R E S E A R C H
R E P O R T
March 2006

Evaluating the
Effectiveness of
Supermax Prisons
Daniel P. Mears, Florida State University

research for safer communities

URBAN INSTITUTE

Justice Policy Center

Acknowledgments
This report was written by Daniel P. Mears, an associate professor at Florida State University
and previously a senior research associate in the Justice Policy Center at the Urban Institute (UI).
Jamie Watson, formerly a research associate at the Institute and now an independent consultant,
provided assistance with key sections of the report, including the discussion of data and methods,
and the creation, summary, and analysis of the matrix of supermax goals and impacts. Jennifer L.
Castro, a research associate at the Institute, conducted the statistical analyses of the survey data
and assisted with summarizing and analyzing the results. Sarah Lawrence, formerly a research
associate at the Institute, assisted with writing a benefit-cost analysis (BCA) policy brief on
supermax prisons and developing an accompanying spreadsheet-based tool to illustrate how
BCAs of supermax prisons can be conducted. Vera Kachnowski, a research associate at the
Institute, provided invaluable assistance in editing the BCA policy brief and finalizing the
development of the BCA tool. John Roman, a senior research associate at the Institute, assisted
with development, testing, and review of the BCA tool. Michael D. Reisig, an associate
professor at Florida State University, collaborated with the author in developing a greater
understanding of supermax prisons and their ability to achieve systemwide order.
Dr. Mears is the principal investigator of Supermax Prisons: Examining the Impacts and
Developing a Cost-Benefit Model, a project at the Urban Institute. The project (# 2002-IJ-CX0019) was funded by the National Institute of Justice (NIJ) and housed within the Institute’s
Justice Policy Center. Andrew Goldberg, the NIJ Program Monitor for the project, provided
invaluable support throughout the project. Many colleagues at the Urban Institute, especially
Lisa Brooks, Adele Harrell, John Roman, Jeremy Travis, and Christy Visher, gave generously of
their time and offered helpful insights and suggestions. Christy graciously reviewed the entire
report and provided many ideas for improving it. The author gratefully acknowledges the support
provided by the correctional administrators and wardens, local and state policymakers, and
researchers who agreed to be interviewed for this study and who provided invaluable insights
about supermax prisons. During all stages of the research study, the UI research team asked for
feedback and assistance from many people and invariably received prompt and helpful advice.
Our thanks to these people, including Gerald Berge, Michele Moczygemba Connolly, James
Gondles Jr., Charles Hinsley, Todd Ishee, Dimitria Pope, Larry Reid, Chase Riveland, Allen
Scribner, William Stickman, and Reginald Wilkinson. We also thank James Austin and Gerald
Gaes for their advice in the early stages of the project. William Johnson, Terry Wells, and
Rodney Henningsen kindly provided a copy of the survey instrument they used in their national
survey of prison wardens. We modified several items from this instrument in the UI survey.

Points of view in this report are those of the author and do not necessarily represent the
official position or policies of the U.S. Department of Justice or of the Urban Institute, its board
of trustees, or its sponsors.

i

Executive Summary1
Twenty years ago, super- maximum-security prisons were rare in America. As of 1996, over
two-thirds of states had “supermax” facilities that collectively housed more than 20,000 inmates.
Based on the present study, however, as of 2004, 44 states had supermax prisons. Designed to
hold the putatively most violent and disruptive inmates in single-cell confinement for 23 hours
per day, often for an indefinite period of time, these facilities have been lightning rods for
controversy. Economic considerations are one reason—supermaxes typically cost two or three
times more to build and operate than traditional maximum security prisons. A perhaps bigger
reason lies in the criticism by some that supermax confinement is unconstitutional and inhumane.
While proponents and opponents of supermax prisons debate such issues, a fundamental set of
questions has gone largely unexamined: What exactly are the goals of supermax prisons? How, if
at all, are these goals achieved? And what are their unintended impacts?
The Urban Institute, with funding from the National Institute of Justice, conducted a study to
help answer these questions with the goal of creating a foundation that would stimulate more
informed and balanced research and policy discussions about supermax prisons. The study drew
on several sources of information—a comprehensive review of correctional agency and
legislative documents, and theoretical and evaluation research on supermax prisons; interviews
with legislators, corrections officials, wardens, and corrections officers; site visits to three states;
and a national survey of state prison wardens. Among the study’s key findings:
•

Despite disagreements among some scholars and practitioners concerning the definition
of a supermax, over 95 percent of state prisons wardens agreed with a modified version
of the definition used by the National Institute of Corrections in its 1996 survey of state
correctional systems. (The definition: A supermax is a stand-alone unit or part of another
facility and is designated for violent or disruptive inmates. It typically involves up to 23hour-per-day, single-cell confinement for an indefinite period of time. Inmates in
supermax housing have minimal contact with staff and other inmates.)

•

In 1996, 34 states reported to the National Institute of Corrections that they had supermax
prisons. Based on the Urban Institute survey respondents who self- identified as supermax
wardens, as of 2004, 44 states had supermaxes housing approximately 25,000 inmates
(Mears 2005, 7).

•

Considerable differences of opinion exist about the stated or perceived goals of supermax
prisons. Among wardens nationally, however, there is substantial (over 95 percent)
agreement that supermax prisons serve to achieve at least four critical goals—increasing
safety, order, and control throughout prison systems and incapacitating violent or
disruptive inmates. There is less agreement about whether they improve inmate behavior
throughout prison systems; decrease riots, the influence of gangs, or escapes; successfully
punish, reduce the recidivism of, or rehabilitate violent or disruptive inmates; or deter
crime in society.

•

The logic by which supermax prisons achieve each of a range of goals remains unclear.
Do such prisons, for example, increase system order, and, if they do, does the effect arise

1

This summary draws on Mears (2005), “A Critical Look at Supermax Prisons,” in Corrections
Compendium.

ii

through incapacitation, general deterrence of nonsupermax inmates, or some other
mechanism? Current theory and research provide little foothold for answering such
questions as they relate to the diverse goals associated with supermaxes.
•

Interview, site visit, and survey respondents, as well as published accounts in the
literature, point to a wide range of unintended effects of supermax prisons. Some of them
may be relatively rare or benign, but many, such as increased mental illness, raise
substantial concerns. At the same time, respondents identified positive unintended effects
of supermaxes, such as improving living conditions and outcomes for general population
inmates; these effects might offset such concerns or at least broaden the justification for
investing in supermaxes.

•

States generally have not conducted benefit-cost analyses of their supermaxes prior to or
after investing in them. It thus remains unclear whether the benefits of these prison
facilities outweigh their costs. That uncertainty increases when unintended effects are
taken into account.

•

Balanced assessments of supermax prisons require reference to their full range of goals
(weighted by the importance of specific goals to which some states may give greater
priority), as well as to their unintended effects; alternatives that may be equally or more
effective; and the political, moral, and economic dimensions of supermaxes as a
correctional policy.

•

Among the most critical unanswered questions about supermaxes is their effect on
prisoner reentry. Are supermax inmates less or more likely to reoffend upon release from
prison? To obtain housing and employment? To successfully reintegrate into families and
communities? The literature to date is largely silent on these and many other critical
supermax issues.

In keeping with the few previous studies of supermax prisons, the Urban Institute’s research
suggests grounds for skepticism as well as concerns about the fiscal and human costs of these
new forms of correctional housing. At the same time, it is clear that states and wardens believe
supermax prisons can be effective correctional management tools, and this belief should not be
lightly dismissed. For these reasons, it is essential that policymakers and corrections executives
support research that can help determine whether supermax prisons are, or are likely to be,
effective. Since the goals may vary by state, evaluations likely should be conducted on a stateby-state basis. Such research need not be extremely costly. Indeed, where funds are minimal,
considerable advances can be made through efforts to clarify the goals and logic of supermax
prisons and to improve appropriate supermax operations.

iii

Evaluating the Effectiveness of Supermax Prisons
Contents
Executive Summary..........................................................................................................................ii
1.

Introduction ............................................................................................................................. 1

2.

Background ............................................................................................................................. 4

3.

Purpose, Goals, and Objectives of Study ................................................................................ 9

4.

Research Design and Methodology: Data and Methods ........................................................ 10

5.

Site Visit Profiles ................................................................................................................... 15

6.

Document Review, Site Visit, and Interview Results.............................................................. 36

7.

National Survey Results ........................................................................................................ 40

8.

Benefit-Cost Analysis of Supermax Prisons.......................................................................... 44

9.

Lessons Learned................................................................................................................... 45

10.

Conclusion ............................................................................................................................ 49

11.

References............................................................................................................................. 50

Appendix: Tables............................................................................................................................ 68

S UPERMAX PRISONS

URBAN INSTITUTE

Evaluating the Effectiveness of Supermax Prisons

1.

Introduction

During the past two decades of prison expansion in the United States, super- maximumsecurity prisons—or so-called “supermax” prisons—have proliferated nationally (Adams 1996;
Austin and Irwin 2001; Mears 2005). In 1984, one prison in the United States fit the description
of a supermax facility, according to a 1996 survey conducted by the National Institute of
Corrections (NIC) (1997). In a study updating the NIC statistics, King (1999) estimated that in
1998, supermax prisons held approximately 20,000 inmates, representing close to 2 percent of all
state and federal inmates serving one or more years. Two-thirds of states had supermax prisons,
with many states planning for or building more during the next decade (King 1999; Riveland
1999b).
Supermax prisons represent a large investment of resources. Despite the costs, states
increasingly are relying on these high-security facilities, even though we as yet have relatively
little information about their goals, impacts, or relative costs and benefits (Briggs, Sundt, and
Castellano. 2003; Kurki and Morris 2001; Pizarro and Stenius 2004; Riveland 1999b; Ward and
Werlich 2003). The commonly expressed view is that supermax prisons serve to house the
“worst of the worst” (NIC 1997). But such an explanation sidesteps a basic question: What is the
reason for housing the “worst of the worst”?
Some sources suggest the primary rationale is to protect other inmates and staff. How this
protection occurs is unclear. The “rotten apple” theory suggests that removing the “bad apples”
(i.e., the most violent inmates) helps prevent other inmates from committing assaults and
infractions (Ward and Werlich 2003). An alternative argument is that supermax prisons
incapacitate the worst inmates, preventing them—but not necessarily their less serious
counterparts—from injuring others. According to this view, there is no “rotten apple” effect per
se. Rather, any overall reduction in prison violence results entirely from incapacitating the most
violent and serious offenders. Others have identified additional goals (discussed in greater detail
in this report) that can be associated with supermax prisons (Kurki and Morris 2001; NIC 1997;
Riveland 1999b).
Not surprisingly, ambiguity about the goals of supermax prisons has led to ambiguity about
who belongs in supermax prisons and what exactly should happen to them once there. Riveland
(1999b, 6) has observed that
The combined best thinking of professionals who have administered, developed, operated,
and/or planned such programs [supermax facilities] would suggest their purpose should be
for extended control of inmates known to be violent, assaultive, major escape risks, or likely
to promote disturbances in a general population prison and that the criteria for admission and
release from such a facility should be explicit and narrow.
Although few people would argue with the importance of controlling such inmates, each group
constitutes a distinct type of risk, and a focus on each group in turn suggests different goals of

1

S UPERMAX PRISONS

URBAN INSTITUTE

supermax prisons (e.g., controlling the most violent inmates vs. preventing escapes vs.
preventing riots). Despite the fact that frequent mention is made of “the worst of the worst,” few
sources explicitly identify the characteristics of this population (Haney 2003a).
The lack of research on the precise goals and underlying theories of supermax prisons is
paralleled by a similar lack of empirical research on the intended and unintended impacts of
supermax prisons (Kurki and Morris 2001). The few available studies (e.g., Austin et al. 1998;
Rocheleau, Forcier, and Jackson 1998; Ward and Werlich 2003) typically adopt a relatively
narrow view of the purpose and effects of supermax prisons, and thus provide an incomplete
picture. As a result, we know little about the full range of goals and impacts of supermax prisons,
which impacts are the most important, how (if at all) they are achieved, or whether the various
impacts offset the costs of building and operating supermax prisons (Austin and Irwin 2001;
Briggs et al. 2003; Elsner 2004; Haney 2003b; King 1999; Kurki and Morris 2001; NIC 1997;
Pizarro and Stenius 2004; Rhodes 2004; Riveland 1999b; Toch 2003; Ward and Werlich 2003).
The considerable costs associated with supermax prisons have led to calls for benefit-cost
analysis (BCA) studies of supermax prisons (Riveland 1999b; Welsh and Farrington 2000). Such
analyses remain rare. One reason may be the ambiguity about what impacts and cost estimates
should be used. Another may be the lack of guidance about how exactly a benefit-cost approach
could be adapted to supermax prisons. Regardless of the reason, corrections officials and
policymakers currently lack sufficient information to determine whether the benefits of existing
or proposed supermax prisons outweigh their costs. If states and prison systems are to allocate
their scarce resources effectively, they will need information on the goals, impacts, and costeffectiveness associated with their most costly prison facilities (Kurki and Morris 2001; Riveland
1999b).
The study discussed in this report, funded by the National Institute of Justice, addresses these
various issues by providing a systematic, empirically based exploration of the full range of goals
and intended and unintended impacts of supermax prisons, by investigating how key impacts are
or may be achieved, and by creating a benefit-cost policy brief and tool to guide practitioners and
researchers in conducting benefit-cost analyses of supermax prisons. These efforts are intended
to provide a foundation on which comprehensive and balanced assessments of supermax prisons
can proceed.
The remainder of the report is divided into the following chapters:
•

Chapter 2 briefly describes the characteristics of supermax prisons, their goals and
impacts, and the need for benefit-cost analyses of these high-security facilities.

•

Chapter 3 describes the purpose and specific goals and objectives of the research study.

•

Chapter 4 describes the research design and methodology used in the study.

•

Chapter 5 presents the descriptive results from site visits to three states (Maryland, Ohio,
and Texas).

•

Chapter 6 presents the descriptive results from the interviews with diverse stakeholders
and the comprehensive literature review.

•

Chapter 7 presents the descriptive results for each of the questions used in the national
survey of state prison wardens.

2

S UPERMAX PRISONS

URBAN INSTITUTE

•

Chapter 8 describes the BCA policy brief (cowritten with Sarah Lawrence and available
on the Urban Institute web site) and BCA tool (cocreated with Sarah Lawrence and Vera
Kachnowski).

•

Chapter 9 distills the lessons learned from the site visits, interviews, and comprehensive
literature review, and the survey of state prison wardens.

•

Chapter 10 concludes with the study’s implications for research and policy discussions.

•

Chapter 11 provides the references used in the report and document review.

•

The appendix provides the tables and figures referenced in the report.

3

S UPERMAX PRISONS

2.

URBAN INSTITUTE

Background

By some estimates, there are 20,000 or more inmates currently in supermax confinement. As
shown in table 1, as of 1998, there reportedly were 19,630 inmates in supermax prisons,
representing close to 2 percent of all state prisoners nationally. Given that state prison
populations have increased during the past six years and that in 1998 many states had supermax
facilities slated to open in the next few years (see table 2), it is reasonable to believe that at least
20,000 or more inmates currently reside in some type of supermax housing (see Briggs et al.
2003). Indeed, based on the Urban Institute survey, it appears that at least 44 states now have
supermax prisons (Mears 2005, 7) housing roughly 25,000 inmates.
The use of supermaxes varies dramatically across states (table 1). As of 1998, some states,
such as Pennsylvania, incarcerated fewer than 1 percent of inmates in supermax facilities, and
some, such as Mississippi, incarcerated up to 12 percent of their inmates in supermaxes. There
are no updated national or state- level estimates of the number of inmates in supermax
confinement.
Characteristics of Supermax Prisons

One challenge to studying supermax prisons lies in the fact that a wide variety of terms are
used to describe them. According to a NIC (1997, 1) report: “It is clear that what is ‘supermax’
in one jurisdiction may not be supermax in another.” Indeed, correctional systems employ many
terms to describe what the media frequently term “supermax” (Henningsen, Johnson, and Wells
1999, 54). These terms include: “special housing unit, maxi- maxi, maximum control facility,
secured housing unit, intensive management unit, and administrative maximum penitentiary”
(Riveland 1999b, 5). Moreover, although some states embrace the “supermax” term, others avoid
it (King 1999).
In the national survey that the NIC (1997, 1) conducted in 1996, the following definition of
supermax was used:
In this survey, “supermax” housing is defined as a free-standing facility or a distinct unit
within a facility that provides for the management and secure control of inmates who have
been officially designated as exhibiting violent or serious and disruptive behavior while
incarcerated. Such inmates have been determined to be a threat to safety and security in
traditional high-security facilities, and their behavior can be controlled only by separation,
restricted movement, and limited direct access to staff and other inmates.
Supermax housing, for purposes of this survey, does not include maximum or close custody
facilities or units that are designated for routine housing of inmates with high-custody needs,
inmates in disciplinary segregation or protective custody, or other inmates requiring
segregation or separation for other routine purposes (emphasis in original).
Despite the disagreement, most state correctional systems have certain high-security prison
facilities that exhibit similar features and constitute what conventionally would be termed
“supermax” prisons (Haney 2003a; see also Pfeiffer 2004). These features, described below,

4

S UPERMAX PRISONS

URBAN INSTITUTE

have been identified in the NIC (1997) survey and are described in detail in a number of articles
and reviews (e.g., Austin and McGinnis 2004; Briggs et al. 2003; Elsner 2004; Haney 2003b;
Henningsen et al. 1999; Hershberger 1998; King 1999; Kurki and Morris 2001; Pizarro and
Stenius 2004; Rhodes 2004; Riveland 1999b).
Supermax facilities typically are either newly created or retrofitted freestanding facilities or
distinct units within a new or existing general prison facility (see table 2). A central facet of
supermax facilities is the more restrictive management of particular inmates and, more generally,
the recourse to a more restrictive security environment than typically associated with traditional
maximum-security prisons (King 1998, 620; Kurki and Morris 2001, 390; Haney 2003a, 125–
126; Riveland 1999b, 5-6). Inmates in these facilities generally are handcuffed during any
contact with staff, eat and go to recreation individually, are confined in their cells for up to 23
hours a day, receive one-on-one or individually based programming (e.g., self-study courses,
staff visits), have only noncontact visitation privileges, and are supervised more closely and by
more staff than inmates in general population facilities.
The operation of supermax prisons varies across states and facilities, according to the NIC
(1997) survey. Many states place the supermax housing decision authority at the institutional
level, while others place it with department of corrections (DOC) directors or deputy directors.
Some supermax prisons are used for “routine segregation purposes (e.g., discipline, protective
custody, and program segregation)” (NIC 1997, 3), though most are not. In some DOCs, inmates
may complete their sentences in supermax housing and then are released directly to the
community, while in others a transition to nonsupermax housing is required. In most DOCs with
supermax prisons, specialized staffing approaches have been developed. Programming in
supermax prisons also varies, with most offering a wide range of services and, because of the
restrictive environment, few able to follow through in providing the services (Haney 2003b).
Although accounts of supermax prisons generally concur with these assessments, it should be
emphasized that, as with many other dimensions of supermax prisons, relatively little research
exists that documents the level, types, and exposure of programming received by supermax
inmates (King 1999; Kurki and Morris 2001; Riveland 1999b).
Goals of Supermax Prisons

The goals of supermax facilities vary greatly. There appears to be general agreement that one
purpose is to promote order and to protect staff and inmates from the “worst of the worst” (King
1999; Kurki and Morris 2001; NIC 1997; Riveland 1999b). A related purpose sometimes
mentioned is to “normalize” general prison conditions, allowing for more effective management
and rehabilitation of other inmates. Still another is to better protect the public or to make the
public feel safer. As Hershberger (1998, 54) has written
These facilities are designed to hold the most violent, disruptive or escape-prone offenders.
By isolating the “worst of the worst,” these facilities increase the safety of staff, other
inmates and the general public. They also allow inmates in other institutions to live in a more
normalized prison environment, with greater freedom of movement and access to
educational, vocational and other correctional programs.
It is important not to overlook other implicit, and perhaps equally important, goals of
supermax prisons. For example, supermax prisons may reduce the recidivism of inmates placed
5

S UPERMAX PRISONS

URBAN INSTITUTE

in them, either through a deterrent effect or through better service delivery than in traditional
prisons (King 1999). It is possible, too, that general population inmates will recidivate less due to
fear of being placed in supermax facilities, and that they will be more orderly because their needs
are more effectively addressed (Riveland 1999b, 5).
Still other goals can be envisioned. For example, supermax prisons may improve the
economy in communities in which they are placed (Riveland 1999b, 19). Although unlikely to
constitute an explicit goal, it may nonetheless be a critical reason that supermax prisons are built.
Such prisons also may help community residents, or citizens in general, feel safer, even if the
perception is not based on knowledge about the actual effects of supermax prisons on recid ivism.
For each of the implicit or explicit goals of supermaxes, there is little evidence as to which
ones are the most important or whether some states view some goals as more important than
others, how supermaxes achieve these goals, or whether the intended and unintended impacts are
sufficient to warrant their costs. Moreover, few attempts have been made to link research on
supermaxes to the theoretical and empirical accounts about the conditions of order in prisons
despite the central importance of order in managing prisons and in corrections research (Adams
1992; Bottoms 1999; DiIulio 1987; Reisig 1998; Sparks, Bottoms, and Hay 1996; Toch 1992).
With rare exception (e.g., Briggs et al. 2003), research to date largely ignores the variability
in the goals of supermax prisons and focuses primarily on supermax inmates—whether they
recidivate back into supermax confinement (e.g., Ward and Werlich 2003), the conditions they
face during supermax confinement (e.g., Kurki and Morris 2001), and their mental health (e.g.,
Haney 2003a)—rather than other potentially affected populations and stakeholders, such as
general population inmates, prison officers, prison systems, and the communities and states in
which supermax prisons reside (Clare and Bottomley 2001; King 1999; Riveland 1999b).
Research thus is needed that examines the reasons that, and purposes for which, supermax
prisons have been built. Such research in turn can be used to help determine the criteria by which
the effectiveness of supermax prisons can be assessed.
Measuring the Impacts of Supermax Prisons

As noted earlier, there have been few studies of the goals of supermax prisons. Because of
ambiguity about the precise goals of supermax prisons, it therefore is unclear what impacts
should be assessed since goals essentially provide the framework for determining which specific
impacts are relevant (Rossi, Freeman, and Lipsey 1999). For example, if improving systemwide
order is considered a goal of supermax prisons, a wide range of specific measures could be used
(e.g., counts or rates of homicides, assaults, infractions, inmate participation in programming,
physical conditions of prisons, including amount of graffiti), depending on the definition of order
used. By contrast, if systemwide safety is a primary goal, then impacts measuring safety, not
safety and order, would be more relevant (e.g., counts and rates of homicides and assaults). If
reducing crime in society or making the public feel safer were goals, then these clearly suggest a
range of different measurable impacts (e.g., reduced offending, reduced fear of victimization
among the public, increased willingness to walk alone or with others in public areas). Of course,
if a given state asserts that all of these dimensions represent goals of their supermax prisons, then
a broad array of impacts must be measured to assess the effectiveness of these prisons.
Because of the limited research on the goals of supermax prisons, there also is limited

6

S UPERMAX PRISONS

URBAN INSTITUTE

research on their specific impacts, both intended and unintended. Nonetheless, a number of
sources provide lists of potential measures that can be used in corrections research. The guidance
provided by Burt (1981) and Lynch (1994), for example, establishes a foundation on which to
begin identifying the potential impacts of supermax prisons (Adams 1992; DiIulio et al. 1993;
King 1998; Marsden and Orsagh 1983; Reisig 1998; Sparks et al. 1996; Useem and Goldstone
2002). For example, Burt (1981) identified a range of corrections-related impacts that included
security (e.g., escape rates); living and safety conditions (e.g., victimization, prison atmosphere,
overcrowding, sanitation); inmate physical and mental health; program and services impacts
(e.g., improvements in basic skills, education, vocational training); and postrelease success (e.g.,
recidivism, employment). A study of supermax prisons might focus on additional impacts, such
as changes in inmate and staff victimization in general population prisons, economic conditions
in communities with supermax prisons, public safety, and perceptions of safety. Focusing on
communities is warranted given research showing links between incarceration policies and
community crime and quality of life (Lynch and Sabol 2001).
In short, given the scant attention given to the impacts of supermax prisons, research is
needed that identifies the range of impacts that should be investigated to ensure that undue
weight is not given to any one or another impact to the exclusion of others that may be as or
more important. Moreover, because of research that suggests a range of negative unintended
effects of supermax prisons (e.g., creation or aggravation of mental illness among supermax
inmates), research is needed that examines the full range of potential unintended effects, positive
and negative, that may be associated with supermax prisons since such effects can directly bear
on any overall assessment of the effectiveness of such prisons.
Benefit-Cost Analysis Assessments of Supermax Prisons

Despite calls for more research on supermax prisons, and especially for benefit-cost
assessments (e.g., Riveland 1999b, 22), little progress has been made in this area. The interest in
benefit-cost analyses of supermax prisons stems in part from the fact that these prisons can have
diverse impacts. They also represent a considerable investment of resources, typically costing
more to build and operate than do general population prison facilities (Austin and Irwin 2001;
Elsner 2004; King 1999; Kurki and Morris 2001; Pizarro and Stenius 2004; Riveland 1999b).
Benefit-cost analyses are especially appropriate when comparing interventions that have
different goals and impacts. They differ from cost-effectiveness analyses, which identify returns
per a given outcome that is common to two interventions (e.g., cost per averted crime) and
require that comparisons among interventions use similar outcomes. When examining supermax
prisons as a type of policy, BCAs are especially useful because of the range of potential goals
and impacts. If all supermax prisons were built solely to produce more systemwide order, then
cost-effectiveness analyses might be sufficient. In essence, this approach would allow one to
compare different approaches to achieving the same goal (systemwide order). But, in fact,
supermax prisons are associated with a range of possible goals and which goals are emphasized
may vary depending on one’s frame of reference (e.g., legislators might focus exclusively on
public safety whereas corrections executives might focus primarily on systemwide safety). In
such a context, BCAs simplify the ability to make meaningful comparisons.
To be as accurate and as useful as possible, BCAs require specification of all possible
impacts to account adequately for the full range of relevant costs and benefits (Gramlich 1981;

7

S UPERMAX PRISONS

URBAN INSTITUTE

Welsh and Farrington 2000). It also is necessary to examine information on different types of
costs, including fixed or capital costs as well as variable or operational costs (e.g., hiring and
training of staff) associated with supermax prisons (Camp and Camp 1999).
The ability to monetize impacts is central to conducting BCAs. But it is not always possible
to monetize every impact. Opponents of supermax prisons might argue, for example, that no
moral basis exists for placing a dollar value on the inhumane conditions supermax inmates are
believed to face. For BCAs, such exceptions actually are quite common. They are dealt with by
noting those impacts that can be monetized versus those that cannot (Gramlich 1981). The final
assessment relies on impacts for which monetary assignments can be made, excluding (and
noting) those that cannot be monetized. Benefit-cost analyses cannot provide “the” correct
answer in situations involving value-based considerations. But they can provide quantified and
monetized assessments of measurable dimensions thought to be appropriate for evaluation, and
thus highlight more clearly the benefit-cost context in which value-based considerations may
affect policies (Gramlich 1981; Welsh and Farrington 2000).
There are, as noted earlier, many potential impacts of prisons that are intended or unintended
and that can be positive or negative. For example, supermax prisons may improve the ability of
general population prison wardens to control inmates. They also may allow prisons to manage
better a resource- intensive inmate population (i.e., the most disruptive inmates), yielding
efficiencies both within the new facilities and in general inmate prisons. But they also may have
no impact on general prison conditions, and operational conditions within supermax prisons may
reduce family visitation, the ability to provide educational and vocational services to supermax
inmates (due, for example, to frequent “lockdowns”), or induce or aggravate mental disorders.
Such impacts in turn may hamper the ability of supermax parolees to transition successfully into
society. In addition, for supermax officers, there may be higher rates of stress, which might result
in increased sick leave, medical care for injuries, decreased work performance, and decreased
inmate safety due to understaffing (Finn 2000; Riveland 1999b).
Different impacts may generate costs or benefits, depending on the direction of impact
(positive or negative). If an impact of a supermax prison results in added costs, or averted
benefits of other aspects of prison operations, it can be viewed as a cost. If the impact results in
an averted cost (e.g., fewer funds expended on medical care), it can be viewed as a benefit.
Impacts that do not result in either added or averted costs can be viewed as benefits when they
constitute desirable outcomes (e.g., greater public satisfaction with correctional policy).
Because the impacts of supermax prisons may vary tremendously, it is critical that a BCA be
preceded by an attempt first to identify what the full range of potential impacts are. Ideally, these
impacts can be quantified and classified, depending on the direction of effect (positive or
negative), as costs or benefits. The costs and benefits then can be monetized and incorporated
into a model that both includes capital and operational costs and adjusts for the scale of the
intervention and level of impacts. This model in turn can be compared with opportunity costs and
benefits associated with investing in alternative interventions.
Not only is research on the goals and impacts of supermax prisons limited, but BCAs are
rarely conducted before or after these prisons are built. It is clear, then, that a compelling need
exists for research that not only can inform benefit-cost analyses but also can help corrections
executives and stakeholders to conduct and use such analyses to inform decisionmaking.

8

S UPERMAX PRISONS

3.

URBAN INSTITUTE

Purpose, Goals, and Objectives of Study

Supermax prisons have become increasingly popular and they also are expensive. Yet we
know little about their specific goals, intended and unintended impacts, how these impacts are
achieved, or whether from a benefit-cost perspective they represent a good investment of
resources.
Against this backdrop, this study was designed to assist policymakers and corrections
officials make more informed decisions about investing in supermaxes. To this end, the study
had two specific goals. The first was to create an empirically based conceptual framework to
guide research on the goals and impacts of supermax prisons and how these impacts are
achieved. With this foundation, practitioners and researchers alike will be able to identify
specific goals and impacts that merit greater investigation and assessment. They also will be able
to develop and test more precisely the causal logic underpinning supermax prisons.
It should be emphasized that systematic, empirically based exploratory research is a
necessary first step toward assessing supermax prisons or any policy where the goals are wideranging and there has been little assessment of the logic of the policy. For example, the risk of
proceeding to quantify specific impacts (e.g., institutional infractions, recidivism) without careful
consideration of the full range of relevant impacts is that undue influence may be given to the
former (Rossi, Freeman, and Lipsey 1999). Some recent studies (e.g., Briggs et al. 2003) focus,
for example, on whether supermax prisons are associated with decreased homicides and assaults
in prison systems. Such research is critical to informing debates about these prisons. But it
largely ignores a range of other potential measures of supermax prison effectiveness.
The second goal was to assist policymakers and practitioners in conducting benefit-cost
analyses of supermax prisons and to illustrate the uses of BCAs in deciding how to allocate
scarce resources. The study created a BCA policy brief and a BCA tool, which were designed
primarily to show the importance of benefit-cost analyses, the feasibility of conducting them,
and, perhaps most importantly, how key assumptions—including assumptions about the
importance, or lack thereof, of certain impacts—can affect benefit-cost determinations.

9

S UPERMAX PRISONS

4.

URBAN INSTITUTE

Research Design and Methodology: Data and Methods

This study relied on several methodologies: (1) A review of correctional agency and
legislative documents, and theoretical and evaluation research on supermax prisons; (2) site
visits to three states, including in-person and telephone interviews with corrections
policymakers, officials, and practitioners in these states (n=39); (3) telephone interviews with
counterparts to these stakeholders in eight other states (n=21); (4) a national survey of state
prison wardens in fall 2003; and (5) creation of a BCA policy brief and tool. Each stage of data
collection informed the ne xt, and collectively the different sources of data provided a more
complete picture of supermax prisons than any one source alone could provide.
Document Review

The researchers began with a review of correctional agency and legislative documents
bearing on the range and extent of goals, impacts for various populations and stakeholders, and
costs associated with supermax prisons. Figure 1 (see appendix) illustrates the general
framework we used to identify different populations and stakeholder groups; in each instance we
explored a range of potential unintended positive and negative effects. The tables discussed in
chapter 6 reflect the use of this framework in organizing the results not only of the review but
also of our analyses of the site visit and interview data. The agency and legislative documents
were collected from published research and from searches of correctional and legislative web
sites.
The review entailed examination of research on supermax prisons and on prisons generally.
A substantial body of research on prison control and higher custody prison units exists, for
example, that was relevant to explaining the logic of supermax prisons (e.g., Adams 1992; Clare
and Bottomley 2001; Sparks et al. 1996). The review of documents and prison literature provided
the initial foundation on which to begin developing a comprehensive list of potential goals,
impacts, and causal logic models associated with supermax prisons. This information in turn
assisted with development of the benefit-cost analysis policy brief and tool.
Site Visits and Interviews

The site visits to Maryland, Ohio, and Texas involved reviewing agency and legislative
documents and research reports bearing on the goals or impacts of each state’s supermax prisons,
as well as in-person interviews with state corrections officials, wardens, and local and state
policymakers and researchers.
Selection of the three states was made, based on consultation with corrections experts, on
several grounds: access and cooperation of each state; regional variation (King 1999); and
variation in size. In addition, the states have had supermax prisons for different periods of time,
with Maryland having built a new supermax in the late 1980s, Ohio having constructed a new
facility in the late 1990s modeled after the federal supermax in Florence, Colorado, and Texas
having a mix of new and retrofitted supermax facilities, some dating back to the mid-1980s. The
expectation was not that these states would represent all states with supermax prisons, but that
their diversity would be useful in providing a foundation for the subsequent phases of research,

10

S UPERMAX PRISONS

URBAN INSTITUTE

including interviews with practitioners and policymakers in eight other states and a national
survey of wardens.
The interview schedule was developed based on the document review. The questions focused
on the goals of supermax prisons, the range of impacts that may be relevant for different
populations, how these impacts are realized, and what the basis is for respondent assessments of
impacts. Respondents were also asked to provide views about their own experiences and from
their own roles, as well as views they have more generally about supermax prisons in their state
or where they work. The initial use of this schedule during the first site visit provided an
opportunity to identify ways it could be improved. Subsequent to this visit, slight modifications
to the wording and structuring of questions were added. These modifications helped improve the
overall flow of the interview and did not change the content substantially.
For each site visit, approximately 10 one- to two-hour in-person interviews were conducted
by two members of the research team. Respondents included prison administrators and wardens
and local and state policymakers and researchers.
Respondents were identified through a snowball sampling strategy. In each instance, we
contacted people whom we either knew in advance or who were recommended to us. We
explained the goals of the research, whom we hoped to talk to in each state, and why. This
process in turn led to recommendations about others we should contact and efforts on our behalf
to arrange interviews and meetings. The assistance provided was not only helpful, it was
necessary to gain access to and to overcome mistrust among those we hoped to interview.
The in-person interviews enabled the researchers to explore the goals and impacts of
supermax prisons and ways in which supermax prisons achieve specific impacts. They also
enabled us to gain insight into issues that we otherwise might miss or not fully appreciate and
that would be important to policy discussions and evaluations of supermax prisons. The
interviews in Texas, for example, reinforced a critique raised in some research articles about the
considerable variability in what is meant by a “supermax” facility (King 1999; Riveland 1999b).
Although many sources state that Texas has supermax facilities (e.g., NIC 1997; King 1999),
Texas calls their highest security facilities “administrative segregation” (ad seg) prisons. The inperson interviews helped sensitize the researchers to the importance of these distinctions and the
need to identify commonalities that may underlie different terminology.
The research team conducted the site visit interviews together to facilitate the questioning
and coding process so that key findings and issues could be discussed and explored following
each interview. One researcher asked questions and the second recorded responses, occasionally
asking questions to help the primary interviewer to follow up on certain responses. Respondents
were told that their responses would be kept confidential. This was done to ensure that the
respondents candidly expressed their views.
Data from the site visits were compiled to create a profile of the potential goals and impacts
and causal logic of supermax prisons in each state, including descriptions of the more general
policy context in which these prisons arose. These profiles then were used to assist with the
mapping of the full range of potential goals and impacts associated with supermax prisons. The
site visit material and interviews helped provide a more complete picture than could be obtained
through reliance on publicly available documents. And, unlike the subsequent telephone
interviews, they generally provided a foundation on which to understand how the perspectives of
diverse stakeholders within each state differed.
11

S UPERMAX PRISONS

URBAN INSTITUTE

The telephone interviews were conducted with similar sets of respondents in other states.
They afforded the Urban Institute (UI) researchers the opportunity to document a fuller range of
potential goals and impacts associated with supermax prisons. Given the small sample of three
states for the site visits, it was critical to conduct additional interviews to ensure that as full and
as nationally representative a range of goals, impacts, and so on were identified as possible. As
with the site visit materials, the telephone interview data were used to develop further the
mapping of potential goals and impacts of supermax prisons.
Respondents in the site visit and telephone interviews included correctional staff at many
organizational levels, allowing us to obtain a range of perspectives about supermax prisons.
These included prison administrators, supermax and maximum-security prison wardens,
corrections officers, program and legal staff, budget officers, mental health professionals, and
researchers. Each site visit involved interviews, mostly in person, with 10 or more people. Most
of the site visit interviews were conducted as part of focus groups that lasted from one to two
hours. As part of each visit, the authors also interviewed three to four state legislators who were
members of criminal justice committees. These interviews helped provide insight into the
political dimensions of supermax prisons. The authors interviewed Republicans and Democrats,
most of whom were elected prior to the creation of supermax prisons. The interviews generally
were conducted privately in the legislator’s office and lasted 15 minutes. Each telephone
interview was approximately 20 to 30 minutes. Although we did not interview prisoners, we
reviewed many documents that attempted to present prisoners’ perspectives, including lawsuits
filed by prisoners, summaries of prison investigations, and research, such as the recent
ethnography by Rhodes (2004).
For the purposes of this study, we defined a goal as the intended purpose of a supermax
prison and an impact as the more specific measure or manifestation of that goal. For example, a
goal of supermax prisons may be to increase public safety. The intended impacts of that goal
may include reducing prisoner escapes, reducing recidivism, and deterring crime.
The coding of goals and impacts followed from and expanded on those developed during the
document review. This coding involved creating categories of goals and impacts that could be
identified for different stakeholder populations (supermax prison inmates, officers, and wardens;
general population prison inmates, officers, and wardens; correctional systems, including
executive administrators, health providers, and parole officers; communities, including local
government and policy leaders, businesses, and residents; and state policymakers, businesses,
and residents). Following the suggestion of methodology texts (e.g., Babbie 1995; Caudle 2004;
Mohr 1995; Scheirer 1994), UI researchers examined the literature and site visit and telephone
interviews for themes until it was felt that the full range of possible goals and impacts,
explanations of these impacts, and key policy issues had been identified. The UI researchers
created the initial categories and coded the site visit material together, thus ensuring that a
consistent approach to coding occurred for the analysis of the telephone interviews.
It should be emphasized that the goal of this study was to identify possible goals and impacts
of supermax prisons, and not to evaluate any one prison or state correctional system. For that
reason, the researchers did not see any particular advantage in identifying statements from
specific respondents about supermax prisons or the correctional systems in their states. Where
helpful, we identify the occupation of respondents, since this information can help explain the
perspective behind and potential motivation of a particular comment.

12

S UPERMAX PRISONS

URBAN INSTITUTE

Chapter 5 summarizes the case study profiles of the three states that were visited (Maryland,
Ohio, and Texas). Chapter 6 summarizes the list of goals and intended and unintended impacts
(positive and negative) identified in our analyses of the literature review and site visit and
interview data. The goals and impacts are organized according to key stakeholder groups
(supermax prisons, general population prisons, criminal justice system, local communities, and
populations at state and national levels) and subpopulations specific to each (e.g., prisoners, staff,
and wardens in supermax and general population prisons). The researchers entered all goals,
impacts, and identified causal logics (how supermax prisons contribute to the goals and impacts)
in a Microsoft Excel file (summarized in chapter 6) and in accompanying tables to make the
results more accessible.
National Survey of Wardens

The national survey focused on adult state prison facilities housing men or women. We
obtained the initial name and address list from a directory available through the American
Correctional Association (ACA) (2003). We excluded juvenile detention, medical, and
psychiatric facilities, as well as hospitals, community corrections facilities, and halfway houses.
We then checked corrections agency web sites to update name and address information, adding
new facilities and deleting closed ones where appropriate, and ensuring that wardens were
correctly matched to the specific facilities they supervised. In instances where the corrections
web sites differed from the ACA directory, we used the web site information on the assumption
that it would be more up-to-date. When a warden supervised several facilities, we sent a survey
only to his or her highest-level facility. If a warden supervised two facilities with similar security
levels, we sent the survey to the facility with the larger population.
The survey questions were designed to focus on goals and impacts of supermax prisons and
related issues about which wardens would likely have knowledge. Based on a review of the
literature and analysis of the site visits and interviews, the UI researchers created a set of closedand open-ended questions and structured these with assistance from survey methodologists at UI.
Project staff field-tested the instrument with correctio nal officials and researchers. The final
version of the instrument was distributed to all wardens along with a letter explaining the
purpose of the survey and the research project, and an accompanying letter of support from the
ACA. Follow- up mail-outs were sent six weeks later to ensure a high response rate. The final
sample was 601, representing 69 percent of the total that were distributed, excluding one state,
New York, which refused to participate. (If New York is included, the total universe was 948
and the response rate, accordingly, would be 63 percent.)
To obtain this response rate we undertook the following steps: obtained the assistance of
wardens in the development and pretesting of the survey; kept the length of the survey to four
pages so it could be completed in 15 minutes or less; included a letter of support from the ACA’s
executive director; assured respondents that their responses would be confidential and
anonymous; provided self-addressed, stamped envelopes; immediately provided materials to
departments of corrections who requested more information about the study before allowing
wardens to compete the surveys; conducted a second mailing; and conducted follow- up phone
calls to nonrespondents encouraging them to complete and return their surveys.
We were aware that response rates to mail surveys had been declining during the past two
decades (Ayidiya and McClendon 1990; Baim 1991; Bradburn 1992). And we had been

13

S UPERMAX PRISONS

URBAN INSTITUTE

forewarned that response rates to mail surveys among corrections departments, and wardens in
particular, had been declining in recent years. In correspondence, the first author, Dr. Wesley
Johnson, associate dean of administration at the Sam Houston State University College of
Criminal Justice, mentioned that his national survey of wardens in 1995 yielded just over a 70
percent response rate, but that a later survey in 1998 yielded just over a 50 percent response rate
(personal communication; see also Wells, Johnson, and Henningsen 2002, 175). The response
rate obtained in our national survey of state prison wardens thus appears to be in keeping with
recent trends. Moreover, response rates of 50 to 60 percent generally are viewed as acceptable
(Mangione 1998), although Dillman (2000) has advocated for higher response rates. Notably, a
recent review of social science research found that the average response rate in survey-based
studies was 55 percent (Baruch 1999).
To assess whether systematic nonresponse occurred with certain states, which would reduce
our ability to generalize to state prison wardens nationally, we examined state-by-state response
rates. All states but New York responded, and the response rates among all wardens in all but
two states exceeded 40 percent.
The above caveats aside, this survey offers an opportunity to explore the views of a large
number of wardens nationally about a critical issue—the goals and impacts of supermax
prisons—for which wardens have unique insights. Juxtaposed against their views is a lack of
empirical studies of supermax prisons in specific states, to say nothing of national studies. For
this reason, even if we assume that there is some type of nonrandom dimension to the response
rates, the results should be of considerable interest both to researchers and policymakers.
Benefit-Cost Analysis Brief and Tool

The UI researchers created a policy brief and a Microsoft Excel spreadsheet–based tool to
assist policymakers and researchers in understanding, conducting, and using benefit-cost
analyses of supermax prisons. The document review, site visits, telephone interviews, and
national survey all provided information that assisted with creating these products. The policy
brief (Lawrence and Mears 2004) is available on the UI web site.

14

S UPERMAX PRISONS

5.

URBAN INSTITUTE

Site Visit Profiles

The following site visit profiles are intended to give readers a sense of the historical and
social contexts surrounding the construction of supermax prisons in three states, and to convey
the range of issues that arise in attempting to assess the effectiveness of these prisons. In each
profile, the following dimensions are discussed: definitional issues, the history associated with
construction of one or more supermax prisons, the characteristics of the state’s supermax
prison(s), intended impacts and unintended positive impacts, unintended negative impacts,
evaluations that have been conducted or issues relevant to attempts to evaluate each state’s
supermax prison(s), alternatives to supermax prisons suggested by respondents, and issues each
state faces in coming years pertaining to their supermax prisons.
It should be emphasized that in contrast to what some researchers have suggested in the
literature concerning the barriers to conducting studies in and of supermax prisons (e.g., Ward
and Werlich 2003), officials and practitioners in each state were uniformly welcoming and
disarmingly open about their views, both positive and negative, of supermax prisons.
Maryland

As of January 1, 2001, Maryland housed nearly 23,145 prisoners, the 19th largest number of
state prisoners in the United States (Beck, Karberg, and Harrison 2002, 2–3). As of that date, the
state’s supermax prison, the Maryland Correctional Adjustment Center (MCAC), held 246
prisoners (Maryland Department of Public Safety and Correctional Services [MDPSCS] 2003) or
1 percent of Maryland’s prison population (Beck et al. 2002, 2–3).
Definitional Issues

The MCAC policies and practices are consistent with those in other supermax prisons (NIC
1997, 1). The MCAC is a freestanding facility that indefinitely houses prisoners who are violent
or seriously disruptive (MDPSCS 2003). All supermax prisoners at the MCAC are in single cells
for up to 23 hours a day, and their contact with staff and other prisoners is significantly restricted
(Will 1997). The facility’s design places great emphasis on security. For example, the MCAC
relies heavily on technology in its operations (such as electronically operated cell doors) and has
a centralized control center that allows correctional officers to observe all the unit’s cells from a
single location (James 2002; MDPSCS 2003).
The MDPSCS references the facility as a “Level II Maximum Security” prison on its web
site, rather than as a supermax prison (MDPSCS 2003). Interview respondents, however,
concurred that the MCAC is a supermax prison, and that staff regularly refer to it as such. In
addition, media coverage frequently refers to the MCAC as a supermax prison (e.g., James
2002).
History

Reports and the interviews indicate that a number of factors may have contributed to the
decision to create the supermax prison. One respondent reported that Maryland had considered
building a supermax prison in 1972 but was delayed in building it for budgetary reasons. The
Maryland DPSCS proposed the idea again after an inmate killed a correctional officer in the mid1980s. Adding to the argument to build the facility, the state’s maximum-security prison was
15

S UPERMAX PRISONS

URBAN INSTITUTE

overcrowded and fights amongst prisoners were occurring regularly. Nearly all respondents
noted that the state turned to a supermax facility as a tool to manage inmates who had little or no
incentive to follow institutional rules. For example, the prison reportedly had experienced
difficulty managing prisoners who were sentenced to life in prison or who had received so many
disciplinary reports that they were permanently in administrative segregation. Respondents stated
that the lack of incentive to comply with rules might have contributed to the death of the
correctional officer since the prisoner who killed him had been sentenced to life in prison.
One respondent suggested another reason for the supermax prison: Prisoners had cha nged
demographically and required a new management approach; in particular, prisoners were
younger, less receptive to programs, and less responsive to traditional punishment. Respondents
also indicated that a response was needed to help correctional officers cope in an environment
where a fellow officer was murdered and prisoner fights were commonplace.
The public did not appear to oppose the MCAC’s construction. One legislator interviewed
for the study explained that some key policymakers at the time gained public support by
promoting the prison as an “economic tool” that would bring jobs to the area and as a means to
constrain violent criminals. None of the four legislators with whom we spoke reported that their
constituents had complained about the supermax prison in recent years.
Echoing comments we heard in other states, respondents explained that Maryland did not
conduct a study to determine the number of beds that the MCAC supermax would need. Rather
they constructed the number of beds that were equal to the number of administrative segregation
beds in Maryland’s prison system. Construction of the MCAC began in 1986, and the facility
opened in January 1989, when Maryland’s prison population totaled approximately 14,000, with
288 beds (MDPSCS 2003). A little over a decade later, respondents reported that staff often
struggle to keep beds filled in spite of a near doubling of Maryland’s prison population to nearly
25,000. To help fill the beds, the MCAC also houses prisoners who are not necessarily
disciplinary problems, including federal pretrial detainees and death-row prisoners. Respondents
suggested that about half the beds were actually needed and that research would have helped
avoid this problem. They also explained that the political dimensions of the issue and the
resulting media coverage created a “lock them up” atmosphere that enabled this overestimation
to occur and contributed to the decision not to support programming.
Characteristics

Like many states, Maryland has one supermax facility, the MCAC (NIC 1997, 4–6).
However, unlike other states, the MCAC is located in a metropolitan city—downtown Baltimore
(MDPSCS 2003). The facility’s design was based on the federal penitentiary in Marion, but the
MCAC did not include the types of programs administered in the Marion facility. The MCAC
can hold up to 286 prisoners and houses 246 prisoners according to the MDPSCS web site
(MDPSCS 2003). The average length of time spent in the MCAC is one year and five months
(Will 1997). While incarcerated at the MCAC, prisoners are not eligible for parole (Moran
1994).
MCAC cells are 7 by 10 feet and are single cells; they have solid doors except for a small
rectangular food slot and a small window. Prisoners’ cells have exterior windows wit h a small
view of the outside (Associated Press 2003). For an hour once every two to three days, MCAC

16

S UPERMAX PRISONS

URBAN INSTITUTE

prisoners are released from their cells to bathe and to exercise alone in a “windowless prison
dayroom” (Gavora 1996). Otherwise, prisoners are only released from their cells for visits and
for medical reasons (Patrick 1996, 4). Although the MCAC originally allowed prisoners to
recreate together, they ended this practice after several prisoners killed one another (Gavora
1996).
Prisoners sent to the MCAC include those “who are found to cause a great deal of violence or
have a destructive influence (e.g., gang leaders)” (Townsend 1999) and death-row prisoners
(Gavora 1996). In July 2000, the MCAC also held 90 persons in the custody of the U.S. marshall
for the District of Maryland (either federal pretrial detainees or fugitives) because “no other
facility is capable of handling the high volume of prisoners for the Baltimore Federal
Courthouse” (McKinney 2000). Federal prisoners are the only ones double celled in the MCAC.
As of July 1994, the MCAC had a slightly disproportionate share of black inmates compared
with the general prison population as a whole: 89.7 percent of MCAC inmates were black and
9.9 percent were white, while 77.3 percent of MDOC inmates were black and 22.5 percent were
white (Moran 1994).
In response to disruptions, such as throwing (“chunking”) bodily excretions at staff, the
MCAC instituted an incentive program, referred to as the Quality of Life program (Gavora
1996). This program entails rewards for positive behavior, such as additional recreation time,
visits, and library privileges (MDPSCS 2002, 4–7). It also entails reductions in rewards given
negative behavior. MCAC prisoners who are “particularly disruptive” can also be sent to other
states as an exchange for the other states’ most unmanageable prisoners (Townsend 1999).
Intended Impacts and Unintended Positive Impacts

Most respondents indicated that the MCAC’s primary goal is to make the general population
prisons safer and more manageable by confining those prisoners who are violent or who
repeatedly disrupt the order of the prison system. Respondents thought that the MCAC had been
successful in accomplishing this goal, and some noted that they expected the most significant
impacts to have occurred in the three prisons that supplied most prisoners to the MCAC.
Expected impacts included a drop in use-of- force incidents, assaults, homicides, and other
violent incidents in general population prisons.
Another related goal of the MCAC is to manage those prisoners who cannot be safely and
securely managed in another facility. Respondents viewed the MCAC, under current policy and
practice, as a safe environment for staff and prisoners. Indeed, they reported that many staff
desire to work at the MCAC because it is a safer and more predictable working environment than
other facilities. A few legislators also asserted that a goal of the supermax facility is to enhance
public safety by reducing prisoner escapes or to increase their feelings of safety and security.
Several respondents relayed that another goal of the prison is to change MCAC prisoners’
behavior and to return them to the maximum-security prison. One respondent explained that
deprivations (e.g., less freedom, fewer privileges, and hindrance of family involvement) help the
prisoner better understand the costs of committing infractions and may reduce the likelihood that
he would commit infractions in the future. One legislator also considered a purpose of the
supermax facility to be an incentive for general population prisoners to comply with institutional
rules as well.

17

S UPERMAX PRISONS

URBAN INSTITUTE

Andrew C. White, a lawyer in Baltimore and an ex–federal prosecutor, also noted an
additional unintended positive impact. Having questioned more than a dozen MCAC prisoners,
he found the following:
It is not your typical prison. It can have a psychological effect on people. I have had
a number of people go into Supermax for the weekend and come out and be very
willing to cooperate with law enforcement. For the untrained prisoner it can be quite
a shock. (James 2002)
Unintended Negative Impacts

Critics have asserted that the facility has had a number of negative impacts on prisoners’
well-being. Most notably, in 1995, the U.S. Department of Justice (DOJ) launched a probe of the
MCAC (Corrections Digest 1995). In 1996, DOJ reported that MCAC conditions violated
prisoners’ constitutional rights (Patrick 1996). The MCAC has since established policies and
practices to address each of these conditions. Below are some of the identified conditions.
Medical Care. DOJ found several barriers inhibiting MCAC prisoners’ access to medical
care, most of which violated MDOC policy and generally accepted standards of practice. First,
correctional officers rather than medical personnel were administering sick-call requests. Second,
a physician was not regularly scheduled to be on site on a weekly basis. Third, medical personnel
were only available five days a week rather than seven. Fourth, prisoners who were not indigent
were required to make copayments for medical services and nonpsychiatric prescriptions when
the prisoner requested them. And, finally, prisoners were not receiving “face-to- face intake
screenings” by medical personnel upon their arrival.
Mental Health Care. DOJ also found that the MCAC did not adequately screen or treat
prisoners for mental illness. First, staff members were not screening all incoming prisoners for
mental health issues within the first 12 hours of their arrival, as policy required. Second, MCAC
mental health care providers were not actively monitoring prisoners’ mental health conditions;
rather they were relying on correctional officers’ recommendations and on prisoners’ requests for
care. As evidence that MCAC’s screening and monitoring process was insufficient, DOJ cited
the findings of an impromptu, MDOC prisonwide screening that identified 20 MCAC prisoners
for immediate transfer to a facility designed for the care of mentally ill prisoners and identified
35 prisoners who needed mental health care. Third, a correctional officer was always present in
mental health assessments of prisoners, which DOJ contended could prohibit a thorough
assessment of the prisoners’ mental state given any reservations the prisoner may have in talking
openly with the mental health care provider. Fourth, few opportunities were available to
prisoners to participate in mental health programs or to receive counseling; rather mental health
care primarily consisted of prescription management. Fifth, prisoners’ mental health records
were often incomplete and not organized in a systematic way. Sixth, no quality assurance system
was in place to ensure that the MCAC provided adequate mental health care. In addition, the
MCAC administered lithium to a number of prisoners who were not screened properly
beforehand, which could potentially cause serious health problems for some.
DOJ argued that these practices were not only a violation of prisoners’ cons titutional rights to
adequate access to mental health care and services, but also they could prohibit the improvement
of prisoners’ mental health or exacerbate their conditions. Andrew White, an attorney, also
explained that his experience working with MCAC prisoners had led him to believe that the
18

S UPERMAX PRISONS

URBAN INSTITUTE

conditions in the MCAC “can have a psychological effect on people” (James 2002).
Exercise. DOJ also found that MCAC prisoners’ limited access to the opportunities to
exercise represented a violation of both MDOC policy and the prisoners’ constitutional rights.
Due to a shortage of staff, prisoners were only able to leave their cell for less than an hour once
every two or three days for exercise and showering. In addition, MCAC prisoners were never
able to see the outdoors or to go outside.
Indefinite Segregation. DOJ asserted that the MCAC needed a more standardized, objective
classification system by which to assess the appropriateness and readiness of MCAC prisoners’
transfer to a general population facility. At the time of the investigation, staff could not make any
transfer decision using only objective criteria. Prisoners reported their view that many transfer
decisions were unfair. After studying a sample of transfer decisions, DOJ also found
inconsistency in these decisions; for example, officials reported that one prisoner remained in the
MCAC for 4.5 years, during which time he had a record clean of disciplinary citations. Further,
DOJ officials note that: “. . . Inmates know that they cannot obtain a transfer out of the Supermax
program, which may be causing disincentives to comply” (Patrick 1996, 9). Thus, these practices
may increase the likelihood that prisoners will exhibit disruptive behavior in the MCAC.
Abuse. Finally, DOJ alleged that MCAC officials use of the “pink room” constituted abuse.
Although DOJ officials were unable to view the pink room because MCAC transformed this
room into office space before their visit, the pink room was reportedly an unheated cell that
included no furniture or running water, provided only a hole in the floor for a toilet, and was
covered with old feces and urine. Prisoners were confined in the room in only their underwear
for up to four days. Because they were often placed in restraints, their hands were not free to
allow them to eat with their hands or to remove their underwear before performing bodily
functions.
DOJ allowed Maryland 49 days to make the report’s recommended changes. Maryland
officials contested the argument that these violated prisoners’ constitutional rights (Gavora
1996). The MCAC came into compliance with the DOJ’s recommendations after making a
number of changes, including, for example, implementation of a screening process to ensure that
mentally ill prisoners are not classified to that facility.
Increases Disorder. Interview respondents indicated that, despite the stringent conditions of
the MCAC, the institution could create an incentive for prisoners to commit offenses that would
result in their placement into the facilities. An estimated 10 to 15 percent of all MCAC prisoners
are perpetual protective custody prisoners who fear living in general population but who also
desire to avoid the stigma of protective custody, typically used for prisoners who are targeted by
other prisoners (e.g., sexual offenders). These prisoners will also engage in behavior in the
MCAC that will prevent them from qualifying to return to a general population prison. Since
most prisoners’ families and friends reside in Baltimore, respondents also explained that
prisoners might commit an offense that will result in their placement into the MCAC to receive
additional visits.
Staff. One legislator expressed concern that the work environment would prove difficult for
MCAC staff given that staff would be more likely to face harassment from the prisoners.
Another respondent echoed this concern for staff and argued that the facility should address the
problem by rotating staff regularly, which would also ensure that “fresh ideas” come into the
facility.
19

S UPERMAX PRISONS

URBAN INSTITUTE

Recidivism. Respondents were concerned that prisoners were released from supermax prisons
without the bene fit of rehabilitative programs or of social interaction. One respondent explained
that these facilities “take away the tools necessary to manage human beings,” with treatment,
programs, and a strong incentive system viewed as critical components of that toolbox.
Respondents explained that the MCAC instead relies more on the “stick” to change prisoner
behavior but noted that this approach fails to produce a lasting change in the prisoner and may
actually cont ribute to negative behavior in the long term. As one interviewee explained,
I’m concerned that we’re making hermits. You really can only make the system safer
if you change the thinking of the inmate—not just by having a big stick. Without
help, these prisoners, who are the most disruptive in the system, will not get better
and could actually get worse. I feel like many are more antisocial when they are
returned to general population. Five years out, even if they haven’t returned to the
MCAC, they may not be getting involved. We make loners here, not better people.
The result: Former MCAC prisoners may be less capable of interacting with others and
may therefore experience greater difficulty reintegrating successfully into the community.
In turn, they may be more likely to recidivate.
To offset this potential problem, the MCAC attempts to transition most supermax
prisoners back into general population prison where they can benefit from programs and
services prior to release. However, the facility still releases approximately three to four
prisoners per year directly to the street.
Fiscal Costs. Critics have cited the MCAC’s high fiscal costs as having a negative impact on
the state’s budget. On average, it costs three times as much to house prisoners in the MCAC as it
does to house place them in a non-supermax facility (McKinney 2000). The MCAC’s increased
cost results in part from its high staff- to-prisoner ratio. For example, policy requires at least two
correctional officers when escorting prisoners; the MCAC has 1.1 staff members for every 1
prisoner (or 263 to 246) (MDPSCS 2003). At the same time, one legislator explained that the
facility’s additional cost is small in proportion to the budget and therefore does not register on
legislators’ “radar screens” as a potential means to save funds.
Evaluation

Although a formal evaluation of the MCAC has not yet been conducted, violence persisted in
the facility until several changes were instituted. First, the MCAC ended group recreation in
response to a series of incidents when prisoners were killing other prisoners (Gavora 1996).
Also, due to a prevalence of staff injuries, the MCAC instituted security enhancements in fiscal
year 1999; following this change, staff injuries fell by 32 percent (MDPSCS 2003). In 2001, new
cameras and a digital recording system were installed in the MCAC as well; the number of uses
of force by correctional officers over the subsequent six- month period dropped in half (from 58
incidents to 24), potentially because officials were able to hold both prisoners and staff more
responsible for their actions (Access Control and Security Systems 2002). Another important
change is the MCAC’s establishment of the Quality of Life program “that encourages positive
behavior” through a system of penalties and rewards (Access Control and Security Systems
2002).
Several respondents were asked whether a benefit-cost study of the MCAC would be helpful.
One respondent asserted that this type of analysis would have been most helpful in deciding

20

S UPERMAX PRISONS

URBAN INSTITUTE

whether to build the MCAC, especially considering that a number of unexpected costs have
arisen since its creation. Most other respondents, however, expected the utility of a benefit-cost
study to be limited. One legislator thought that an evaluation’s impact would be restricted by the
extent to which the community housing the supermax facility was dependent on the prison for
jobs. This legislator explained that a study alone would not likely elevate the issue in the political
arena; a major incident would likely be required to do so. Several other legislators thought that a
study demonstrating the cost ineffectiveness of the supermax would not result in the closure of
the facility, but it could be used to prevent another supermax prison from being constructed or to
improve the efficiency of the facility’s operations. As one legislator explained, the supermax
prison represents a “philosophical punishment that people would support or reject irrespective of
a study’s findings.”
Alternatives

One respondent thought that the supermax facility might not be necessary if the MDPSCS
implemented the Quality of Life program throughout all prisons. Others indicated that using the
Quality of Life program in conjunction with close management units in all prisons could serve as
an effective alternative to a single supermax prison. At the same time, some respondents
explained that in the absence of a supermax facility, prisons often transferred prisoners from one
facility to another in order to control them, an approach that they viewed as inefficient. Several
others saw no alternative to a supermax prison.
Future Issues

Respondents indicated that discussions of building a new supermax prison in Maryland were
underway since some argue that a new supermax facility would be necessary to provide
programs to prisoners. At the same time, the newly appointed state correctional secretary, Mary
Ann Saar, suggested closing the MCAC given its lack of emphasis on rehabilitation (Fesperman
2003; Nitkin 2003). The fiscal challenges that Maryland, like many other states, currently faces
increases the importance of the state’s decision, especially considering that the facility cost $21
million to build and costs $15 million per year to operate (Maryland Chamber of Commerce
2003). Research that examines the goals, costs, and benefits of the MCAC as well as those of its
alternatives could provide critical information that would assist Maryland to optimize use of its
resources as it decides to modify or eliminate the MCAC or to build a new supermax facility.
Ohio

Ohio currently houses nearly 46,000 prisoners, the fifth largest number of prisoners in the
United States (Beck, Karberg, and Harrison 2002, 3). As of May 1, 2003, the state’s supermax
prison, the Ohio State Penitentiary (OSP), held 260 prisoners (OSP 2003, 1), less than 1 percent
of Ohio’s prison population (Beck et al. 2002, 3).
Definitional Issues

The OSP policies and practices are consistent with those in supermax prisons, as described
by the National Institute of Correction (NIC 1997, 1). The OSP is a freestanding facility that
indefinitely houses prisoners who are violent or seriously disruptive (Davis 1999, 3). All
supermax prisoners at the OSP remain isolated in their cells for up to 23 hours a day, and their
contact with staff and other prisoners is significantly restricted (Davis 1999, 3).

21

S UPERMAX PRISONS

URBAN INSTITUTE

The Ohio Department of Rehabilitation and Correction (ODRC) itself refers to the OSP as a
supermax prison (http://www.drc.state.oh.us/web/prisprog.htm). Respondents interviewed for
this study considered a number of the OSP’s characteristics to fit those of a supermax. The
building is designed to minimize prisoner movement as well as contact among prisoners and with
staff; more staff are required for such tasks as cell extraction and prisoner transfers; and
prisoners are classified to the OSP because of their in-prison conduct. The OSP itself was built
using a modified version of the architectural blueprints used for the federal supermax facility—
the United States Penitentiary Ad ministrative Maximum Facility (ADX)—in Florence, Colorado.
Respondents explained that using the supermax terminology helped initially to rally support
for building the prison. Policymakers and the public demanded a “get tough” response to prison
violence, especially after a prison riot in 1991. In recent years, however, ODRC has moved away
from using this terminology—preferring to classify prisons based on the level of inmates (from
Level 1, least restrictive custody, to Level 5, most restrictive custody)—to draw attention to the
greater emphasis they now place on programming and humane conditions within the state’s
prisons, including the OSP. Officials more frequently refer to the OSP by name (i.e., the OSP) or
as a Level 5 facility.
History

In 1993, a riot broke out in the Southern Ohio Correctional Facility (SOCF), Ohio’s
maximum-security prison in Lucasville. During the riot, correctional officers were held hostage;
ultimately, one officer and nine prisoners were killed (Pietras 2001). Immediately following the
riot, administrators locked down the prison and transferred four prisoners to other prisons.
In the wake of the riot, the ODRC argued that a supermax prison was required to “control the
most dangerous inmates” (Abramsky 2002). Five years later, the OSP opened. During the time in
between the riot and the OSP’s opening, the ODRC used a 20-bed unit as a temporary supermax
facility (Davis 1999, 8). No riots have occurred since 1993, including the five- year period before
the OSP opened.
The precise cause of the riot remains unclear. One respondent suggested that gang violence,
which had been escalating in the weeks preceding the riot, was a critical factor. Another
indicated that the high ratio of prisoners to correctional officers could have been a contributing
factor. Ohio’s Correctional Institution Inspection Committee (CIIC) asserted that overcrowding
was likely to blame (Davis 1999).
If overcrowding caused the prison riot, it remains unclear why the reliance on a temporary
20-bed supermax unit proved sufficient to stem further riots in the five years after the 1993 riot.
Other changes at OSP, including an increased focus on prison control generally—especially in
the absence of an available supermax facility—could explain the lack of riots since 1993.
Public views about the proposed supermax initially were mixed. Many neighboring residents
adamantly opposed the facility because they viewed it as a public safety risk. There reportedly
was little opposition based on the costs of building a more expensive type of prison facility.
Interview respondents believed that the supermax was built because of the perception that it
would help prevent another riot. They also pointed to additional factors, such as the demand for
jobs in Youngstown, which was in an economic decline; a rising number of violent crimes that
contributed to a general “appetite for punishment” among the public; and a large-scale prisonbuilding expansion that already was underway in Ohio and across the country.
22

S UPERMAX PRISONS

URBAN INSTITUTE

Soon after the OSP’s opening, opposition dropped significantly. The 500 jobs created in the
economically depressed area reportedly created a powerful constituency for the facility that
included local politicians, residents, and unions. Some respondents explained that many
community residents now believe that the supermax increases public safety because it is more
secure than other types of prisons and so helps to prevent escapes.
Despite a budget shortfall of more than $650 million in fiscal year 2003–2004 (Federal
Reserve Bank of Cleveland 2003, 2), Ohio reportedly is unlikely to consider substantially
modifying its investment in its supermax prison. In interviews for this study, respondents,
including legislators from both ends of the political spectrum, said the State would not consider
scaling back or closing the facility for several reasons. First and foremost, as one legislator put it,
“There is no constituency for violent, recidivist criminals.” By contrast, there are many powerful
constituents, including legislators and residents in Youngstown, where the supermax is located,
that support the supermax. Second, the OSP is only five years old; the State would be more
likely to shut down other, older facilities. Third, the ODRC explicitly argued, and continues to
do so, that it needs the supermax to maintain order throughout the prison system. Fourth, many
residents, politicians, and prison administrators in Lucasville, where the next-highest securitylevel prison resides, might oppose any such effort on the grounds that most of the supermax
prisoners would be sent there, increasing the risk to public safety there. Fifth, although the OSP’s
operational costs are greater than those of other types of prisons, the difference is viewed as
relatively trivial in the context of a multibillion-dollar state budget. Further, closing or
substantially modifying the facility itself would entail costs. Finally, respondents explained that
most legislators are unfamiliar with the supermax, its characteristics, and its costs; as a result,
they are less likely to focus on it.
Despite the lack of awareness among policymakers about the state’s supermax prison, several
legislators interviewed for this study expressed several concerns. One questioned the
constitutionality of the OSP’s conditions and was concerned about the stress officers face in
working in a setting “where prisoners have so little to lose.”
Another legislator was concerned generally about the possibility of prisoner abuse and the
ability of supermax inmates to receive treatment or be rehabilitated in such a restrictive
correctional setting. The legislator also cited concern about correctional officers in the supermax
possibly “burning out” faster than those in other facilities due to the reportedly stressful working
conditions. (ODRC staff reported that the conditions officers in the OSP and non-OSP facilities
vary, and that the stress is comparable, though the sources of stress may differ.)
An additional set of concerns was expressed by one legislator, who noted that there was a
risk of overclassifying prisoners to the facility, that the institution is only half full, and that the
mental health needs of supermax inmates may not always be met.
Characteristics

Like many states, Ohio only has one supermax facility, the OSP (NI C 1997, 4–6). The OSP
can hold up to 504 prisoners (Davis 1999, 3) and, as of May 1, 2003, it housed 260 (OSP 2003).
Of these prisoners, 32 percent are supermax prisoners (Level 5 prisoners). The remaining
prisoners are maximum- security prisoners (Level 4 prisoners) who requested to be transferred to
the OSP from the maximum-security prison in Lucasville or whose security level was reduced
from Level 5 to Level 4 but who requested to remain at the OSP.

23

S UPERMAX PRISONS

URBAN INSTITUTE

According to ODRC Policy Statement 1361, Level 5 prisoners are those who “commit or
lead others to commit violent, disruptive, predatory, riotous actions, or who otherwise pose a
serious threat to the security of the institution” (OSP 2003). The difference between Level 5 and
Level 4 prisoners is that the latter would have been “involved” in such acts but would not have
“led” others to commit them (Policy Statement 1361).
Interview respondents explained that the lawsuit was to blame for the low number of Level 5
prisoners in the OSP. In their view, the OSP would be comprised almost entirely of Level 5
prisoners and would be at capacity if the court allowed administrators to classify chronic
disrupters as Level 5 inmates. Also, the court now requires that the ODRC complete extensive
paperwork when classifying an inmate as Level 5, which has reportedly impeded the assignment
of inmates to Level 5 classifications.
The average length of time spent in the OSP is 2.6 years (Davis 1999, 13). OSP prisoners
spend 23 hours each day in single-bed cells and, for at least five days out of the week, one hour
in recreation (Davis 1999, 3). OSP cells have solid doors and no external (outside- facing)
windows (Associated Press 2003). To stop prisoners from contacting each other or from
chunking (i.e., throwing feces and urine), the ODRC placed steel strips over gaps in the cell
doors (Associated Press 2003). Until the OSP completes a full outdoor recreation facility in
2004, as mandated by the court, prisoners will continue to recreate in either an indoor or an
outdoor cell that includes a chin- up bar and a sit-up bench (Davis 1999, 10–11). When out of
their cells, supermax prisoners are kept in full restraints, are escorted by at least two correctional
officers (Davis 1999, 3), and are strip-searched (ACLU 2002a). Supermax prisoners receive
programming through a closed-circuit television located in each prisoner’s cell, at the cell door,
in an interview booth (i.e., an iron-barred box), or in a congregation of six prisoners in interview
booths lined up side by side (OSP 2003).
All prisoners’ visits are noncontact (Davis 1999, 3). Visits are limited to two weekdays
during working hours, which contrasts with the visiting hours of the state’s maximum- security
prison for male prisoners, where visits are permitted six days each week, including both weekend
days (ODRC 2003). Within the OSP, prisoners can move to lower security levels that allow them
greater access to “published materials received, recreation and shower opportunities beyond
those mandated, phone call frequency and duration, and commissary expenditure limits” (Davis
1999, 9). Respondents commented that the OSP is relatively quiet, and that unlike other
supermax prisons, inmates do not generally scream or yell constantly.
Intended Impacts and Unintended Positive Impacts

The OSP’s mission is to promote safety and security “by confining those inmates who pose a
threat to staff, other inmates, or institutional security in a controlled setting that is conducive to
self- improvement” (ODRC 2003). More specifically, the ODRC’s general counsel, Greg Trout,
has stated that the supermax enhances the safety of general population prisons (Associated Press
2002).
One of the OSP’s primary purposes is to punish prisoners who violate safety and security. At
the prison’s opening, Dr. Reginald A. Wilkinson, the ODRC director, described the OSP as “a
minimum-privilege and maximum-control facility [to] house the state’s worst offenders” and
explained that it “was created to serve as punishment for prisoners who are dangerous and
disruptive” (ODRC 1998, 1). Elsewhere, Wilkinson (1997) has commented extensively on the
benefits of supermax prisons:
24

S UPERMAX PRISONS

URBAN INSTITUTE

Many of us have determined that removing predatory and other dangerous offenders from the
population improves safety and security system wide. By developing “supermax” prisons, we
can isolate problem offenders in one facility. While this concept has come under some
criticism, the benefits appear to outweigh perceived problems. Recalcitrant offenders
participate in targeted in-cell programming rather than sitting idle in a disciplinary cell.
Supermax staff receive specialized training in working with dangerous prisoners, and policies
and procedures are specific to problem offenders without restricting the privileges of the
general population. Segregation cells in other facilities can be used more effectively when
not clogged by repeat offenders, and finally, the “supermax” facility becomes a disincentive
for seriously negative behavior. Furthermore, technological advancements have made high
security more “user friendly.” Ohio's new $65 million supermax prison in Youngstown will
feature sophisticated surveillance, security and fire control systems. Technology will benefit
supermax prisoners by allowing them, even under high-security conditions, to earn academic
credits through computer-based educational programming.
In short, the OSP supermax prison is described as contributing to a wide range of goals,
including increased and better programming for difficult inmates and increased general prison
system order through general deterrence. The OSP’s goal of punishment is of interest because it
contrasts with the goals articulated by other states, such as Texas, that explicitly say that
supermax prisons are not to be used as punishment (Riveland 1999b, 22).
Several respondents stated that the supermax discourages serious incidents at Ohio’s
maximum-security prison, which is the source of most of the OSP’s prisoners. That view was
countered by statements from other respondents indicating that two-thirds of OSP’s inmates
voluntarily chose to remain at or to be placed in the OSP. That inmates would choose the OSP
appears counter to common sense. Some explanations proffered in news accounts (Associated
Press 2003) and in the interviews included are the following: Inmates at OSP live in their own
cells and thus have greater privacy; are not required to work or participate in as many programs;
enjoy better food; have television sets and air conditioning; can eat in their cell; may be closer to
home; have less chance of being injured and injuring others; have access to one of the best
libraries in the Ohio prison system; and enjoy a cleaner and newer facility.
Despite these potential attractions, the OSP clearly has many less-than-desirable features,
from the perspective of inmates. These include the 23-hour-per-day solitary confinement, the
ineligibility for parole, the reduced access to programming, and the limited access to and contact
with family.
Many respondents viewed the supermax as an important prison management tool, not only
for preventing riots but for enhancing the day-to-day operations. The OSP provides a more
controlled environment where “problem prisoners” can be placed, which can limit the problem of
their frequent transfers from one institution to another. Respondents explained that the
incapacitation and isolation of supermax prisoners has reduced incidents throughout the system.
One respondent also argued that the ODRC has saved money by reducing staff and inmate
injuries.
Respondents suggested that the supermax makes general population staff feel safer and their
jobs easier to fulfill. They emphasized that although correctional staff in the supermax must do
more work, many prefer to work in the supermax because it provides a more controlled work
environment, which increases their safety and the security of the facility. Several respondents
25

S UPERMAX PRISONS

URBAN INSTITUTE

noted that staff also prefer working in the supermax because prisoners are rarely on drugs. As
one respondent explained, “Here, you know what you’re getting,” as compared with other
prisons where the presence of drugs can result in spontaneous or unpredictable assaults or serious
disruptions.
The OSP reportedly enhances public safety by keeping escape-prone prisoners (i.e., prisoners
who have attempted to escape previously) more securely confined. One respondent stated that
the supermax increases public safety through a general deterrent effect, suggesting that potential
offenders refrain from committing crimes out of fear that they will be placed in a supermax
facility.
Apart from improving prison order, the OSP was said to improve conditions for general
population prisoners. When dangerous or disruptive inmates are removed from the general
population, the remaining inmates can more easily access programs and services because they
are not concerned about their safety. Restrictions on prisoners’ freedom can also be relaxed. One
respondent explained, “If you have to deal with the bad seeds, then you’re dealing with
disruptions and that reduces the time devoted to programs and recreation. [The supermax]
definitely has an impact.” If program participation increases (whether in quantity or quality),
then recidivism may decrease among these prisoners and an increase in public safety may follow.
Several respondents emphasized that inmates within the OSP improve their behavior through
the OSP programming (e.g., the anger-management program). The facility’s low recidivism
rate—6 of the more than 500 OSP inmates released to the state’s general population prisons were
returned to the OSP—was cited as evidence of this impact. (The rate is difficult to interpret in
part because many inmates who may have engaged in serious or disruptive acts may not have
been transferred back to the OSP because of the ongoing class action lawsuit.)
Unintended Negative Impacts

Criticisms of Ohio’s supermax prison have focused on a number of issues, including fiscal
costs, prison conditions, and care of the mentally ill. These and other issues, such as the
argument that the supermax has little impact on general prison order, are discussed below.
Fiscal costs. Although not necessarily unintended, the operating costs of the OSP are
significantly greater than those of Ohio’s other prisons. On average, it costs $149 per day to
house a supermax prisoner, $101 per day to house a maximum-security prisoner (ODRC 2003),
and $63 per day to house the average nonsupermax prisoner (ODRC 2001, 28). Public officials
express concern about these costs, but it is unclear whether most policymakers or the public at
large know or care about the differential costs of supermax versus nonsupermax prisons.
This increased cost of the OSP is due, in part, to the fact that it has a staff- to-prisoner ratio 50
percent higher than that of the state’s maximum-security prison. The OSP has one staff member
for every 1.2 prisoners (431 to 497), including one correctional officer for every 1.7 prisoners
(289 to 497) (ODRC 2003). By contrast, the state’s maximum- security prison has one staff
member for every 1.8 prisoners (765 to 1,361), including one correctional officer for every 2.5
prisoners (536 to 1,361) (ODRC 2003).
A cost not included in these figures is the expense of housing maximum-security prisoners in
the more costly supermax security prison (Associated Press 2003). Today, only 32 percent of all
OSP prisoners are classified as supermax prisoners (or Level 5 prisoners); the remaining
prisoners are all classified as maximum-security prisoners (ODRC 2003).
26

S UPERMAX PRISONS

URBAN INSTITUTE

Another added cost to the operation of the OSP is the cost involved in the litigation levied
against the facility. Interview respondents explained that lawsuits filed against the state’s
nonsupermax prisons have tended to be much smaller in scope and dismissed more frequently
than the federal lawsuit that has been brought against the OSP. They emphasized, however, that
the likelihood that a lawsuit will be filed against a supermax depends on the views of the
particular courts of jurisdiction.
The respondents for this study explained that greater disaggregation of the fiscal costs by
housing type is needed. For example, to provide an appropriate comparison of the costs of
providing health care for supermax versus general population inmates, one would need fiscal
cost information about health care services and staffing.
Misplacement of supermax inmates. Although the OSP was intended to house the most
dangerous and disruptive prisoners, critics argue that many of its prisoners do not fit this
description due to an ill-defined classification system that produces inconsistent and unfair
placement and release decisions. Raymond Vasvari, the Legal Director of the Ohio American
Civil Liberties Union (ACLU), contended: “This prison was supposed to house the worst and
most violent inmates in the state, but that seems never to have been the case. Who gets in, who
stays in and for how long is all a matter of luck. That is more than unfair, it is unconstitutional”
(ACLU 2002b).
This argument—difficult to support or refute without solid statistics on admissions and
release practices—is bolstered, in part, by the findings of a 1999 report from the investigation of
the CIIC, whose members consisted of members of the Ohio House and Senate (Davis 1999).
The report concluded that the classification criteria were too vague to justify many classification
decisions (Davis 1999, 8). Investigators found there was no clear reason why some prisoners
were classified to the OSP when other prisoners who exhib ited similar behavior remained at a
general population prison (Davis 1999, 8). In March 2002, the federal district judge mandated
that Ohio develop new guidelines for classifying prisoners into the OSP and provide prisoners
sufficient opportunity to appeal the decision (Associated Press 2002).
The CIIC report also found that there may be pressure to overclassify prisoners into the
supermax prison because the state’s sole maximum- security facility, SOCF, lacks sufficient bed
space (Davis 1999, 8). The Committee highlighted the fact that, between 1993 and 1998, the
state could not keep its temporary 20-bed, supermax- like unit half full, raising questions about
the need for a 500-bed supermax facility (Davis 1999, 8).
Violations of constitutional rights. In January 2001, the ACLU and the Center for
Constitutional Rights filed what would become a class-action lawsuit (Austin et al. v. Wilkinson
et al.) heard by a federal court charging that the classification process was arbitrary and that
other conditions in the OSP violated prisoners’ constitutional protections against cruel and
unusual punishment (ACLU 2002a). The suit alleged the following conditions: (1) Isolation in 7by-14-foot cells for 23 hours a day, with fluorescent cell lights on at all times and punishment if
prisoners attempted to cover their eyes; (2) shackling and strip-searching prisoners when out of
their cells; (3) noncontact visitation; (4) outdoor recreation for only one hour a day in an
unheated cell with only a 4-by-6 foot screened window; (5) conducting mental health interviews
through cell doors, requiring prisoners to explain their cases in front of correctional officers and
other prisoners; and (6) conducting psychotherapy sessions with the prisoner “chained to a pole”
(ACLU 2002a). The ACLU also filed a lawsuit arguing that closing the gaps in cell doors limits

27

S UPERMAX PRISONS

URBAN INSTITUTE

the “fresh air in the cells where inmates are confined for 23 hours a day” (Corrections Digest
2002b, 6). Finally, ACLU’s Vasvari also argued that some of the discipline enacted on OSP
prisoners is cruel, including “taking away inmates’ clothing and bedding…, combining and
cooking up meals into a ‘food loaf’ brick” (Pietras 2001).
The ACLU contended that the harsh conditions of the OSP “led to” prisoner suicides (ACLU
2001). OSP prisoners account for 15 percent of all ODRC prisoner suicides in spite of the fact
that OSP prisoners constitute only 1 percent of the whole ODRC prisoner population (Pietras
2001). No study has, however, examined whether this is the consequence of the OSP ho using a
disproportionate number of prisoners who were already mentally ill and likely to commit suicide,
or whether the conditions of the OSP independently contributed to a higher suicide rate.
Increased aggression and mental illness. The ACLU has claimed that the conditions in the
OSP cause increased aggression and mental illness (ACLU 2001). Although the OSP’s policy
prohibits the placement of seriously mentally ill prisoners into the OSP, the CIIC’s 1999 report
indicated that the OSP had 23 prisoners “on the psychiatric outpatient caseload” and 43 “on the
general outpatient caseload” (Davis 1999, 10). At the end of 2001, a federal district judge
prohibited the ODRC from placing mentally ill prisoners into the OSP (Associated Press 2002).
A year later, prosecutors and the ODRC reached a settlement regarding psychiatric and physical
health care, the application of physical restraints, and outdoor activity (Associated Press 2002).
Increased prison disorder or no effect on order. Ohio’s supermax prison was built in large
part to help create greater systemwide prison order. By some accounts, including those of the
individuals interviewed for this study, the OSP has achieved this goal, by eliminating riots and
removing violent and disruptive inmates. But because the ongoing lawsuit has impeded the
ODRC’s ability to place violent and disruptive inmates in the OSP, there is a question about
whether the supermax can have much of an impact on prison order. Criticisms about the
misplacement and mistreatment of supermax inmates (ACLU 2001, 2002b) raise additional
questions about the possibility that the OSP may actually increase prison disorder in two ways.
First, inmates in the OSP may be more likely to engage in misconduct. Second, inmates
throughout the Ohio prison system may perceive the ODRC to be unfair and thus may become
more likely to violate prison rules.
Increased risk to public safety. The CIIC report expressed concern about the OSP
compromising public safety through the release of supermax inmates who reportedly received
little or no programming:
The prospect of inmates [alleged to be “the worst of the worst”] being released to the free
world straight from an extended period of solitary confinement at OSP raises some serious
concerns of public safety and serious issues of whether such inmates have been afforded due
consideration and preparation for a successful re-entry into free society. (Davis 1999, 13)
The OSP warden stated that some prisoners are released to the streets directly from this facility
(Davis 1999, 13). Unless these prisoners are released under probation, they will not be
supervised once released to the community because Ohio law prohibits prisoners from being
considered for parole while they are in OSP (ACLU 2002a).
Reduced ability to manage violent and disruptive inmates. One interview respondent said that
it was easier to manage prisoners—including violent and disruptive inmates—in a general

28

S UPERMAX PRISONS

URBAN INSTITUTE

population facility because the staff are able to interact with the prisoners more easily. This
informal interaction can create a greater ability to steer these inmates from violent or disruptive
behavior. By contrast, in the OSP, the prisoners are almost always in their cells behind solid
doors, and so staff- inmate interactions are largely precluded.
Evaluation

A formal evaluation of the OSP has not yet been conducted. Interview respondents suggested
a number of potential indices for measuring the OSP’s effectiveness in promoting prison order.
These include riots, rule infractions, inmate-on-staff violence, inmate-on- inmate violence,
recidivism once returned to general population, positive drug tests, gang activity, requests for
mental health assistance, medical interventions for injuries and use of force, and requests for
transfer to protective custody. As with many other states that have supermax prisons, it remains
unclear how these diverse impacts should be measured and weighted. And researchers to date
have provided little specific guidance about this issue.
Some information on changes in the assault rates in Ohio prisons is available. But the impact
of Ohio’s supermax facility on inmate assaults remains unclear. Assault rates fell in the state’s
lower-security prisons but grew in the overall prison population. Between 1997 and 2000, the
Ohio prison system assault rate increased from 8 to 10 assaults for every 1,000 prisoners
(Abramsky 2002). One respondent noted that the reduction in assaults is partly due to the
reduction in the prison population and partly due to the OSP. Other changes in the prison system
could have also contributed to this result, including the hiring of additional staff, changes in the
classification system, differences in the treatment of mentally ill prisoners, and changes in
sentencing laws and practices.
Several respondents were asked whether a benefit-cost study would help policymakers with
deciding whether to open, modify, or close Ohio’s supermax prison. In most instances, the
response was the same: A benefit-cost study would be interesting, but it would have little impact
on policymaker decisions and ultimately would serve as only one part of a larger set of factors to
consider. The primary reason identified was that the public would not tolerate any substantial
change to, and certainly not the closing of, the supermax facility. The OSP was billed as a “get
tough” strategy for handling violent and disruptive inmates, which continues to engender wide
public support. One respondent emphasized that the public strongly demanded appropriate
punishment of the most serious criminals and were willing to pay almost any price to ensure that
occurred. Further, closing it several years after it opened would not generally be viewed as
sensible. Policymakers advocating for closure would risk their reelection chances.
Alternatives

Both in written reports and in the interviews conducted with Ohio stakeholders, there is little
evidence of a clear, effective alternative to the OSP supermax facility. Every prison in Ohio has a
control (or isolation) unit that could conceivably serve as alternatives to the OSP—this was
reported to be the approach adopted in the past. Respondents strongly emphasized, however, that
the prison system would not operate as well without the OSP. They reported that the main
advantage of the OSP is its design, which allows staff to view prisoners’ cells from a centralized
position, limits prisoners’ ability to chunk (i.e., throw food, bodily fluids, or feces out of their
cell doors), and has a smaller number of cells to manage in each unit. Another cited advantage of
the OSP is that it has bed space available so that responses to certain violent or disruptive
offenses can be immediate and certain.
29

S UPERMAX PRISONS

URBAN INSTITUTE

Future Issues

The lawsuit Austin et al. v. Wilkinson et al. and other such lawsuits ultimately may restrict
the State’s ability to place maximum-security (Level 4) prisoners in the supermax, allowing only
the placement of Level 5 inmates to the facility. As a result, prison administrators may need to
develop alternative ways to use the OSP’s capacity. If administrators are able to continue to
classify maximum-security (Level 4) prisoners to the OSP, they likely will need to continue to
work to provide these lower security–level prisoners with the programs, services, and freedoms
that would be available to them in the State’s maximum-security prison.
Providing supermax prisoners with more jobs, classroom interaction, additional
programming, and more humane conditions appears to be an ongoing priority for the OSP.
Providing additional congregate programming has proved to be a particularly difficult challenge
to prison administrators because, as one respondent succinctly explained, “You can’t provide
programming when people are killing each other.” When there is no guarantee that certain
precautions will work to ensure the security and safety of inmates and staff, prison administrators
and staff are reluctant to provide additional freedoms, programs, and services.
Another persistent challenge for administrators will be finding a reliable classification system
for placing and releasing supermax prisoners. One respondent reported that the OSP’s
classification process had improved greatly since the facility’s opening. But the challenge of
making fair decisions and predicting which prisoners will succeed after release to the general
population prison facilities will likely be particularly difficult.
Texas

Texas incarcerates the largest number of prisoners in the nation, with nearly 165,000
prisoners in custody in 2001 (Beck et al. 2002, 3). According to a recent survey by Camp and
Camp (2002), Texas also leads the nation in housing inmates in administrative segregation (ad
seg) units: It held roughly one-third (9,148 of 28,975) of all ad seg inmates in U.S. prisons as of
January 1, 2001. The National Institute of Corrections, in its 1996 survey of state correctional
systems, found that close to one-third (16 of 57) of all supermax facilities in the nation are
located in Texas (NIC 1997, 3).
Definitional Issues

Although the NIC categorized Texas’ ad seg units as supermax facilities, the Texas
Department of Criminal Justice (TDCJ) does not. The NIC acknowledged the difficulty in
defining supermax prisons since the use and meaning of the term varies across jurisdictions (NIC
1999:3). NIC’s classification of the Texas ad seg facilities as supermax prisons stems from the
fact that these facilities exhibit characteristics found in NIC’s definition: The facilities hold highrisk prisoners for lengthy periods of time, prisoners are housed 23 hours per day in single-bed
cells, and their movement and contact with staff are significantly restricted (Henningsen et al.
1999; Hershberger 1998; Irwin and Austin 1997; King 1998).
Interview respondents expressed different views about this issue. Several attributed the
distinction between Texas’ ad seg units and supermax facilities to a simple difference in
terminology. They noted the similarity between Texas’ ad seg housing and supermax housing in
other states. Others did not consider the ad seg housing to be supermax- like since there are few
stand-alone ad seg facilities, and it is not only the most disruptive and violent prisoners who are
30

S UPERMAX PRISONS

URBAN INSTITUTE

placed in them. Sometimes, for example, inmates who repeatedly violate rules, but who are not
necessarily violent or do not necessarily incite others to violence, are placed in ad seg housing.
For some of the respondents, the federal “administrative maximum” (ADX) penitentiary in
Florence, Colorado, typifies a supermax prison.
History

Since the 1980s, the total number of prisoners in Texas ad seg housing has risen
dramatically. In 1987, over 3,000 Texas prisoners resided in ad seg units (Camp and Camp 1987,
18) and by 2001 the total had tripled to more than 9,000 (Camp and Camp 2002, 38). The total
number of Texas prisoners also increased dramatically during this same time period, rising from
38,534 to 134,574 (Camp and Camp 1987, 3; Camp and Camp 2002, 2). Since 1992, ad seg
prisoners have consistently constituted approximately 7.8 percent of all TDCJ prisoners. (TDCJ
does not have data on its ad seg population before 1992.)
Texas expanded its use of these facilities in the mid-1980s following a marked increase in
prison violence. TDCJ spokesman Larry Fitzgerald explained that, during this time, Texas
prisons faced high homicide rates and, in some cases, “violent prison gangs virtually ruled” the
prison system (Johnson 2002, 5A). TDCJ responded by expanding its use of ad seg units to
“restore order and break up criminal groups” (Johnson 2002, 5A), with the overarching goal of
enhancing institutional safety and security in the general prison population (Austin et al. 1998,
1). Interview respondents explained that TDCJ locked down all institutions overnight, identified
the prisoners causing the problems, and put them in ad seg units.
The consistent rise in ad seg prisoners has several possible explanations. It may simply
reflect the overall growth in the prison population, especially gang members. Respondents
suggested that the prison population has “hardened” as a result of several policy changes. The
Texas legislature in the 1980s, for example, eliminated early release through good-time credits
for violent and serious offenders. In 1997, the Legislature prohibited this type of early release for
all prisoners; the sole exception was for nonviolent prisoners, who could be released early with a
review board’s approval. As a result, prisoners spent more time in prison with fewer incentives
for positive behavior. Some observers believe that this made prisoners more difficult to manage,
as reflected in the rise in disciplinary infractions. One respondent observed, “There’s a different
kind of offender today, one who will turn to violence more than before. Prisoners come in
without hope and some die in prison and there’s no incentives for them to behave well.”
Characteristics

Most of Texas’ ad seg units hold between 400 and 800 prisoners and are located inside or
attached to a general population facility (TDCJ 1999). Almost all of Texas’ ad seg prisoners (99
percent) are deemed to be security risks (Austin et al. 1998, 1)—that is, individuals who are
likely to escape or who are involved in gangs and/or violence (Henningsen et al. 1999, 56).
Around half of all ad seg prisoners are gang members (Austin et al. 1998, 4, 26). Unlike other
states that confine gang members to supermax facilities only when they are disruptive, Texas
confines gang members to ad seg units solely for gang membership. Some prisoners not
considered to be dangerous or a threat to security are placed in ad seg for chronic violation of
institutional rules; staff could find no other way to control their behavior. As one official
explained, “They’re not horrible criminals, just bad inmates.”
Ad seg is the highest security level in Texas prisons and places the most restrictions on

31

S UPERMAX PRISONS

URBAN INSTITUTE

prisoners’ freedom and activities. Few ad seg prisoners, for example, are able to participate in
any programs. Indeed, in 1995, the Texas legislature passed a law that prohibits the use of state
appropriations to provide rehabilitative programs. Recently, however, TDCJ secured federal
funds to provide in-cell programming to ad seg prisoners to facilitate an improved reentry back
into communities. In spite of ad seg’s restrictions, Texas provides some incentives for good
behavior. Prisoners who exhibit negative behavior can be moved to one of two more-restrictive
levels within ad seg housing, which reduces their privileges and the number of days they can
recreate for one hour outside of their cells. Following good behavior, some ad seg prisoners are
transferred to general population housing, but the length of time varies greatly among different
types of ad seg prisoners. Gang members, for example, can only be transferred out if they
undergo a two-year gang denunciation process.
TDCJ has three types of design prototypes for ad seg facilities: telephone pole, Michael
prototype, and expansion cell block (or, a high-security prison). Telephone pole facilities were
designed so that the correctional officers were stationed at the end of a long hall of cells stacked
several stories high. Portions of these facilities are sometimes designated as ad seg. TDCJ began
building ad seg units using the Michael prototype, which improved upon the previous model by
allowing correctional officers the ability to view all cells within the unit from a centralized
location. In the late 1990s, TDCJ began designing ad seg facilities, most notably the Estelle Unit
that was profiled in a Frontline episode (Koppel 1998a–d). These facilities are based on
expansion cell block prototype, which are semiautonomous buildings located next to main prison
facilities. Although the rules and guidelines are the same for these newer facilities, they are more
reliant on technology to manage prisoners, less well lit, and more costly to build and operate.
They also allow prisoners less human contact, and the only view into the unit is through a small
window on each cell’s steel door. Between 2000 and 2002, TDCJ built four additional expansion
cell blocks to accommodate the growing number of prisoners receiving major disciplinary
citations.
Although no transitional program has been in place for ad seg prisoners, TDCJ will soon
begin providing a step-down process for gang members and a reentry program for all prisoners
leaving custody directly from ad seg.
Intended and Unintended Positive Impacts

Interview respondents generally agreed ad seg’s primary purpose is to ensure staff and
prisoners’ safety. They contended that by locking up the riskiest prisoners, TDCJ has a safer
prison environment.
Respondents also emphasized that ad seg serves as a tool for managing disruptive behavior
and protecting security. When asked, most respondents did not, however, believe that ad seg
housing deters general population prisoners from violent or disruptive behavior. Several
individuals emphasized that TDCJ does not use ad seg for punitive purposes. One respondent
argued that it would not be effective as a deterrent because so few prisoners who commit major
disciplinary infractions are sent to ad seg. By contrast, one respondent argued that there may be a
deterrent effect only for certain groups of prisoners. For example, gang members have a high
probability of being classified to ad seg and, as a result, they may choose not to commit or incite
behavior that could identify them as gang members.
Some respondents suggested that ad seg may also work to normalize the environment for
prisoners: By removing the “bad seeds,” general population prisoners can more easily go about
32

S UPERMAX PRISONS

URBAN INSTITUTE

their routines and access services.
Respondents also explained that ad seg can provide a positive working and living
environment for some staff and prisoners. Although officers of ad seg units must contend with
more verbal abuse and chunking (i.e., prisoners throwing feces and urine), the units tend to be
safer, involve less contact with prisoners, and provide greater routine and structure than general
population facilities. Similarly, respondents reported that some prisoners prefer to live in ad seg,
which may help them stay out of trouble, avoid participating in programs, and avoid conflict
with other prisoners and correctional officers.
Unintended Negative Impacts

In interviews, respondents identified three main unintended impacts that ad seg could
produce. First, ad seg housing is more costly to operate than general population facilities. The
Texas Criminal Justice Policy Council reported that TDCJ’s ad seg units cost an average of
$61.63 per prisoner per day in 2002—45 percent more than general population units’ average
cost of $42.46 per prisoner per day (Hook 2003, 12). These costs in part stem from the fact that
ad seg units require more staff to maintain security and to deliver services. Confining only one
prisoner per cell also adds to ad seg operational costs. One respondent explained that ad seg
prisoners frequently are denied parole hearings and the opportunity to work, which can
potentially result in lengthier terms of incarceration and thus additional costs to the prison
system.
Second, some respondents expressed concern that ad seg confinement can aggravate or cause
mental health problems. One respondent pointed out a potentially contributing problem: A
prisoner may be placed in ad seg because of his behavior, but the behavior may be caused by a
mental health problem. TDCJ, however, has in-patient psychiatrists to assist in the identification
and treatment of these prisoners. TDCJ currently is developing a policy so that staff will be able
to more quickly identify mentally ill prisoners and transfer them out of ad seg. In addition, TDCJ
is working with NIC on a technical assistance project relating to the treatment and handling of ad
seg prisoners who may be mentally ill.
Third, respondents were concerned that ad seg could negatively impact public safety. TDCJ
releases 1,400 prisoners from ad seg to the street each year. The prisoners are released without
any kind of step-down process to improve the transitio n and, while incarcerated, have had little
to no access to programming. Further, they generally are poor candidates for parole and are, as a
result, not likely to be under any supervision when they return to the community. At the same
time, the administrators note that they have not observed many of these prisoners returning for
serious crimes.
Texas has faced several legal challenges resulting from these latter issues. The court cited
“current and ongoing constitutional violations regarding administrative segregation [in] the
conditions of confinement and the practice of using administrative segregation to house mentally
ill inmates” (TDCJ 2002a). The court ordered TDCJ to remove all mentally ill prisoners from ad
seg units. TDCJ has also been subject to litigation concerning their placement of gang members
in ad seg units, but TDCJ’s gang confirmation process has reportedly allowed it to withstand
these challenges. To address a finding that prisoners remain too long in ad seg, TDCJ established
criteria that the state classification review board uses to review ad seg prisoners’ cases every
quarter to assess their readiness for release. Previously, a prisoner could remain in ad seg for as
much as 10 years if he did not ask to be released or if a correctional officer did not advocate for
33

S UPERMAX PRISONS

URBAN INSTITUTE

his release.
Despite these additional costs and impacts, respondents generally viewed ad seg housing as
critical to the management of difficult prisoners. This benefit was viewed as sufficient to offset
any costs or negative impacts that may result from the use of ad seg.
Evaluation

What are the results of Texas’ investme nt in ad seg facilities? Almost all respondents
believed that the dramatic decline in prison homicides and the increased control of gangs that
occurred in the late 1980s was due to the use of ad seg housing. But few empirical studies have
actually been conducted, and those that exist focus on relatively narrow sets of impacts and rely
on data of questionable utility.
Austin and colleagues’ (1998) study found, for example, that prisoner homicides and
stabbings dropped sharply following an increased use of ad seg. However, the study concluded
that data limitations severely hampered their ability to rigorously assess the effectiveness of
these high-security prisons. The authors could not, for instance, rule out other factors—such as
changes in programming or the composition of inmates—that might have also explained this
trend (Austin et al. 1998, 4). Respondents for this study explained that increases in the number of
correctional officers, the growth of a more experienced population of correctional staff, and a
decline in prison crowding could have confounded the results as well. A study assessing the
impact of each of these changes on prison safety and security has not yet been conducted (see,
however, Crouch and Marquart 1989).
Interestingly, the study by Austin and his colleagues also found that in the late 1990s
prisoner-on-staff assaults rose by over 50 percent (Austin et al. 1998, 4), a disproportionate
percentage of which happened in ad seg housing. Although ad seg prisoners represented 6
percent of all Texas prisoners at the time of the study, 40 percent of all prisoner-on-staff assaults
occurred in ad seg units (TDCJ 1999). In some respects, the disproportionate number of assaults
should not be surprising—ad seg facilities are intended to house the worst of the worst. At the
same time, the near total lock-down nature of the facilities suggests that such assaults should be
minimal.
Alternatives

Respondents were unable to identify any alternatives to ad seg units. Respondents suggested
that without ad seg housing, protecting prisoner and staff safety would require locking down a
larger number of prisoners and ending many programs.
Some respondents suggested that TDCJ should not eliminate ad seg but simply change
certain conditions within ad seg facilities. For example, ad seg prisoners could spend more time
out of their cells, have more control over in-cell lighting, and have bars on their cell doors to
humanize the environment. Several respondents thought that TDCJ should provide rehabilitative
and reentry programs to ad seg prisoners. At the same time, they acknowledged the unique
challenge of delivering programs and services to this population—namely, in-cell programming
is very costly and out-of-cell programming raises security and safety risks.
Respondents also highlighted several conditions of ad seg that TDCJ should not change. For
example, they emphasized that ad seg prisoners should not be allowed to recreate in groups. One
respondent commented, “We did allow group recreation [once], but we picked up dead bodies

34

S UPERMAX PRISONS

URBAN INSTITUTE

out of the recreation room.” Another respondent asserted that ad seg’s restrictions on prisoners’
movement is critical to protecting institutional safety: “Before long, if these prisoners are
walking, then there’s going to be an increase in assaults and homicides. No doubt.” Virtually all
the respondents thought that ha ving one prisoner per cell is essential to ensuring prisoners’ safety
and safeguarding security because it limits collusion among prisoners.
Future Issues

Recent developments have presented TDCJ with both opportunities and challenges. An
opportunity arose for TDCJ when it recently received federal funds to provide a reentry program
to ad seg prisoners at one of the expansion cell blocks. The lessons learned from this effort will
likely help inform other states of the barriers, along with strategies for overcoming these barriers,
associated with administering programs to supermax prisoners.
At the same time, TDCJ, like other state correctional institutions, has faced significant budget
cuts even as the prison population continues to grow. Because state budget deficits may prohibit
the construction of new prisons, TDCJ will need alternative strategies to control the most
disruptive inmates and maintain prison order. The costliness of ad seg—both in terms of staffing
and the loss of bed space—may place these facilities under greater scrutiny in the coming years.
The need to understand the purposes and benefits of ad seg and other high-security, or
“supermax,” facilities is, as a consequence, becoming increasingly important. These facilities
may prove to be the most efficient and effective approach to managing the most violent inmates
and reducing their effects on other inmates. But further research and comparisons with
alternative strategies will be needed.

35

S UPERMAX PRISONS

6.

URBAN INSTITUTE

Document Review, Site Visit, and Interview Results 2

As discussed in chapter 4, the UI researchers systematically reviewed a range of documents
and research directly or indirectly relevant to identifying the goals and intended and unintended
impacts, as well as the causal logic, of supermax prisons. The causal logic analysis focuses on
identifying how exactly a supermax prison might contribute to a given goal or impact.
In this chapter, the summary descriptive results of this effort are presented, focusing
exclusively on supermax goals and unintended effects as these pertain to five categories:
supermax prisons, general population prisons, the criminal justice system, local communities,
and states and the country as a whole (see table 3). To facilitate systematic exploration of goals
and impacts, as well as various causal logics associated with each goal and impact, the UI
researchers created a matrix in Microsoft Excel matrix. The Excel file essentially constitutes a
database that provides considerably more detail about specific goals, impacts, and causal logics,
including the sources (e.g., specific articles or interviews) to substantiate each. (The references
section of this report provides a listing of documents that were used to create the matrix. Not all
are cited in this report, but are included to provide readers a comprehensive listing of the
material drawn on in developing the matrix.)
The central point of presenting the descriptive findings in this chapter is to provide an initial
foundation for the study’s central contention: evaluations of supermax prisons that focus on only
one goal, or one or two impact measures, are likely to gravely misrepresent the range of goals
and impacts relevant to assessing the effectiveness of supermax prisons.
Supermax Prisons

Supermax prisoners may be affected by supermax prisons and, indeed, two goals were
associated with these prisoners—modifying the behavior of these prisoners during and after
release from supermax confinement and punishing them (see table 3.1). In each instance, a range
of specific measurable impacts associated with these goals were identified in our analyses,
including greater compliance with rules and reduced violent and disruptive behavior upon release
from a supermax. At the same time, supermax prisoners may experience a range of unintended
effects. Many positive effects were identified, including but not limited to improved quality of
life for supermax inmates (e.g., a greater sense of safety and calm), as well as higher quality
medical and psychiatric care, reduced fear and stress, increased rule compliance, and reduced
intimidation by gang members. Many more negative unintended effects were identified. These
included increased disciplinary infractions while in supermax confinement (the expectation is
that these would be lower), increased tensions among supermax inmates and staff, increased
violent and aggressive tendencies, alleged violations of human rights, and increased recidivism
upon release to society. As the table shows, a larger number of unintended effects were
identified. In no instance did the literature, site visits, or interviews identify the extent of these
impacts; they indicated only that at least in some instances these impacts existed or were strongly
suspected.
2

This discussion is elaborated upon in a forthcoming article (Mears and Watson 2006), “Towards a Fair
and Balanced Assessment of Supermax Prisons,” in Justice Quarterly.

36

S UPERMAX PRISONS

URBAN INSTITUTE

Supermax prisons may also affect the staff who work in them, as table 3.2 shows. Indeed,
supermax prisons are associated with several goals that benefit staff, including the goal of
increasing correctional staff safety and the control of prisoners, and reducing the influence of
gangs. In each instance, these goals are associated with specific impacts directly or indirectly
affecting supermax staff, including reduced homicides and increased sense of safety, reduced
use-of- force incidents, and reduced ability of gangs to harm or intimidate staff. In addition,
supermax staff may experience a number of unintended benefits, including higher rates of
promotion (which several states identified), greater job satisfaction, and reduced stress.
Unintended negative impacts include the potential for staff to feel less safe and satisfied with
their jobs, greater stress and turnover, and a greater risk of injury.
Finally, supermax wardens may be affected by their work at a supermax (see table 3.3). They
are not obvious beneficiaries of specific goals, but at least one unintended positive impact was
mentioned—namely, supermax wardens may garner increased prestige within the correctional
system and in the community. At the same time, there is the unintended risk that they may also
bear the brunt of criticisms about supermax prisons, and thus experience greater stress than they
otherwise would managing a facility with a different security level.
General Population Prisons

General population prisoners are potentially the primary intended beneficiaries of supermax
prisons, given the goals of improving systemwide safety, order, and control and reducing gang
influence (see table 3.4). Here, the intended impacts parallel those identified above for supermax
prisoners, including, among others, reduced murders, assaults, and riots; increased rule
compliance; reduced lockdowns and escapes; and reduced gang involvement and gang-related
violence and disruptions. Some unintended positive impacts include improved living conditions
and access to programs and services, decreased stress and fear, and greater freedoms and
privileges. Unintended negative impacts include decreased safety (if, for example, inmates view
supermax placement as arbitrary and unfair) and reduced resources available for programming.
General population prison staff may experience similar, although not identical benefits (see
table 3.5). Supermax prisons in many instances have been built to benefit not only general
population inmates but also staff, with benefits that fall largely under the same dimensions
(improved safety, order, and control over inmates, and reduced gang influence). They also may
experience a range of positive and negative impacts that parallel those experienced by supermax
staff, including the potential for better or worse work conditions, improved or deteriorated
inmate-staff relations, and increased or decreased job satisfaction and retention.
Finally, general population prison wardens may benefit primarily from experiencing fewer
management challenges and, in turn, from enhanced prestige within and external to the prison
system for operating safer, better-run facilities (see table 3.6). At the same time, the presence of a
supermax may reduce the resources available to general population prison wardens to hire staff
and provide programming, thus creating the potential for increased management challenges.
Criminal Justice System

As table 3.7 shows, the criminal justice system as a whole may potentially benefit from
supermax prisons. The primary goal mentioned in interviews was that supermax prisons were
37

S UPERMAX PRISONS

URBAN INSTITUTE

built in part to improve the operational efficiencies of the prison system, which in turn was
associated with impacts affecting both the prison system and the broader criminal justice system.
For example, one impact mentioned in association with this goal included simplified travel
logistics for transitioning difficult inmates from one prison to another. In contrast to a dispersion
strategy, the supermax concentration approach allows prison systems to send all inmates to one
(or more) prisons rather than sending them strategically to facilities throughout the prison
system. Other impacts were mentioned or identified as well, including simplified and more
consistent training of staff (e.g., how to manage disruptive and violent inmates), reducing the
number of staff needed in some facilities (e.g., because these facilities might be more easily
controlled with fewer staff). Few unintended impacts were mentioned or identified. One impact
suggested in interviews was that prison systems incurred, in aggregate, fewer per- inmate costs,
and that public and political support increased due to impressions that the prison systems were
better and more professionally run. Other interviews suggested precisely the opposite unintended
impacts—namely, that costs increased dramatically and that public and political support was
diminished.
Table 3.8 identifies an additional population potentially affected by supermax prisons—
parole officers. In some interviews, respondents suggested that supermax inmates pose unique
postprison release challenges and threats to parole officer safety. Others suggested that many
supermax inmates in fact may be more easily managed than they otherwise would have been but
for the supermax experience.
Local Communities

Local communities may benefit, primarily unintentionally, from supermax prisons. These
prisons may, for example, increase tax revenues to local governments (insofar as state
corrections agencies pay local taxes for land used) (table 3.9), support for local politicians who
advocate for the prisons (table 3.10), and revenues for local businesses (table 3.11). Some
respondents indicated that local politicians who took positions supportive of supermax prisons
sometimes suffered for it during reelection campaigns (table 3.10). In addition, different sources
indicated that local residents may benefit from supermax prisons (table 3.12). Prison escapes
may be less likely and citizens may, for example, be less fearful in the belief that the supermax
prison constitutes a symbol of the extent of control the prison system has over its inmates. In
addition, when the local economy improves, residents in general are expected to benefit through
access to more employment opportunities and through greater local government tax revenues
from the prison system and these businesses.
States and Country

Supermax prisons may also have potential impacts on state governments and nonlocal
businesses (those that are not located in the towns or cities where supermax prisons are situated),
as well as the country as a whole. In most instances, these impacts were mentioned in the
literature or interviews, but did not appear to be particularly likely. For example, some sources
indicate that supermax prisons can decrease state and national governme nt expenditures by
reducing crime and, in turn, by reducing the costs associated with aggregate- level crime control
policies and victimization (tables 3.13 and 3.14). Nonlocal businesses may benefit through
contracts with state corrections agencies to provide specialized services to supermax facilities
38

S UPERMAX PRISONS

URBAN INSTITUTE

(table 3.15). Others, including a number of reports, pointed to the potential for increased strain
between the United States and other countries who view supermax prisons as violating human
rights. Finally, nonlocal residents may benefit from supermax prisons because of overall
reductions in recidivism and escapes from prison systems (table 3.16). On the other hand, if
supermax prisons create greater prison disorder and violence, as some sources suggested, the n
public safety could be compromised.

39

S UPERMAX PRISONS

7.

URBAN INSTITUTE

National Survey Results 3

This chapter summarizes the results of the national survey of wardens. As with chapter 6, the
intent of the chapter is simply to provide a description of the empirical findings, focusing on
percentage and descriptive (mean and median) statistics for each question in the survey.
Additional discussion of the findings are provided in Mears (2005) and Mears and Castro (2006).
Nearly all wardens believed that the role of the criminal justice system is to achieve
deterrence (91 percent) and rehabilitation (87 percent), while a slightly lesser, albeit substantial,
majority believed that incapacitation (79 percent) and punishment (69 percent) are primary goals.
A small fraction (8 percent) of wardens identified additional goals, which included public
protection and community safety, as well as offender retribution, resocialization, and
reintegration into society. Overall, the majority of wardens believed that the role of the criminal
justice system is to serve all four goals—deterrence, rehabilitation, incapacitation, and
punishment.
The study’s definition of a supermax derived in part from a 1996 survey by the National
Institute of Corrections (1997), in which supermax facilities are defined as places where inmates
“officially designated as exhibiting violent or serious and disruptive behavior” are confined in a
“free-standing facility” or “distinct unit” in a setting that involves “separation, restricted
movement, and limited direct access to staff and other inmates” (1). In the UI survey, a supermax
is defined as “a stand-alone unit or part of another facility and is designated for violent or
disruptive inmates. It typically involves up to 23- hour per day, single-cell confinement for an
indefinite period of time. Inmates in supermax housing have minimal contact with staff and other
inmates.” The definition acknowledges media and research depictions of supermax facilities,
which typically focus on the notion of single-cell, 23- hour-per-day confinement, generally for an
indefinite period of time, with few if any visitation privileges or access to programming or
services (Briggs et al. 2003; Nitkin 2003). Although there has been some disagreement as to
whether this definition is sufficient (King 1999; Kurki and Morris 2001), most wardens we
surveyed (95 percent) agreed with the definition provided.
Analysis of responses from respondents who self- identified as supermax wardens indicated
that, as of 2004, 44 states had supermaxes, up from the 34 states identified in the 1996 National
Institute of Corrections (1997) survey. Collectively, the 44 states—including Washington, D.C.,
and New York (the latter of which did not participate in the study but is known to have
supermaxes; [see, e.g., Pfeiffer 2004])—house approximately 25,000 inmates.
King (1999) provided an estimate of 19,630 inmates in supermax prisons in 34 states.
Because some wardens supervised multiple facilities, or facilities that were not entirely dedicated
to supermax inmates, a precise estimate of the number of inmates in supermax confinement is
not possible. Nonetheless, it is reasonable, and likely conservative, to project King’s (1999)
average per-state supermax inmate count (577) to arrive at an estimate of at least 25,000 inmates
in supermax confinement among the 44 states. The estimate is likely conservative become some
wardens may supervise facilities that many observers would call supermaxes but that these
3

This discussion is elaborated upon in a forthcoming article (Mears and Castro 2006), “Wardens’ Views
on the Wisdom of Supermax Prisons,” in Crime and Delinquency.

40

S UPERMAX PRISONS

URBAN INSTITUTE

wardens do not.
Table 4 presents the percentage, ranked from highest to lowest, of prison wardens who
agreed or strongly agreed that inmates with each of 12 possible characteristics should be placed
in a supermax. Wardens overwhelmingly indicated that inmates who exhibit violence or the
potential to instigate violence in others belong in supermax confinement. Almost all (99 percent)
wardens agreed that inmates who assault staff or other inmates repeatedly or cause injury should
be placed in a supermax. Roughly 80 percent or more agreed that inmates who instigate others to
be violent, are prison gang leaders, and present escape risks warrant such confinement.
Wardens were less certain about other inmate characteristics. Approximately half of all
wardens believed that drug dealers, chronic rule-violators, and prison gang members belong in a
supermax, while less than one-third felt that “high-profile” inmates and inmates at risk of being
attacked should be held in such confinement. Less than 20 percent of wardens felt that inmates
incarcerated for a serious offense or those with a serious mental illness belong in a supermax.
Four percent of respondents also listed other types of inmates who should be placed in
supermax facilities: sexual predators, terrorists, death row inmates, inmates who kill others while
in prison, and inmates who make or possess weapons.
We then asked supermax wardens to list the most common reasons for which inmates are
placed in their facilities. They listed virtually all of the same behavioral characteristics identified
in the first sections of table 4. The response is striking because of the relatively high levels of
disagreement among wardens regarding the appropriateness of supermax for many different
types of inmates. Perhaps more striking is the list of additional reasons wardens identified for
which inmates are placed in supermax prisons. Although several characteristics associated with
the different types of inmates may appear to conform generally with the logic of a supermax
(e.g., being a constant threat to staff, the public, and other inmates; inciting, leading, or
participating in riots), other characteristics are substantially less so (e.g., a failure to adjust to
prison life; having major medical problems; being a repeat offender; being a young adult
offender; refusing to live elsewhere) (see table 4, bottom section). For states that view supermax
confinement as appropriate for the so-called “worst of the worst,” the underlying question is
whether the inmates actually being placed in supermax confinement truly meet this standard.
Prison wardens were asked to indicate the extent to which they agreed that supermax prisons
serve to achieve each of 12 possible goals, as well as the extent to which they agreed supermax
states have been successful at achieving each goal. Over 95 percent of prison wardens agreed
that supermax prisons exist to increase safety, order, and control throughout the prison system
and to incapacitate violent and disruptive inmates. Approximately 80 percent believed that the
goals of supermax states are to improve inmate behavior throughout the prison system (83
percent) and to decrease riots (82 percent), the influence of gangs (79 percent), and escapes (71
percent). Close to half agreed that supermax prisons are used to punish (49 percent) and reduce
recidivism (45 percent) among violent and disruptive inmates. More than one-third of wardens
agreed that supermax prisons serve to rehabilitate these inmates (36 percent), and less than onequarter (24 percent) agreed that they deter crime in society.
The range of warden agreement regarding the effectiveness of supermaxes in achieving each
goal was similarly distributed. Ninety- five percent agreed that supermax prisons had successfully
increased safety, order, and control throughout the prison system and incapacitated violent and
disruptive inmates. Nearly 80 percent believed that supermax prisons improved inmate behavior
41

S UPERMAX PRISONS

URBAN INSTITUTE

and decreased both riots, the influence of gangs throughout the prison system, and escapes.
Sixty-one percent of wardens felt that supermax prisons effectively punished violent and
disruptive inmates, but substantially less (42 percent) believed the recidivism of such inmates
had been reduced. Nearly one-third agreed that supermax prisons successfully rehabilitated
inmates, and less than one-quarter believed that supermax prisons successfully deterred crime in
society.
Figure 2 presents the percentage of wardens who stated that supermax prisons had increased,
decreased, or not affected each of 20 potential areas of impact. Out of the 20 items listed, only 7
positive impacts were identified by a majority of respondents. More than 80 percent of wardens
indicated that supermax prisons had increased staff safety and order within prison institutions,
and three-quarters believed inmate safety had increased as well. In addition, over two-thirds of
wardens felt that supermax prisons decreased the number of inmate violent acts, and nearly 60
percent believed supermaxes decreased inmate fear of victimization. Almost half of wardens
believed that supermaxes decreased staff use-of- force incidents and staff fear of victimization.
For the other 13 areas, most wardens believed that supermax prisons have had no impact.
Over 70 percent of wardens indicated that supermax facilities had not affected inmate recidivism
after release, staff disciplinary actions, inmate mental health, local business development,
community residents’ fear of crime, staff turnover, and support for local politicians. More than
60 percent of wardens believed supermax had not affected inmate complaints against staff, local
government tax revenues, and inmate perception of the legitimacy of the prison system. Finally,
over half of all wardens felt that local employment, inmate access to programs, and inmate
infractions had not changed as a result of the presence of supermax prisons.
When evaluating any program or policy, we want information not only about whether
specific goals are achieved, but also about whether unintended effects arise that might affect the
overall assessment (Rossi et al. 1999). Perhaps the most commonly leveled compla int about
supermax prisons is that they cause or increase mental illness among the inmates housed within
them (e.g., Haney and Lynch 1997). Apart from this presumably unintended effect, few other
effects have been explored by researchers. Yet wardens in this study identified a range of
unintended effects, some positive and some negative.
Positive unintended effects included improving staff effectiveness by increasing the amount
and quality of staff training, teamwork, and professionalism, and creating better staff working
conditions, which, in turn, contribute to reduced staff burnout and turnover. Wardens also noted
that supermax prisons increase inmate morale and perceptions among inmates that prison
authority is legitimate. Supermax prisons also reportedly make it easier to deliver programming
to general population inmates. Last but not least, wardens identified supermax effects that fell
outside of the correctional system. They suggested, for example, that supermax prisons increase
public perceptions of safety, enhance the correctional system’s relationships with local
communities, improve local economies, and, more generally, heighten the prestige of the
correctional system among corrections agencies in other states.
Wardens also identified many negative unintended effects. They cited increases in staff abuse
of authority, disciplinary actions, and use-of- force incidents. Some wardens indicated that the
presence of supermaxes creates a false sense of security among staff, which in turn lulls them
into greater complacency and less vigilance. They suggested that these prisons actually increase
staff and inmate fear of victimization, and argued that supermax confinement constitutes cruel

42

S UPERMAX PRISONS

URBAN INSTITUTE

and unusual punishment because some inmates, such as mentally ill and nuisance inmates
receive little to no appropriate treatment or services. The wardens also highlighted systemwide
effects, such as increased inmate violence and decreased perceptions among inmates that prison
authority is legitimate. As with the positive unintended effects, wardens identified negative
effects external to the prison system, including concerns about increased recidivism and reentry
failure among released supermax inmates, decreases in local business development and property
values, and increases in the public’s fear of crime.
The extent of any of these effects is largely unknown. It appears likely that some effects may
occur in specific states and not in others, and that the way in which correctional systems utilize
their supermax prisons influences which unintended outcomes arise.
Finally, we grouped wardens into categories according to the highest level of security at the
institution they supervised. Our final categorization of wardens showed that nearly one-quarter
(23 percent) supervised supermax facilities, more than one-third supervised
maximum/close/high-security institutions, and 27 percent supervised medium security
institutions. Slightly more than 10 percent supervised minimum security prisons, and a few
wardens supervised other institutions (e.g., reception centers).
Three-quarters of wardens came from states with a supermax facility. Of these wardens,
nearly 70 percent had sent an inmate to a supermax, and more than 50 percent had received an
inmate from a supermax. Supermax wardens reported that their institutions were operational as
early as 1950 and as recently as 2004. However, wardens from only two states said they had a
supermax that became operational before 1980 (one in 1950, and the other in 1976).
Cross-classifying respondents who indicated that they were supermax wardens with states
revealed that 44 states have one or more supermax facilities. The estimate includes Washington,
DC, and New York. The latter is reported to have supermax facilities, according to King (1999)
and others (e.g., Pfeiffer 2004). Clearly, more states have invested in supermax prisons since the
last survey of states, which was conducted in 1996 and indicated that 34 states had supermaxes
(NIC 1997).

43

S UPERMAX PRISONS

8.

URBAN INSTITUTE

Benefit-Cost Analysis of Supermax Prisons

For the supermax project, the Urban Institute created two stand-alone products, a policy brief
(Lawrence and Mears 2004) and a Microsoft Excel spreadsheet–based benefit-cost analysis tool.
This chapter briefly describes these two products.
Benefit-Cost Analysis Policy Brief

As noted earlier, the UI research team learned early on in the project that corrections
executives and practitioners, as well as policymakers focused on corrections issues, know little
about how BCAs are conducted, how they can best be used, and the critical role that assumptions
of various kinds play in BCAs. Based on conversations with UI researchers and practitioners in
the field, the research team determined that a report was needed that corrections executives,
practitioners, and policymakers could use in making more informed requests for BCAs,
interpreting BCA results, and conducting their own BCAs.
To this end, the team created a policy brief that was reviewed by researchers, BCA experts,
and practitioners. The brief provides examples of practical applications of benefit-cost analysis,
introduces the logic of this analytic tool, describes the specific steps involved in conducting a
benefit-cost analysis, and then shows how these steps apply to supermax prisons. The brief
emphasizes the critical role that informed judgments and assumptions, along with empirical
research, play in affecting the results of benefit-cost analyses.
Benefit-Cost Analysis Tool

As part of the supermax project, the researchers also created a benefit-cost analysis tool,
which is meant to help corrections executives and practitioners understand the basic steps
involved in conducting benefit-cost analyses of supermax prisons. The tool serves to illustrate
the critical decisions that must be made for benefit-cost analyses to be useful, emphasizes the
need for accurate monetization of costs and impacts, and highlights how assumptions made by
analysts or executives (e.g., about the perspective of analysis) can affect the outcomes of a BCA.
Both the tool and the BCA policy brief provide examples of and guidance on how to conduct
benefit-cost analyses. The tool also allows users to practice conducting a BCA within the
constraints of a predetermined scenario. The tool identifies one question to be considered
(whether building a supermax prison would be cost-beneficial) and specifies the perspective of
analysis (the perspective of a department of corrections), and then allows the user to enter data
and walk through each of the steps needed to answer this question. Similar tools would be
needed for other frames of reference (e.g., the perspective of society). Although the basic
structure of BCA remains the same in each instance, alternative perspectives can entail
fundamentally different sets of impacts and thus different monetized values for generating the
benefit-cost ratio and net difference between benefits and costs.
It should be emphasized that the tool has not been designed to replace a professional BCA
tailored to a project’s specific nuances. Rather, the tool, along with the BCA policy brief, is
designed to help policymakers and corrections executives and practitioners become better
educated requesters and consumers of benefit-cost analyses of supermax prisons.

44

S UPERMAX PRISONS

9.

URBAN INSTITUTE

Lessons Learned

The analyses above point to several generalizations that have important ramifications for
research and policy. Here, several of the more striking findings are presented, and two critical
issues are given special focus—the causal logic of supermaxes and the criteria for determining
whether states should invest in them.
First, contrary to what many observers have argued, definitional issues may not be as
problematic as they are sometimes held to be (see, e.g., King 1999). For example, despite
disagreements among some scholars and practitioners concerning the definition of a supermax,
over 95 percent of state prisons wardens agreed with a modified version of the definition of a
supermax used by the National Institute of Corrections in its 1996 survey of state correctional
systems. To be clear, definitional issues can be critical, especially for ensuring that discussions
are about “apples and apples” rather than “apples and oranges.” However, no amount of precise
details can substitute for a general definition that allows commonalities to be identified.
The 1996 National Institute of Corrections (1997) survey, updated by King (1999) two years
later, indicated that 34 states had supermax prisons and that more states had supermaxes in
planning stages. That assessment clearly was correct. Six years later, at the time of the UI survey,
44 states were identified as having at least one supermax. That growth is striking and suggests
that supermax prisons are likely to remain a common feature of criminal justice in the United
States for the indefinite future.
In contrast to some criminal justice policies that have narrowly defined goals, supermaxes are
associated with a remarkable diversity of goals that policymakers, corrections officials,
practitioners, and research attach to them, goals that range from increasing safety and order to
reducing escapes to deterring crime in society. Notably, wardens overwhelmingly (over 95
percent) agree that supermaxes serve to increase safety, order, and control in prison systems and
function to incapacitate violent or disruptive inmates. That consistency is important from the
standpoint of evaluation, as it suggests that these goals should be given priority. However, states
may give different weightings to certain other goals, and so any fair or balanced assessment of
supermaxes should likely take such weightings into account on a state-by-state basis.
Wardens generally believe that supermax prisons are effective in achieving the four key
goals of safety, order, control, and incapacitation of violent or disruptive inmates, but agree less
about other goals. That assessment should not be taken lightly because it is wardens, especially
general population wardens, who stand to benefit from supermaxes achieving such goals. At the
same time, such beliefs stand in stark contrast to the dearth of empirical studies assessing these
or any of the other goals associated with supermaxes. Two concerns thus emerge: Supermaxes
may not be contributing to these goals, even as states continue to invest in them, and so may
constitute a questionable investment, or, conversely, they may be contributing substantially to
many of them, suggesting that perhaps greater investment is warranted.
An additional concern that many respondents raised in the study centered on negative
unintended effects of supermax prisons. Examples well exceeded the typical concern about the
mental health of supermax inmates, and included the possibility that supermaxes actually
increase systemwide disorder and violence and contribute to staff turnover. Further probing
revealed, however, many positive unintended effects, such as improving the living conditions

45

S UPERMAX PRISONS

URBAN INSTITUTE

and outcomes for general population inmates. Here again, without systematic empirical
investigation of such effects, supermax advocates and opponents alike stand on a less-than- firm
foundation for promoting or criticizing supermaxes.
At a time when governments increasingly are calling for state agency accountability
(Campbell 2003), the absence of benefit-cost analyses of supermaxes is surprising. Such analyses
may be difficult to undertake and involve considerable complexity. Yet supermaxes arguably
represent a close to $1 billion investment over 30 to 40 years, the typical life span of a prison. In
that context, even a crude BCA might well provide important guidance about whether
supermaxes merit less or more investment.
Juxtaposed against these observations are two critical issues—the causal logic of supermaxes
and the criteria for assessing them—that have largely been ignored in research and debates about
supermax prisons. (The discussion here draws on Mears [2005], as well as Mears and Castro
[2006], Mears and Reisig [2006], and Mears and Watson [2006].] The causal logic question is
important because absent a coherent logical foundation, it is unclear why policymakers would
invest in supermaxes. Upon first glance, the logic seems inescapable—incarcerate the putative
“worst of the worst” and prison systems will improve. However, investigation of this logic raises
questions about whether supermaxes are, in theory or practice, likely to be effective in achieving
any of a range of goals.
Consider systemwide order as one example. For supermaxes to achieve this goal, they would
presumably need to deter other, nonsupermax inmates who create disorder. But how likely are
general population inmates to be deterred if they know that fewer than 1 to 2 percent of inmates
are placed in a supermax? Similarly, are supermax inmates typically confined in a supermax long
enough to produce a deterrent effect? (What in fact would the length of time need to be to
produce a substantial effect?) If the effect arises through incapacitation, then it must be true that
most systemwide disorder results from a few troublesome inmates. There is, however, little
research to suggest that such is the case. Even if it were, the question arises: How many such
inmates are there in a given system? If there are more than a supermax or two can hande, then
would there likely be a substantial improvement in order? Viewed differently, a large literature
suggests that prison order results primarily from the consent of inmates to adhere to rules, and
that if the legitimacy of prison authority comes into question, then adherence to these rules will
diminish (Sparks et al. 1996). Respondents in this study echo concerns raised in other studies
(e.g., Kurki and Morris 2001) that inmates believe supermax facilities are sometimes used in a
capricious or arbitrary manner. To the extent that such observations are true, supermaxes actually
could contribute to more not less disorder. 4
Similar observations can be made about other goals. For example, for supermaxes to reduce
escapes, officials must be able to readily and accurately identify inmates who might escape. Yet
validated instruments for this purpose that do not entail a large number of false positives (i.e.,
that do not entail identification of certain inmates as being at risk of escape who really are not at
risk), do not yet exist. Supermaxes might also deter would-be offenders in society, but that
premise requires that such offenders know that supermaxes exist and that such a fact concerns
them.
4

This discussion is elaborated upon in a forthcoming article (Mears and Reisig 2006), “The Theory and
Practice of Supermax Prisons,” in Punishment and Society .

46

S UPERMAX PRISONS

URBAN INSTITUTE

In short, important questions exist about not only whether but how supermax prisons might
be effective. To date, the presumption—evidenced by the fact that 44 states now have
supermaxes—appears to be that the logic, the “how,” is obvious and likely to be effective. Future
studies may reveal coherent logics that are grounded both in theory and in practice, and that
buttress arguments for supermaxes. There is little evidence to date that such exist. Yet examining
the logic of supermax prisons is important for a simple and pragmatic reason—if we know how
supermaxes contribute to specific outcomes, we then can modify them to increase the likelihood
of such outcomes. If, for example, general population inmates assault each other and staff less
often out of a fear of supermax placement, then measures might be taken—so long as they were
legal and ethical—to capitalize on that fear. For example, greater effort could be expended
advertising the fact that supermax facilities exist or that placement in them is likely in the event
of misconduct. If, however, reduced assaults result primarily from incapacitating specific
inmates, then such measures would not be needed and would be unlikely to be effective.
The second critical issue concerns the lack of systematic assessments of supermax prisons.
To date, advocates and opponents alike have largely focused on a delimited set of dimensions—
supermax prisons control unruly inmates or they cause or aggravate mental illness, for example.
Such debates risk shifting public and policymaker support either for or against, and do so with
little reference to the range of issues identified in this study that are relevant to any general
assessment of supermaxes.
As emphasized above, supermaxes are associated with many goals, not just one, and so
assessments should reflect that diversity. They also are associated with many unintended effects,
some of which alone might suffice to generate considerable support or opposition for them.
Supermax prisons represent a policy designed to achieve particular goals, and so assessments
should be comparative. How well does a supermax achieve particular goals and minimize
unintended effects compared with doing nothing? Or with investing in other types of policies?
Debates about supermaxes should have reference to the potential impacts associated with these
alternatives if they are to contribute to reasoned decisionmaking. Not the least, supermaxes raise
political, moral, and economic questions that would be relevant to any overall assessment. To
what extent are supermaxes supported or opposed primarily on political grounds? To what extent
do they raise moral or humanitarian concerns, and are such sufficient to affect decisions about
supermaxes, no matter how effective they might be? And economically, are supermaxes
sensible? Do they generate a return that merits the investment, especially as compared with
alternatives?
All of these dimensions are relevant to arriving at balanced policy decisions (Mears and
Watson 2006), yet few states appear to have given explicit and systematic attention to them. It is,
therefore, essential that policymakers and corrections executives support research that can help
determine whether supermax prisons are, or are likely to be, effective. Since the goals vary by
state, evaluations should be conducted on a state-by-state basis. Such research need not be costly.
Indeed, where funds are minimal, considerable advances can be made in efforts to clarify the
goals and logic of supermax prisons and to improve appropriate supermax operations.
For researchers, a veritable raft of questions remains to be empirically investigated. Clearly,
investigation into the actual effectiveness of supermaxes in achieving specific goals is needed.
But little is known about the use of supermaxes. For example, few studies document the extent to
which the criteria states articulate for the placement and release of prisoners are followed. Even
less is known about how long inmates stay on average in a supermax and how long they then are
47

S UPERMAX PRISONS

URBAN INSTITUTE

in traditional prisons before release to society. How many are released from a supermax straight
into communities? What is the behavior of released supermax inmates upon reentry into other
prisons or into society? What are the characteristics (e.g., age, sex, race/ethnicity, prior record
and length-of-stay, and behavior that led to supermax confinement) of inmates placed in
supermax facilities and have these characteristics changed over time? Investigations into such
questions would yield considerable insight into the operations of prison systems generally, but
also would provide directly relevant and useful operational information to corrections officials.

48

S UPERMAX PRISONS

URBAN INSTITUTE

10. Conclusion
Supermax prisons represent a marked change in how correctional systems attempt to achieve
a number of goals, most notably improving systemwide order and safety and control over the
most violent and disruptive inmates. These and other goals identified in this study are eminently
defensible. It is, for example, difficult to argue with the desire to create a prison system that is
orderly and safe. At the same time, supermax prisons appear, at least to state prison systems, to
be a reasonable response to management needs they face.
Supermax prisons may in fact prove to be an effective corrections management tool, one that
is cost-effective and that achieves outcomes that no other approach can. The results of this study
suggest otherwise, however. There is little research—including the present study’s analyses of
interviews with correctional policymakers, executives, and practitioners, and a survey of state
prison wardens—to suggest that supermax prisons effectively achieve any of a range of goals,
including improving systemwide order and safety; however, much research, including the
present study, suggests that these prisons are unlikely to be able to achieve these goals.
In addition, the present study has highlighted a range of considerations that might help
increase the chances that this prison management tool can be effective. Of course, to critics who
view supermax confinement as inherently inhumane and unconstitutional, no improvements
would justify supporting supermax prisons. And their concerns are unlikely to be alleviated by
the results of empirical studies or bene fit-cost analyses.
Setting aside such arguments, supermax prisons are “here to stay” for the indefinite future,
and many states are continuing to invest in them. Thus, as this report has suggested, there is a
pressing need for research and discussions of the issues presented in the different chapters. Not
least, considerably more empirical research is needed on whether supermax prisons can and do
achieve a range of specific goals. Such research should be complemented by studies that explore
whether other strategies can achieve comparable goals at a lower cost and with fewer potential
unintended effects (see, e.g., Briggs et al. 2003; Gendreau and Keyes 2001; Haney 2003a).
Indeed, without such research, correctional policymakers and executives will have little basis for
knowing how to improve their supermax operations, and they will have little incentive or
foundation to forego what many of them see as a necessary, even if costly, tool in effectively
managing prison systems.

49

S UPERMAX PRISONS

URBAN INSTITUTE

11. References
Abramsky, Sasha. 1999. “When They Get Out.” The Atlantic Monthly 283: 30–36. June.
http://www.theatlantic.com/issues/99jun/9906prisoners.htm. (Accessed June 2, 2003.)
———. 2002. “Return to the Madho use.” The American Prospect 13: 26–29. February 11.
http://www.prospect.org/print/V13/3/abramsky-s.html. (Accessed June 2, 2003.)
Access Control and Security Systems. 2002. “Video History Increases Accountability at
Maryland’s Supermax Correctional Facility.” Bellevue, NE: Primedia Business Magazine
and Media, Inc., September 1. http://www.securitysolutions.com. (Accessed June 24, 2003.)
Adams, Kenneth. 1992. “Adjusting to Prison Life.” In Crime and Justice: A Review of Research,
vol. 16, edited by Michael H. Tonry (275–359). Chicago, IL: University of Chicago Press.
———. 1996. “The Bull Market in Corrections.” The Prison Journal 76: 461–467.
Akers, Ronald L., and Christine Sellers. 2004. Criminological Theories: Introduction,
Evaluation, and Application, 4th ed. Los Angeles, CA: Roxbury.
American Civil Liberties Union (ACLU). 2001. “ACLU Class Action Lawsuit Targets Brutal
Conditions at Ohio’s ‘Supermax’ Prison.” New York: American Civil Liberties Union,
January 9. http://www.aclu.org. (Accessed March 12, 2003.)
———. 2002a. “Federal Judge Says Ohio ‘Supermax’ Prison Violates Constitutional Rights.”
New York: American Civil Liberties Union, February 26. http://www.aclu.org. (Accessed
February 26, 2002.)
———. 2002b. “Ohio ‘Supermax’ Prison Case Goes to Trial Monday; Focus Will be on Brutal
Conditions at Penitentiary.” New York: American Civil Liberties Union, January 4.
http://www.aclu.org. (Accessed March 12, 2003.)
American Correctional Association. 2003. 2003 Directory of Adult and Juvenile Departments,
Institutions, Agencies, and Probation and Parole Authorities. Lanham, MD: American
Correctional Association.
Amnesty International. 2001. “Amnesty International Condemns Housing Minors in Wisconsin
Supermax Prison.” London: Amnesty International News Release, July 26.
http://www.amnestyusa.org/news/2001/usa07262001.html.
Anderson, Lisa. 1998. “Is ‘Supermax’ Too Much? Strict Isolation of Inmates Called Cruel,
Ineffective.” Chicago Tribune, August 2, C1.
Associated Press. 2002. “Ohio Ordered to Revise Inmate-Transfer Rules.” The Cincinnati
Enquirer, March 28, p. C6. Online: enquirer.com. (Accessed 3/12/03.)

50

S UPERMAX PRISONS

URBAN INSTITUTE

———. 2003. “Inmates Volunteer to Stay at ‘Supermax’ Prison.” News Channel 5, March 4.
http://www.newsnet5.com. (Accessed March 12, 2003.)
Atherton, Eugene. 2001. “Incapacitation with a Purpose.” Corrections Today 63(4): 100–103.
Austin, James. 2003. Findings in Prison Classification and Risk Assessment. Washington, DC:
National Institute of Corrections.
Austin, James, and John Irwin. 2001. It’s About Time: America’s Imprisonment Binge. Belmont,
CA: Wadsworth.
Austin, James, and Kenneth McGinnis. 2004. Classification of High-Risk and Special
Management Prisoners: A National Assessment of Current Practices. Washington, DC:
National Institute of Corrections.
Austin, James, Stan Repko, Robert Harris, Ken McGinnis, and Susan Plant. 1998. Evaluation of
the Texas Department of Criminal Justice Administrative Segregation Population.
Washington, DC: National Council on Crime and Delinquency.
Ayidiya, Stephen A., and McKee J. McClendon. 1990. “Response Effects on Mail Surveys.”
Public Opinion Quarterly 54: 229–247.
Babbie, Earl. 1995. The Practice of Social Research, 7th ed. Belmont, CA: Wadsworth.
Baim, Julian. 1991. “Response Rates: A Multinational Perspective.” Marketing and Research
Today 19: 115–119.
Baruch, Yehuda. 1999. “Response Rate in Academic Studies: A Comparative Analysis.” Human
Relations 52: 421–438.
Beck, Allen J. 2001. Prisoners in 2000. Washington, DC: U.S. Department of Justice, Bureau of
Justice Statistics.
Beck, Allen J., Jennifer C. Karberg, and Paige M. Harrison. 2002. Prison and Jail Inmates at
Midyear 2001. Washington, DC: Bureau of Justice Statistics.
Bidna, Howard. 1975. “Effects of Increased Security on Prison Violence.” Journal of Criminal
Violence 3: 33–46.
Bloomberg, Seth A. 1977. “Participatory Management: Toward a Science of Correctional
Management.” Criminology 15: 149–163.
Blumstein, Alfred, and Allen J. Beck. 1999. “Population Growth in U.S. Prisons.” In Crime and
Justice: A Review of Research, edited by Michael H. Tonry and Joan Petersilia (17–61).
Chicago: University of Chicago Press.
Boin, Arjen. 2001. “Securing Safety in the Dutch Prison System: Pros and Cons of a Supermax.”
The Howard Journal 40: 335–346.

51

S UPERMAX PRISONS

URBAN INSTITUTE

Bonczar, Thomas P., and Tracy L. Snell. 2003. Capital Punishment, 2002. Washington, DC:
Bureau of Justice Statistics.
Borg, Marian J. 1997. “The Southern Subculture of Punitiveness? Regiona l Variation in Support
for Capital Punishment.” Journal of Research in Crime and Delinquency 34: 25–45.
Bottoms, Anthony E. 1999. “Interpersonal Violence and Social Order in Prisons.” In Crime and
Justice: A Review of Research, edited by Michael H. Tonry and Joan Petersilia (205–282).
Chicago: University of Chicago Press.
Bradburn, Norman M. 1992. “Presidential Address: A Response to the Nonresponse Problem.”
Public Opinion Quarterly 56: 391–397.
Branham, Lynn S. 1997. Limiting the Burdens of Pro Se Inmate Litigation: A TechnicalAssistance Manual for Courts, Correctional Officials, and Attorneys General.” Chicago, IL:
American Bar Association, Criminal Justice Section.
Briggs, Chad S., Jody L. Sundt, and Thomas C. Castellano. 2003. “The Effect of Supermaximum
Security Prisons on Aggregate Levels of Institutional Violence.” Criminology 41: 1341–
1376.
British Broadcasting Corporation News (BBC). 2000. “US Rebuked over Torture.” May 15.
http://news.bbc.co.uk/1/hi/world/americas/749606.stm. (Accessed April 10, 2004.)
Brook, Daniel. 2003. “A History of Hard Time: Solitary Confinement, Then and Now.” Legal
Affairs 2: 39–42.
Bruton, James H. 2004. The Big House: Life Inside a Supermax Security Prison. Stillwater, MN:
Voyageur Press.
Burns, Tom, and George M. Stalker. 1994. The Management of Innovation, 3rd ed. New York:
Oxford University Press.
Burt, Marti. 1981. Measuring Prison Results: Ways to Monitor and Evaluate Corrections
Performance. Washington, DC: National Institute of Justice.
Burt, Martha R., Nancy Pindus, and Jeffrey Capizzano. 1999. The Social Safety Net at the
Beginning of Federal Welfare Reform: Organization of and Access to Social Services for
Low Income Families. Washington, DC: Urban Institute Press.
Camp, Camille G., and George M. Camp. 1987. The Corrections Yearbook 1987: Instant
Answers to Key Questions in Corrections. South Salem, NY: Criminal Justice Institute.
———. 1999. The Corrections Yearbook 1999: Adult Corrections. Middletown, CT: Criminal
Justice Institute.
———. 2002. The Corrections Yearbook 2001: Adult Systems. Middletown, CT: Criminal
Justice Institute.

52

S UPERMAX PRISONS

URBAN INSTITUTE

Camp, Scott D., and Gerald G. Gaes. 2002. “Growth and Quality of U.S. Private Prisons:
Evidence from a National Survey.” Criminology and Public Policy 1: 427–450.
Campbell, Robin. 2003. Dollars and Sentences: Legislators' Views on Prisons, Punishment, and
the Budget Crisis. New York: Vera Institute of Justice.
Caplow, Theodore, and Jonathon Simon. 1999. “Understanding Prison Policy and Population
Trends.” In Prisons, edited by Michael H. Tonry and Joan Petersilia (63–120). Chicago, IL:
University of Chicago Press.
Caudle, Sharon L. 2004. “Using Qualitative Approaches.” In Handbook of Practical Program
Evaluation, 2nd ed., edited by Joseph S. Wholey, Harry P. Hatry, and Kathryn E. Newcomer
(417–438). San Francisco: Jossey-Bass.
Chaptman, Dennis. 2001. “Data Shows Majority in Supermax Assaultive.” JSOnline Milwaukee
Journal Sentinel, October 29. http://www.jsonline.com/news/state/oct01/super30102901.asp.
(Accessed January 6, 2003.)
Chua-Eoan, Howard, and Victoria Rainert. 1999. “The Bomber Next Door.” Time 153(11): 55.
March 22.
Clare, Emma, and Keith Bottomley. 2001. “Evaluation of Close Supervision Centres.” Research,
Development and Statistics Directorate Research Study 219. London: Home Office of the
Research, Development, and Statistics Directorate.
Clarke, Matthew T. 2001. “Ohio ACLU Challenges Supermax.” Prison Legal News 12(10): 28.
October.
Clear, Todd R., and George F. Cole. 2003. American Corrections, 6th ed. Belmont, CA:
Wadsworth.
CNN.com. 2000. “New Prison Brings Money, Jobs, and Criticism.” CNN.com, February 15.
http://www.cnn.com/2000/US/02/15/wis.prison.economy. (Accessed March 29, 2004.)
Cohen, Mark A. 2000a. “The Economics of Crime and Punishment: Implications for Sentencing
of Economic Crimes and New Technology Offenses.” George Mason University Law Review
9: 503–28.
——–. 2000b. “Measuring the Costs and Benefits of Crime and Justice.” In Criminal Justice
2000, vol. 4, Measurement and Analysis of Crime and Justice, edited by David Duffee (263–
315). Washington, DC: National Institute of Justice.
Cohen, Mark A., Ted R. Miller, and Shelli B. Rossman. 1994. “The Costs and Consequences of
Violent Behavior in the United States.” In Understanding and Preventing Violence:
Consequences and Control of Violence, vol. 4, edited by Albert J. Reiss, Jr., and Jeffrey A.
Roth (67–166). Washington, DC: National Academy Press.
Cohen, Fred. 2002. “Supermax, Segregation, and Seriously Mentally Ill Inmates: Wisconsin

53

S UPERMAX PRISONS

URBAN INSTITUTE

Facility under Fire but Basic Questions Remain.” Correctional Law Reporter 13(6): 85–6.
April/May.
Collins, William C. 2004. Supermax Prisons and the Constitution: Liability Concerns in the
Extended Control Unit. Washington, DC: National Institute of Corrections.
Conover, Ted. 2000. Newjack: Guarding Sing Sing. New York: Random House.
——–. 2001. “Behind Bars with 1,000 Male Convicts.” New York Times Magazine, September
9.
Cooke, David J. 1989. “Containing Violent Prisoners: An Analysis of the Barlinnie Special
Unit.” British Journal of Criminology 29:129–143.
——. 1991. “Criminal Victimization in Male Federal Prisons.” Howard Journal of Criminal
Justice 30:95–107.
Correctional Law Reporter. 2002. “Supermax Placement Demands Wolff Hearing, Says Ohio
Judge.” Correctional Law Reporter 14(1): 1–2.
Corrections Digest. 1995. “DOJ Launches Probe of Maryland’s ‘Supermax’ Prison.” Corrections
Digest 26(6): 2–3.
——–. 2001. “Connecticut Ends Pact with Virginia.” Corrections Digest 32(30): 4.
——–. 2002a. “Supermax Puts Seals at Doors for Security.” Corrections Digest 33(4): 4.
——–. 2002b. “Wisconsin Settles Supermax Lawsuit.” Corrections Digest 33(2): 6.
Corrections Today. 2002. “Private Sector News.” Corrections Today, October.
Corzine, Jay, Lin Huff-Corzine, and Hugh P. Whitt. 1999. “Cultural and Subcultural Theories of
Homicide.” In Homicide: A Sourcebook of Social Research, edited by Dwayne Smith and
Margaret Zahn (42–57). Thousand Oaks, CA: Sage.
Crouch, Ben M., and James W. Marquart. 1989. An Appeal to Justice: Litigated Reform of Texas
Prisons. Austin: University of Texas Press.
Cullen, Francis T. 2005. “The Twelve People Who Saved Rehabilitation: How the Science of
Criminology Made a Difference.” Criminology 43: 1–42.
Cunningham, Mark D., and Mark P. Vigen. 2002. “Death Row Inmate Characteristics,
Adjustment, and Confinement: A Critical Review of the Literature.” Behavioral Sciences and
the Law 20: 191–210.
Currie, Elliott. 1998. Crime and Punishment in America. New York: Henry Holt.
Davis, Peter. 1999. Inspection Report: Ohio State Penitentiary. December 8. Columbus, OH:

54

S UPERMAX PRISONS

URBAN INSTITUTE

Correctional Institution Inspection Committee of the 123rd General Assembly.
DeMaio, Jerry R. 2001. “If You Build It, They Will Come: The Threat of Overclassification in
Wisconsin’s Supermax Prison.” Wisconsin Law Review 2001: 207–248.
Department of Maryland State Police and Maryland Department of Public Safety and
Correctional Services. 2001. “Department of Maryland State Police Charge Inmate with
Murder and Department of Public Safety Criminal Investigators Charge Same Inmate with
Attempted 1st Degree Murder of a Correctional Officer.” State of Maryland News Release,
January 26. Pikesville, MD: Department of Maryland State Police.
DiIulio, John J., Jr. 1987. Governing Prisons. New York: Free Press.
DiIulio, John J., Jr., Geoffrey P. Alpert, Mark H. Moore, George F. Cole, Joan Petersilia, Charles
H. Logan, and James Q. Wilson. 1993. Performance Measures for the Criminal Justice
System. Washington, DC: Bureau of Justice Statistics.
Dillman, Don A. 2000. Mail and Internet Surveys: The Tailored Design Method. New York:
John Wiley.
DiMaggio, Paul J., and Walter W. Powell. 1983. “The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizational Fields.” American Sociological
Review 48: 147–160.
Ekland-Olson, Sheldon. 1986. “Crowding, Social Control, and Prison Violence: Evidence from
the Post-Ruiz Years in Texas.” Law and Society Review 20: 389–421.
Elsner, Alan. 2004. Gates of Injustice: The Crisis in America’s Prisons. New York: Pearson.
Fabelo, Tony. 1996. “Whatever Is Next After the Prison-Building Boom Will Be Next in Texas.”
The Prison Journal 76: 475–483.
Federal Reserve Bank of Cleveland. 2003. The Ohio Budget. Cleveland, OH: Federal Reserve
Bank of Cleveland. http://www.clevelandfed.org/Research/ETCurrent/Current/ohiobud.pdf.
(Accessed May 22, 2003.)
Fellner, Jamie. 2000. Out of Sight: Super-Maximum Security Confinement in the United States,
vol. 12, no. 1. New York: Human Rights Watch. http://www.hrw.org/reports/2000/supermax.
(Accessed January 6, 2003.)
Fesperman, Dan. 2003. “State to Shift Prison Policy toward Rehabilitation.” The Baltimore Sun,
November 6, 1B.
Finn, Peter. 2000. Addressing Correctional Officer Stress: Programs and Strategies.
Washington, DC: U.S. Department of Justice, National Institute of Justice.
Florida Corrections Commission. 2000. 2000 Annual Report. Tallahassee, FL: Florida
Corrections Commission. http://www.fcc.state.fl.us/fcc/reports/final00/5smu.htm. (Accessed

55

S UPERMAX PRISONS

URBAN INSTITUTE

March 23, 2003.)
Franklin, Richard H. 1998. “Assessing Supermax Operations.” Corrections Today 60: 126–127.
Freeze, Colin. 2002. “Once a Prisoner…” Globe News, May 11.
http://www.globebooks.com/Justice_MichaelJackson.html. (Accessed January 6, 2003.)
Gaes, Gerald, Timothy J. Flanagan, Laurence L. Motiuk, and Lynn Stewart. 1999. “Adult
Correctional Treatment.” In Prisons, edited by Michael H. Tonry and Joan Petersilia (361–
426). Chicago: University of Chicago Press.
Gavora, Jessica. 1996. “The Prisoners’ Accomplice.” Policy Review 79: September–October.
http://www.policyreview.org/sept96/abuses.html. (Accessed July 14, 2003.)
Geier, Peter. 2002. “Supermax Death Suit Settled for $150,000.” The Daily Record 2(175),
January 31. http://www.mddailyrecord.com/pub/2_175_thursday/legalnews/73929-1.html.
(Accessed July 14, 2003.)
Gendreau, Paul, Claire Goggin, and Paula Smith. 2000. “Generating Rational Correctional
Policies.” Corrections Management Quarterly 4: 52–60.
Gendreau, Paul, and David Keyes. 2001. “Making Prisons Safer and More Humane
Environments.” Canadian Journal of Criminology 43: 123–130.
Gibbs, Jack. 1975. Crime, Punishment, and Deterrence. New York: Elsevier.
Gonnerman, Jennifer. 1999. “The Supermax Solution.” The Village Voice, May 19–25, 36–45.
http://www.villagevoice.com/issues/9920/gonnerman.php. (Accessed October 16, 2002.)
Good, Regan. 2003. “The Supermax Solution.” The Nation, March 3, 7.
Gramlich, Edward M. 1981. Benefit-Cost Analysis of Government Programs. Englewood Cliffs,
NJ: Prentice-Hall.
Grann, David. 2004. “The Brand: How the Aryan Brotherhood Became the Most Murderous
Prison Gang in America.” The New Yorker LXXX: 157–171. February 16.
Grassian, Stuart. 1983. “Psychopathological Effects of Solitary Confinement.” American Journal
of Psychiatry 140: 1450–1454.
Grassian, Stuart, and Nancy Friedman. 1986. “Effects of Sensory Deprivation in Psychiatric
Seclusion and Solitary Confinement.” International Journal of Law and Psychiatry 8: 49–65.
Hammack, Laurence. 2004. “Minimum-Security Prisons Bulging.” The Roanoke Times, April
11, B1.
Hanes, Stephanie. 2004. “State Lacks Federal Prison.” The Baltimore Sun, September 6, 1B.
http://www.baltimoresun.com/news/local/bal-

56

S UPERMAX PRISONS

URBAN INSTITUTE

md.detention06sep06,1,5717927.story?coll=bal- local- headlines. (Accessed September 13,
2004.)
Haney, Craig. 1998. “Riding the Punishment Wave: On the Origins of Our Devolving Standards
of Decency.” Hastings Women’s Law Journal 9: 27–78.
———. 2002. “Making Law Modern: Toward a Contextual Model of Justice.” Psychology,
Public Policy, and Law. 8: 3–63.
——. 2003a. “Mental Health Issues in Long-Term Solitary and ‘Supermax’ Confinement.”
Crime and Delinquency 49: 124–156.
——. 2003b. “The Psychological Impact of Incarceration: Implications for Postprison
Adjustment.” In Prisoners Once Removed: The Impact of Incarceration and Reentry on
Children, Families, and Communities, edited by Jeremy Travis and Michelle Waul (33–66).
Washington, DC: Urban Institute Press.
Haney, Craig, and Mona Lynch. 1997. “Regulating Prisons of the Future: A Psychological
Analysis of Supermax and Solitary Confinement.” New York University Review of Law and
Social Change 23: 477–570.
Harrison, Paige J., and Allen J. Beck. 2003. Prisoners in 2002. Washington, DC: Bureau of
Justice Statistics.
———. 2002. Prisoners in 2001. Washington, DC: Bureau of Justice Statistics.
Harry, Jennifer. 2002. “Wisconsin Ruling Defines Mentally Ill.” Corrections Today 64(3): 16.
Hatry, Harry. 1989. Comparison of Privately and Publicly Operated Corrections Facilities in
Kentucky and Massachusetts. Washington, DC: The Urban Institute.
Hellmuth, Obata, and Kassabaum, Inc. 2002. “Wisconsin SuperMax Prison.” January 17.
http://www.hok.com/ne w/newsOLD/. (Accessed January 6, 2003.)
Henningsen, Rodney J., W. Wesley Johnson, and Terry Wells. 1999. “Supermax Prisons:
Panacea or Desperation?” Corrections Management Quarterly 3: 53–59.
Hershberger, Gregory L. 1998. “To the Max: Supermax Facilities Provide Prison Administrators
with More Security Options.” Corrections Today 60: 54–57.
Hollister, H.W. 1955. “Why Prisoners Riot.” The Atlantic Monthly. 196(4): 65–67. October.
http://www.theatlantic.com/issues/95nov/prisons/whyriot.htm. (Accessed January 31, 2003.)
Hook, Julie. 2003. Mangos to Mangos: Comparing the Operational Costs of Juvenile and Adult
Correctional Programs in Texas. Austin, TX: Criminal Justice Policy Council.
Human Rights Watch. 1999. Red Onion State Prison: Supermaximum-Security Confinement in
Virginia. New York: Human Rights Watch.

57

S UPERMAX PRISONS

URBAN INSTITUTE

Hunter, Albert. 1985. “Private, Parochial, and Public Social Orders: The Problem of Crime and
Incivility in Urban Communities.” In The Challenge of Social Control, edited by Gerald D.
Suttles and Mayer N. Zald (230–242). Norwood, NJ: Ablex.
Irwin, John. 2005. The Warehouse Prison: Disposal of the New Dangerous Class. Los Angeles,
CA: Roxbury.
Irwin, John, and James Austin. 1997. It’s About Time: America’s Imprisonment Binge, 2nd ed.
Belmont, CA: Wadsworth.
Irwin, John, and Donald R. Cressey. 1962. “Thieves, Convicts and the Inmate Culture.” Social
Problems 10: 142–155.
Jackson, Joe, and William F. Burke, Jr. 1999. Dead Run: The Untold Story of Dennis Stockton
and America’s Only Mass Escape from Death Row. New York: Random House.
Jackson, Michael. 2002. Justice Behind the Walls: Human Rights in Canadian Prisons.
Vancouver: Douglas and McIntyre.
James, Michael. 2001. “Death Row’s Computers Link Up Inmates and Law.” The Baltimore Sun.
March 30, 1A.
———. 2002. “Malvo Could Go to School at Supermax.” The Baltimore Sun. November 1, A8.
http://www.sunspot.net. (Accessed July 14, 2003.)
Johnson, Kevin. 2002. “From Extreme Isolation, Waves of Felons Are Freed.” USA Today.
December 12, A1, A5.
Johnson, Robert. 1987. Hard Time. Belmont, CA: Wadsworth.
Johnson, W. Wesley, Katherine Bennett, and Timothy J. Flanagan. 1997. “Getting Tough on
Prisoners: Results from the National Corrections Executive Survey, 1995.” Crime and
Delinquency 43: 24–41.
Jones, Richard P. 2001a. “Not All at Supermax are ‘Real Bad Actors.’” JSOnline Milwaukee
Journal Sentinel, October 21. http://www.jsonline.com/news/state/oct01/super22102101.asp.
(Accessed July 22, 2003.) (Published in the Milwaukee Journal Sentinel, October 22, 2001,
B1, B3.)
———. 2001b. “Supermax Comes Under Scrutiny.” JSOnline Milwaukee Journal Sentinel,
October 20. http://www.jsonline.com/news/state/oct01/super21102001.asp. (Accessed
February 15, 2004.) (Published in the Milwaukee Journal Sentinel, October 20, 2001, B1.)
Kalman, Gary. 2001. “Avoiding Prison Policy by Sound Bite.” Corrections Today 63(3): 106–
107.
Kantrowitz, Nathan. 1996. Close Control. Albany, NY: Harrow and Heston.

58

S UPERMAX PRISONS

URBAN INSTITUTE

King, Roy. 1998. “Prisons.” In The Handbook of Crime and Punishment, edited by Michael H.
Tonry (589–625). New York: Oxford University Press.
———. 1999. “The Rise and Rise of Supermax: An American Solution in Search of a
Problem?” Punishment and Society 1: 163–186.
Knight, Kevin N. 2000. “The Evolution of Decisionmaking among Prison Executives, 1975–
2000.” In Policies, Processes, and Decisions of the Criminal Justice System, edited by Julie
Horney (177–224). Washington, DC: National Institute of Justice.
Koppel, Ted. 1998a. “Crime and Punishment: Doing Time in Texas.” Part I: Prisoners.
ABCNEWS.com, March 25.
———. 1998b. “Crime and Punishment: Sitting on Top of the Volcano.” Part II: Guards.
ABCNEWS.com, March 26.
———. 1998c. “Crime and Punishment: ‘Blood In, Blood Out.’” Part III: Gangs.
ABCNEWS.com, March 27.
———. 1998d. “Crime and Punishment: Is This Where We Want to Go?” Part IV: Solitary.
ABCNEWS.com, March 30.
Krohn, Marvin. 1995. “Control and Deterrence Theories of Criminality.” In Criminology, edited
by Joseph F. Sheley (329–347). New York: Wadsworth.
Kurki, Leena, and Norval Morris. 2001. “Supermax Prisons.” In Crime and Justice: A Review of
Research, vol. 28, edited by Michael H. Tonry (385–424). Chicago, IL: University of
Chicago Press.
Land, Kenneth, Patricia McCall, and Lawrence Cohen. 1990. “Structural Covariates of Homicide
Rates: Are There Any Invariances across Time and Social Space?” American Journal of
Sociology 95: 922–963.
Lawrence, Sarah, and Daniel P. Mears. 2004. Benefit-Cost Analysis of Supermax Prisons:
Critical Steps and Considerations. Washington, D.C.: The Urban Institute.
Light, Stephen C. 1990. “Measurement Error in Official Statistics: Prison Rule Infraction Data.”
Federal Probation 54: 63–68.
Logan, Charles H. 1993. “Criminal Justice Performance Measures for Prisons.” In Performance
Measures for the Criminal Justice System, edited by John J. DiIulio, Jr. and James Q. Wilson
(15–59). Washington, DC: Bureau of Justice Statistics.
Lovell, David, Kristin Cloyes, David Allen, and Lorna Rhodes. 2000. “Who Lives in SuperMaximum Custody?” Federal Probation 64: 33–38.
Lowenkamp, Christopher T., and Edward J. Latessa. 2005. “Developing Successful Reentry
Programs: Lessons Learned from the ‘What Works’ Research.” Corrections Today 67: 72–

59

S UPERMAX PRISONS

URBAN INSTITUTE

77.
Lynch, Allen K. 1994. The Cost Effectiveness of Incarceration: A Critical Review of the
Literature. Tallahassee, FL: Florida Corrections Commission.
Lynch, James P., and William J. Sabol. 2001. Crime, Coercion, and Community: The Effects of
Arrest and Incarceration Policies on Informal Social Control in Neighborhoods.
Washington, DC: The Urban Institute.
MacKenzie, Doris L. 1997. “Criminal Justice and Crime Prevention.” In Preventing Crime:
What Works, What Doesn’t, What’s Promising, edited by Lawrence W. Sherman, Denise
Gottfredson, Doris L. MacKenzie, John Eck, Peter Reuter, and Shawn Bushway (9.1–9.76).
Washington, DC: Office of Justice Programs.
Mangione, Thomas. 1998. “Mail Surveys.” In Handbook of Applied Social Research Methods,
edited by Leonard Bickman and Debra Rog (399–428). Newbury Park, CA: Sage.
Marsden, Mary Ellen, and Thomas Orsagh. 1983. “Prison Effectiveness Measurement.” In
Evaluating Performance of Criminal Justice Agencies, edited by Gordon P. Whitaker and
Charles D. Phillips (211–235). Beverly Hills, CA: Sage.
Maryland Chamber of Commerce. 2003. “2003 Business Agenda.” Legislative Issues.
Annapolis, MD: Maryland Chamber of Commerce.
http://www.mdchamber.org/leg_issues/2003_agenda.htm. (Accessed July 17, 2003.)
Maryland Department of Public Safety and Correctional Services (MDPSCS). 2002. Division
Correction Directive: Case Management Services. Towson, MD: Maryland Department of
Public Safety and Correctional Services.
———. 2003. Maryland Correctional Adjustment Center. Towson, MD: Maryland Department
of Public Safety and Correctional Services. http://www.dpscs.state.md.us/doc/mcac.shtml.
(Accessed June 24, 2003.)
McClain, Meredith. 1996. “Three Strikes and You’re Out: The Solution to the Repeat Offender
Problem?” Seton Hall Legislative Journal 20: 97.
McCorkle, Richard C., Terance D. Miethe, and Kriss A. Drass. 1995. “The Roots of Prison
Violence: A Test of the Deprivation, Management, and Not-So-Total Institution Models.”
Crime and Delinquency 41: 317–331.
McKinney, George K. 2000. “Statement of George K. McKinney, United States Marshall for the
District of Maryland, Before the Subcommittee on Crime of the Committee on the Judiciary
of the House of Representatives.” Annapolis, MD: Subcommittee on Crime of the Committee
on the Judiciary of the House of Representatives, July 13.
http://www.house.gov/judiciary/mcki0713.htm. (Accessed July 14, 2003.)
McNulty, Jennifer. 2000. “Social Scientist Craig Haney Exposes the Psychologically
Devastating Conditions Inside Our Nation’s ‘Supermax’ Prisons.” Crime and Punishment

60

S UPERMAX PRISONS

URBAN INSTITUTE

Review summer: 8–13. Santa Cruz, CA: University of California.
http://review.ucsc.edu/summer.00/crime_and_punishment.html. (Accessed January 31,
2003.)
Mears, Daniel P. 2005. “A Critical Look at Supermax Prisons.” Corrections Compendium 30: 6–
7, 45–49.
Mears, Daniel P., and Jennifer L. Castro. 2006 (forthcoming). “Wardens’ Views on the Wisdom
of Supermax Prisons.” Crime and Delinquency.
Mears, Daniel P., Gretchen E. Moore, Jeremy Travis, and Laura Winterfield. 2003. Improving
the Link between Research and Drug Treatment in Correctional Settings. Urban Institute,
Justice Policy Center, Washington, DC.
Mears, Daniel P., and Michael D. Reisig. 2006. “The Theory and Practice of Supermax Prisons.”
Punishment and Society 8:33-57.
Mears, Daniel P., and Jamie Watson. 2006 (forthcoming). “Towards a Fair and Balanced
Assessment of Supermax Prisons.” Justice Quarterly.
Merton, Robert K. 1957. Social Theory and Social Structure. New York: Free Press.
Meyer, John W., and Brian Rowan. 1977. “Institutionalized Organizations: Formal Structure as
Myth and Ceremony.” American Journal of Sociology 83: 340–363.
Miller, Holly A., and G. R. Young. 1997. “Prison Segregation: Administrative Detention
Remedy or Mental Health Problem?” Criminal Behavior and Mental Health 7: 85–94.
Miller, Nan D. 1995. “International Protection of the Rights of Prisoners: Is Solitary
Confinement in the United States a Violation of International Standards?” California
Western International Law Journal 26: 139–172.
Mohr, Lawrence B. 1995. Impact Analysis for Program Evaluation, 2nd ed. Thousand Oaks,
CA: Sage.
Monahan, John, and Henry J. Steadman. 1983. “Crime and Mental Disorder: An
Epidemiological Approach.” In Crime and Justice: An Annual Review of Research, vol. 4,
edited by Michael H. Tonry and Norval Morris (145–189). Chicago, IL: University of
Chicago Press.
Moran, Nancy. 1994. Maryland Correctional Adjustment Center: Allegation of Brutality and
Allegation of Discrimination. Baltimore, MD: Prisoners Aid Association of Maryland, Inc.
Morris, Norval, and David J. Rothman, eds. 1995. The Oxford History of the Prison. New York:
Oxford University Press.
Murton, Tom. 1971. “Inmate Self- Government.” University of San Francisco Law Review 6: 87–
102.

61

S UPERMAX PRISONS

URBAN INSTITUTE

National Institute of Corrections (NIC). 1997. Supermax Housing: A Survey of Current Practice.
Washington, DC: National Institute of Corrections.
National Institute of Corrections Information Center. 1987. Direct Supervision Jails: Interviews
with Administrators. Washington, DC: National Institute of Corrections.
Nitkin, David. 2003. “Maryland Wants to Demolish Supermax as Obsolete at 14: State Shifts
from Lockup to Rehabilitation in Prison.” The Baltimore Sun, October 16, A1.
Ohio Department of Rehabilitation and Correction (ODRC). 1998. “Director Officially Opens
Ohio State Penitentiary.” News Release. April 9.
———. 2001. Fiscal Year 2001 Annual Report. Columbus, OH: Ohio Department of
Rehabilitation and Correction.
———. 2002. Population Projections, December 2002. Columbus, OH: Ohio Department of
Rehabilitation and Correction, Bureau of Research.
———. 2003. Ohio State Penitentiary. Columbus, OH: Ohio Department of Rehabilitation and
Correction. http://www.drc.state.oh.us/Public/osp.htm. (Accessed March 1, 2004.)
Ohio State Penitentiary (OSP). 2003. Operations Manual. Youngstown, OH: Ohio Department
of Rehabilitation and Correction, Ohio State Penitentiary.
Patrick, Deval L. 1996. Notice of Findings of Investigation: Maryland Correctional Adjustment
Center. May 1. Washington, DC: U.S. Department of Justice, Civil Rights Division.
Pfeiffer, Mary Beth. 2004. “A Death in the Box.” The New York Times Magazine, October 31,
48–53.
Pietras, Jamie. 2001. “Cruel and Unusual: The ACLU Takes Youngstown’s Supermax Prison to
Court.” Columbus Alive News. http://www.columbusalive.com/2001/20010118/news1.html
(Accessed March 1, 2004).
Pizarro, Jesenia, and Vanja M. K. Stenius. 2004. “Supermax Prisons: Their Rise, Current
Practices, and Effect on Inmates.” The Prison Journal 84: 248–264.
Poole, Eric D., Robert M. Regoli, and Charles W. Thomas. 1980. “The Measurement of Inmate
Social Role Types: An Assessment.” The Journal of Criminal Law and Criminology 71:
317–324.
Preer, Robert. 2004. “The Solitary Men.” The Boston Globe Magazine, April 4, 30–37.
Rajkumar, Andrew S., and Michael T. Frenc h. 1997. “Drug Abuse, Crime Costs, and the
Economic Benefits of Treatment.” Journal of Quantitative Criminology 13: 291–323.
Ralph, Paige H., and James W. Marquart. 1991. “Gang Violence in Texas Prisons.” The Prison
Journal 71: 38–49.

62

S UPERMAX PRISONS

URBAN INSTITUTE

Rebman, Christine. 1999. “Comment: The Eighth Amendment and Solitary Confinement: The
Gap in Protection from Psychological Consequences.” DePaul Law Review 49: 567–619.
Reisig, Michael D. 1998. “Rates of Disorder in Higher-Custody State Prisons: A Comparative
Analysis of Managerial Practices.” Crime and Delinquency 44: 229–244.
———. 2001. “The Champion, Contender, and Challenger: Top-Ranked Books in Prison
Studies.” The Prison Journal 81: 389–407.
———. 2002. “Administrative Control and Inmate Homicide.” Homicide Studies 6: 84–103.
Rhodes, Lorna A. 2004. Total Confinement: Madness and Reason in the Maximum Security
Prison. Los Angeles, CA: University of California Press.
Riveland, Chase. 1999a. “Prison Management Trends, 1975–2025.” In Crime and Justice: A
Review of Research, edited by Michael H. Tonry and Joan Petersilia (163–203). Chicago:
University of Chicago Press.
———. 1999b. Supermax Prisons: Overview and General Considerations. Washington, DC:
National Institute of Corrections.
Rocheleau, Ann M., Michael Forcier, and Jennifer Jackson. 1998. Evaluation of the
Massachusetts Department of Correction Department Disciplinary Unit. Cambridge, MA:
BOTEC.
Roman, John, and Adele Harrell. 2001. “Assessing the Costs and Benefits Accruing to the Public
from a Graduated Sanctions Program for Drug-Using Defendants.” Law and Policy 23: 237–
268.
Roman, John, Adele Harrell, Jennifer Woodard, and Stacia Riggs. 1998. A Methodology for
Measuring Costs and Benefits of Court-Based Drug Intervention Programs Using Findings
from Experimental and Quasi-Experimental Evaluations. Washington, DC: The Urban
Institute.
Rossi, Peter H., Howard E. Freeman, and Mark W. Lipsey. 1999. Evaluation: A Systematic
Approach, 6th ed. Thousand Oaks, CA: Sage.
Rothman, David J. 1995. “Perfecting the Prison: United States, 1789–1865.” In The Oxford
History of the Prison, edited by Norval Morris and David J. Rothman (111–129). New York:
Oxford University Press.
Russo, Maria. 2001. “Psycho Factories.” Salon.com. March 29.
http://archive.salon.com/books/review/2001/03/29/ha llinan/index2.html. (Accessed
November 22, 2002.)
Sampson, Robert J., and Stephen Raudenbush. 1999. “Systematic Social Observation of Public
Spaces: A New Look at Disorder in Urban Neighborhoods.” American Journal of Sociology
105: 603–651.

63

S UPERMAX PRISONS

URBAN INSTITUTE

Scheirer, Mary Ann. 1994. “Designing and Using Process Evaluation.” In Handbook of Practical
Program Evaluation, edited by Joseph S. Wholey, Harry P. Hatry, and Kathryn E.
Newcomer (40–68). San Francisco: Jossey-Bass.
Shelden, Randall G., Sharon K. Tracy, and William B. Brown. 1997. Youth Gangs in American
Society. New York: Wadsworth.
Sheppard, Robert A., Jeffrey A. Geiger, and George Welborn. 1996. “Closed Maximum
Security: The Illinois SuperMax.” Corrections Today 58: 84–87.
Simon, Jonathan. 2000. “The ‘Society of Captives’ in the Era of Hyper- incarceration.”
Theoretical Criminology 4: 285–308.
Singer, Paul. 2003. “Inmates Volunteer to Stay at ‘Supermax’ Prison.” Associated Press, March
3. http://www.newsnet5.com. (Accessed March 14, 2003.)
Skogan, Wesley. 1990. Disorder and Decline: Crime and the Spiral of Decay in American
Cities. Los Angeles, CA: University of California Press.
Smith, Jordan. 2002. “Landmark Prison Oversight Case Ends.” The Austin Chronicle 21: 42.
June 21. http://www.austinchronicle.com/issues/dispatch/2002-06-21/pols_feature6.html.
(Accessed February 3, 2003.)
Sparks, Richard, Anthony E. Bottoms, and Will Hay. 1996. Prisons and the Problem of Order.
Oxford, UK: Oxford University Press.
Steinke, Pamela. 1991. “Using Situational Factors to Predict Types of Prison Violence.” Journal
of Offender Rehabilitation 17: 119–132.
Stickrath, Thomas J., and Gregory A. Bucholtz. 2003. “Supermax Prisons: Why?” In Supermax
Prisons: Beyond the Rock, edited by Donice Neal (1–14). Lanham, MD: American
Correctional Association.
Suedfeld, Peter, Carmen Ramirez, John Deaton, and Gloria Baker-Brown. 1982. “Reactions and
Attributes of Prisoners in Solitary Confinement.” Criminal Justice and Behavior 9: 303–340.
Sykes, Gresham M. 1958. The Society of Captives. Princeton, NJ: Princeton University Press.
Texas Department of Criminal Justice (TDCJ). 1999. Texas Department of Criminal Justice:
Institutional Division Overview. Austin, TX: Texas Department of Criminal Justice.
——-. 2001. Modern Corrections in the 21st Century: Annual Review 2001. Austin, TX: Texas
Department of Criminal Justice.
——. 2002a. “The Final Chapters of the Ruiz Case.” Criminal Justice Connections Newsletter.
September/October.
http://www.tdcj.state.tx.us/mediasvc/connections/SeptOct2002/agency2_v10no1.html.
(Accessed January 31, 2003.)

64

S UPERMAX PRISONS

URBAN INSTITUTE

———. 2002b. Texas Department of Criminal Justice: Administrative Segregation Plan. Austin,
TX: Texas Department of Criminal Justice.
Timberg, Craig. 2001. “Documents Show Use of Force at Wallens Ridge.” The Washington Post,
July 31, B1.
Toch, Hans. 1992. Mosaic of Despair: Human Breakdowns in Prison. Washington, DC:
American Psychological Association.
——. 1994. “Democratizing Prisons.” The Prison Journal 73: 62–72.
——. 2001. “The Future of Supermax Confinement.” The Prison Journal 81: 376–388.
——. 2003. “The Contemporary Relevance of Early Experiments with Supermax Reform.” The
Prison Journal 83: 221–228.
Townsend, Kathleen Kennedy. 1999. Minutes: Meeting of the Maryland State Commission on
Criminal Sentencing Policy. Annapolis, MD: Maryland State Commission on Criminal
Sentencing Policy, October 8. http://www.msccsp.org/commission/minutes/10_99.html.
(Accessed June 4, 2003.)
United Nations (no author listed). “Committee Against Torture Concludes Twenty- fourth
Session.” Press Release HR/4472. New York: United Nations.
http://www.un.org/News/Press/docs/2000/20000522.hr4472.doc.html. (Accessed March 31,
2003.)
U.S. Bureau of Justice Statistics. 2003. Correctional Populations in the United States, 1997, and
Prisoners in 2002. Washington, DC: Bureau of Justice Statistics.
http://www.ojp.usdoj.gov/bjs/glance/tables/corrtyptab.htm. (Accessed September 9, 2003.)
U.S. News and World Report. 1995. “A New Home for the Incorrigibles.” U.S. News and World
Report 118(2): 15.
Useem, Bert, Camille G. Camp, and George M. Camp. 1996. Resolution of Prison Riots. New
York: Oxford University Press.
Useem, Bert, and Jack A. Goldstone. 2002. “Forging Social Order and Its Breakdown: Riot and
Reform in U.S. Prisons.” American Sociological Review 67: 499–525.
Useem, Bert, and Peter Kimball. 1991. States of Siege. New York: Oxford University Press.
Useem, Bert, Raymond V. Liedka, and Anne M. Piehl. 2003. “Popular Support for the Prison
Build-Up.” Punishment and Society 5: 5–32.
Useem, Bert, and Michael D. Reisig. 1999. “Collective Action in Prisons: Protests, Disturbances,
and Riots.” Criminology 37: 735–759.
Walters, Bruce A., and Richard L. Priem. 1999. “A CEO-Issues Model of Firm Performance for

65

S UPERMAX PRISONS

URBAN INSTITUTE

Single-Business Manufacturers.” Academy of Strategic and Organizational Leadership
Journal 3: 1–13.
Walters, Steven. 2002. “Settlement Calls for Changes to Supermax.” JSOnline Milwaukee
Journal Sentinel, January 3. http://www.jsonline.com/news/state/jan02/9928.asp. (Accessed
March 15, 2004.)
Ward, David A. 1995. “A Corrections Dilemma: How to Evaluate Super-max Regimes.”
Corrections Today 57: 104–108.
Ward, David A, and Thomas G. Werlich. 2003. “Alcatraz and Marion: Evaluating SuperMaximum Custody.” Punishment and Society 5: 53–75.
Warr, Mark. 1996. “Organization and Instigation in Delinquent Groups.” Criminology 34: 11–
37.
Webb, Lonny R. 2001. Addressing Severe Behavior Problems in a “Super-Max” Prison Setting.
Salem, OR: Oregon State Penitentiary.
Welch, Michael. 1999. Punishment in America: Social Control and the Ironies of Imprisonment.
Thousand Oaks, CA: Sage.
Wells, Terry L., W. Wesley Johnson, and Rodney J. Henningsen. 2002. “Attitudes of Prison
Wardens toward Administrative Segregation and Supermax Prisons.” In Correctional
Perspectives, edited by Leanne F. Alarid and Paul F. Cromwell (171–180). Los Angeles:
Roxbury.
Welsh, Brandon C., and David P. Farrington. 2000. “Monetary Costs and Benefits of Crime
Prevention Programs.” In Crime and Justice: Review of Research, vol. 27, edited by Michael
H. Tonry (305–361). Chicago, IL: University of Chicago Press.
Wilkinson, Reginald A. 1997. “The Future of Adult Corrections.” Testimony provided to the
National Council on Crime and Delinquency’s Conference, Reducing Crime in America: The
Agenda for the 21st Century, Dec. 5. http://www.drc.state.oh.us/web/Articles/article38.htm.
(Accessed April 2, 2003.)
Will, George. 1997. “‘Supermaxes’ Make Prisons Safer.” The Post Standard, January 14, AZ6.
Wisconsin Department of Corrections. 2003. “Wisconsin Secure Program Facility.”
http://www.wi-doc.com/supermax.htm. (Accessed January 6, 2003.)
———. 2001. “Corrections Implements Supermax Changes.” News Releases, November 15.
http://www.wi-doc.com/corrections_implements_supermax_.htm. (Accessed Nove mber 13,
2002.)
Wolfgang, Marvin E., and Franco Ferracuti. 1967. The Subculture of Violence. London:
Tavistock.

66

S UPERMAX PRISONS

URBAN INSTITUTE

Wooldredge, John, T. Griffin, and Travis Pratt. 2001. “Considering Hierarchical Models for
Research on Inmate Behavior: Predicting Misconduct with Multilevel Data.” Justice
Quarterly 18: 203–231.
Zahn, Mary. 2003. “Supermax Death Spurs Probe of Inmate Care.” JSOnline Milwaukee Journal
Sentinel. May 10. http://www.jsonline.com/news/metro/may03/140055.asp. (Accessed May
16, 2003.)
Zimring, Franklin E., and Gordon Hawkins. 1995. Incapacitation: Penal Confinement and the
Restraint of Crime. New York: Oxford University Press.

67

S UPERMAX PRISONS

URBAN INSTITUTE

Appendix: Tables

68

S UPERMAX PRISONS

Table 1.

URBAN INSTITUTE

States with Supermax Facilities, 1997–1998
Supermax Beds

Sentenced
Prison Pop.

Incarceration Rate
per 100,000

Percent of
Total Beds

Northeast
Connecticut
Maine
Massachusetts
New Jersey
New York
Pennsylvania
Rhode Island

3,214
586
100
124
96
2,000
200
108

163,836
13,005
1,542
10,847
28,361
70,026
34,963
2,100

397
123
278
351
386
291
213

2.0
4.5
6.5
1.1
0.3
2.9
0.6
5.1

Midwest
Illinois
Indiana
Michigan
Minnesota
Nebraska
Ohio
Wisconsin

2,290
500
85
421
120
164
500
500

216,391
40,788
17,730
44,771
5,306
3,329
48,002
14,682

342
301
457
113
200
429
283

1.1
1.2
0.5
0.9
2.3
4.9
1.0
3.4

South
Florida
Georgia
Louisiana
Maryland
Mississippi
North Carolina
Oklahoma
South Carolina
Texas
Virginia
West Virginia

7,584
1,000
10
1,048
286
1,756
300
392
200
1,229
1,267
96

480,061
64,540
35,722
29,265
21,088
14,548
27,726
20,542
20,264
140,729
27,524
3,160

437
472
672
413
531
370
617
536
717
407
174

1.6
1.5
0.3
3.6
1.4
12.0
1.1
1.9
1.0
0.9
4.6
3.0

West
Arizona
California
Colorado
Idaho
Montana
Nevada
Oregon
Washington
Wyoming

6,542
1,728
2,942
750
96
64
430
196
300
36

242,315
22,353
154,368
13,461
3,946
2,242
8,884
7,589
13,198
1,566

484
475
342
323
25
518
232
233
326

2.7
7.7
1.9
5.6
2.4
2.9
4.8
2.6
2.3
2.3

19,630

1,102,603

—

1.8

All states

Sources: King 1999, updating figures from Riveland 1999b.

S UPERMAX PRISONS

Table 2.

URBAN INSTITUTE

Types of Supermax Housing
New Construction as
Supermax Housing

Separate
facility

Arizona
California
Colorado
Connecticut
Florida
Georgia
Idaho
Illinois
Indiana
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Montana
Nebraska
Nevada
New Jersey
North Carolina
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
Texa s
Virginia
Washington
Wisconsin
Wyoming

Unit in a
new facility

Retrofitted Construction as
Supermax Housing

Unit in an
existing
facility

Separate
facility

v
v
v
v
v*

Unit in an
existing
facility

v

v
v
v
v
v

v
v
v

v*

v*

v
v*

v

v

v
v

v
v*

v

v

v
v*

v
v
v*

v
v
v
v
v

v*
v

v

v

v*
v*

v

Source: Adapted from table 2 in NIC (1997:7).
Note: Asterisks indicate facilities that the NIC (1997) survey identified as being under Department of Corrections
consideration at the time of the survey or had not yet received funding approval. The table is meant purely to convey
the variety of housing approaches that can constitute a “supermax” facility. Some states that reportedly have
supermax prisons were not included in the NIC study because of the definition NIC used (King 1999).

S UPERMAX PRISONS
Figure 1.

URBAN INSTITUTE

Conceptual Framework for Examining the Goals and Impacts of Supermax Prisons

Different Populations/
Stakeholders

Supermax Prison

Intended and Unintended Impacts
(Positive/Negative)

Supermax Prison
• Inmates
• Guards
• Wardens

Supermax Prison
• Inmates — ?
• Guards — ?
• Wardens — ?

General Inmate Prisons
• Inmates
• Guards
• Wardens

General Inmate Prisons
• Inmates — ?
• Guards — ?
• Wardens — ?

Correctional System
• Administrators
• Treatment providers
• Parole officers

Correctional System
• Administrators — ?
• Treatment providers — ?
• Parole officers — ?

Communities
• Local government
• Businesses
• Residents

Communities
• Local governments — ?
• Businesses — ?
• Residents — ?

State
• State government
• Businesses
• Residents

State
• State governments — ?
• Businesses — ?
• Residents — ?

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3. Summary of the Matrix of Goals and Impacts
To document and organize the findings from our literature review and interviews, we developed a Microsoft Excel file with a matrix that categorized goals and
impacts according to (1) key stakeholder units of analysis (supermax prisons, general population prisons, criminal justice system, local communities, and states
and country), and subpopulations specific to each (see the matrix outline below) and (2) whether each impact was intended or unintended (positive or negative).
A goal or impact was included if at least one source cited it.
The tables presented in this appendix summarize the main findings from the matrix. The tables are organized into the five general categories of stakeholder
populations and the subpopulations for each. Each of the 16 subpopulation tables lists the goals and impacts of supermax prisons associated with the specific
stakeholder subpopulation. In the first line of table 3.4, for example, the table shows that a goal of supermax prisons is to increase the safety of general
population prisoners, an unintended positive impact may be improved living conditions for general population prisoners, and an unintended negative impact may
be decreased safety for general population prisoners.
Not included in these tables but included in the matrix is (a) a more in-depth discussion of each goal and impact; (b) a discussion of how each goal and impact is
or may be causally associated with supermax prisons; and (c) citations to sources, including interviews, that mentioned various goals, impacts, or causal logics.
Matrix Outline
Key stakeholders and subpopulations
Supermax Prisons
Supermax prisoners
Supermax staff
Supermax wardens
General Population Prisons
General populat ion prisoners
General population staff
General population wardens
Criminal Justice System
Executive admin. of departments of corrections
Parole officers
Local Communities
Local government
Local politicians
Local businesses
Local residents
States and Country
National government
State politicians
Nonlocal businesses
Nonlocal residents

Goals and intended impacts

Unintended positive impacts

Unintended negative impacts

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3.1. Supermax Prisoners
•

•

Goals and intended impacts
modifies the behavior of supermax (SM)
prisoners during and after incarceration
o greater reintegration into general population
(GP) and/or the community upon release
o greater rule compliance upon release to GP
prisons
o reduced levels of violent activity in GP
and/or in the community
o low rate at which SM prisoners return to SM
prison
punishes violent and disruptive prisoners
o increases restrictions of freedom and
privileges (e.g., less out-of-cell time and
fewer visits)
o increases perception that they receive more
severe punishment

•
•
•
•
•
•
•
•

Unintended positive impacts
improves quality of life for some
supermax prisoners
improves care for mentally ill
prisoners
reduces fear and stress
increases safety
reduces number or seriousness of
disruptions or outbursts
reduces escapes
increases rule compliance (e.g.,
number of infractions)
reduces intimidation by gang
members

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Unintended negative impacts
increases disciplinary infractions of
supermax prisoners
increases tensions with staff
decreases rehabilitation
increases violent and aggressive tendencies
causes or exacerbates psychological
problems
increases abuse by staff
violates right to due process
constitutes cruel and unusual punishment
violates human rights
disproportionate punishment
increases risks to physical health
hinders relationship with family members
exacerbates reintegration challenges
increases recidivism
decreases safety

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3.2. Supermax Staff
•

•

•

Goals and intended impacts
increases safety of supermax staff
o reduces murders
o reduces assaults
o reduces riots
o increases the sense of safety
increases order and control of prisoners
o increases rule compliance (e.g.,
number of infractions)
o reduces number or seriousness of
disruptions or outbursts
o reduces use-of-force incidents
o reduces escapes
reduces gang influence
o reduces the number or percentage of
inmates who are gang members
o reduces the frequency and amount of
drug trafficking by gang members
o reduces the intimidation of staff and
prisoners by gang members
o reduces gang-initiated murders and
assaults

•
•
•
•

Unintended positive impacts
improves work conditions of supermax
staff
increases promotion rate
reduces stress and fear
increases job satisfaction

•
•
•
•
•
•
•
•

Unintended negative impacts
increases the number and/or seriousness of
disciplinary infractions committed by
supermax staff
decreases safety
increases likelihood of a riot
heightens stress
increases tensions with staff
increases turnover
creates unpleasant work environment
decreases job satisfaction

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3.3. Supermax Wardens
Goals and intended Impacts
(none identified)

•

Unintended positive impacts
increases supermax warden’s public image
and political support

•
•
•

Unintended negative impacts
increases supermax warden’s stress
increases risk of public image and political
support
reduces time devoted to managing the
prison

S UPERMAX PRISONS

URBAN INSTITUTE

General Population Prisons
Table 3.4. General Population Prisoners
•

•

•

Goals and intended Impacts
increases safety of general population
prisoners
o reduces murders
o reduces assaults
o reduces riots
o increases the sense of safety
increases order and control of prisoners
o increases rule compliance (e.g.,
number of infractions)
o reduces number or seriousness of
disruptions or outbursts
o reduces lockdowns
o reduces use-of-force incidents
o reduces warning shots fired by staff
o reduces escapes
reduces gang influence
o reduces the number or percentage of
inmates who are gang members
o reduces the frequency and amount of
drug trafficking by gang members
o reduces the intimidation of staff and
prisoners by gang members
o reduces gang-initiated murders and
assaults

•
•
•
•
•
•
•

Unintended positive impacts
improves living conditions of general
population prisoners
increases freedoms and privileges
improves access to programs and services
improves outcomes
normalizes the prison environment
decreases stress and fear
improves prisoner-staff relations

•
•
•
•
•

Unintended negative impacts
decreases safety of general population
prisoners
increases the number and/or seriousness of
the disruptions to manage
reduces resources
increases chance of a riot
creates a less pleasant living environment

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3.5. General Population Staff
•

•

•

Goals and intended impacts
increases safety of general population staff
o reduces murders
o reduces assaults
o reduces riots
o increases the sense of safety
increases order and control of prisoners
o increases rule compliance (e.g.,
number of infractions)
o reduces number or seriousness of
disruptions or outbursts
o reduces lockdowns
o reduces use-of-force incidents
o reduces warning shots fired by staff
o reduces escapes
reduces gang influence
o reduces the number or percentage of
inmates who are gang members
o reduces the frequency and amount of
drug trafficking by gang members
o reduces the intimidation of staff and
prisoners by gang members
o reduces gang-initiated murders and
assaults

•
•
•
•

Unintended positive impacts
improves work conditions of general
population staff
improves prisoner-staff relations
reduces stress and fear
increases job satisfaction

•
•
•
•
•
•

Unintended negative impacts
decreases safety of general population
staff
increases the number and/or seriousness of
the disruptions to manage
increases challenges in managing and
working with prisoners
reduces resources
increases chance of a riot
increases staff turnover

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3.6. General Population Wardens
Goals and intended impacts
(none identified)

•
•

Unintended positive impacts
increases public image and political
support of general population wardens
reduces management challenges

•

Unintended negative impacts
reduces resources of general population
wardens

S UPERMAX PRISONS

URBAN INSTITUTE

Criminal Justice System
Table 3.7. Executive Administration of Department of Correction
•

Goals and intended impacts
improves the operational efficiencies of
prison system
o simplifies travel logistics
o simplifies staff training
o reduces costs
o increases ease of developing and
implementing rules and policies
helps SM staff develop specialized
skills specifically designed for
managing a certain type of prisoner
improves management of the prison
system’s most problematic prisoners
as well as GP prisoners
reduces staff members required
increases number of days under
normal conditions (not lockdown)
reduces restrictions on GP prisoners,
which improves their access to
available programs and services
o reduces wait-time for placing inmates
in segregation
o reduces staff time devoted to
transporting prisoners from one
facility to another

•
•

Unintended positive impacts
increases public image and political
support for prison system
reduces costs

•
•

Unintended negative impacts
increases risk to public image and political
support for prison system
increases expenses

S UPERMAX PRISONS

URBAN INSTITUTE

Table 3.8. Parole Officers
Goals and intended impacts
(none identified)

•
•

Unintended positive impacts
increases safety of parole officers
decreases management challenges

•
•

Unintended negative impacts
decreases safety of parole officers
increases management challenges

S UPERMAX PRISONS

URBAN INSTITUTE

Local Communities
Table 3.9. Local Government
Goals and intended impacts
(none identified)

•
•
•

Unintended positive impacts
increases tax revenue of local government
reduces costs
assists prosecutors

Unintended negative impacts
(none identified)

•

Unintended positive impacts
increases support of local politicians

•

•

Unintended positive impacts
increases revenue of local businesses

Unintended negative impacts
(none identif ied)

Table 3.10. Local Politicians
Goals and intended impacts
(none identified)

Unintended negative impacts
decreases support of local politicians

Table 3.11. Local Businesses
Goals and intended impacts
(none identified)
Table 3.12. Local Residents
•

Goals and intended impacts
increases public safety
•
o reduces escape attempts and number of •
successful escape attempts
o lowers crime rates of potential
offenders in society
o lowers recidivism among ex-prisoners
o reduces the public’s fear of crime

Unintended positive impacts
increases employment of local residents
improves the local economy

•

Unintended negative impacts
decreases public safety

S UPERMAX PRISONS

U R B A N I NS T I T U T E

States and Country
Table 3.13. National Government
Goals and intended impacts
(none identified)

•

Unintended positive impacts
reduces costs of national government

•

Unintended negative impacts
strains relations between national
government and other nations

Table 3.14. State Government
Goals and intended impacts
(none identified)

•
•

Unintended positive impacts
reduces costs of state government
increases tax revenue

•

Unintended positive impacts
increases revenue of nonlocal businesses

•
•

Unintended negative impacts
increases expenses of state government
increases risk to political support

•

Unintended negative impacts
increases taxes of nonlocal businesses

•

Unintended negative impacts
decreases public safety

Table 3.15. Nonlocal Businesses
Goals and intended impacts
(none identified)
Table 3.16. Nonlocal Residents
•

Goals and intended impacts
Unintended positive impacts
(none identified)
increases public safety
o reduces escape attempts and number of
successful escape attempts
o lowers crime rates of potential
offenders in society
o lowers recidivism among ex-prisoners
o reduces the public’s fear of crime

S UPERMAX PRISONS

Table 4. Wardens’ Views of Types of Inmates Who Should Be Placed in Supermaxes

Agree or strongly agree (%)

Inmates who should be placed in a supermax are those who . . .
assault staff repeatedly or cause injury
assault other inmates repeatedly or cause injury
instigate other inmates to be violent
are prison gang leaders
are an escape risk
are drug dealers while in prison
are chronic rule violators
are prison gang members
are “high profile”
are at risk of being attacked
have been incarcerated for a serious offense
have a serious mental illness

99.5
99.3
89.3
82.5
79.2
55.6
51.0
46.9
30.9
23.6
18.5
10.1

Other inmates who should be placed in supermax
are sexual predators
are terrorists
are on death row
kill others while in prison
make or possess weapons
Inmates who are placed in supermax
are all of the above (except terrorists)
are a constant threat to staff, the public, and other inmates
are disorderly and disruptive “discipline problems”
fail to adjust to prison life
require control, isolation, and separation
manipulate restraints
receive a felony charge while in prison
incite, lead, or participate in riots
engage in sexual activity with known diagnosis of hiv or hepatitis c
are repeat offenders
are young adult offenders
have lengthy or multiple life sentences
have major medical problems
refuse to live elsewhere

Source: Mears (2005).
Notes: Ns range from 577 to 600. Four percent of respondents also identified other types of inmates,
listed in the second section. The N for the last section, which focuses on responses to a question asked
only of supermax wardens, was 213.

S UPERMAX PRISONS

URBAN INSTITUTE

Figure 2. Wardens’ Views of the Specific Impacts of Supermaxes

Staff safety
Level of order in institution
Inmate safety
Inmate violent acts
Inmate fear of victimization
Staff use- of- force incidents
Staff fear of victimization
Local employment
Specific Impacts of
Supermax Prisons

Inmate infractions
Local government tax revenues

Supermax = beneficial

Support for local politicians

Supermax = harmful
Supermax = no effect

Inmate perception of prison legitimacy
Local business development
Community residents' fear of crime
Staff turnover
Inmate access to programs
Staff disciplinary actions
Inmate complaints against staff
Inmate mental health
Inmate recidivism after release to society
0

10

20

30

40

50

60

Percent

Source: Mears (2005).
Note: Ns range from 580 to 592.

70

80

90

100