Full Terms & Conditions of access and use can be found at
Download by: [University of Cincinnati Libraries] Date: 05 April 2017, At: 07:22
ISSN: 0741-8825 (Print) 1745-9109 (Online) Journal homepage: http://www.tandfonline.com/loi/rjqy20
Solitary Confinement as Punishment: Examining
In-Prison Sanctioning Disparities
Joshua C. Cochran, Elisa L. Toman, Daniel P. Mears & William D. Bales
To cite this article: Joshua C. Cochran, Elisa L. Toman, Daniel P. Mears & William D. Bales
(2017): Solitary Confinement as Punishment: Examining In-Prison Sanctioning Disparities, Justice
To link to this article: http://dx.doi.org/10.1080/07418825.2017.1308541
Published online: 05 Apr 2017.
Submit your article to this journal
View related articles
View Crossmark data
Solitary Conﬁnement as Punishment:
Examining In-Prison Sanctioning
Joshua C. Cochran, Elisa L. Toman,
Daniel P. Mears and William D. Bales
Drawing on prior sentencing and prison scholarship, this study examines the
use of solitary conﬁnement as a form of punishment. Speciﬁcally, it assesses
whether, given a prison infraction, minority inmates—and young, male,
minority inmates in particular—are more likely to be placed in solitary and
to be placed in it for longer durations. Multilevel regression analyses of
state prison data suggest little support for the hypothesis that minority
males, or young minority, males, are sanctioned more harshly than other
inmates. The analyses identify, however, that males are more likely than
Joshua C. Cochran, PhD, is an Assistant Professor at the University of Cincinnati, School of Criminal
Justice, P.O. Box 210389, Cincinnati, OH 45221, USA. E-mail: firstname.lastname@example.org. His research
interests include theory, imprisonment, prisoner reentry, and sentencing. His work has appeared in
Criminology, the Journal of Quantitative Criminology,Justice Quarterly, the Journal of Research
in Crime and Delinquency and in a recent book, with Daniel P. Mears, Prisoner Reentry in the Era
of Mass Incarceration (Sage). Elisa L. Toman is an Assistant Professor at Sam Houston State
University’s Department of Criminal Justice and Criminology, Huntsville, TX 77341, USA. E-mail:
email@example.com. Her research interests include theories of punishment, trends in criminal
sentencing, and the implication of individuals’ experiences with the corrections system. She has
published recently in Journal of Quantitative Criminology,Justice Quarterly, and Journal of Crimi-
nal Justice. Daniel P. Mears, PhD, is the Mark C. Stafford Professor of Criminology at Florida State
University’s College of Criminology and Criminal Justice, 112 South Copeland Street, Eppes Hall,
Tallahassee, FL 32306-1273, USA. E-mail: firstname.lastname@example.org. His work has appeared in leading crime
and policy journals and American Criminal Justice Policy (Cambridge University Press), which
received the Academy of Criminal Justice Sciences Outstanding Book Award, Out-of-Control Crimi-
nal Justice (Cambridge University Press), and, with Joshua C. Cochran, Prisoner Reentry in the Era
of Mass Incarceration (Sage). William D. Bales, PhD, is a Professor at Florida State University’s Col-
lege of Criminology and Criminal Justice, 112 South Copeland Street, Eppes Hall, Tallahassee, FL
32306-1127, USA. E-mail: email@example.com. Dr. Bales focuses on a range of crime and policy topics,
including factors that contribute to recidivism, the effectiveness of electronic monitoring, and
tests of labeling theory. He has published in Criminology,Criminology and Public Policy,Justice
Quarterly, and other crime and policy journals. Correspondence to: Joshua C. Cochran, School of
Criminal Justice, University of Cincinnati, P.O. Box 210389, Cincinnati, OH 45221-0389, USA. E-mail
Ó2017 Academy of Criminal Justice Sciences
Justice Quarterly, 2017
females to be placed in solitary as a form of disciplinary punishment and
that younger females are more likely to be placed in it than older females.
The ﬁndings highlight that age and sex may interact to inﬂuence punishment
decisions and raise questions about the precise roles of race and ethnicity
in affecting punishment decisions. Implications of the ﬁndings for theory,
research, and policy are discussed.
Keywords prison; sanctioning; race; ethnicity; solitary conﬁnement;
Scholarship on the exercise of formal social control, punishment in particular,
consistently ﬁnds evidence of racial, ethnic, and gender disparities in criminal
sanctioning decisions (e.g. Baumer, 2013; Spohn, 2015). Studies typically ﬁnd,
for example, that young minority males are more likely to receive harsh pun-
ishments from the courts (e.g. Brennan, 2006; Jordan & Freiburger, 2015;
Vogel & Porter, 2016; Wang, Mears, Spohn, & Dario, 2013; Warren, Chiricos, &
Bales, 2012). Prior scholarship has drawn on focal concerns theory and
research on attributional stereotyping to argue that implicit biases in court
actors’ perceptions of offenders inﬂuence sanctioning decisions. This line of
work suggests that racial and ethnic minorities—especially those who are
young and male—are more likely to be perceived as a threatening and culpable
group, giving rise in turn to more severe sentencing (see, e.g. Brennan, 2006;
Demuth, 2003; Feldmeyer, Warren, Siennick, & Neptune, 2015; Frenzel & Ball,
2008; Harris, 2009; Zatz, 2000).
Limited research exists, however, that examines punishment decisions, such
as the use of solitary conﬁnement,
that occur after court sentencing. As Butler
and Steiner (in press) and others (e.g. Frost & Monteiro, 2016; Morris, 2016)
have argued, this gap is conspicuous for several reasons. Solitary conﬁnement
1. For the purposes of this paper, “solitary conﬁnement” refers to what sometimes is referred to
as “disciplinary conﬁnement,” which is conﬁnement in response to inmates’ disciplinary infrac-
tions. Other terms—such as “isolation,” “supermax,” “segregation,” “administrative segregation,”
and “restrictive housing,”—can be and are interchangeably used with “solitary conﬁnement”
(Mears, 2016). In each instance, the conﬁnement of focus typically centers on (1) an inmate spend-
ing entire days, such as 20 hours or more, by themselves, (2) doing so for varying periods of time,
and (3) for different goals, such as punishment, protection, or some general managerial purpose.
There is, at present, no consistent terminological usage (Mears, 2016). For example, some scholar-
ship may use “supermax” to refer to long-term stays whereas others may view “supermax” housing
to encompass short- or long-term stays. We recognize this inconsistency and have opted here to
refer to “solitary conﬁnement” because the term has been in use for many decades and has been
the term used in prominent presidential and legislative discussions. We use it instead of “disci-
plinary conﬁnement” because this latter term does not clearly capture the fact that inmates may
be alone during such conﬁnement. Ultimately, regardless of terminology, our focus is on the use of
placing inmates in a cell, alone, typically for 20 hours or more for one or more days, as a punish-
ment for infractions.
2 COCHRAN ET AL.
arguably constitutes the harshest mechanism of formal social control that
prisons can employ. Policymakers, practitioners, and human rights groups have
expressed concerns that it is used capriciously and that its harms outweigh its
putative beneﬁts (American Civil Liberties Union, 2014; Amnesty International,
2012; Haney, 2003; Mears & Castro, 2006; Metzner & Fellner, 2010; Obama,
2016). Concerns exist, too, that solitary conﬁnement may be used dispropor-
tionately with some groups, such as minorities (Mears & Bales, 2010). Few
empirical studies, however, have examined the factors that contribute to the
use of solitary conﬁnement as punishment (see Beck, 2015; Butler & Steiner, in
press; Crouch, 1985; see also Mears & Reisig, 2006; Morris, 2016). Little is
known, for example, about potential disparities that may exist in solitary
conﬁnement placement decisions (Frost & Monteiro, 2016).
The goal of this paper, then, is to address these research gaps and to
advance scholarship on incarceration and the exercise of formal social control
by examining in-prison punishment disparities. Speciﬁcally, and drawing on
focal concerns perspective, this study seeks to test the hypothesis that solitary
conﬁnement, as a form of punishment, is used more for minorities and, in par-
ticular, for young minority males. To test this hypothesis, we use seven years
of infraction event data for all state prison inmates incarcerated in the state
of Florida from 2005 to 2011. The data provide an opportunity to examine the
use of solitary conﬁnement as punishment and whether its use serves, as some
research intimates, to create racial and ethnic inequalities. We begin with a
discussion of prior literature on disparities in criminal justice decision-making.
Next, we turn to research on sentencing and prison misconduct to guide analy-
ses on the use of solitary conﬁnement as punishment, and then discuss the
data, methods, and ﬁndings.
Focal Concerns and Disparities in Punishment Decisions
Prior empirical research identiﬁes racial and ethnic disparities at almost every
decision point in the criminal justice process. Blacks and, in many instances,
Hispanics are, for example, more likely to be contacted by police, searched
upon contact, arrested, held in jail pretrial, convicted, receive a prison
sentence, and serve a lengthy term of incarceration (Baumer, 2013;
Kutateladze, Andiloro, Johnson, & Spohn, 2014; Schlesinger, 2005; Tonry &
Melewski, 2008; Ulmer, 2012; Wang et al., 2013; Western, 2006; Wooldredge,
2012; Wooldredge, Frank, Goulette, & Travis, 2015). A large body of research
has centered, in particular, on explaining incarceration sentencing decisions
and has focused on the ways that race, ethnicity, sex, and age intersect to
create a higher risk of more severe sanctioning. For example, sentencing stud-
ies consistently ﬁnd that young black and Hispanic males are at a particularly
heightened risk for tougher sentences as compared to their older, white, male
IN-PRISON SANCTION DISPARITIES 3
or female counterparts (Brennan, 2006; Demuth, 2003; Feldmeyer & Ulmer,
2011; Frenzel & Ball, 2008; Harris, 2009; Jordan & Freiburger, 2015; Vogel &
Focal concerns theory provides insight for anticipating and understanding
such decisions and the mechanisms that lead to racial and ethnic disparities in
sentencing (Steffensmeier, 1980). Building on broader attributions research
(e.g. Albonetti, 1997; Kautt & Spohn, 2002; Kramer & Ulmer, 2009; Levinson,
2007), it argues that court actors, such as judges and prosecutors, make sanc-
tioning decisions based on cognitive heuristics, or what might be viewed as
perceptual shorthands (Steffensmeier, Ulmer, & Kramer, 1998). Speciﬁcally,
the theory proposes that court actors consider three key dimensions that moti-
vate or otherwise inﬂuence punishment decisions: (1) a convicted individual’s
blameworthiness or perceived culpability, (2) the risk or danger that a given
individual is perceived to pose to a community, and (3) the organization and
practical constraints of a given court or jurisdiction.
From a focal concerns perspective, these three dimensions—blameworthi-
ness, dangerousness, and practical constraints—are central to the decision-
making processes of prosecutors and judges. In the context of a typical state
court with a high volume of cases and where court actors must make rapid
sanctioning determinations, court actors’ judgments will be based on both
conscious and unconscious cues about individuals (Spohn & Holleran, 2000;
Steffensmeier & Demuth, 2000; see also, Kahneman, 2011). Court actors rely
on these cues to guide their sentencing decisions. When these cues are in
error, they lead courts to sanction some groups in a more severe manner than
others, which in turn can lead to disparities in sentencing.
Scholarship to date suggests that court actors are more likely to associate
danger with minority defendants due to underlying perceptions of racial threat
(e.g. Crawford, Chiricos, & Kleck, 1998; Demuth & Steffensmeier, 2004;
Feldmeyer & Ulmer, 2011; Feldmeyer et al., 2015; Hagan, 1974; Harris, 2009;
Wang & Mears, 2010; Warren et al., 2012). Focal concerns then are channeled
through a racial threat lens. Speciﬁcally, members of the majority group are
prone to feel threatened by members of outgroups (i.e. racial and ethnic
minorities) and then rely on race or ethnicity—or certain groups, such as
young, minority males—to guide their interpretation of culpability and risk
and, in turn, their view about the need for or appropriateness of punishment.
Focal concerns and perceived racial threat then can lead court actors to
apply more formal control towards minority defendants or particular sub-
groups, such as young blacks or young black males. Importantly, the perceived
concerns and threats may vary. For example, in a given jurisdiction, court
actors may perceive that drug offenders represent a greater threat to
community safety than do low-level violent or property offenders. In turn,
sanctioning of drug offenders may be consistently more severe in that
jurisdiction. A similar outcome can arise if court actors hold similar views
about younger defendants versus older ones, males versus females, or, again,
sub-groups, such as young black males or young Hispanic males.
4 COCHRAN ET AL.
Prior research has not typically measured perceived focal concerns or threat
directly (see, however, Albonetti & Hepburn, 1996; Bridges & Steen, 1998;
Harris, 2009). Yet, ﬁndings from studies of court sentencing lend support to
the argument that minorities—young minority males in particular—are per-
ceived to be more threatening and deserving of severe punishment (Baumer,
2013; Brennan, 2006; Mitchell, 2005; Vogel & Porter, 2016; Warren et al.,
Focal Concerns and Disparities in In-Prison Sanctioning
Do similar punishment disparities emerge inside prisons? Although considerable
research has focused on inmate misconduct and its causes, as well as on prison
social disorder (Gonc¸alves, Gonc¸alves, Martins, & Dirkzwager, 2014; Useem &
Piehl, 2008), limited attention has been paid to how prisons respond to, or
sanction, misconduct. The small number of studies that do exist identify little
racial variation in in-prison punishment decisions (e.g. Butler & Steiner, in
press; Crouch, 1985). It is, however, theoretically plausible that bias in
in-prison sanctioning decisions exist and parallel those that occur throughout
earlier criminal justice decision-making points. Indeed, in-prison sanctioning
processes may be viewed as analogous to court sentencing. In a typical prison,
inmates are charged, or “written up,” for a formal infraction. As with a law
violation, the disciplinary infraction represents a formal violation of rules and
a “write up” in prison is akin to being arrested and charged. In some instances,
prison rule violations address behaviors that are also illegal and could be
viewed as criminal. Once charged, the infraction is investigated, evidence and
testimony are provided, and a disciplinary team hands down a verdict. If a
guilty verdict is rendered, the team issues a sanction.
In addition, just as criminal caseloads have increased and many state courts
have become overburdened in recent decades (e.g. Ulmer & Johnson, 2004),
state prison populations have expanded (Spohn, 2015; Useem & Piehl, 2008;
Western, 2006). The resulting prison growth creates the potential for more
infractions and caseload pressures that, as with court decision-making
(Steffensmeier et al., 1998; Ulmer & Johnson, 2004), cause prison staff to rely
more heavily on perceptual shorthands to efﬁciently process “cases.”
To the extent that in-prison sanctioning entails a decision-making process
that parallels court sentencing processes, it provides grounds for anticipating
similar patterns in punishment. For example, prison ofﬁcers assigned to disci-
plinary teams may be inﬂuenced by racial cues or stereotypes when making
sanctioning determinations. The premise is that racial bias or perceptions of
threat extend from society to the prison environment (Irwin, 1980,2005;
Marquart, 1986; Steiner & Wooldredge, 2009; Trulson & Marquart, 2002). More
generally, corrections ofﬁcers may perceive, much as court actors might,
racial or ethnic status, as well as age and gender, as markers of increased cul-
pability and dangerousness. The result then would be tougher, more severe
IN-PRISON SANCTION DISPARITIES 5
sanctions for minorities, younger inmates, males, and young minority males in
It is, of course, possible that the prison context negates the types of focal
concerns cues that occur in society. For example, racial threats that exist in
society may not transfer into prisons. From the perception of ofﬁcers, most
inmates, once admitted to prison, may be viewed simply as “cons.” Unlike
court sentencing actors, prison ofﬁcers are removed from the deliberations
about individuals’ guilt or innocence related to their original crime. Instead,
inmates arrive at the prison predesignated as convicted felons. Crouch (1985)
suggests, for example, that in the eyes of prison ofﬁcers, those who are
incarcerated are perceived primarily as “convicts;” all other designations,
such as racial or ethnic characteristics, are secondary. As a result, ofﬁcers
may in practice simply view all inmates the same or as similarly threatening
and culpable. In such a context, if all inmates are viewed equally as
criminals, “extralegal” characteristics that otherwise might serve as cues for
perceptual shorthands that guide determinations of guilt may become irrele-
vant. By contrast, in society, focal concerns processes—such as the applica-
tion of racial attributions—may be more likely to occur precisely because of
the uncertainty that court actors have about an individual’s potential guilt or
This study seeks to contribute to scholarship on incarceration and formal social
control by examining whether minority inmates, and whether young, minority,
males in particular are more likely to be placed in solitary conﬁnement in
response to inmate disciplinary infractions (i.e. disciplinary conﬁnement), net
of “legal” factors, such as prior record and type of infraction committed. We
focus here on disciplinary conﬁnement because it is a frequently used in-prison
sanction (Beck, 2015) and it is analogous to being assigned a prison sentence in
Despite a growing body of literature on solitary conﬁnement and its poten-
tial impacts, existing studies have paid limited attention to in-prison sanction-
ing decisions and the role of race and ethnicity in them. Of these studies, all
but one (Butler & Steiner, in press) focus on relatively small samples that limit
their generalizability. In each instance, no evidence of racial or ethnic dispari-
ties are found. For example, Flanagan (1982), examining an inmate sample
from a northeastern state, used bivariate analyses to test associations between
in-prison punishments and inmate characteristics and found no signiﬁcant
correlation between race and in-prison punishment variation. Later, Crouch
(1985) examined 121 inmates in a single Texas prison facility and found no
impact of race or ethnicity on the severity of in-prison punishments. Similarly,
Howard, Winfree, Mays, Stohr, and Clason (1994) examined 390 disciplinary
6 COCHRAN ET AL.
events in federal prison and, again, found no association between race/
ethnicity and sanction decisions.
Only two other relevant, prior studies exist—they both examined a recent,
large sample of inmates who committed infractions. Speciﬁcally, Butler and
Steiner (in press) used inmate self-report data from the Bureau of Justice
Statistics 2004 Survey of Inmates in State and Federal Correctional Facilities to
identify predictors of self-reported placement in disciplinary segregation
among inmates who reported being found guilty of a rule violation. Their anal-
yses estimated the impact of demographic characteristics and found no statis-
tically signiﬁcant association between race and ethnicity with the likelihood of
placement in solitary conﬁnement. A study by Olson (2016) used the same BJS
data-set but found that black inmates in fact were more likely to report having
spent time in disciplinary segregation. Olson’s analysis differed substantially,
however, in that it did not account for self-reported in-prison misconduct.
Thus, the model estimation did not account for a key confounding inﬂuence—
whether or not an inmate was reported or written up for an in-prison infrac-
tion by ofﬁcers—that has been linked to racial disparities in prior research
(e.g. Poole & Regoli, 1980; for review, see Gonc¸alves et al., 2014).
This discrepancy between the two analyses highlights that, for prison mis-
conduct and any attendant responses, there are at least three points of discre-
tion and, in turn, opportunities for disparities: (1) the decision made by prison
ofﬁcers to write an infraction report when they observe inmate behavior,
(2) the decision, or determination, about whether an inmate is found to be
guilty of the alleged infraction, and (3) the decision to sanction (e.g. to place
an inmate in solitary conﬁnement or not) given a ﬁnding of guilt. Our analysis,
like Butler and Steiner (in press), focuses on the latter, or third, decision. In
the conclusion, we return to these distinctions and emphasize the need for
research on disparities that may arise at each decision point.
Other studies of solitary conﬁnement exist, but have focused primarily on
describing inmates placed in some form of segregation, regardless of whether
it was for punishment, protection, or managerial reasons. For example, Mears
and Bales (2010) explored factors associated with placement in supermax
prisons, but they did not limit their analysis to inmates who committed an
infraction (see also Beck, 2015; Mears, 2013; Schlanger, 2013).
In short, to our knowledge, apart from Butler and Steiner’s (in press) work,
no research has been undertaken that utilizes recent national or statewide
samples and employs rigorous analytical designs to examine predictors of
in-prison punishment decisions (see, however, Steiner & Cain, 2017). Butler
and Steiner (in press) called for work that builds on their analysis by examining
2. One other study by McClellan (1994) examined over 500 infraction sentencing events in two
Texas state prison facilities and examined gender differences in sanctions, but did not examine the
impact of race or ethnicity on sanctioning. In addition, Houser and Belenko (2015) examined in
prison sanctioning outcomes for 211 female prison inmates in Pennsylvania prisons but did not
differentiate use of solitary conﬁnement from other serious in-prison sanctions.
IN-PRISON SANCTION DISPARITIES 7
prison systems’ administrative sanctioning records, as opposed to self-report
data, to understand patterns and potential disparities that emerge in prison
system responses to rule violations. This study seeks to respond to that call
and, more broadly, to advance efforts to understand the exercise of formal
social control and, in particular, the use of disciplinary conﬁnement.
Data and Methods
This study uses prison administrative records data from the Florida Department
of Corrections. The sample includes all inmates admitted to all state prison facil-
ities in Florida between 1 January 2005 and 30 December 2011 who recorded a
disciplinary misconduct between those dates. Thus, we focus our analyses on
sanctioning events linked to all members of this sample for all formal disci-
plinary infraction (DI) events and subsequent formal sanctions assigned by prison
ofﬁcer disciplinary teams in Florida state prisons during this seven-year time per-
iod. The data provide detailed DI information as well as information about the
inmates, including their demographic characteristics and prior criminal records.
In instances when a given inmate contributed more than one DI event, we
focused on the ﬁrst event and removed subsequent infractions from the data ﬁle
so that no individual appeared more than once in the data. The total sample
used in these analyses consists of 89,133 inmate DI events. These events
occurred—are nested—within 167 prison facilities. To account for this cluster-
ing, we utilize multilevel modeling across all of our analyses. Facilities are the
level 2 unit of analysis and inmates are the level 1 unit of analysis.
The focus of our analysis is on inmate placement in solitary conﬁnement in
response to a formal DI being assigned to an inmate. In Florida, if an inmate is
written up for a DI by an ofﬁcer, the infraction report is passed onto a prison
disciplinary team. The disciplinary team typically consists of several ofﬁcers
and is led by a hearing ofﬁcer. Together, the team of ofﬁcers weigh the evi-
dence and related testimony surrounding a given incident. If an inmate is
found guilty, the disciplinary report stays on the inmate’s record and the disci-
plinary team determines appropriate actions, which includes a range of sanc-
tioning options. The most severe sanction is solitary conﬁnement. Alternative
sanctions can include a range of other punishments of lesser severity, including
extra work duty during leisure hours, restricted labor squad, payment for
damages, suspension of work release, visitation restriction, and loss of privi-
leges. [For further details of Florida’s prison sanctioning processes, along with
information on the maximum restrictions state administrative codes set on
8 COCHRAN ET AL.
how solitary conﬁnement is used in response to infractions, see FL Administra-
tive Code rules, including 33.601.307 and 33.601.314 (http://www.dc.state.ﬂ.
The primary dependent variable for our analysis is a dichotomous measure
of placement in solitary conﬁnement for inmates found guilty of a DI
(1 = placement in solitary conﬁnement, and 0 = an alternative, non-solitary
conﬁnement sanction). All DI events included in the data ﬁle and our analyses
were those that resulted in a guilty charge.
Our second dependent variable is a dichotomous measure of length of time
in solitary conﬁnement for inmates who received placement in such conﬁne-
ment as their in-prison sanction. Florida state administrative code provides
guidelines for in-prison sanctioning decisions in the form of maximum limits on
the length of solitary conﬁnement (see FL 33.601.307). These maximums vary
across infraction types and, for the most part, range in value from 15 to
90 days, thus providing opportunities for variation to emerge in sentence
lengths within and across infraction types. Preliminary analyses identiﬁed that
when inmates in Florida receive solitary conﬁnement, disciplinary teams regu-
larly assign sentences of set values. For example, 19 percent of solitary sen-
tence lengths were exactly 15 days, 57 percent were exactly 30 days, and 14
percent were exactly 60 days. Our primary goal for these analyses was to
examine whether disparities emerge in the use of longer versus shorter lengths
of conﬁnement. Thus, to simplify the analysis, we created a dichotomous indi-
cator of “longer” versus “shorter” sentence lengths, where 0 = 30 days or less
and 1 = greater than a 30-day sentence. Views about what constitutes a
“short” versus a “long” stay in conﬁnement vary, but 30 days is a commonly
used threshold (see, e.g. Liman Program & Association of State Correctional
Administrators, 2015). We undertook a series of ancillary analyses using alter-
native outcome measurements and modeling procedures, including continuous
measures of sentence length and various categorical measures of sentence
length as outcome variables. Findings across these analyses were consistent
and substantively the same as those shown below.
Focal Independent Variables
The focus of our analysis is on inmate demographic measures and their
potential inﬂuence on the likelihood of placement and length of time in soli-
tary conﬁnement. Our analyses include three dichotomous measures of race
and ethnicity using the following dichotomous measures: inmate is white
(1 = yes, 0 = no), black (1 = yes, 0 = no), or Hispanic (1 = yes, 0 = no). Black is
the reference category. More recent sentencing work has highlighted the
importance of considering not only race but also ethnicity (Feldmeyer &
Ulmer, 2011; Feldmeyer et al., 2015; Warren et al., 2012). Florida provides an
ideal context in this regard because it contains a large enough Hispanic
population to consider ethnicity in analyses. We also include a dichotomous
IN-PRISON SANCTION DISPARITIES 9
measure of sex (1 = male, 0 = female) and a continuous measure of age.To
examine potential interaction effects, we created multiplicative terms of
race/ethnicity ×age, race/ethnicity ×sex, and race/ethnicity ×age ×sex.
We are able to incorporate a diverse range of covariates useful for accounting
for potential confounding inﬂuences on the association between demographic
characteristics and the likelihood of receiving a solitary conﬁnement place-
ment in response to a DI. The models include several measures of inmates’
prior criminal record. Offense information for each inmates’ primary offense
that led to their prison sentence is measured with the following dichotomous
variables where 1 = yes and 0 = no: primary offense—violent,primary offense
—drug,primary offense—sex, and primary offense—other.Primary offense—
property is used as the reference category. We also include a count measure
of the number of prior prison commitments for each inmate, the number of
months for each inmate’s prison sentence length, and the number of months
of each inmates’ time served prior to their ﬁrst DI.
Not least, we include DI measures for each event. First, we include a count
of DI total charges, which is a count of the number of speciﬁc infractions
associated with each infraction event. This approach mirrors the creation of
multiple criminal charge measures for arrest events in sentencing studies.
Second, we include detailed indicators of the type of infraction each inmate
received. Our measures of DI type stem directly from the infraction designa-
tions utilized by the Florida prison system. Prior studies of inmate misconduct
typically differentiate only between violent and non-violent infractions or use
other limited sets of categories (e.g. violent, property, drug, other; see
Gonc¸alves et al., 2014). These data include detailed infraction information
including, for some infraction types, a differentiation between more (“major”)
or less (“minor”) severe infractions. Each infraction was designated as one of
the following types: DI Violent (major),DI Violent (minor),DI Property
(major),DI Property (minor),DI Contraband (major),DI Contraband (minor),
DI Deﬁance (major),DI Drug,DI Sex,DI Disorder, and DI Regulation Violation.
DI Deﬁance (minor) serves as the reference category.
We conduct a series of analyses that examine the potential role of race, eth-
nicity, age, and gender on the likelihood and length of time in solitary. The
analyses proceed in the following stages. First, we begin with multilevel
regression models that examine the main effects of race, ethnicity, age, and
sex, as well as other potentially relevant covariates, on the likelihood of place-
ment in solitary conﬁnement as a punishment for engaging in misconduct.
10 COCHRAN ET AL.
Second, consistent with prior sentencing research, we conduct a stepwise
series of multilevel regression analyses aimed at examining the inﬂuence of
membership in speciﬁc demographic subgroups. We employ a series of two-
way and three-way interactions between race/ethnicity, sex, and age, which
assess whether there is a youth penalty among males and females and then
whether there is a young, minority penalty among males and females.
Third, we conduct infraction-speciﬁc analyses. For these analyses, we run
separate models for each infraction type that include only those inmates who
committed that type of infraction. The goal of these analyses is to determine
if race and ethnicity, and membership in speciﬁc demographic subgroups,
exert a different effect depending on the type of infraction that inmates com-
mitted. Prior sentencing research suggests, for example, that court actors
have greater discretion in how to charge less severe offense types than they
do more severe offenses, such as violent and sexual offenses. In turn, greater
room for bias may emerge in sentencing individuals convicted of less severe
offenses (e.g. Unnever & Hembroff, 1988). We anticipate that a similar process
may occur in prison sanctioning decisions.
Fourth, we assess whether disparities emerge in the sentence length of
inmates placed in solitary conﬁnement. For these analyses, we employed a
similar set of models to those described above, but examined only those
inmates who were sentenced to solitary conﬁnement. We also conducted addi-
tional analyses using two-stage tobit regression analysis (Bushway, Johnson, &
Slocum, 2007; Johnson & Kurlychek, 2012). The ﬁndings were the same as
those shown below.
Descriptive statistics for the sample and measures are included in Table 1.
Across the 89,133 sanctioning events included in the analysis, 68 percent
resulted in a solitary conﬁnement sentence. Of those who received solitary
conﬁnement, 15 percent of placements had sentences of more than 30 days.
The fact that 68 percent of ﬁrst-time infractions result in solitary placement is
somewhat surprising. Notably, there are no national estimates of how fre-
quently solitary placements are used in response to inmate infractions (see,
however, Butler & Steiner, in press; Morris, 2016). Many accounts of solitary
conﬁnement, however, highlight that it is used frequently in prison and that it
serves many purposes, including not only punishment but also managerial goals
and the protection of certain inmates (Beck, 2015; Browne, Cambier, & Agha,
2011; Frost & Monteiro, 2016; Pizarro & Narag, 2008).
The data and sample consist only of inmates who were found guilty for for-
mal prison rule violations, but the composition of the inmates is largely similar
in composition to most state prison populations (Petersilia, 2003; Travis,
Western, & Redburn, 2014). It is, for example, primarily male (91 percent) and
majority black (51 percent). Thirty-nine percent of the sample is white and 10
IN-PRISON SANCTION DISPARITIES 11
percent are Hispanic. The average age at the time of infraction was 30 years
old. Most inmates were incarcerated for a violent (28 percent), property
(29 percent), or drug crime (24 percent) and the average sentence length for
this population was 71 months. The average DI’s in the sample included a sin-
gle charge (1.10). The most common DI types were deﬁance minor (23 per-
cent), regulation violation (18 percent), deﬁance major (15 percent), disorder
(12 percent), violent minor (9 percent), and contraband minor (9 percent).
We now turn to a series of multivariate, multilevel logistic regression mod-
els. These models account for clustering of inmate infraction events within
prison facilities and predict the likelihood of placement in solitary conﬁnement
Table 1 Descriptive statistics (N= 89,133)
Mean SD Min. Max.
Solitary (disciplinary) conﬁnement 0.68 0.47 0 1
Solitary sentence length (“longer”=1, “shorter”=0) 0.15 0.36 0 1
Black 0.51 0.50 0 1
White 0.39 0.49 0 1
Hispanic 0.10 0.30 0 1
Sex (1 = male, 0 = female) 0.91 0.29 0 1
Age (continuous) 30.27 10.26 13 87
Primary offense—violent 0.28 0.45 0 1
Primary offense—property 0.29 0.46 0 1
Primary offense—drug 0.24 0.43 0 1
Primary offense—sex 0.05 0.22 0 1
Primary offense—other 0.13 0.33 0 1
Prior prison commitment 0.84 1.46 0 16
Sentence length (months) 71.12 106.36 0 600
Time served 7.03 7.59 0 75.28
DI Total charges 1.10 0.41 1 16
DI Violent (major) 0.01 0.10 0 1
DI Violent (minor) 0.09 0.29 0 1
DI Property (major) 0.02 0.13 0 1
DI Property (minor) 0.03 0.17 0 1
DI Contraband (major) 0.03 0.16 0 1
DI Contraband (minor) 0.09 0.29 0 1
DI Deﬁance (major) 0.15 0.36 0 1
DI Deﬁance (minor) 0.23 0.42 0 1
DI Drug 0.05 0.21 0 1
DI Sex 0.01 0.08 0 1
DI Disorder 0.12 0.33 0 1
DI Regulation violation 0.18 0.38 0 1
12 COCHRAN ET AL.
as a result of a DI. Table 2presents three models. Model 1 focuses solely on
inmate demographic characteristics. In line with predictions of focal concerns
theory, we ﬁnd that black inmates are signiﬁcantly more likely than white
inmates to be sanctioned to solitary conﬁnement; no difference surfaced
between Hispanic and black inmates in the likelihood of placement. Similarly,
we ﬁnd that male inmates are substantially more likely than female inmates to
be punished through placement in solitary. And younger inmates were more
likely than older inmates to be sentenced to solitary conﬁnement.
These demographic effects could be explained, however, by variation in
other characteristics, such as prior record and variation in the type of miscon-
duct in which inmates engage (or, for which they are written up). Models 2
and 3 add measures to account for these possibilities. In model 2, we account
for inmates’ prior record and the total charges associated with a given infrac-
tion event. Here we ﬁnd that the race, age, and sex effects hold. In addition,
inmates incarcerated for violent offenses, inmates who have been incarcerated
before, and inmates who have multiple charges associated with their infraction
events are signiﬁcantly more likely to receive solitary conﬁnement placement.
Turning to model 3 in Table 2, we now include the full model speciﬁcation
and account for variation in the type of DI in addition to the other covariates.
In model 3, we ﬁnd that each infraction type has a statistically signiﬁcant asso-
ciation with placement in solitary. Variation in the effect sizes suggests that
substantial heterogeneity exists in the likelihood of solitary conﬁnement place-
ment across infraction types. For example, inmates charged with a violent
major DI are 45 times more likely to be placed in solitary conﬁnement than
those charged with deﬁance minor. Inmates charged with violent minor infrac-
tions are approximately 16 times more likely to be placed in solitary conﬁne-
ment as compared to such inmates. Commission of other infraction types
results in a lower likelihood of being sentenced to solitary conﬁnement, as
compared to commission of acts characterized as “deﬁance minor.” These
infractions include property infraction charges, engaging in minor contraband,
and violating regulations.
These coefﬁcient estimates align with a bivariate analysis of the percent of
inmates placed in solitary conﬁnement across DI categories (not shown). For
example, at the bivariate level, 98 percent of inmates charged with violent
major infractions were placed in solitary compared to 66 percent of inmates
charged with deﬁance minor. Regulation violation, on the other hand, is the
most leniently treated infraction. Only 30 percent of those charged with a
regulation violation are placed in solitary conﬁnement as a result.
Our main focus, however, is on the effects of race, ethnicity, sex, and age.
Notably, the coefﬁcient for race, which is statistically signiﬁcant in models 1
and 2, is no longer signiﬁcant once DI categories are included in the model. It
appears, then, that variation in DI charges explains the statistically signiﬁcant
racial variation in the likelihood of solitary conﬁnement placement. This ﬁnd-
ing parallels that of Mears and Bales (2010), though their analysis focused on
placement in segregation in general, not placement in it as a punishment for
IN-PRISON SANCTION DISPARITIES 13
Table 2 Mixed effects logistic regression of disciplinary conﬁnement on measures of inmate characteristics, prior record, and disciplinary
infractions (N= 89,133 inmates, 167 facilities)
Model 1 Model 2 Model 3
bSE OR bSE OR bSE OR
White −0.211*** 0.02 0.810 −0.191*** 0.02 0.827 −0.012 0.02 0.989
Hispanic −0.025 0.03 0.975 0.000 0.03 0.999 0.084* 0.03 1.087
Sex 0.965*** 0.24 2.643 0.981*** 0.24 2.713 1.110*** 0.28 3.051
Age −0.004*** 0.00 0.996 −0.003*** 0.00 0.997 −0.005*** 0.00 0.995
Primary offense—violent – – – 0.072*** 0.02 1.076 0.031 0.03 1.033
Primary offense—drug – – – −0.060*** 0.02 0.943 −0.043 0.03 0.958
Primary offense—sex – – – −0.021 0.04 0.978 −0.072 0.05 0.930
Primary offense—other – – – 0.041 0.03 1.040 0.035 0.03 1.034
Prior prison comm. – – – 0.021*** 0.01 1.021 0.013 0.01 1.012
Sentence length – – – 0.0004*** 0.00 1.000 0.001*** 0.00 1.001
Time served – – – −0.008*** 0.00 0.992 −0.009*** 0.00 0.991
DI Total charges – – – 1.404*** 0.04 4.070 1.297*** 0.04 3.654
DI Violent (major) – – – – – – 3.812*** 0.23 45.215
DI Violent (minor) – – – – – – 2.794*** 0.06 16.350
DI Property (major) – – – – – – −0.444*** 0.06 0.642
DI Property (minor) – – – – – – −0.387*** 0.05 0.680
DI Contraband (major) – – – – – – 2.100*** 0.08 8.166
DI Contraband (minor) – – – – – – −0.657*** 0.03 0.518
DI Deﬁance (major) – – – – – – 1.469*** 0.03 4.344
DI Drug – – – – – – 3.004*** 0.10 20.169
DI Sex – – – – – – 1.339*** 0.12 3.815
DI Disorder – – – – – – 1.230*** 0.03 3.420
DI Regulation violation – – – – – – −1.577*** 0.03 0.207
Constant 0.433 0.23 – −1.068*** 0.24 – −1.239 0.27 –
Log likelihood −49530.960 −48388.851 −37454.997
Notes. Black, primary offense—property, DI deﬁance (minor), serve as reference variables. ***p< .001, **p< .01, *p< .05.
14 COCHRAN ET AL.
committing an infraction. Statistically signiﬁcant effects for sex and age
persist in model 3, although the effect of one-year increases in inmate age is
substantively small (OR = .995). The effect of gender, however, remains sub-
stantively large—males are roughly three times more likely than females to be
placed in solitary conﬁnement for a DI. A statistically signiﬁcant estimated
effect of Hispanic status emerges in model 3, but the magnitude of effect is
small (OR = 1.087).
Taken together, and contrary to what was hypothesized, the ﬁndings suggest
that race does not directly affect prison system decisions to place inmates in
solitary in response to violations; Hispanics are, though, somewhat more likely
than whites to be sentenced to solitary. Prior research highlights the impor-
tance, however, of investigating the intersection of race and ethnicity with
sex and age to assess whether racial effects may be conditional on the speciﬁc
demographic subgroup to which an individual belongs and to assess whether
young, minority males are particularly more likely to receive placement in soli-
Table 3explores these possibilities by adding two- and three-way interac-
tions between race/ethnicity, sex, and age to the fully speciﬁed model from
above. Inspection of model 1 in Table 3reveals no evidence to support the
three-way interaction or the hypothesis that younger black males or younger
Hispanic males are more likely to be placed in solitary. However, a two-way
interaction between sex and age was statistically signiﬁcant. Model 2 removes
the insigniﬁcant two- and three-way interactions to assess whether the sex-age
interaction remains. Inspection of the model reveals that it does.
To this point, then, the analyses suggest that race and ethnicity have lim-
ited to no effect on prison system decisions to sanction inmates to solitary
conﬁnement in response to rule violations. Similarly, we ﬁnd no evidence of a
young minority male penalty—that is, we do not ﬁnd that young black males
or that young Hispanic males are more likely to be placed in solitary conﬁne-
ment than other demographic subgroups. This ﬁnding runs counter to court
sentencing research that consistently ﬁnds that race and ethnicity, together
with sex and age, appreciably inﬂuence sanctioning decisions. That sex and
age interact, however, points to the possibility that perceptual shorthands, or
cues, may still guide sanctioning decisions.
Figure 1provides two panels that illustrate the limited impact of race and
ethnicity on in-prison sanctioning and, at the same time, they highlight the sal-
ience of sex (males are sanctioned more severely than females) and of the
sex-age interaction that emerges in the models. Panel A is based on estimates
from model 1 in Table 3. It provides the separate estimated probabilities of
placement in solitary conﬁnement, by age, for black, white, and Hispanic
males and females, respectively, holding all other covariates at their means.
The plot lines illustrate the limited variation across race and ethnicity for both
males and females. At the same time, inspection of panel A underscores the
substantively large differences between males and females in the likelihood of
solitary conﬁnement, by age, and also a modest effect of age that is speciﬁc
IN-PRISON SANCTION DISPARITIES 15
to females. Panel B focuses on this two-way interaction (sex ×age) more clo-
sely and provides estimated probabilities based on estimates from model 2 in
Table 3. Here again we see evidence of the gap between males and females
Table 3 Interaction models, logistic regression of disciplinary conﬁnement on
demographics, prior record, and disciplinary infractions covariates (N= 89,133 inmates,
Model 1 Model 2
bSE OR bSE OR
Sex ×Age 0.010* 0.01 1.010 0.007* 0.00 1.007
Sex ×White 0.124 0.22 1.132 – – –
Sex ×Hispanic −0.165 0.44 0.848 – – –
White ×Age 0.007 0.01 1.007 – – –
Hispanic ×Age −0.003 0.01 0.997 – – –
White ×Sex ×Age −0.006 0.00 0.994 – – –
Hispanic ×Sex ×Age 0.002 0.01 1.002 – – –
White −0.183 0.21 0.834 −0.012 0.02 0.988
Hispanic 0.260 0.43 1.295 0.082* 0.03 1.086
Sex 0.810* 0.32 2.256 0.882** 0.30 2.415
Age −0.015** 0.01 0.985 −0.011*** 0.00 0.989
Primary offense—violent 0.029 0.03 1.031 0.032 0.03 1.032
Primary offense—drug −0.044 0.03 0.958 −0.042 0.03 0.959
Primary offense—sex −0.074 0.05 0.928 −0.074 0.05 0.928
Primary offense—other 0.033 0.03 1.032 0.034 0.03 1.035
Prior prison comm. 0.013 0.01 1.012 0.011 0.01 1.011
Sentence length 0.001*** 0.00 1.001 0.001*** 0.00 1.001
Time served −0.009*** 0.00 0.991 −0.009*** 0.00 0.991
DI Total charges 1.295*** 0.04 3.649 1.295*** 0.04 3.651
DI Violent (major) 3.812*** 0.23 45.232 3.811*** 0.23 45.190
DI Violent (minor) 2.795*** 0.06 16.364 2.793*** 0.06 16.335
DI Property (major) −0.442*** 0.06 0.643 −0.443*** 0.06 0.642
DI Property (minor) −0.386*** 0.05 0.680 −0.387*** 0.05 0.679
DI Contraband (major) 2.101*** 0.08 8.176 2.101*** 0.08 8.176
DI Contraband (minor) −0.655*** 0.03 0.519 −0.657*** 0.03 0.518
DI Deﬁance (major) 1.467*** 0.03 4.337 1.467*** 0.03 4.338
DI Drug 3.006*** 0.10 20.213 3.005*** 0.10 20.185
DI Sex 1.327*** 0.12 3.769 1.328*** 0.12 3.772
DI Disorder 1.230*** 0.03 3.422 1.230*** 0.03 3.421
DI Regulation violation −1.578*** 0.03 0.206 −1.577*** 0.03 0.207
Constant −0.941** 0.31 −1.033*** 0.29
Log likelihood −37451.268 −37453.053
Notes. Black, primary offense—property, DI deﬁance (minor), serve as reference variables.
***p< .001, **p< .01, *p< .05.
16 COCHRAN ET AL.
and the sex-speciﬁc inﬂuence of age. For example, all else equal, at the
average age of conﬁnement (age 30), males have an 84 percent probability of
placement in solitary as a punishment for committing an infraction; for
females, this probability is 63 percent. In addition, among females, age exerts
a non-trivial effect on in-prison punishment. Younger female inmates are more
likely than older females to be placed in solitary conﬁnement.
Table 4provides an additional test of potential race and ethnic inﬂuences
on in-prison sanctioning decisions by examining the inﬂuence of demographic
subgroup membership across disciplinary infraction types. Ancillary analyses
(described above) indicated that substantial variation exists across infraction
types in the percentage of infractions that lead to solitary conﬁnement. For
this reason, we use the same models from the earlier tables, but repeat them
for each infraction type. (Accordingly, infraction types are not included as con-
trols.) Sample sizes are noted for each model. Not all infraction types could be
examined due to sample size limitations, which led to exclusion of violent
major, contraband major, deﬁance major, drug, and sex violations. Across the
remaining eight infraction types for which sample sizes permitted the interac-
tional analyses, we ﬁnd limited to no evidence of a statistically or substan-
tively signiﬁcant effect of race or ethnicity on prison punishment decisions or
of an effect of the interaction of speciﬁc race, ethnicity, sex, and age sub-
groups. In two instances (violent minor and disorder), there is a statistically
signiﬁcant coefﬁcient estimate for a Hispanic x sex x age interaction. Plots of
these interactions indicated, however, that these effects were substantively
We turn now to our ﬁnal analysis, which examines variation in the sentence
length—the amount of time—inmates were sanctioned to solitary conﬁne-
ment. We conducted two sets of analyses. First, we created a sample of only
those inmates who received a solitary conﬁnement placement (N= 60,540) and
examined a similar progression of models to those above and that focused on
the dichotomous measure of longer versus shorter sentence lengths. Results
from these analyses are shown in Table 5and largely parallel the ﬁndings
above. The analyses indicate that males and younger inmates sentenced to
solitary conﬁnement are more likely to receive lengthy placements. In addi-
tion, and somewhat surprisingly, white and Hispanic inmates were more likely
to receive lengthy placements than black inmates. The substantive differences
in sentence lengths between racial and ethnic groups are, however, modest.
Similar to the previous sets of models focused on whether inmates received
solitary conﬁnement, no statistically signiﬁcant two- or three-way interactions
3. Our analyses focus on lengths of sentence, which may differ from the amount of time individu-
als actually stayed in solitary conﬁnement. The data did not allow for examining differences
between sentence length and time served in solitary. Such differences constitute an additional
potential disparity that may arise and that warrants attention in future research.
IN-PRISON SANCTION DISPARITIES 17
Second, we conducted a second set of analyses using two-stage tobit regres-
sion analysis (not shown), which is sometimes used in sentencing studies to
model simultaneously the decision to incarcerate and the assignment of sen-
tence length. Substantively similar results as those shown in the previous sets
of analyses emerged.
Panel A: Predicted probabilities of disciplinary confinement, by race/ethnicity, sex, and age (model 1)
Panel B: Predicted probabilities of disciplinary confinement, by gender, and age (model 2)
Figure 1 Predicted probabilities of disciplinary conﬁnement.
18 COCHRAN ET AL.
Table 4 Assessment of race/ethnicity, sex, and age interactions within infraction types
Violent (Minor) Property (Major) Property (Minor) Contraband (Minor)
bSE OR bSE OR bSE OR bSE OR
Sex ×White −1.545 1.01 0.213 1.571 2.80 4.813 −1.911 1.78 0.148 0.966 0.69 2.627
Sex ×Hispanic −4.593 2.57 0.010 5.191 4.72 179.732 −0.080 3.42 0.923 −0.079 1.28 0.924
Sex ×Age −0.046 0.02 0.955 −0.024 0.06 0.976 −0.038 0.04 0.963 0.008 0.02 1.008
White ×Age −0.037 0.03 0.963 0.051 0.08 1.053 −0.041 0.05 0.960 0.017 0.02 1.017
Hispanic ×Age −0.122 0.08 0.885 0.147 0.14 1.159 0.037 0.10 1.037 0.013 0.04 1.013
White ×Sex ×Age 0.063 0.03 1.065 −0.042 0.09 0.959 0.070 0.05 1.073 −0.021 0.02 0.979
Hispanic ×Sex ×Age 0.163* 0.08 1.178 −0.164 0.14 0.848 0.004 0.10 1.004 −0.013 0.04 0.987
White 0.645 0.95 1.905 −1.711 2.78 0.181 0.982 1.75 2.670 −0.619 0.66 0.538
Hispanic 3.411 2.51 30.288 −4.652 4.68 0.010 −0.888 3.37 0.412 0.407 1.25 1.502
Sex 3.196** 1.01 24.437 1.853 2.17 6.381 1.711 1.49 5.535 0.937 0.67 2.552
Age 0.021 0.02 1.021 0.031 0.06 1.031 0.025 0.04 1.025 −0.020 0.02 0.980
Primary off—violent 0.011 0.13 1.011 −0.146 0.16 0.864 −0.077 0.12 0.926 −0.041 0.07 0.960
Primary off—drug 0.159 0.15 1.172 −0.349 0.20 0.705 −0.087 0.13 0.917 0.005 0.07 1.005
Primary off—sex −0.144 0.29 0.866 0.116 0.31 1.123 0.032 0.21 1.032 −0.161 0.12 0.851
Primary off—other −0.320 0.17 0.726 0.094 0.23 1.099 0.064 0.16 1.066 0.079 0.09 1.082
Prior prison commits 0.001 0.05 1.001 0.042 0.07 1.043 0.020 0.03 1.020 0.050* 0.02 1.051
Sentence length 0.003** 0.00 1.003 0.002* 0.00 1.002 0.000 0.00 1.000 0.001** 0.00 1.001
Time served −0.001 0.01 0.999 −0.009 0.01 0.991 −0.017** 0.01 0.984 −0.005 0.00 0.995
DI Total charges 0.941*** 0.20 2.564 1.998*** 0.18 7.373 1.432*** 0.17 4.188 1.851*** 0.10 6.366
Constant 1.003 0.96 −3.570 2.16 −2.049 1.49 −2.358*** 0.66
N(inmates) 8,238 1,545 2,681 8,291
N(facilities) 151 133 150 159
Log likelihood −1492.825 −825.091 −1561.095 −4735.029
IN-PRISON SANCTION DISPARITIES 19
Deﬁance (Major) Deﬁance (Minor) Disorder Regulation Violation
bSE OR bSE OR bSE OR bSE OR
Sex ×White 0.955 0.58 2.599 0.367 0.43 1.443 −1.016 0.56 0.362 0.179 0.52 1.196
Sex ×Hispanic −0.450 1.30 0.638 0.276 0.75 1.317 −1.998 1.08 0.136 1.415 1.14 4.116
Sex ×Age 0.033* 0.01 1.034 −0.012 0.01 1.010 0.003 0.01 1.003 0.015 0.01 1.015
White ×Age 0.026 0.02 1.026 0.010 0.01 1.010 −0.020 0.02 0.980 0.011 0.02 1.012
Hispanic ×Age −0.009 0.04 0.991 0.007 0.02 1.007 −0.048 0.03 0.953 0.036 0.03 1.037
White ×Sex ×Age −0.031 0.02 0.970 0.010 0.01 0.986 0.022 0.02 1.022 −0.003 0.02 0.997
Hispanic ×Sex ×Age 0.009 0.04 1.009 0.007 0.02 0.988 0.073* 0.03 1.076 −0.034 0.03 0.966
White −0.703 0.54 0.495 −0.393 0.41 0.675 0.907 0.52 2.478 −0.407 0.50 0.666
Hispanic 0.447 1.25 1.563 −0.135 0.73 0.874 1.325 1.01 3.764 −1.372 1.11 0.254
Sex 0.501 0.59 1.651 0.621 0.47 1.861 1.489* 0.56 4.433 0.129 0.57 1.138
Age −0.030* 0.01 0.971 −0.012 0.01 0.988 −0.003 0.01 0.997 −0.021 0.01 0.979
Primary off—violent 0.207* 0.08 1.230 0.000 0.05 1.000 0.063 0.08 1.065 0.020 0.06 1.020
Primary off—drug 0.025 0.08 1.025 −0.077 0.05 0.926 0.069 0.08 1.071 −0.093 0.06 0.911
Primary off—sex 0.063 0.15 1.065 −0.112 0.08 0.894 −0.124 0.15 0.883 −0.123 0.10 0.884
Primary off—other 0.273** 0.10 1.314 0.023 0.06 1.023 −0.048 0.10 0.953 −0.014 0.07 0.986
Prior prison commits −0.016 0.03 0.984 0.020 0.01 1.020 0.029 0.02 1.029 −0.008 0.02 0.992
Sentence length 0.001 0.00 1.001 0.000* 0.00 1.000 0.001 0.00 1.001 0.000* 0.00 1.000
Time served −0.023*** 0.00 0.977 −0.014*** 0.00 0.987 −0.017*** 0.00 0.983 0.007* 0.00 1.007
DI Total charges 1.073*** 0.16 2.923 0.957*** 0.09 2.604 1.361*** 0.12 3.899 1.090*** 0.11 2.973
Constant 1.096 0.58 −0.366 0.46 −0.476 0.55 −1.709** 0.56
N(inmates) 13,464 20,209 10,855 15,990
N(facilities) 165 164 158 162
Log likelihood −4112.487 −10964.148 −3848.128 −7799.960
Notes. Black and primary offense—property serve as reference variables. ***p< .001, **p< .01, *p< .05.
Table 4 (Continued)
20 COCHRAN ET AL.
Finally, we conducted a range of ancillary analyses to test the robustness of
these ﬁndings. For example, we explored polynomial speciﬁcations of age and
used different infraction categories, including the aggregation of major and
minor categories of infraction types. In addition, we explored a series of
Table 5 Mixed effects logistic regression of sentence length (1 = “Longer,”
0 = “Shorter”) on inmate characteristics (N= 60,540 inmates, 166 facilities)
Model 1 Model 2
bSE OR bSE OR
Sex ×Age – – – 0.010 0.02 1.010
Sex ×White – – – 0.002 0.62 1.003
Sex ×Hispanic – – – 0.764 1.26 2.141
White ×Age – – – 0.019 0.02 1.019
Hispanic ×Age – – – 0.013 0.04 1.013
White ×Sex ×Age – – – −0.012 0.02 0.988
Hispanic ×Sex ×Age – – – −0.009 0.04 0.991
White 0.259*** 0.03 1.295 0.039 0.61 1.039
Hispanic 0.342*** 0.05 1.408 −0.529 1.25 0.589
Sex 0.894*** 0.21 2.445 0.804 0.54 2.237
Age 0.005** 0.00 1.005 −0.009 0.02 0.991
Primary offense—violent −0.030 0.04 0.971 −0.034 0.04 0.968
Primary offense—drug 0.041 0.04 1.042 0.036 0.04 1.036
Primary offense—sex 0.070 0.07 1.073 0.068 0.07 1.072
Primary offense—other −0.033 0.05 0.967 −0.042 0.05 0.955
Prior prison commitment −0.041** 0.01 0.960 −0.034** 0.01 0.966
Sentence length 0.000 0.00 1.000 0.000 0.00 1.000
Time served −0.001 0.00 0.999 −0.001 0.00 0.999
DI Total charges 0.578*** 0.03 1.783 0.577*** 0.03 1.783
DI Violent (major) 3.796*** 0.09 44.534 3.803*** 0.09 44.832
DI Violent (minor) −1.667*** 0.09 0.189 −1.664*** 0.09 0.189
DI Property (major) 1.450*** 0.08 4.262 1.459*** 0.08 4.299
DI Property (minor) −2.522*** 0.26 0.080 −2.518*** 0.26 0.081
DI Contraband (major) −2.095*** 0.13 0.123 −2.089*** 0.13 0.124
DI Contraband (minor) 3.765*** 0.05 43.147 3.766*** 0.05 43.217
DI Deﬁance (major) 0.475*** 0.04 1.607 0.476*** 0.04 1.609
DI Drug 3.796*** 0.09 44.534 3.803*** 0.09 44.832
DI Sex −1.224*** 0.30 0.294 −1.225*** 0.30 0.294
DI Disorder −3.012*** 0.15 0.049 −3.010*** 0.15 0.049
DI Regulation violation 2.965*** 0.06 19.401 2.970*** 0.06 19.486
Constant −3.988*** 0.21 −3.780*** 0.54
Log likelihood −15362.402 −15354.82
Notes. Black, primary offense—property, DI deﬁance (minor), serve as reference variables.
***p< .001, **p< .01, *p< .05.
IN-PRISON SANCTION DISPARITIES 21
analyses examining subsequent infractions, as opposed to the ﬁrst infraction.
(Doing so requires accounting for the nesting of infraction sentencing events
within individuals and for the prior sanctions associated with earlier infrac-
tions.) We found that variation in whether an individual receives solitary con-
ﬁnement is reduced with each subsequent infraction because the likelihood of
receiving solitary conﬁnement increases with each subsequent infraction.
Across all of these ancillary analyses, the pattern of results were substantively
similar to those identiﬁed in the tables.
Prior formal social control and sentencing theories, along with a large body of
empirical research on criminal court decision-making, underscores the salience
of race and ethnicity in punishment. This paper sought to apply prior punish-
ment research to understand sanctioning that occurs inside prisons and, at the
same time, to contribute to efforts to understand better the use of solitary
conﬁnement. The main ﬁndings can be summarized brieﬂy.
Contrary to what focal concerns and related research on sentencing would
anticipate, we found no consistent evidence of a racial or ethnic penalty in
prison sanctioning decisions. Initial models revealed that black inmates were
more likely to be placed in solitary conﬁnement. However, this effect was
eliminated after controlling for variation in reported commission of infractions.
This ﬁnding thus supports the limited body of prior empirical work (Butler &
Steiner, in press; Crouch, 1985) that indicates that, unlike punishment deci-
sion-making that occurs outside of prison, race and ethnicity may have little
impact at this particular decision point.
At the same time, consistent evidence of gender disparities surfaced. All
else equal, females were less likely than males to receive solitary conﬁnement
as a disciplinary punishment. This ﬁnding aligns with prior punishment research
that identiﬁes patterns of leniency in the sanctioning of females. When viewed
through the lens of focal concerns theory, for example, the ﬁndings accord
with the view that males may be perceived by criminal justice actors as more
dangerous or threatening and thus as warranting harsher sanctions (see, e.g.
Steffensmeier et al., 1998). There is little research that directly examines this
issue as it relates to in-prison sanctioning, but some accounts suggest some
warrant for its salience (for a recent review, see Gonc¸alves et al., 2014). It is
possible, too, that facility bed space in solitary conﬁnement units—a practical
constraint—might explain the gender gap. Female prison facilities, for exam-
ple, tend to operate with less solitary conﬁnement capacity.
Age exerted a consistent, but modest inﬂuence on in-prison sanctioning
decisions for females but not males. The probability of placement in solitary
conﬁnement for young females was greater than for older females and
approached, but still remained lower than, the probability of such placement
among younger males. Although we were not able to test empirically the
22 COCHRAN ET AL.
causes of the age effect for females or the differences in the effect of age
between males and females, a focal concerns perspective can shed light on a
possible cause. For example, in female prisons, younger inmates may be per-
ceived as posing a greater threat to social order and thus may be sanctioned
more harshly. In male prisons, however, age may not be viewed in such a way.
The study ﬁndings run counter to what some studies of sentencing have
found. In particular, for example, no consistent or strong evidence of a three-
way interaction between race and ethnicity, age, and sex emerged. That is,
there was little indication that race or ethnicity exerted a strong direct or
interactive effect, as would be anticipated from some focal concerns or threat
perspectives, on prison punishment decisions.
Not least, the analyses revealed that considerable variation exists in the use
of solitary conﬁnement in response to different types of infractions. For exam-
ple, although 68 percent of infraction events resulted in a placement in soli-
tary conﬁnement, more than 90 percent of violent major, violent minor, and
contraband major infractions resulted in solitary. By contrast, only 30 percent
of regulation violations resulted in solitary conﬁnement. This ﬁnding—that
infraction type is perhaps the primary contributor to in-prison sanctioning
decisions—aligns with earlier empirical work by Crouch (1985) and others (e.g.
Flanagan, 1982), which underscored relative uniformity in sanctioning decisions
across race and ethnic status, but substantial variation based on the severity
of misconduct for which an individual was reported.
Several implications ﬂow from this study. First, the ﬁndings do not support
the hypothesis that race or ethnicity feature prominently in prison sanctioning.
Minority inmates in this study appeared to receive substantively similar punish-
ments as whites. What might explain the seemingly reduced salience of race
and ethnicity in the prison context? Although it goes beyond our analysis, one
possibility is that race and ethnicity provide limited impact on ofﬁcers’ percep-
tions of who is most dangerous or guilty because all prison inmates are con-
victed felons and thus, in the eyes of prison ofﬁcers, already culpable. Crouch
(1985), in an earlier study of in-prison sanctioning, referred to this possibility
as “universalism” in prison ofﬁcers’ perceptions of prison inmates, and sug-
gested that inmates’ convict status may supersede any stigma stemming from
minority status. More research is needed, then, that measures how prison ofﬁ-
cers view different groups of inmates and what characteristics or actions
among these inmates may most affect ofﬁcers’ perceptions. Such research
would provide direct measurements of the mechanisms theorized by focal con-
cerns and other attributions perspectives (Albonetti & Hepburn, 1996; Bridges
& Steen, 1998), and in doing so, provide further insight into whether focal con-
cerns can be applied to in-prison sanctioning, and if it can, test further
whether race operates differently across sentencing contexts.
Second, the ﬁndings here should not be construed to suggest that racial and
ethnic disparities do not exist in prisons. There is, for example, a large litera-
ture that suggests that important inequalities exist in prisons (see, e.g. Case &
Fasenfest, 2004; Cochran, Mears, Bales, & Stewart, 2016; Goodman, 2008;
IN-PRISON SANCTION DISPARITIES 23
Massoglia, 2008; Patterson, 2015). It may well be that the disparities surface,
for example, prior to sanctioning and in ways that administrative records data
do not reveal. To illustrate, prison conditions may disadvantage black inmates
in ways that contribute to misconduct: Ofﬁcers may discriminate against them,
the programs and services available to them may be minimal, their prison
placements may be farther from their home communities, and so on (Mears &
In the context of prison misconduct and prison administrative responses to
it, the punishment decision, which was the focus of this paper, is one of at
least three decisions, or opportunities, where race, ethnicity, gender, and age
may have impacts and thus lead to disparate experiences. Research is needed
that can also focus on actions that occur prior to in-prison sentencing deci-
sions, including research on disparities in ofﬁcers’ decision to record an infrac-
tion for any given inmate as well as disparities in whether inmates are found
guilty after being written up for infractions. In addition, there is a need for
research that parallels policing research (see, e.g. Lytle, 2014; Reisig,
McCluskey, Mastrofski, & Terrill, 2004; Visher, 1983) in examining how inmate
characteristics shape prisons’ and prison ofﬁcers’ decisions. Ofﬁcers may, for
example, discriminate against minorities by recording certain behaviors as
infractions that would not be recorded, or may be less likely to be recorded,
as such for white inmates (e.g. Poole & Regoli, 1980). Here, inequalities exist
and are masked by seemingly race-neutral sanctioning that occurs given an
infraction (Mears & Bales, 2010). The fact of a prior sanction to solitary then
may create a record that is used to justify subsequent punishment decisions.
In so doing, cumulative disadvantages may arise that are concentrated among
minority inmates (see, e.g. Frase, 2009; Kutateladze et al., 2014; Tonry &
Melewski, 2008; Wooldredge et al., 2015).
Third, this study identiﬁes that there may not only be variation in the use of
solitary conﬁnement in general (Beck, 2015; Frost & Monteiro, 2016), there
also may be variation in the use of solitary conﬁnement as a form of punish-
ment. Solitary, or segregation, has been used for many purposes, such as the
protection of inmates or as a management strategy for creating greater order
in prisons (Mears, 2013). Studies are needed that identify whether the factors
that contribute to placement in solitary for these other reasons vary from
those associated with placement in it for punishment.
Fourth, future research should assess the extent to which different contexts
inﬂuence in-prison punishment. Recent sentencing research has highlighted the
potential importance of contextual factors on sentencing (e.g. Johnson, 2006;
Ulmer & Johnson, 2004) and prison studies have emphasized the potential
impact of prison facility characteristics on individuals’ experiences in prison
(e.g. Steiner & Wooldredge, 2008; Wooldredge, Grifﬁn, & Pratt, 2001). Going
forward, studies are needed that integrate these two areas of research. That
is, research is needed that can assess whether variation in facility characteris-
tics, such as security level, racial composition, solitary conﬁnement bed space
or capacity, and inmate-to-ofﬁcer ratio inﬂuence prisons’ usage of solitary
24 COCHRAN ET AL.
conﬁnement in response to infractions. In our analyses, for example, we used
multilevel modeling to account statistically for nesting of inmates in different
prison facilities, but we could not measure speciﬁc characteristics of each
facility. Estimates of contextual inﬂuences might, for example, reveal circum-
stances under which racial and ethnic disparities in sanctioning do emerge and
they might also help to explain the gender and age differences identiﬁed in
Finally, although this study did not examine whether solitary conﬁnement
exerts a beneﬁcial or harmful impact on inmates, additional research is needed
that can better identify any such impacts. Such research is needed to assess
whether solitary conﬁnement for purposes of punishment is effective. It also is
needed because, to the extent that solitary conﬁnement helps or harms, such
effects will systematically occur more among those groups for whom conﬁne-
ment-as-punishment disproportionately occurs.
Potential beneﬁts include reduced misconduct and recidivism (see, e.g.
Lovell, Johnson, & Cain, 2007; Mears & Bales, 2009; Medrano, Ozkan, & Morris,
2017; Morris, 2016; Pizarro, Zgoba, & Haugebrook, 2014; see, generally, Garcia,
2016). Solitary conﬁnement for purposes of punishment would appear unlikely
to improve inmate mental health. However, debate exists about whether it
harms, and this debate in part has made the use of solitary conﬁnement a light-
ning rod for controversy. On the one hand, a large literature exists that suggests
that such conﬁnement may adversely affect inmates’ mental health (see, e.g.
Andersen et al., 2000; Arrigo & Bullock, 2008; Cloyes, Lovell, Allen, & Rhodes,
2006; Grassian, 1983; Haney, 1993, 2003; Kaba et al., 2014; Romano, 1996;
Singer, 1971). On the other hand, recent reviews and studies raise questions
about the methodological rigor of much of the prior work and, at the same time,
suggest that solitary conﬁnement may not harm the mental health of inmates
(see, e.g. Gendreau & Labrecque, in press; Morgan et al., 2016).
Although not the focus of this study, it bears emphasizing that uncertainty
about the effects of solitary conﬁnement for punishment is paralleled by uncer-
tainty about the potential effects of solitary conﬁnement for purposes of pro-
tecting inmates or of achieving any of a range of other managerial goals, such as
promoting systemwide safety and order or reducing the inﬂuence of gangs
(Mears, 2016). There is, in short, a considerable need for more research aimed
at identifying disparities in the use of solitary conﬁnement for any of a range of
purposes and in determining the effects of such conﬁnement. There is, too, a
need for policies and oversight that can ensure that prison punishments are fair,
warranted, and effective and that, at the same time, solitary conﬁnement in
general occurs in a way that also is fair, warranted, and effective.
We thank Sonja Siennick for providing helpful guidance with the analyses, Eric
Stewart for his suggestions, and the Editor and anonymous reviewers for
IN-PRISON SANCTION DISPARITIES 25
recommendations to improve the paper. We also thank the Florida Department
of Corrections (FDC) for providing the data. All views expressed here are those
of the authors only, not of the FDC. A version of this paper was presented at
the 2016 annual meeting for the Academy of Criminal Justice Sciences held in
No potential conﬂict of interest was reported by the authors.
Albonetti, C. A. (1997). Sentencing under the federal sentencing guidelines: Effects of
defendant characteristics, guilty pleas, and departures on sentence outcomes for
drug offenses, 1991-1992. Law and Society Review, 31, 789-822.
Albonetti, C. A., & Hepburn, J. R. (1996). Prosecutorial discretion to defer criminaliza-
tion: The effects of defendant’s ascribed and achieved status characteristics. Journal
of Quantitative Criminology, 12, 63-81.
American Civil Liberties Union. (2014). The dangerous overuse of solitary conﬁnement
in the United States. New York, NY: Author.
Amnesty International. (2012). Cruel isolation: Amnesty international’s concerns about
conditions in Arizona maximum security prisons. London: Amnesty International.
Andersen, H. S., Sestoft, D., Lillebaek, T., Gabrielsen, G., Hemmingsen, R., & Kramp, P.
(2000). A longitudinal study of prisoners on remand: Psychiatric prevalence, incidence
and psychopathology in solitary vs. non-solitary conﬁnement. Acta Psychiatrica
Scandinavica, 102, 19-25.
Arrigo, B. A., & Bullock, J. L. (2008). The psychological effects of solitary conﬁnement
on prisoners in supermax units: Reviewing what we know and recommending what
should change. International Journal of Offender Therapy and Comparative Crimi-
nology, 52, 622-640.
Baumer, E. (2013). Reassessing and redirecting research on race and sentencing. Justice
Quarterly, 30, 231-261.
Beck, A. J. (2015). Use of restrictive housing in U.S. prisons and jail, 2011-12. Washing-
ton, DC: Bureau of Justice Statistics.
Brennan, P. K. (2006). Sentencing female misdemeanants: An examination of the direct
and indirect effects of race/ethnicity. Justice Quarterly, 23, 60-95.
Bridges, G. S., & Steen, S. (1998). Racial disparities in ofﬁcial assessments of juvenile
offenders: Attributional stereotypes as mediating mechanisms. American Sociological
Review, 63, 554-570.
Browne, A., Cambier, A., & Agha, S. (2011). Prisons within prisons: The use of segrega-
tion in the United States. Federal Sentencing Reporter, 24, 46-49.
Bushway, S., Johnson, B. D., & Slocum, L. A. (2007). Is the magic still there? The use of
the heckman two-step correction for selection bias in criminology. Journal of
Quantitative Criminology, 23, 151-178.
Butler, H. D., & Steiner, B. (in press). Examining the use of disciplinary segregation
within and across prisons. Justice Quarterly,34, 248-271.
Case, P., & Fasenfest, D. (2004). Expectations for opportunities following prison educa-
tion: A discussion of race and gender. Journal of Correctional Education, 55, 24-39.
26 COCHRAN ET AL.
Cloyes, K. G., Lovell, D., Allen, D. G., & Rhodes, L. A. (2006). Assessment of psychoso-
cial impairment in a supermaximum security unit sample. Criminal Justice and
Behavior, 33, 760-781.
Cochran, J. C., Mears, D. P., Bales, W. D., & Stewart, E. A. (2016). Spatial distance,
community disadvantage, and racial and ethnic variation in prison inmate access to
social ties. Journal of Research in Crime and Delinquency, 53, 22-254.
Crawford, C., Chiricos, T., & Kleck, G. (1998). Race, racial threat, and sentencing of
habitual offenders. Criminology, 36, 481-512.
Crouch, B. M. (1985). The signiﬁcance of minority status to discipline severity in prison.
Sociological Focus, 18, 221-233.
Demuth, S. (2003). Racial and ethnic differences in pretrial release decisions and out-
comes: A comparison of Hispanic, black, and white felony arrestees. Criminology,
Demuth, S., & Steffensmeier, D. (2004). Ethnicity effects on sentence outcomes in large
urban courts: Comparisons among white, black, and Hispanic defendants. Social
Science Quarterly, 85, 994-1011.
Feldmeyer, B., & Ulmer, J. T. (2011). Racial/ethnic threat and federal sentencing.
Journal of Research in Crime and Delinquency, 48, 238-270.
Feldmeyer, B., Warren, P. Y., Siennick, S. E., & Neptune, M. (2015). Racial, ethnic, and
immigrant threat. Journal of Research in Crime and Delinquency, 52, 62-92.
Flanagan, T. J. (1982). Discretion in the prison justice system. Journal of Research in
Crime and Delinquency, 19, 216-237.
Frase, R. S. (2009). What explains persistent racial disproportionality in minnesota’s
prison and jail populations? Crime and Justice, 38, 201-280.
Frenzel, E. D., & Ball, J. D. (2008). Effects of individual characteristics on plea negotia-
tions under sentencing guidelines. Journal of Ethnicity in Criminal Justice, 5, 59-82.
Frost, N., & Monteiro, C. E. (2016). Administrative segregation in U.S. prisons.
Washington, DC: U.S. Department of Justice, Ofﬁce of Justice Programs, National
Institute of Justice.
Garcia, M. (Ed.). (2016). Restrictive housing in the U.S.: Issues, challenges, and future
directions. Washington, DC: U.S. Department of Justice, Ofﬁce of Justice Programs,
National Institute of Justice.
Gendreau, P., & Labrecque, R. M. (in press). The effects of administrative segregation:
A lesson in knowledge cumulation. In J. Wooldredge & P. Smith (Eds.), Oxford hand-
book on prisons and imprisonment. New York, NY: Oxford University Press.
Gonc¸alves, L. C., Gonc¸alves, R. A., Martins, C., & Dirkzwager, A. J. E. (2014). Predict-
ing infractions and health care utilization in prison. Criminal Justice and Behavior,
Goodman, P. (2008). “It’s just black, white, or hispanic”: An observational study of
racializing moves in California’s segregated prison reception centers. Law & Society
Review, 42, 735-770.
Grassian, S. (1983). Psychopathological effects of solitary conﬁnement. American
Journal of Psychiatry, 140, 1450-1454.
Hagan, J. (1974). Extra-legal attributes and criminal sentencing: An assessment of a
sociological viewpoint. Law and Society Review, 8, 357-383.
Haney, C. (1993). Infamous punishment: The psychological effects of isolation. National
Prison Project Journal, 8, 3-21.
Haney, C. (2003). Mental health issues in long-term solitary and “supermax” conﬁne-
ment. Crime and Delinquency, 49, 124-156.
Harris, A. (2009). Attributions and institutional processing: How focal concerns guide
decision-making in the juvenile court. Race and Social Problems, 1, 243-256.
IN-PRISON SANCTION DISPARITIES 27
Houser, K., & Belenko, S. (2015). Disciplinary responses to misconduct among female
prison inmates with mental illness, substance use disorders, and co-occurring disor-
ders. Psychiatric Rehabilitation Journal, 38, 24-34.
Howard, C. J., Winfree, T. L., Jr., Mays, G. L., Stohr, M. K., & Clason, D. L. (1994).
Processing inmate disciplinary infractions in a federal correctional institution:
Legal and extralegal correlates of prison-based legal decisions. The Prison Journal,
Irwin, J. (1980). Prisons in turmoil. Boston, MA: Little, Brown.
Irwin, J. (2005). The warehouse prison: Disposal of the new dangerous class. Los
Angeles, CA: Roxbury Press.
Johnson, B. D. (2006). The multilevel context of criminal sentencing: Integrating judge-
and county-level inﬂuences. Criminology, 44, 259-298.
Johnson, B. D., & Kurlychek, M. C. (2012). Transferred juveniles in the era of sentenc-
ing guidelines: Examining judicial departures for juvenile offenders in adult criminal
court. Criminology, 50, 525-564.
Jordan, K. L., & Freiburger, T. L. (2015). The effect of race/ethnicity on sentencing:
Examining sentence type, jail length, and prison length. Journal of Ethnicity in
Criminal Justice, 13, 179-196.
Kaba, F., Lewis, A., Glowa-Kollisch, S., Hadler, J., Lee, D., Alper, H., … Venters, H.
(2014). Solitary conﬁnement and risk of self-harm among jail inmates. American
Journal of Public Health, 104, 442-447.
Kahneman, D. (2011). Thinking, fast and slow. New York, NY: Farrar, Straus, and
Kautt, P., & Spohn, C. (2002). Crack-ing down on black drug offenders? Testing for
interactions among offenders’ race, drug type, and sentencing strategy in federal
drug sentences. Justice Quarterly, 19, 1-35.
Kramer, J. H., & Ulmer, J. T. (2009). Sentencing guidelines: Lessons from Pennsylvania.
Boulder, CO: Lynne Rienner.
Kutateladze, B. L., Andiloro, N. R., Johnson, B. D., & Spohn, C. C. (2014). Cumulative
disadvantage: Examining racial and ethnic disparity in prosecution and sentencing.
Criminology, 52, 514-551.
Levinson, J. D. (2007). Forgotten racial equality: Implicit bias, decisionmaking, and mis-
remembering. Duke Law Journal, 57, 345-424.
Liman Program & Association of State Correctional Administrators. (2015). Time-in-cell:
ASCA-Liman 2014 National Survey of Administrative Segregation in Prison.
New Haven, CT: Yale Law School.
Lovell, D., Johnson, L. C., & Cain, K. C. (2007). Recidivism of supermax prisoners in
Washington state. Crime & Delinquency, 53, 633-656.
Lytle, D. J. (2014). The effects of suspect characteristics on arrest: A meta-analysis.
Journal of Criminal Justice, 42, 589-597.
Marquart, J. W. (1986). Prison guards and the use of physical coercion as a mechanism
of prisoner control. Criminology, 24, 347-366.
Massoglia, M. (2008). Incarceration, health, and racial disparities in health. Law &
Society Review, 42, 275-306.
McClellan, D. S. (1994). Disparity in the discipline of male and female inmates in Texas
prisons. Women & Criminal Justice, 5, 71-97.
Mears, D. P. (2013). Supermax prisons. Criminology and Public Policy, 12, 681-719.
Mears, D. P. (2016). Critical research gaps in understanding the effects of prolonged
time in restrictive housing on inmates and the institutional environment. In
M. Garcia (Ed.), Restrictive housing in the U.S.: Issues, challenges, and future direc-
tions (pp. 233-295). Washington, DC: National Institute of Justice.
28 COCHRAN ET AL.
Mears, D. P., & Bales, W. D. (2009). Supermax incarceration and recidivism. Criminol-
ogy, 47, 1131-1166.
Mears, D. P., & Castro, J. L. (2006). Wardens’ views on the wisdom of supermax
prisons. Crime & Delinquency, 52, 398-431.
Mears, D. P., & Bales, W. D. (2010). Supermax housing: Placement, duration, and time
to reentry. Journal of Criminal Justice, 38, 545-554.
Mears, D. P., & Reisig, M. D. (2006). The theory and practice of supermax prisons.
Punishment & Society, 8, 33-57.
Medrano, J. A., Ozkan, T., & Morris, R. (in press). Solitary conﬁnement exposure and
capital inmate misconduct. American Journal of Criminal Justice.
Metzner, J. L., & Fellner, J. (2010). Solitary conﬁnement and mental illness in U.S. pris-
ons: A challenge for medical ethics. Journal of the American Academy of Psychiatry
and the Law, 38, 104-108.
Mitchell, O. (2005). A meta-analysis of race and sentencing research: Explaining the
inconsistencies. Journal of Quantitative Criminology, 21, 439-466.
Morgan, R. D., Gendreau, P., Smith, P., Gray, A. L., Labrecque, R. M., MacLean, N., …
Mills, J. F. (2016). Quantitative syntheses of the effects of administrative segregation
on inmates’ well-being. Psychology, Public Policy, and Law, 22, 439-461.
Morris, R. G. (2016). Exploring the effect of exposure to short-term solitary conﬁnement
among violent prison inmates. Journal of Quantitative Criminology, 32, 1-22.
Obama, B. B. (2016, January 25). Barack Obama: Why we must rethink solitary conﬁne-
ment. The Washington Post. Retrieved May 1, 2016, from https://www.washington
Olson, J. C. (2016). Race and punishment in American prison. Journal of Public Adminis-
tration Research and Theory, 26, 758-768.
Patterson, E. J. (2015). Hidden disparities: Decomposing inequalities in time served in
California, 1985-2009. Law & Society Review, 49, 467-498.
Petersilia, J. (2003). When prisoners come home: Parole and prisoner reentry. New
York, NY: Oxford University Press.
Pizarro, J. M., & Narag, R. E. (2008). Supermax prisons: What we know, what we do not
know, and where we are going. The Prison Journal, 88, 23-42.
Pizarro, J. M., Zgoba, K. M., & Haugebrook, S. (2014). Supermax and recidivism: An
examination of the recidivism covariates among a sample of supermax ex-inmates.
The Prison Journal, 94, 180-197.
Poole, E., & Regoli, R. (1980). Race, institutional rule breaking, and disciplinary
response: A study of discretionary decision making in prison. Law & Society Review,
Reisig, M. D., McCluskey, J. D., Mastrofski, S. D., & Terrill, W. (2004). Suspect disre-
spect toward the police. Justice Quarterly, 21, 241-268.
Romano, S. M. (1996). If the SHU ﬁts: Cruel and unusual punishment at California’s Peli-
can Bay State prison. Emory Law Journal, 45, 1089-1090.
Schlanger, M. (2013). Prison segregation: Symposium introduction and preliminary data
on racial disparities. Michigan Journal of Race and Law, 18, 1-12.
Schlesinger, T. (2005). Racial and ethnic disparity in pretrial criminal processing.
Justice Quarterly, 22, 170-192.
Singer, R. G. (1971). Conﬁning solitary conﬁnement: Constitutional arguments for a
“new penology”. Iowa Law Review, 56, 1251-1258.
Spohn, C. (2015). Evolution of sentencing research. Criminology and Public Policy, 14,
Spohn, C., & Holleran, D. (2000). The imprisonment penalty paid by young, unemployed
black and Hispanic male offenders. Criminology, 38, 281-306.
IN-PRISON SANCTION DISPARITIES 29
Steffensmeier, D., Ulmer, J., & Kramer, J. (1998). Interaction of race, gender, and age
in criminal sentencing: The punishment cost of being young, black, and male.
Criminology, 36, 763-798.
Steffensmeier, D. J. (1980). Assessing the impact of the women’s movement on
sex-based differences in the handling of adult criminal defendants. Crime and Delin-
quency, 26, 344-357.
Steffensmeier, D. J., & Demuth, S. (2000). Ethnicity and sentencing outcomes in U.S.
federal courts: Who is punished more harshly? American Sociological Review, 65,
Steiner, B., & Cain, C. M. (2017). Punishment within prison: An examination of the inﬂu-
ences of prison ofﬁcials’ decisions to remove sentencing credits. Law and Society
Review, 51, 70-98.
Steiner, B., & Wooldredge, J. (2008). Inmate versus environmental effects on prison
rule violations. Criminal Justice & Behavior, 35, 438-456.
Steiner, B., & Wooldredge, J. (2009). The relevance of inmate race/ethnicity versus
population composition for understanding prison rule violations. Punishment & Soci-
ety, 11, 459-489.
Tonry, M., & Melewski, M. (2008). The malign effects of drug and crime control policies
on black Americans. Crime and Justice, 37, 1-44.
Travis, J., Western, B., & Redburn, S. (Eds.). (2014). The growth of incarceration in
the United States. Washington, DC: The National Academies Press.
Trulson, C. R., & Marquart, J. W. (2002). Inmate racial integration: Achieving racial
integration in the Texas prison system. The Prison Journal, 82, 498-525.
Ulmer, J. T. (2012). Recent developments and new directions in sentencing research.
Justice Quarterly, 29, 1-40.
Ulmer, J. T., & Johnson, B. (2004). Sentencing in context: A multilevel analysis. Crimi-
nology, 42, 137-178.
Unnever, J. D., & Hembroff, L. A. (1988). The prediction of racial/ethnic sentencing
disparities: An expectation states approach. Journal of Research in Crime and Delin-
quency, 25, 53-82.
Useem, B., & Piehl, A. M. (2008). Prison state. New York, NY: Cambridge University
Visher, C. A. (1983). Gender, police arrest decisions, and notions of chivalry.
Criminology, 21, 5-28.
Vogel, M., & Porter, L. C. (2016). Toward a demographic understanding of incarceration
disparities: Race, ethnicity, and age structure. Journal of Quantitative Criminology,
Wang, X., & Mears, D. P. (2010). Examining the direct and interactive effects of
changes in racial and ethnic threat on sentencing decisions. Journal of Research in
Crime and Delinquency, 47, 522-557.
Wang, X., Mears, D. P., Spohn, C., & Dario, L. (2013). Assessing the differential effects
of race and ethnicity on sentence outcomes under different sentencing systems.
Crime and Delinquency, 59, 87-114.
Warren, P., Chiricos, T., & Bales, W. (2012). The imprisonment penalty for young black
and Hispanic males: A crime-speciﬁc analysis. Journal of Research in Crime and
Delinquency, 49, 56-80.
Western, B. (2006). Punishment and inequality in America. New York, NY: Russell Sage
Wooldredge, J. (2012). Distinguishing race effects on pre-trial release and sentencing
decisions. Justice Quarterly, 29, 41-75.
30 COCHRAN ET AL.
Wooldredge, J., Frank, J., Goulette, N., & Travis, L., III (2015). Is the impact of cumu-
lative disadvantage on sentencing greater for black defendants? Criminology and
Public Policy, 14, 187-223.
Wooldredge, J., Grifﬁn, T., & Pratt, T. (2001). Considering hierarchical models for
research on inmate behavior: Predicting misconduct with multilevel data. Justice
Quarterly, 18, 203-231.
Zatz, M. S. (2000). The convergence of race, ethnicity, gender, and class on court deci-
sion making: Looking toward the 21st century. In J. Horney (Ed.), Policies, processes
and decisions in the criminal justice system (pp. 503-552). Washington, DC: U.S.
Department of Justice, Ofﬁce of Justice Programs, National Institute of Justice.
IN-PRISON SANCTION DISPARITIES 31