Criminal records and college admissions: A modified experimental audit

• Date: 01 February 2020
• Author: Robert Stewart,Christopher Uggen
• DOI: http://doi.org/10.1111/1745-9125.12229
• Published date: 01 February 2020

Received: 26 April 2018 Revised: 5 August 2019 Accepted: 7 August2019

DOI: 10.1111/1745-9125.12229

ARTICLE

Criminal records and college admissions: A modified

experimental audit∗

Robert Stewart Christopher Uggen

Department of Sociology, University of Minnesota

Correspondence

RobertStewart, Department of Sociology,

Universityof Minnesota, 909 Social Sciences,

26719t h AvenueS, Minneapolis, MN 55455.

Email:stewa640@umn.edu

Fundinginformation

SpencerFoundation, Grant/Award Number:

201600091

∗Thisresearch was supported by a grant from

the Spencer Foundation. Earlierversions of this

article werepresented at t he annual meetings

(2017and 2016, respectively) of the American

SociologicalAssociation and t he American

Societyof Cr iminology.We are grateful to

MikeVuolo and Ryan Larson for their generous

adviceand feedback, as well as to Chelsea

Carlson,Eleanor Plaunt, Lesley Schneider,

TaylorDay,Alex Thibido, Susan Balto, and

EmilyKucharski for their invaluable research

assistance.Thanks also to the anonymous

reviewersand Jody Miller for their careful

reviewand valuable comments.

Abstract

In this article, we consider the effect of criminal records on

college admissions. Nearly 72 percent of colleges require

criminal history information during their application pro-

cesses, which indicates that an applicant’s criminal his-

tory could be a significant impediment to achieving the

benefits associated with higher education. We conducted

a modified experimental audit to learn whether and to

what extent criminal records affect admissions decisions.

Matched same-race pairs of tester applications were sent

to a national sample of nonelite 4-year colleges, with both

testers applying as either Black or White. Within each pair,

one application signaled a prior low-level felony convic-

tion only when required by the application. Consistent with

the findings of research on employment, we find the rejec-

tion rate for applicants with felony convictions was nearly

2.5 times the rate of our control testers. Relative to the

large racial differences observed in previous studies of hir-

ing decisions, we find smaller racial differences in admis-

sions decisions. Nevertheless, Black applicants with crim-

inal records were particularly penalized when disclosing a

felony record at colleges with high campus crime rates. We

address implications for reentry, racial progress, and the

college “Ban the Box” movement.We suggest colleges con-

sider narrowing the scope of such inquiries or removing the

question altogether – particularly when it conflicts with the

goals of these institutions, including reducing the underrep-

resentation of students of color.”

156 © 2019 American Society of Criminology wileyonlinelibrary.com/journal/crim Criminology.2020;58:156–188.

STEWARTAND UGGEN 157

KEYWORDS

collateral consequences, college, criminal records, education,

experimental

Criminal records are a pervasive, acutely restrictive feature of American social life (Shannon et al.,

2017; Travis, Western, & Redburn, 2014). They can limit access to employment (Pager, 2003; Pager,

Bonikowski,& Western, 2009; Uggen, Vuolo, Lageson, Ruhland, & Whitham, 2014), public assistance

(Gustafson, 2011), housing (Thacher, 2008), dating (Evans, 2019; Massoglia, Remster,& King, 2011),

family planning options (Children’s Bureau, 2015), and civic participation (Manza & Uggen, 2006).

These restrictions can perpetuate cycles of crime, inequality, and lost opportunity, especially for low-

income people and people of color.

Higher education has long been considered an instrument of social mobility and social cohesion in

American society, as well as a potentiallycompelling mechanism for facilitating desistance (Blomberg

& Pesta, 2017; Ford & Schroeder, 2010; Runell, 2017). College attendance and completion are asso-

ciated with lower rates of unemployment and with higher relative earnings (Arum & Hout, 1998;

Mayhew, Bowman, Rockenbach, Seifert, & Wolniak, 2016; U.S. Bureau of Labor Statistics, 2017).

Receiving a higher education can also unlock access to valued opportunities, develop human capital,

and foster civic membership (Becker, 1994; Boli, Ramirez, & Meyer, 1985; Brown, 2001). Yet, most

colleges require applicants to disclose detailed criminal history information as part of the application

process, and some evidence indicates that applicants are being rejected on the basis of their records

(Custer, 2018; Pierce, Runyan, & Bangdiwala, 2014; Rosenthal, NaPier, Warth, & Weissman, 2015;

Weissman, Rosenthal, Warth, Wolf, & Messina-Yauchzy, 2010). Thus, the benefits of higher education

may not accrue for students with criminal records.

The increasing scrutiny of criminal records in college admissions is especially consequential for

groups most subject to the criminal justice system, particularly young Black males (Brame, Bushway,

Paternoster, & Turner, 2014). Considering the historic underrepresentation of Black Americans in

higher education (Davis & Otto, 2016) and their overrepresentation in justice-involved populations

(Kaeble & Glaze, 2016), criminal history disclosure requirements could lead to additional barriers to

racial progress, student learning, and democracy. Officials cite campus safety as the principal reason

for requiring criminal history information, especially in the aftermath of widely publicized on-campus

incidents (Dickerson, 2008; Pierce et al., 2014). Significantly, however, these questions may be ill-

suited for predicting future criminality (Runyan, Pierce, Shankar, & Bangdiwala, 2013) and reducing

campus crime rates (Olszewska, 2007).

The use of criminal records in the college admissions process has rarely been considered in crimi-

nology, so little is known about the prevalence of criminal history disclosure questions and the degree

to which a criminal record might affect college acceptance. In this article, we provide the first estimates

of the effect of a criminal record on college admissibility of young men, as well as of the race differ-

ences in these effects. We begin by reviewing research on the relationship between higher education

and punishment, discrimination on the basis of criminal records, and the use of criminal records in the

higher education context. We then discuss the likely impact of policy remedies such as the “Ban-the-

Box” (BTB) movement beforeasking an empirical question with policy implications: whether and how

felony records affect college rejection rates. We use a modified experimental audit method to generate

a specific estimate of the impact of a single felony convictionfor robbery or burglary when applying to

nonelite 4-year colleges, and we briefly highlight other forms of differential treatment that applicants

with records encountered in our study. We find that even though a felony record is not a categorical

158 STEWARTAND UGGEN

disqualifier, it can be a barrier to college access. We conclude by placing these results in the context

of research on employment discrimination and the nascent BTB movement for colleges that parallels

recent “fair chance” legislation governing job applications.

1PUNISHMENT AND EDUCATION

In a considerable amount of research literature, scholars have examined the effects of employment on

desistance and recidivism (see, e.g., Berg & Huebner, 2011; Denver,Siwach, & Bushway, 2017; Laub

& Sampson, 2001; Pager, Western, & Sugie, 2009; Skardhamar & Savolainen, 2014; Uggen, 2000),

but there have been few rigorous studies in which the focus was on the relationship between higher

education and desistance/recidivism, particularly for those not in prison. In existing observational and

quasi-experimental studies of prison education, researchers have generally found that higher educa-

tion (whether in or out of prison) is associated with positive outcomes, including lower recidivism

and higher employment and earnings (Duwe & Clark, 2014; Lockwood, Nally, Ho, & Knutson, 2012).

With respect to recidivism, in studies in which propensity score matching is employed to adjust for

selectivity, scholars have generally found that prison-based college programs reduce recidivism mea-

sured by rearrest, reconviction, reincarceration, and revocation (Duwe & Clark, 2014; Kim & Clark,

2013). In a meta-analysis of more than 50 studies, the researchers estimated that participants in prison

education programs were 43 percent less likely to recidivate than were nonparticipants (Davis, Boz-

ick, Steele, Saunders, & Miles, 2013). Although these outcomes can be highlighted to speak to prison

education, they are of limited usefulness because people who are incarcerated are at much greater like-

lihood of reoffending than is a general population sample of college applicants. Still, these research

findings should encourage further study of education and postcriminal justice involvement and its link

to employability and desistance from crime. Given the public and policy interest in higher education,

desistance, and employability, an examination of barriers to college is merited.

Obtaining a higher education may facilitate desistance by serving as a bridge to cultivate social

bonds with prosocial peers and role models, develop self-efficacy, and acquire human capital (Becker,

1994; Brown, 2001; Ford & Schroeder, 2010; van Dinther, Dochy, & Segers, 2011). Furthermore,

receiving an education likely reduces crime by increasing the likelihood of better earnings and the

quality of employment (Arum & Hout, 1998; Carnevale, Rose, & Cheah, 2013; Mayhew et al., 2016;

U.S. Bureau of Labor Statistics, 2017), whichcould help counterbalance the negative effects of criminal

justice involvement on employment, earnings, and mobility (Western, 2002).

Nevertheless, just as in the labor market, people with criminal records may be denied access to

the benefits of higher education precisely because of their criminal records. A criminal record can be

interpreted as a “negative credential” that can be used to restrict opportunities in much the same way

that a college degree can be interpreted as a “positive credential” to expand opportunities (Gaddis,

2015; Pager, 2003). Where having a college degree might indicate honesty, a strong work ethic, and

responsibility to an employer, forexample, having a criminal record might indicate dishonesty, danger-

ousness, and irresponsibility (Kurlychek, Brame, & Bushway, 2006; Lageson, Vuolo, & Uggen,2015),

constraining job opportunities. Thus, if having a criminal record poses a similar challenge to higher

education access as it does in the labor market, then the presence of prior criminal justice involvement

could serve as a limitation to educational attainment and its beneficial effects on desistance and the

transition to adulthood. As Lovenheim and Owens (2014) found by exploitinga federal policy change

that restricted financial aid for a certain time period for people convicted of drug felonies, even though

there was little evidence the policy change was an effective deterrent, there were notable consequences

for affected students who experienced a delay in college enrollment and attainment, likely stunting