions of, 137–142, 145, 149, Cool, homophily based on, 96–97,
151–153, 208n34 104–106, 156, 204n5, 204n12
Chemistry, 117, 205n9 Correll, Shelley, 204n20
Cheslock, John, 134 Council of Graduate Schools, 194n6
China, 9, 10, 12, 17, 29, 61, 134, 137–140, Counterscripts, 17, 96–97, 110, 114,
142–143; applications from, 133–134, 174–175
207n3; evaluating applicants from, Courant, Paul, 175
137–153; test preparation, 137–141, 148, Creativity, 48, 57–59, 119, 128, 130, 140,
155, 207n14 151, 161, 210, 212
Civil Rights Act, 159 Credential Inflation, 2, 6, 53
Civil Rights Movement, 159 Crocker, Jennifer, 116
Civil Rights Project, 160 Cultural capital, 101, 103, 171, 204n8
Clark, Burton, 23, 196n37 Curriculum, 52, 193n3
Classics, 2, 100–104, 143, 154, 202n13;
admissions committee dynamics observed Decision making, 71, 157, 176, 206n25; in
in, 85, 156; conceptualizations of the context of ambiguity, 148, 162–164,
potential and diversity, 57, 64–65; 173; and diversity considerations, 62, 64,
description of departments included in 72, 166; and interpretations of student
study, 16, 23, 25, 180, 182; evidence of research experience, 126, 171; faculty
homophily in, 97, 100–104, 106, 112; aversions in, 162–164, 166; in high-status
judgments of intelligence in, 123–124, organizations, 49, 163; methodological
156 approaches to studying decision making,
Index 245
180, 182, 187; nature of, 20–22, 39, interest, 67, 160; and conceptions of
44–45; as part of the admissions process, merit, 47–48, 50, 71; definition of, 59–60,
18, 22–23, 29, 195n23, 198n11; through 62, 67, 157, 196n40, 200n24; as a
quantification, 31, 38, 40, 55, 71, dimension of quality and merit, 56, 172;
195n29, 199n23; relationship to and enrollment management, 23, 39, 171,
disciplinary logics, 75–77, 87–88, 93; and 212n43; faculty commitment to, 92, 164,
risk taking, 53, 67, 162–163; through 177; fellowships, 66, 67, 89, 136, 209n4;
serial impressions, 84–86 geographic, 61, 135–136, 201n28;
Decision theory, 162–164 graduate diversity officers, 39, 169;
Delegation, 21–23, 28–30, 37–38, 44 intellectual, 49, 203n19; as obligation,
Deliberation, 39–43, 68–70, 143–146, 156, 62–63; as opportunity, 63–65; as
161–162; about borderline cases, 40, 111; organizational interest, 9, 114–115,
disciplinary logics reflected in, 84–94; 159–161, 200–201n26; as part of
discussions of intelligence during, 199, admissions review, 18, 22, 55, 72, 155,
123–124, 128, 130; nature of, 37, 41, 43, 157–158; on research teams, 210n18, 40,
79, 112–113, 131; value of, 37, 39, 45 47, 72; and risk aversion, 67–70; as a
Deliberative democracy, 18, 20, 40, 44, 47, value, 92, 157–158
162, 164, 197n2, 206n25 Doctoral education, 3–9, 14, 200n17,
Democracy, 206n18 202n2, 210n8; and academic depart-
Department, 8, 11, 54, 115, 145, 150; ments, 23, 198n8; diversity consider-
description of departments in study, ations for, 155, 161, 170–172, 177;
24–27; political dynamics within, 47–48; relationship to disciplines, 74; and risk
relationship to admissions committees, aversion, 70; and student research
23, 28–30 experience, 58, 127, 171
Disagreement, 21, 29–30, 34–38, 117, 150; Dowd, Alicia, 72
caused by homophily, 98; disguised
through ratings of applicants, 34–35, Economics, 2, 4, 32, 57, 200n13, 202n7;
40–41; over diversity considerations, 59, admissions practices and priorities in,
62; as part of academic life, 43–45; 75–78, 82–84, 93; beliefs about diversity
relationship to disciplinary consensus, 75, in, 60–61, 64, 135; description of
82, 86–91, 93; and status hierarchies, 29, departments included in study, 14, 16, 27;
93; over use of GRE scores, 89–91, 93 evaluation of GRE scores in, 50, 142,
Disciplines, 7–9, 13, 101, 157–158, 177, 195n30; evidence of homophily in, 97,
212n45; classification, 202n5; construc- 106; interpretations of intelligence, 125,
tions of merit in, 50–52, 56–59; 129
disciplinary cultures, 117–118, 135, Educational Testing Service, 71, 130–131,
203n20; disciplinary logics, 17, 74–94, 141, 173, 195n31, 197n43, 198n19,
205n6; intellectual consensus in, 74–76, 200n19, 201n40, 206nn19–20, 212n36,
84–94, 96, 106, 203n5, 203n19, 204n6; 212n40
and interpretations of diversity, 60, 62, Efficiency, 9, 28, 37, 44
65, 72; judgments of intelligence in, 118, Elitism, 54, 103, 107
120, 121, 124, 126, 129; prestige Ellsberg, Daniel, 207n11
hierarchies within, 103–104, 113, 114; Emotions, 70, 80–81, 112, 122, 202n13
rationale for disciplines included in study, Engineering, 43
14, 19, 182; relationship to university English, 137, 141–146, 151, 163, 205n7,
departments, 23, 28, 45 205n9
Discrimination, 12, 13, 67, 149, 159, 166, Enrollment management, 23, 134–136,
175, 197n46, 199n1, 210n14 207n5
Disparate impact, 199n1 Enthusiasm, 80–81, 86; of applicants, 57,
Diversity, 11–12, 13, 72; on admissions 121; and homophily, 96, 99, 121; of
committees, 28, 41, 198n12; ambivalence reviewers, 32, 96, 124
about, 164–166; benefits of, 159–160, Environmental conditions, 196n33, 206n23,
200n25; as compelling governmental 210n24, 211n35
246 Index
Equity, 3, 12, 13, 22, 47, 161, 165–166, scores, 47, 155; international students’
169, 172, 175, 196n41 scores, 137–141, 141–142, 151–152;
Espeland, Wendy, 198n20 interpretation of, 53, 55, 71, 87–93,
Ethical considerations, 137–142 118–120, 174, 200n19; meanings
Ethnicity, 47, 113, 210n18 associated with, 49, 50, 53–54, 118–120,
Evaluation, 17, 18, 204n15. See also Rating 130, 139, 158, 161, 211n35; misuse of,
Evaluative culture, 9 37, 172–174, 198n19, 201n40; practice
Excellence, 5, 132, 176, 177; inclusive increases scores, 205n6; revisions to test,
excellence, 18, 48, 171–172; and 130, 132, 200n13, 202n9, 212n40; score
intellectual consensus, 106; in organiza- disparities in, 12, 56, 71, 131, 197n43,
tions, 48–49; in relation to diversity, 56, 201n40, 206n23, 209n3, 212n39,
59, 64, 72, 161, 165 212n45; as a tool for efficiency and
clarity, 28, 31–32, 55, 78, 162; validity
Faculty hiring, 51, 64, 126, 154 of, 7–8, 71, 131, 195nn29–30, 201n41,
Fairness, 45, 195n24, 206n23 203n21, 206n21, 211n35, 152–153;
Fellowships. See Financial aid value attributed to, 49, 55, 111, 155
Financial aid, 22, 23, 60, 66, 101, 135, 167. Grant, 23
See also Diversity: fellowships Gratz v. Bollinger, 160, 209n4, 210n12
First-generation college students, 107–108, Griffin, Kimberly, 39
124 Guinier, Lani, 11, 48, 130, 206n18
Fisher v. University of Texas, 41, 160, Gumport, Patricia, 126, 198n10
109n4 Gutmann, Amy, 20, 197n1
Fisk University, 170, 172, 211n30
Fit, 13, 49, 54, 59, 109, 112, 118, 158 Haro, Roberto, 65
France: education, 204n8, 210n8; philos- Harvard University, 50, 60, 120
ophy, 208n20 Hatt, Beth, 117
Fraud in applications. See Cheating Heimer, Carol, 39–40
Friendly debate, 41–43 Heterarchy, 10, 157
Hiring, 209n2
Garces, Liliana, 61 Hispanic Americans. See Latinos
Gatekeeping, 5, 19, 70; and assessments of Holistic review, 32, 40, 56, 71, 136, 153,
merit and diversity, 48, 64; faculty 156, 167
gatekeeping, 22, 45, 72, 93, 176; Homophily, 17, 80, 84, 95–116, 204n5
gatekeeping course, 206n23; organiza- Human capital, 100
tional, 176–177 Humanities, 3, 52; admissions committee
Geertz, Clifford, 74 dynamics observed in, 28, 30, 162; and
Gender, 3, 4, 48, 113, 131, 144–145, 155, the academic labor market, 5, 11,
157, 194n9, 204n20, 206nn23–24, 194n16; description of departments
207n4; and GRE, 47, 173; consideration included in study, 14, 16, 75, 180, 182;
in admissions, 65; inequality, 210n24, evaluation of GRE scores in, 50, 119;
212n45; on research teams, 210n18 evaluation of writing samples in, 125,
Georgia Tech University, 211n30 127, 129, 205n13; intellectual consensus
Ginsburg, Ruth Bader, 41, 48 and disciplinary logics within, 75, 79, 84,
Goffman, Erving, 48, 131, 179 93, 202n5; and international students’
Grades, in college, 53, 122, 197n43, English skills, 141, 152; judgments of
211n35; as evaluation criterion, 50–51, intelligence in, 119, 120–121, 122–125
54–55, 71; grade point average, 31, 40, Humor, among reviewers, 42–43
50–51, 109, 111, 155; in graduate school,
131, 195, 204n9; meanings of, 49, 53–54 Identity, 47–48, 96, 99, 111–114, 119, 158
Graduate Record Examination (GRE), 2, IDP Education, 207n4
11, 12, 197n43, 200n22, 212n41; Implicit bias, 41–42, 153. See also
ambivalence about, 36–37; and decision Discrimination
making, 42, 44–45; distributions of Inclusive excellence, 48
Index 247
Inequality, 3, 12, 15, 174, 209n3; in the disposition, 121–122; organizational, 45,
academy, 64, 157, 196–197n41; and 112; student, 48, 101, 108, 120, 126,
diversity efforts, 61, 62, 72, 136, 161; 132, 159; subject matter, 120, 122
institutionalized, 13; maintained by Legitimacy, 52–54, 132, 164, 198n16; of
decision making preferences, 20, 45, 166; admissions practices, 9, 41, 45, 52, 76,
and postracial discourse, 165–166, 197n49; 82; through disciplinary logics, 75, 93;
in relation to beliefs about intelligence, and diversity, 62, 63; and norms of
129, 132; reproduced by homophily, 95 academic work, 21; and perceived
Innovation, 56, 57, 59, 66, 200 intelligence, 54, 130; of race-conscious
Institutional affiliations. See Pedigree admissions, 159–160
Institutional theory, 159 Lemann, Nicholas, 31, 205n7
Intelligence, 17, 18, 116–132, 200n19, Lesbian, gay, bisexual, transgender, and
201n37, 205n2, 206n18; and admissions queer (LGBTQ) students, 153
tests, 53–55, 118–120, 124, 130–131, Letters of recommendation, 55, 58, 68, 111,
161, 205n7; context-specific definitions, 128, 137, 156, 163; cultural variation in,
117, 205n4; as disciplinary cultural 147–148, 197n43
practice, 118, 124–130; as general Linguistics, 2, 199n21, 202n4; admissions
impression, 118–120, 128–132; committee dynamics observed in, 28,
intellectual dispositions, 121–122; 29, 41–44, 67; and admissions inter-
intellectual virtues, 121–4; judgments of, views, 143; description of departments
47, 53–55, 69–70 included in study, 14, 16, 26, 180, 182;
Interdisciplinarity, 177, 196n37 diversity considerations in, 67–70;
International students, 17, 46–47, 133–153, evidence of homophily in, 97; intellec-
204n5, 206n2, 207n4, 211n35; distin- tual consensus in, 75, 85; judgments of
guishing among, 146–148; organizational intelligence in, 54
interests in, 134–136, 207nn6–7; Locks, Angela, 167
preference for, 96–97
Interviews, for admissions, 118, 143–146 March, James, 163
Ivy League, 52, 102, 107, 110, 113 Master’s degrees, 3, 198n8, 211n30
Mathematics, 206n23
James, William, 5, 13, 116 MCAT, 205n9
Johns Hopkins University, 173 McNair Postbaccalaureate Achievement
Johnson, Lyndon, 159 Program, 211n32
Member checking, 200n13
Kahneman, Daniel, 70, 211n27 Menand, Louis, 165
Kiang, Peter, 149 Mentoring, 90–92, 165–166
Klitgaard, Robert, 50, 195n32 Merit, 7, 8, 9, 10, 11, 12, 46–49, 62, 67,
Kroc, Rick, 134 92–93, 99, 110, 153; ambiguity of, 132,
Kuncel, Nathan, 8n30 156; assumptions about, 171–172; as
conventional achievement, 50–56,
Labor market, 2, 11, 177; for professoriate, 155–157; democratic, 48; as future of
3, 5, 195n23, 196n41 discipline, 56–58, 113, 118, 126,
Lamont, Michele, 48, 106, 117, 158, 186, 128–129, 156–157; meritocracy, 30, 149,
200n7 205n7; organizational perspective on,
Latinos, 3, 12, 18, 47, 63–64, 135, 152, 161, 196n37
159, 163, 171, 200, 212n39; baccalau- Meyer, John, 28
reate origins of, 169–170 Meyerhoff Scholars Program, 211n32
Leaders, 28, 171; admissions committee Milkman, Katherine, 12
chair, 29, 30, 68, 97–89, 128, 165, Minority-serving institutions, 169
198n15, 202n13 Moss-Kanter, Rosabeth, 110
Learning: contexts for learning, 153; curves, Moss-Racusin, Corrine, 12
101, 108; environments, 9, 66, 206n23, Muñiz, Marcela, 39
210nn23–24, 211n35; and intellectual Museus, Samuel, 149
248 Index
Narrow framing, 45, 211n27 62, 63, 201n27; rating and ranking of
National Center for Fair and Open Testing, applicants in, 32–34, 35, 37
212n41 Physics, 2, 4, 117, 158, 199n24, 205n9; and
National Endowment for the Humanities, admissions interviews, 143; description of
205n13 departments included in study, 14, 16, 24,
National origin, 60, 212n45 180, 182; disciplinary logics of, 75, 82,
National Research Council, 6, 194, 209n7 83; evaluation of GRE scores in, 50, 55,
National Science Foundation, 169–170, 140, 142; evidence of homophily in, 97,
211n32 104, 106; faculty perspectives on diversity
Native Americans, 3, 18, 135, 163, 171; and and equity, 60, 135; faculty views on risk
GRE scores, 47 aversion, 52; and international students,
Natural sciences, 52, 107, 129, 152, 202n5; 140, 142, 146
description of departments included in Plato, 176
study, 14, 16, 180, 182; judgments of Policy capturing, 211n28
intelligence in, 125, 127 Political science, 2, 165, 199n24; and
Neumann, Anna, 122 admissions interviews, 143; decision-
Nixon, Richard, 159 making dynamics observed in, 40, 41, 43;
Noblesse oblige, 92, 97, 111, 114 description of departments included in
Noncognitive admissions criteria, 57, 172, study, 14, 16, 27, 28, 180, 182; disci-
212n36 plinary logics in, 75, 85, 87–94; evidence
Nonresident students, 207n7 of homophily in, 97, 105
Norms, 36, 39, 43, 44, 202 Postracial discourse, 165, 210n25
Novelty, 105–106 Potential, 7, 9, 10, 45, 56, 58, 96, 101–102,
126, 129, 132, 153, 164, 174
Oakes, Jeannie, 116, 205 Powell, Lewis, 160
Obama, Barack, 92, 203n23 Preferences, 44–45, 98, 102, 113–114, 156,
Organizational change, 164–166, 171, 158, 176, 196n39, 204n5, 205n6,
174–175 205n22, 211n28; revealed, 56
Organizational culture, 114, 164, Prestige, 2, 8, 78, 201n45, 207n4; of
174–175 graduate program, 49, 135; as organiza-
Organizational learning, 45, 112 tional interest, 9, 13–14, 157–159, 161;
Outreach, 23, 169 of institutional affiliations, 51, 54, 77,
99–104, 107–113, 155, 175, 209n2
Park, Julie, 202n46 Privilege, 111, 166, 171, 204n8
Passion, 121–122 Professional doctorate, 194n9, 194n19
Pedigree, 67–71, 84, 95–97, 99–104, Psychology, 42, 201n41, 211n34
107–113, 154, 158–159, 163, 204n5
Peer review, 21 Quality, 2, 5–11, 158; of academic research,
Pennock-Roman, Maria, 7 78, 161, 203n21; of applicant, 39–40,
Pericles, 45 54–56, 63, 82, 132, 153; challenges in
Personal statement, 42, 50, 55, 118, 127, assessing, 29–31; of dissertation, 203n21;
141–142, 145 of graduate program, 49, 134, 177,
PhD, 2, 3, 5; completion, 200n17; rising 210n23; judgments of, 47, 56, 195n23;
pursuit of, 193n5, 194n6; time to degree, personal and intellectual qualities of
200n17 applicants, 116, 118, 123, 125, 185;
Philosophy, 1, 165, 174–175, 196–197n41, politics of, 47–48, 73; quantification of,
208n20; and admissions interviews, 143; 31, 35, 37, 55; slight differences in
description of departments included in student quality, 6, 39; of test preparation
study, 14, 16, 25–26, 180; disciplinary for international students, 138; of
logics of, 78–84; evaluation of writing undergraduate institution, 51–52, 100,
samples in, 30, 127, 132; evidence of 108; of undergraduate research experi-
homophily in, 97–98, 103, 110; faculty ence, 126–127, 129, 148; university, 163,
perspectives on diversity and equity, 60, 171–172
Index 249
Quantification, 30–32, 34–37, 40, 44, 55, Smith, Daryl, 64, 196
77–78, 81–82, 162, 164, 199n23 Smith vs. City of Jackson, 199n1
Social capital, 147
Race, 47–48, 72, 113, 149, 155, 158, 173 Socialization, 126, 128, 156; disciplinary,
Race-conscious admissions, 23, 59, 61, 67, 74–76; experiences formative to, 100,
157, 159–160, 209n4, 210n14 112; in low-consensus disciplines, 93
Racial inequality, 210n24, 212n45 Social mobility, 204n5; homophily based
Racial isolation, 210n23 on, 96–97, 106–110
Racism, 12, 13, 92–93, 149, 163, 197n49 Social reproduction, 74, 95, 103–104, 113,
Rankings, 135 158
Rating, 1, 29, 31–37, 40–41, 131, 155, Social sciences, 3, 34, 105, 107; description
199n21; use of rubrics, 168–169 of departments included in study, 14, 16,
Recruitment, 5, 22, 69–172, 175, 212n43; 28, 57–58, 180, 182; intellectual
challenges in student, 21, 135, 165, 171; consensus and disciplinary logics within,
diversity, 23, 39, 68; events, 14, 69; in 75–76, 82, 93, 202n5; and international
industry, 154; practices, 22, 25–26, 121, students, 141, 152; prestige hierarchies
167, 211n34; recruitment networks and within, 78, 83
relationships, 39, 170, 172, 211n30; of Socioeconomic status, 50, 60, 99, 135, 155,
study participants, 182–184, 189; 157
through undergraduate research Sociology, 2, 36; assessments of merit in,
experiences, 170–171, 211n32; of 33–35, 51, 54–55, 101, 103; description
underrepresented students, 11, 63, 68–69, of departments included in study, 16, 27;
72, 135 evidence of homophily in, 97; views of
Reed College, 52 admissions exams in, 119, 142; sociolog-
Regents of the University of California v. ical research informing the study, 48,
Bakke, 67, 160–161, 210n12 72–74, 81, 95, 113, 133; views on
Religion, 41–42, 158, 193n4 diversity in, 60, 62, 64, 66
Research experience, 124–130, 211n32; as Solidarity, 97, 111, 174
admissions criteria, 55, 109, 158, 171; Spelman College, 211n30
evaluation of, 57–58, 66, 118; in Spillman, Lyn, 197n52
master’s-PhD bridge program, 170; of Statement of purpose. See Personal
international students, 143, 148; statement
Research universities, 23, 102, 113, 129, Status, 72, 76, 159, 204n20, 105, 110–111,
193, 211n30 150
Risk, 112, 211n35; aversion to, 52, 68–72, Steele, Claude, 90, 131, 206n22
76–78, 89–90, 101, 111, 162–163 Stereotypes, 42, 131, 144, 146; model
Rivera, Lauren, 81, 99, 112–113, 154, minority, 137, 148–150, 166, 208n27;
209n2 stereotype lift, 131, 206n24; stereotype
Rowan, Brian, 28 threat, 90, 131, 173, 203n22, 206n22,
212n39; and availability bias, 163
SAT, 205n7 Sternberg, Robert, 88, 205n3
Science, technology, engineering, and math Stevens, Mitchell, 40, 195n32, 198n20,
(STEM), 3, 127, 135, 170, 172, 206n23 200n21, 201n45
Scripts, 48–49, 52–53, 61, 71–72, 161, 163, Subjectivity, 80
200n7; counterscripts, 96, 110–112, 131, Summer Research Opportunities Program,
174 211n32
Selectivity, of colleges, 14, 22, 209nn2–3 Survey of Earned Doctorates, 3, 4
Self-concept, 4, 116–117, 132, 205n2
Self-esteem. See Self-concept Taste, 35, 80, 94, 112
Self-similarity. See Homophily Technology, 34, 105
Sexism, 12–13 Test of English as a Foreign Language
Shared governance, 21 (TOEFL), 7, 28, 37, 141–142, 152,
Smart. See Intelligence 211n35
250 Index
Test-optional admissions, 173–174, Vanderbilt University, 170, 172, 194n20,

212nn41–42 211n30
Test scores, 31, 40, 45, 50–51, 55, 155,
206n24, 212n39 Walker, George, 202
Thevenot, Laurent, 157 Wealth, 207n2
Thomas, William and Dorothy, 176 Weber, Max, 22
Thompson, Dennis, 20 Wechsler, Howard, 11
Tierney, William, 144, 186, 196n37, Weick, Karl, 174–175, 212n44
208n17 Wesleyan College, 52
Tilly, Charles, 13 White privilege, 197n47
Transparency, 40, 47, 154, 169 Whites, 4, 12, 29, 159, 210n14
Trowler, Paul, 117 Williams, Wendy, 88
Williams College, 52
Undergraduate admissions, 6, 7 Wilson, James, 28
Undergraduate research. See Research Winkle-Wagner, Rachelle, 167
experience Women, 56, 143–145, 153, 175, 204n20,
Undermatching, 107 211n211. See also Gender
United Kingdom, 7. See also British Worldview moments, 86–94. See also
universities Disagreement
University of California, 160 Worthington, Roger, 166
University of Chicago, 211n32 Writing, 50, 99, 205n13
University of Maryland-Baltimore County, Writing sample, 30, 79, 118, 123, 125, 127,
211n32 132
University of Michigan, 173
University of Pennsylvania, 103, 173 Yale University, 87, 203n21
U.S. Census, 208n23 Yield, 63
U.S. News and World Report, 6, 194n20,
209n2, 209n7 Zhang, Liang, 135
U.S. Supreme Court, 41, 60, 67, 160 Zwick, Rebecca, 119–120, 205n6

Posselt, Julie R - Inside Graduate Admissions - Merit, Diversity, and Faculty Gatekeeping-Harvard University Press (2016)

Uploaded by

Copyright:

Available Formats

Posselt, Julie R - Inside Graduate Admissions - Merit, Diversity, and Faculty Gatekeeping-Harvard University Press (2016)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Posselt, Julie R - Inside Graduate Admissions - Merit, Diversity, and Faculty Gatekeeping-Harvard University Press (2016)

Uploaded by

Copyright:

Available Formats

Inside Graduate Admissions

Merit, Diversity, and Faculty Gatekeeping

Library of Congress Cataloging-in-Publication Data

Introduction: Gatekeeping Reconsidered 1

one Decision Making as Deliberative Bureaucracy 20

two Meanings of Merit and Diversity 46

three Disciplinary Logics 74

four Mirror, Mirror 95

five The Search for Intelligent Life 116

six International Students and Ambiguities of

Conclusion: Merit beyond the Mirror 154

Methodological Appendix 179

the support, high standards, and academic freedom extended to me by my

F or three hours the philosophy admissions committee had been

Doctoral Students and Their Professional Pathways

Astronomy Male *** *** *** 137 *** *** 10 190

Note: *** = suppressed to avoid disclosure of confidential information.

engineering, and mathematics. And in the humanities, where the glut of

Reform in doctoral education today must better align notions of student

America is thus as a nation rapidly drifting towards a state of things in which

We advertise our “schools” and send out our degree-requirements, knowing

The competitive trends James identified—among doctoral programs for

described the assumptions of this body of research and a fundamental

There is one unidimensional [variable] Y, such as college grade-point average,

The difficulty of reliably predicting long-term outcomes from any infor-

I think from practically everybody’s viewpoint getting talented, motivated

In practice, merit in graduate admissions is not an absolute assessment of

if it is merely the sum total of an applicant’s “deservingness” based on what

Untangling a Paradox of Admissions

I designed this research to untangle an apparent paradox in the research

from underrepresented groups is common practice in areas from undergrad-

gender, because inequalities can also become locked in (or institutional-

Notes on the Research Design

In this section I provide a general overview of my research design; readers

I have masked and/or changed information about applicants and faculty

Reading This Book

% Scholars % Domestic # Graduate students

closed doors. They may be encouraged or dismayed, for example, to learn

participants think and deliberate. Each chapter takes up the consequences

reproduction. In this political pressure cooker, it is no wonder that change

Decision Making as Deliberative

H ow should admissions decisions be made? How are they made

Deliberative bureaucracy is an organizational adaptation to rising num-

and allowed misperceptions about common criteria to go unchecked. It

An Overview of Admissions Procedures

Before discussing the details of deliberative bureaucracy, some readers may

tutions choosing or rejecting students, but with students choosing or rejecting

Nevertheless, in each of the ten programs I studied, a committee mediates

programs that required writing samples assigned them to be reviewed by

Committees were thus often entrusted to connect outsiders’ judgments of

Another bureaucratic practice involved putting numbers to judgments in

In admissions and elsewhere in education and social life, quantifying quality

Applicant ratings. A second form of quantification involves rating indi-

Above a general threshold of admissibility, faculty were forced to make dif-

Another member of this committee affirmed this:

Individual ratings were key ingredients of collective decisions, both in de-

Rank-ordering applicants. The quantification of judgment continued into

In all cases, allowing each member to have different individual ratings of

Faculty committees conducted admissions as a “numbers game” in other

Astronomy Male * * * 137 * *** 10 190