United States General Accounting Office

GAO

Report to the Subcommittee on Crime,
Committee on the Judiciary, House of
Representatives

August 1996

PRIVATE AND PUBLIC
PRISONS
Studies Comparing
Operational Costs
and/or Quality of
Service

G

A

O

years
1921 - 1996

GAO/GGD-96-158

GAO

United States
General Accounting Office
Washington, D.C. 20548
General Government Division

B-261797
August 16, 1996
The Honorable Bill McCollum
Chairman
The Honorable Charles E. Schumer
Ranking Minority Member
Subcommittee on Crime
Committee on the Judiciary
House of Representatives
Since 1991, when we issued our earlier report on private prisons,1 various
jurisdictions and levels of government have made or planned greater use
of this alternative to publicly run correctional facilities. At the federal
level, plans for privatization of correctional facilities have changed in
recent years. For example, the administration’s fiscal year 1996 budget
proposal reflected a commitment to increase the use of privatized
correctional facilities in the federal Bureau of Prisons (BOP). Specifically,
under the BOP privatization initiative presented in the budget request, the
bureau proposed to contract with private firms, where most appropriate,
to operate the majority of all future federal pretrial detention facilities as
well as the majority of all future federal minimum- and low-security
correctional facilities. The Department of Justice’s fiscal year 1997 budget
justification projected the activation of two privatized facilities in that
year. However, in June 1996, the Justice Department reversed its plans for
using private contractors to operate the facilities identified in the 1996 and
1997 budget proposals. In explaining its decision to staff these facilities
with BOP employees, the Justice Department noted that it was “unable
contractually to reduce the risk of a strike or walk out” of correctional
officers employed by private firms.2
Proponents of privatization assert that the experiences of several states
demonstrate that private contractors can operate prisons at less cost than
the government, without reducing the levels or quality of service. In
contrast, other observers say there is little or no valid evidence that
privatization of corrections is a cost-effective alternative to publicly run
facilities.
To help frame the continuing deliberations of the Justice Department’s
privatization plans, we self-initiated this review to (1) identify studies
1

Private Prisons: Cost Savings and BOP’s Statutory Authority Need to Be Resolved (GAO/GGD-91-21,
Feb. 7, 1991).

2

Letter, dated June 5, 1996, from the Assistant Attorney General for Administration, Department of
Justice, to various congressional committee and subcommittee chairmen and ranking minority
members.

Page 1

GAO/GGD-96-158 Private and Public Prisons

B-261797

(completed since 1991) comparing the operational costs and/or the quality
of service3 of private and public correctional facilities; (2) determine,
based on these studies, what could be concluded regarding the operational
costs and/or quality of service of comparable private and public facilities;
(3) assess whether the reported results are generalizable to correctional
systems in other jurisdictions; and (4) identify lessons learned to help
guide future comparative studies of private and public correctional
facilities. Our work basically was a form of evaluation synthesis whereby
we assessed existing studies, particularly with respect to the strength of
evidence supporting the reported findings. Appendix I presents further
details about our objectives, scope, and methodology.4 We are addressing
this report to you because of your Subcommittee’s interest in prison
issues, as exemplified by the hearings you held on June 8, 1995.
We conducted our work from June 1995 to June 1996 in accordance with
generally accepted government auditing standards. We received oral
comments on a draft of this report from BOP and written comments from
the Department of Justice’s Office of Justice Programs; the National
Council on Crime and Delinquency; and a Northeastern University
(Boston, MA) professor of criminal justice, who is a nationally recognized
authority on corrections administration. These comments and our
evaluation of them are discussed at the end of this letter.

Results in Brief

On the basis of literature searches and discussions with correctional
officials and criminal justice researchers, we identified five studies
completed since 1991 that compare private and public correctional
facilities in relation to operational costs and/or quality of service. Three
states sponsored comparative studies of correctional facilities in their
states—Texas, California, and Tennessee. The National Institute of Justice,
the Bureau of Prisons, and the National Institute of Corrections funded a
comparative study that focused mainly on facilities in New Mexico.
Washington state studied facilities in Tennessee, Louisiana, and
Washington state. The correctional facilities that were the focus of these
studies varied in terms of geographic location and the types of inmates

3

As discussed in appendix I, we did not attempt to generically define “quality of service.” Rather, we
accepted the definition and/or evaluation criteria used in each applicable study that we reviewed.
Some studies, for instance, used a variety of quality measures or outcomes, such as safety and incident
data and the extent of treatment programs for inmates.
4

Although not addressed in this report, the privatization of corrections has, at times, raised various
other issues besides cost and quality of service issues. For example, there has been some debate over
the issue of whether administration of justice (which includes the operation of prisons) is an
inherently governmental function not appropriately delegable to the private sector.
Page 2

GAO/GGD-96-158 Private and Public Prisons

B-261797

housed. There was also variation in the methodologies and measurements
employed by the researchers.
Three of the studies we reviewed (California, Tennessee, and Washington)
made comparisons of costs between reasonably matched private and
public facilities that were operating within each state that was studied. Of
the four private/public comparisons reported in these three studies, two
showed no significant differences in operational costs, one showed a
7-percent difference in favor of the private facility, and the other reported
the private facility to be more costly than one public facility but less costly
than another public facility. One additional study (Texas) reported a 14- to
15-percent savings from privatization; however, the analysis for the Texas
study was problematic because the comparison was based on hypothetical
public facilities, not existing ones. We could not conclude from these
studies that privatization of correctional facilities will not save money.
However, these studies do not offer substantial evidence that savings have
occurred.
Two studies (New Mexico and Tennessee) assessed a wide variety of
factors in their reviews of comparative quality of private and public
facilities. These factors, among others, included measures of safety,
personnel qualifications, physical conditions of the facilities, health care,
and inmate activities. One of these two studies (Tennessee) reported no
difference between private and public facilities. The other study (New
Mexico) reported a higher quality score for one private facility compared
with two public facilities. However, on the inmate survey portion of the
assessment, one public facility had higher scores in all of the areas that
were assessed, except one. One additional study (Washington), using a
less detailed approach to assessing comparative quality, found no
differences between private and public facilities.
These studies offer little generalizable guidance for other jurisdictions
about what to expect regarding comparative operational costs and quality
of service if they were to move toward privatizing correctional facilities.
First, several of the studies focused on specialized inmate populations,
such as those in prerelease situations, that limited their generalizability to
a wider inmate population. Second, methodological weaknesses in some
of the comparisons—such as using hypothetical facilities or nonrandom
survey samples—make some findings questionable, even for the study
setting. Third, a variety of differences in other states and regions could
result in experiences far different from those of the states that were
studied. For example, cost of living and a state’s correctional philosophy

Page 3

GAO/GGD-96-158 Private and Public Prisons

B-261797

could affect the comparative costs and quality of private and public
facilities from state to state. Finally, the age or maturity of the private
system could affect the relationship between private and public facilities
in terms of costs and quality.
From a methodological standpoint, the five studies provide some lessons
learned for future comparisons of private and public correctional facilities.
The importance of focusing on both operational costs and quality of
service, comparing institutions (both private and public) that are actually
in operation, and using multiple measures of quality are attributes that
worked well in selected studies.

Background

A generally accepted evaluation criterion is that any comparative study of
private and public prisons should be based upon the selection and analysis
of similar facilities. For example, the private and public prisons selected
for comparison should be as similar as possible regarding design and
capacity, security level, and types of inmates. Otherwise, any comparative
analysis of operational costs or quality of service could be skewed. On a
per-inmate basis, for instance, higher security prisons can be expected to
have higher operating costs than lower security prisons because the
former type of facilities generally have higher staff-to-inmate ratios.
Even if similar private and public prisons are available for study, a
comparison of operational costs can still present difficulties in ensuring
that all costs, direct and indirect, are consistently and fully quantified.
Possible difficulties can arise due, in part, to differences in budgeting and
accounting practices between and even within the private and public
sectors. Determining the appropriate allocation of corporate headquarters
overhead and government agency overhead, for instance, can be
particularly difficult.
Comparing the quality of service at private and public prisons also
presents challenges and, in fact, can be more difficult than comparing
costs. The concept of “quality” is neither easily defined nor measured. For
example, although the American Correctional Association (ACA) sets
accreditation standards for prisons, accredited facilities can vary widely in
terms of overall quality. According to ACA officials, such variances occur
because ACA accreditation means that a facility has met minimum
standards.

Page 4

GAO/GGD-96-158 Private and Public Prisons

B-261797

Generally, however, assessments of quality can take several approaches.
For example, one is a compliance approach, that is, assessing whether or
to what extent the prisons being compared are in compliance with
applicable ACA standards and/or other relevant policy and procedural
guidelines and/or court orders and consent decrees. Another approach is
to assess performance measures.5 For example, measures of safety could
include assault statistics, safety inspection results, and accidental injury
reports.
The difficulties of comparatively assessing private and public
prisons—regarding operational costs and/or quality of service—are further
complicated if the prisons are not located in the same state. Each state and
its correctional system have characteristics and conditions that must be
recognized in conducting interstate analyses. For example, economic
conditions and cost-of-living factors can vary by state and by regions of
the nation. Similarly, each state’s correctional system may be somewhat
unique regarding the extent of overcrowding, the history of court
intervention, the emphasis given to ACA accreditation, and the presence or
influence of various other factors. As an illustration, with respect to the
five studies we reviewed, appendix III presents state-specific details
regarding some relevant factors that could affect interstate comparisons of
prison costs and/or quality of service.

On the basis of extensive literature searches and inquiries with
knowledgeable corrections officials and criminal justice researchers (see
app. I), we identified five studies completed since 1991 that compared
private and public prisons in reference to operational costs and/or quality
of service. The following is a brief overview of each study:6

Five Comparative
Studies Completed
Since 1991
•

Texas study (1991): Conducted by the Texas Sunset Advisory
Commission, this study compared (1) the actual costs of operating four
privately managed prerelease minimum-security facilities for male
prisoners and (2) the estimated costs of operating similar but hypothetical
public facilities in Texas. The study did not empirically assess quality of
service.

5

Charles H. Logan, “Criminal Justice Performance Measures for Prisons,” Performance Measures for
the Criminal Justice System, U.S. Department of Justice, Oct. 1993.
6

As presented, each study is introduced with a short title showing the sponsoring state and/or the state
wherein the compared facilities are located. Also, the studies are listed in chronological order by the
date of completion or publication indicated in parentheses. Appendix II provides a more detailed
overview of these five studies.

Page 5

GAO/GGD-96-158 Private and Public Prisons

B-261797

•

•

•

•

New Mexico study (1991): Funded by the National Institute of Justice,
the Bureau of Prisons, and the National Institute of Corrections, this study
compared the quality of service at three multicustody facilities (minimumto maximum-security levels) for women, i.e., a private prison and a
state-run prison in New Mexico and a federal prison in West Virginia. The
study did not include a detailed analysis of operational costs.
California study (1994): Conducted by California State University with
funding from the California Department of Corrections, this study focused
on three community correctional facilities for males. All three facilities
were operated (under contracts with the state) as for-profit alternatives to
state-operated prisons. One facility (medium-security) was operated by a
private corporation, the second (high-security) by a local police
department, and the third (low- to medium-security) by a city
administration. The study compared operational and construction costs
and quality of service. More specifically, regarding costs, the study
compared the three facilities (1) with one another and (2) with other state
correctional facilities. Both operational and construction costs were
included in the comparison of the three facilities with one another. The
statewide comparison did not include construction costs for the California
Department of Corrections facilities. Regarding quality of service, the
study compared the three facilities with two state facilities.
Tennessee study (1995): Conducted in two parts, one for operational
costs and one for quality of service, by the Tennessee state legislature, this
study compared three of Tennessee’s multicustody (minimum- to
maximum-security) prisons for male inmates. One prison was privately
managed, and the other two were state-run prisons.
Washington study (1996): At the time of this study, the state of
Washington had no privately run prisons but was considering the
feasibility of such. Therefore, the study, conducted by the Washington
State Legislative Budget Committee, analyzed pertinent information
available in other states. Regarding operational costs, for, example, the
study looked at the three Tennessee facilities (mentioned above) as well
as three multicustody male prisons in Louisiana (two private and one
public).7 Regarding quality of service, the study compared the three
Tennessee facilities, the three Louisiana facilities, and two Washington
facilities.

7

As discussed in appendix II, the Washington study also included some interstate comparisons of
operational and construction costs.

Page 6

GAO/GGD-96-158 Private and Public Prisons

B-261797

In summary, the California, Tennessee, and Washington studies assessed
operational costs and quality of service.8 The Texas study analyzed
operational costs only, and the New Mexico study analyzed quality of
service only.

Drawing Conclusions
From the Five Studies

While the five studies varied in terms of methodological rigor, they do, to
differing degrees, offer some indication of comparative operational costs
and/or quality of service in the specific settings they assessed. However,
regarding operational costs, because the studies reported little difference
and/or mixed results in comparing private and public facilities, we could
not conclude whether privatization saved money. Similarly, regarding
quality of service, of the two studies that made the most detailed
comparative assessments, one study (New Mexico) reported equivocal
findings, and the other study (Tennessee) reported no difference between
the compared private and public facilities.

Comparisons of
Operational Costs
Indicated Little Difference
And/or Mixed Results

Four of the five studies (Texas, California, Tennessee, and Washington)
assessed operational costs of private and public correctional facilities. In
three of the studies (California, Tennessee, and Washington), comparisons
of private and public facilities indicated little or some differences in
operational costs. Only the Texas study reported finding substantially
lower (14- to 15-percent) operational costs for private versus public
correctional facilities.
Using fiscal year 1990 data, the Texas study reported average daily
operational costs of $36.76 per inmate for the private facilities, compared
with estimates of $42.70 to $43.13 for the public facilities. However, the
results of the Texas study are not fully based on actual experience. Rather,
the study compared existing private facilities (prerelease institutions) to
hypothetical public facilities. This type of hypothetical comparison does
not allow for consideration of any unanticipated changes in components
such as staffing levels, other expenses, rate of occupied bed space, or
many other factors that could affect actual costs. Changes in any single
assumption, or set of assumptions, for the hypothetical institutions could

8

The authors of the California study noted, however, that their findings regarding quality of service
were based upon a small, nonrandom sample of inmates.

Page 7

GAO/GGD-96-158 Private and Public Prisons

B-261797

change the size or even the direction of the differences in the comparative
operational costs.9
Based upon our experience in designing and assessing evaluation
methodologies, we found the Tennessee study (of the studies we
reviewed) to have the most sound and detailed comparison of operational
costs of private and public correctional facilities. The study compared
three mixed-population (minimum- to maximum-security) institutions
(one private and two public). All three facilities were located in
Tennessee, and all three had relatively comparable inmate populations, in
terms of numbers and most demographics, except race. Also, direct and
indirect costs were considered in the analysis, and representatives from
both the private and the public facilities agreed on the cost components
and relevant adjustments prior to data collection. The analysis showed
very little difference in average inmate costs per day among the three
facilities—$35.39 for the private facility and $34.90 and $35.45,
respectively, for the two public facilities.
The Washington study, which made intrastate comparisons of correctional
facilities (minimum- and maximum-security populations) in Tennessee and
Louisiana, also found very little difference in the operational costs of
private and public facilities. For Tennessee, the private facilities’ average
daily operational costs per inmate ($33.61) were lower (about 7 percent)
than the comparable costs for the two public facilities studied ($35.82 and
$35.28, respectively). It should be noted that the Tennessee facilities,
which were analyzed and reported on in the 1996 Washington state study,
were the same facilities that are discussed in the 1995 Tennessee study
cited above. For Louisiana, the average inmate costs per day for the two
private facilities studied were $23.75 and $23.34, respectively, and the
comparable daily operational costs for the public facility studied were
$23.55 per inmate.10
The 1994 California study compared three for-profit community
correctional facilities located in that state—one run by a private firm and
two run by local governments. The study found that the private facility’s
average annual costs per inmate ($15,578) were higher than comparable

9

Moreover, in response to competition from the private sector, public prisons could, over time, become
more cost efficient. As discussed later, this was one of the conclusions of the Washington study.
10

As presented in the study, the costs for the Tennessee and Louisiana facilities were calculated after
an adjustment to equalize prison inmate population numbers. The adjustment was necessary because,
among other things, comparisons of per-capita costs should be based upon equivalent levels of
capacity. Otherwise, a facility operating at less than full capacity, for example, generally will show
higher per-capita costs than a facility operating above its rated capacity.

Page 8

GAO/GGD-96-158 Private and Public Prisons

B-261797

costs for one of the government-run facilities ($13,195) but were lower
than such costs for the other government-run facility ($16,627). The lower
cost government-run facility had a disproportionate share of drug
offenders, which could have affected overall costs. Further, the authors of
the study noted that the results of this study must be viewed with some
additional caution because of inconsistencies in the underlying or
supporting cost figures obtained from different sources within the state.

Comparisons of Quality
Are Unclear

Although comparative costs are very important, they are not the only
factors considered by policymakers in deciding the direction or extent of
corrections privatization. A principal concern is whether private
contractors can operate at lower costs to the taxpayers, while providing
the same or even a better level of service as the public sector, particularly
with respect to safety and security issues.
Of the studies we reviewed, two (New Mexico and Tennessee) assessed
the comparative quality of service between private and public institutions
in much greater detail than the other studies. Both studies used structured
data-collection instruments to cover a variety of quality-related topics,
including safety and security, management, personnel, health care,
discipline reports, escapes, and inmate programs and activities. The New
Mexico study reported equivocal findings, and the Tennessee study
reported no difference in quality between the compared private and public
institutions.
The findings in the New Mexico study are difficult to interpret. On the
basis of surveys of correctional staff and reviews of institutional records,
the study reported that the private prison “outperformed” the public
facilities on most of the measured quality dimensions. However, the author
noted that the results from one of the data-collection instruments—the
inmate surveys—showed an opposite result, with one of the public
facilities “outperforming” the private facility on every dimension except
inmate activities (e.g., work and training programs).
The Tennessee study, in assessing the quality of service at one private and
two public prisons, reported that “all three facilities were operated at
essentially the same level of performance.” This conclusion was largely
based on the results of an operational audit conducted at each of the
facilities by an inspection team. Composed of private and public sector
members, the team used a structured survey instrument to conduct a

Page 9

GAO/GGD-96-158 Private and Public Prisons

B-261797

detailed review of records, observe operations and practices, and conduct
interviews.
The Texas study did not empirically assess the quality of service at the
private correctional facilities. Rather, the study noted that all four of the
privately operated prerelease facilities were in general compliance with
11 of the 16 mandates of court rulings applicable to Texas prisons. Also,
the study noted that two of the four private facilities had received ACA
accreditation, and the other two were still involved in the accreditation
process. ACA officials told us, however, that ACA accreditation means that a
facility has met minimum standards and that accredited facilities can vary
widely in terms of overall quality. To reiterate, because it was based on
hypothetical public facilities, the Texas study made no attempt to
comparatively assess the quality of service across private and public
facilities in Texas.
The California study, in assessing quality of service, used inmate and staff
surveys to compare the three community correctional facilities with two
state prisons. However, the results could not be generalized to the inmate
or staff populations of the respective facilities because small, nonrandom
samples were used.
The California study also attempted to compare the three community
correctional facilities with the state’s other correctional institutions with
respect to recidivism rates. The study reported that, of the three
community correctional facilities, one of the publicly managed facilities
was “most impressive” in performance based on recidivism rates.
Sufficient data were not available to adequately complete the analysis
comparing the inmates released from the community correctional facilities
with inmates released from other correctional institutions in the state.
The Washington study assessed the quality of service at the three facilities
(one private and two public) in Tennessee, three facilities (two private and
one public) in Louisiana, and two facilities (both public) in Washington.
While not as detailed as the New Mexico and the Tennessee studies, the
Washington study concluded that the private and public prisons studied
within the respective states (Tennessee and Louisiana) were generally
similar in quality of service. However, the study noted that Washington’s
two state-run facilities had more counselors per inmate than the other
states’ facilities.

Page 10

GAO/GGD-96-158 Private and Public Prisons

B-261797

Generalizability of the
Studies’ Results

The few studies that have compared the operational costs and/or the
quality of service of private and public prisons provide little information
that is widely applicable to various correctional settings. For example,
while these studies compared private and public facilities that generally
were similar (in terms of capacities, inmate demographics, etc.), the
selected facilities were not necessarily typical or representative of prisons
in either the state studied or other jurisdictions. Also, a variety of factors
that relate to a given location or correctional system may render the
experience of one jurisdiction with private prisons very different from that
of another. Further, the passage of time could alter the relationship
between private and public correctional facilities in terms of costs and
quality. For these reasons, among others, the few studies that we reviewed
do not permit drawing generalizable conclusions about the comparative
operational costs and/or quality of service of private and public prisons.
Jurisdictions, such as states, vary on several dimensions that could have
an impact on the comparative costs of private versus public prisons and, in
turn, affect the generalizability of a given study’s results. First, other
states’ correctional philosophies could differ from that of the states
studied. Some state correctional philosophies are more punitive in nature
(as reflected, for example, by higher incarceration rates), whereas other
states are less punitive and more inclined toward treatment.11 California,
for example, which is one of the states discussed in the five studies we
reviewed, generally has had incarceration rates above the national
average. Also, the Washington study noted that the adjusted estimated
per-bed costs for a state-run facility in Washington ($60,400) were almost
double Florida’s costs ($33,900) due, in part, to state differences in
operating and programming approaches.
Second, jurisdictions also vary in relation to a variety of economic factors
that could affect the relationship between private and public prison costs.
Differences in the costs of living could affect both private and public
prison costs, but in some jurisdictions, one more than the other. For
example, a labor shortage could result in higher operational costs for
private and public prisons.
Third, in some jurisdictions, the inmate population to be incarcerated in
private facilities may be different from those inmate populations in the five
studies. Three of the five studies focused on inmate populations that were
11

The rate of incarceration is the number of sentenced prisoners (per 100,000 resident population) in
correctional facilities. Since the mid-1980s, according to Bureau of Justice Statistics data, the southern
and western regions of the nation have had higher incarceration rates than the northeastern and
midwestern regions.

Page 11

GAO/GGD-96-158 Private and Public Prisons

B-261797

not representative of the broader prison population—prerelease prisoners
(Texas study), female prisoners (New Mexico), and those housed in
community correctional facilities (California). Only two studies
(Tennessee and Washington) focused on costs in relation to facilities
housing a more mainstream prisoner population.
Finally, regarding both operational costs and quality of service, the
comparative performance of private versus public correctional facilities is
not likely to be static. Changes over time could alter the comparative
performance. For example, the first year of a new prison—either private
or public—could reflect expenses for training inexperienced staff as well
as hiring replacements for those unsuited to the work. Inexperienced staff
could also have a negative effect on some measures of quality. Also, in the
initial years of managing a prison, a private firm may choose to bill for its
services at rates below costs to obtain or extend a contract. As time goes
by, however, to remain a viable business entity, the contractor’s
cost-recovery practices would have to change. Similarly, over time, public
prisons could become more cost efficient in response to competition from
the private sector. For instance, this conclusion was reached by the
Washington study, which was commissioned to help the state determine
the potential benefits of privatization.

Lessons Learned for
Future Comparative
Studies

The results of studies comparing private and public prisons obviously are
of interest to any jurisdiction whose policymakers are deciding whether or
to what extent corrections should be privatized. Ideally, to be most useful,
such studies should be based upon representative samples of prisons, with
sufficient statistical controls in place to measure and account for any
differences. However, because the number of private correctional facilities
is still relatively small (see app. IV)—and, given the fact that each
stand-alone facility (whether private or public) may have some unique
characteristics—conducting a truly optimal comparative evaluation may
be impractical.
Nonetheless, the five studies completed since 1991 offer several lessons
learned to guide future studies, even if such studies focus on comparing
only one private facility and one public facility. In reviewing the relative
strengths and weaknesses of each study to formulate lessons learned, we
largely relied on our extensive experience in designing and assessing
evaluation methodologies—that is, our experience with generally accepted
methodological standards and practices. Specifically, on the basis of our
review of the five studies, we identified the following lessons learned:

Page 12

GAO/GGD-96-158 Private and Public Prisons

B-261797

•

•

•

•

In considering the extent to which corrections should be privatized, a key
question is whether private contractors can operate at lower costs to
taxpayers, while providing at least the same level of service as the public
sector, particularly with respect to security and safety issues. Thus, it is
important that any study focus on both operational costs and quality of
service. Two of the studies we reviewed (Texas and New Mexico) did not
have this dual focus.
The best approach for evaluating operational costs is to study existing
comparable facilities, not hypothetical facilities. One of the studies we
reviewed (Texas) used hypothetical similar public facilities.
Generally, there is more than one way to objectively measure or compare
prison security, safety, order, and various other dimensions that constitute
quality of service. In this regard, it is important to use multiple indicators
or data sources to provide cross-checks. The New Mexico study, for
example, illustrates that divergent results can be reached by using one
data source (e.g., inmate surveys) versus another source (e.g., staff
surveys).
Comparative findings with respect to operational costs and/or quality of
service in any given year may not hold true for other years. Similarly,
because trends are not self-perpetuating, even findings based on multiyear
comparisons must be carefully considered. Nonetheless, all other factors
being equal, comparative evaluations based upon several years’ data
potentially have more value than evaluations based upon 1 or 2 years of
data. Nearly all five of the studies we reviewed were based upon 1 or 2
years of data.
These lessons learned could be particularly applicable to BOP if, in the
future, it resumes its plans for contracting with private companies to
operate selected federal correctional facilities.12 For instance, according to
April 1995 congressional testimony by the BOP Director, BOP’s privatization
initiative, if implemented, would provide an opportunity to undertake
some thorough comparative evaluations:
“I know that the Attorney General and . . . [the Office of Management and Budget] are very
interested in working carefully with us in the Bureau of Prisons to track, on these new
contracts, very carefully, what the cost impact truly is, because there are a lot of hidden
costs in privatization . . . [T]here has never been, we don’t believe, a real good cost analysis

12

For example, the Justice Department’s budget justification for fiscal year 1997 projected the
activation of two privatized facilities in that year—a detention facility (677 beds) in Seattle, WA, and a
multicustody (minimum- and low-security) facility (2,048 beds) in Elkton, OH. Departments of
Commerce, Justice, and State, the Judiciary, and Related Agencies Appropriations for 1997: Hearings
before a Subcomm. of the House Comm. on Appropriations, 104th Cong., 2d Sess. 1655 (1996).
However, in June 1996, the administration announced that plans to privatize these facilities were being
suspended.

Page 13

GAO/GGD-96-158 Private and Public Prisons

B-261797

to determine, apples to apples, what is the cost of a traditional prison system and private
contracting. The private contractors claim they can do it at great savings, and so we are
very interested in monitoring the ones that we have projected for the next few years and
determining . . . how well the taxpayers are being served on either side.”13

The BOP Director noted that by contracting out the management of
selected facilities incarcerating general inmate populations, BOP was
moving to the “next level of privatization,” which would provide a good
basis for comparative evaluations focusing on “like or similar institutions.”
In this regard, the lessons learned from previous comparative studies
should be useful to BOP if the federal privatization initiative is revisited.

Comments and Our
Evaluation

We obtained oral comments on a draft of this report from BOP and written
comments from the Department of Justice’s Office of Justice Programs;
the National Council on Crime and Delinquency; and a Northeastern
University (Boston, MA) professor of criminal justice, who is a nationally
recognized authority on corrections administration.
BOP

commented that the report was accurate, well done, and useful.

The Office of the Assistant Attorney General, Office of Justice Programs,
concurred with the report and noted that additional study of the
privatization of correctional facilities is needed. The National Institute of
Justice, a component agency of the Office of Justice Programs,
commented that the report appeared “to be as comprehensive as the
available data permits.” Also, the Institute commented that the report’s
discussion of the strengths and weaknesses of the five studies “is
excellent.”
In commenting on the draft, the Executive Vice President, National
Council on Crime and Delinquency, said that the report is accurate in
concluding that few studies have been completed to date and that these
studies have methodological problems that limit understanding the actual
cost-benefits of privatization. He noted, however, that our report could
place more emphasis on the Tennessee study, which is the most rigorous
study to date. Although we concur with the reviewer’s assessment of the
study, our objective was to provide similar information for each of the
studies reviewed. Further, he noted that the report could add more
emphasis to evaluating the claims of private providers that they can
13

Departments of Commerce, Justice, and State, the Judiciary, and Related Agencies Appropriations
for Fiscal Year 1996: Hearings on H.R. 2076 before a Subcomm. of the Senate Comm. on
Appropriations, 104th Cong., 1st Sess. 303, 308 (1995).

Page 14

GAO/GGD-96-158 Private and Public Prisons

B-261797

construct new facilities faster and cheaper than public entities. Since the
studies reviewed did not assess these claims, this issue was beyond the
scope of our work.
The Northeastern University reviewer14 commented that (1) our evaluation
synthesis was an important contribution to the corrections field, (2) the
report’s conclusion that the five studies offer little generalizable guidance
for other jurisdictions regarding the comparative cost and quality of
service of private and public correctional facilities was “right on point,”
and (3) our cautions concerning interstate comparisons were
“well-founded.”
However, the reviewer underscored the need also to focus privatization
research on crime reduction and various philosophical questions
underpinning the privatization debate. These issues were beyond the
scope of our work. In addition, the reviewer suggested that it would be
valuable to the corrections field if the report included a short, concise
statement describing the critical dependent and independent variables that
should be considered in comparative analyses of private and public
corrections facilities. Because of variations in available data and possible
measurement adjustments required in specific research situations, we are
hesitant to prescribe what variables should be studied. The studies
reviewed, however, suggest possible variables. Furthermore, citing the
difficulties researchers have in accessing data from private firms, the
reviewer proposed that the report contain a recommendation that would
facilitate researchers’ access to proprietary information needed for
evaluation of private corrections. In the case of federal corrections-related
contracts, we would likely have access to data, but we have considerably
less jurisdiction at the state level.

We are providing copies of this report to the Chairman and Ranking
Minority Member of the House Judiciary Committee, the Attorney General;
the Director, BOP; and other interested parties. Copies will also be made
available to others upon request.

14

Dr. Edith Elisabeth Flynn, College of Criminal Justice, Northeastern University, Boston,
Massachusetts.

Page 15

GAO/GGD-96-158 Private and Public Prisons

B-261797

The major contributors to this report are listed in appendix V. Please
contact me on (202) 512-8777 if you or your staff have any questions.

Laurie E. Ekstrand
Associate Director, Administration
of Justice Issues

Page 16

GAO/GGD-96-158 Private and Public Prisons

Page 17

GAO/GGD-96-158 Private and Public Prisons

Contents

Letter
Appendix I
Objectives, Scope,
and Methodology
Appendix II
Description of Studies
Comparing Private
and Public Prisons

1

Identification of Comparative Studies
Assessment of Studies
Private Sector Contacts

20
20
20
21

Texas Study (1991)
New Mexico Study (1991)
California Study (1994)
Tennessee Study (1995)
Washington Study (1996)

23
26
27
29
31
33

Appendix III
Factors That Could
Affect Interstate
Comparisons of
Prison Costs And/or
Quality of Service

36

Appendix IV
Use of Private
Corrections in States

40

Appendix V
Major Contributors to
This Report

41

Tables

Table II.1: Citation, Evaluation Parameters, and Reported Results
of Five Studies Comparing Private and Public Prisons
Table III.1: Factors That Could Affect Comparisons With States
Studied
Table IV.1: Private Adult Secure Correctional Facilities Operating
or Planned, as of March 1996

Page 18

23
38
40

GAO/GGD-96-158 Private and Public Prisons

Contents

Abbreviations
ACA
BOP

Page 19

American Correctional Association
Federal Bureau of Prisons

GAO/GGD-96-158 Private and Public Prisons

Appendix I

Objectives, Scope, and Methodology

In initiating this review, our specific objectives were to answer the
following key questions:
•
•

•
•

Identification of
Comparative Studies

What studies, completed since 1991, have compared the operational costs
and/or the quality of service of private and public prisons?
From these studies, what can be concluded with respect to the operational
costs and/or the quality of service of comparable private and public
facilities?
Are the results of these studies generalizable to correctional systems in
other jurisdictions?
From these studies, what are the “lessons learned” to help guide future
comparative studies of private and public prisons?

To identify relevant studies, we requested and obtained literature searches
from the National Criminal Justice Reference Service and the National
Institute of Corrections. Also, within the Department of Justice, we
contacted knowledgeable officials of the component agencies responsible
for managing federal correctional facilities—the Federal Bureau of Prisons
(BOP), U.S. Marshals Service, and Immigration and Naturalization Service.
Similarly, to query knowledgeable state agency officials, we first contacted
the Director of the Private Corrections Project at the University of
Florida’s Center for Studies in Criminology and Law to obtain information
about the number of privatized facilities in each state (see app. IV).1 Then,
for each applicable state, we contacted officials at the state’s corrections
department and/or corrections research agency and inquired about any
completed, ongoing, or planned studies comparing private and public
prisons. Further, we contacted the National Council on Crime and
Delinquency and various nationally recognized researchers in academia.

Assessment of Studies

We assessed each of the five relevant studies that we identified. Our
work in assessing the conclusions or results of each study and the
generalizability of same—as well as identifying any lessons learned—can
be characterized as a form of evaluation synthesis. By definition, an
evaluation synthesis is a systematic procedure for organizing findings from
several disparate evaluation studies. That is, the procedure addresses key
questions or issues by assessing existing studies or evaluations, rather
than by conducting primary data collection.
1

Established in 1988 to conduct policy-relevant research on corrections privatization, the Private
Corrections Project publishes (semiannually) the Private Adult Correctional Facility Census.

Page 20

GAO/GGD-96-158 Private and Public Prisons

Appendix I
Objectives, Scope, and Methodology

In reviewing the studies, we focused on the findings and conclusions of
each study and evaluated these in relation to the methodology used by the
respective study.2 As an initial and fundamental inquiry, for instance, we
focused on the similarity of the private and public facilities being
compared in each study. That is, we wanted to determine (1) if the
facilities were reasonably well matched in relation to design and capacity,
security level, inmate demographics, and other relevant institutional
characteristics and (2) whether the facilities were actually in operation
and, if so, for what length of time prior to the comparative evaluation.
In reviewing each applicable study’s comparative evaluation of operational
costs, we focused on whether (1) both direct and indirect cost
components were considered for the private and public facilities,
(2) actual data versus estimates were used, and (3) consistent cost
components were used. We did not independently verify any of the cost
data presented in the studies.
In reviewing each applicable study’s comparative evaluation of quality of
service, we focused on whether the private and the public facilities were
consistently evaluated in the respective study. That is, in reference to both
the private and the public facilities compared in a given study, we were
interested in whether the same or similar methodology and data sources
were used to evaluate quality of service. Thus, we did not attempt to
generically define “quality of service;” rather, we accepted the definition
and/or evaluation criteria used in each applicable study. Also, we did not
independently verify the reported quality measures or outcomes, such as
safety and incident data and the extent of rehabilitation and treatment
programs for inmates.
We reviewed the relative strengths and weaknesses of each study to
formulate lessons learned for future comparative studies. In doing so, we
largely relied on our extensive experience in designing and assessing
evaluation methodologies—that is, our experience with generally accepted
methodological standards and practices.

Private Sector
Contacts

Initially, in September 1995, to obtain a practical understanding of
privatization issues, we visited two privately operated facilities housing
federal inmates. These facilities, which held deportable aliens, are located
in west Texas and were operated by private firms under the general
2

Two of our staff—senior analysts with specialized training and years of experience in designing and
assessing evaluation methodologies—independently reviewed each of the five studies.

Page 21

GAO/GGD-96-158 Private and Public Prisons

Appendix I
Objectives, Scope, and Methodology

authority of intergovernmental agreements entered into by BOP and the
respective city governments of Big Spring and Eden. We toured the
facilities and interviewed managers and staff of the private firms. Also, we
interviewed the on-site federal monitors.
Further, to obtain additional overview information on privatization issues,
we interviewed a senior executive (the Director for Strategic Planning) of
one of the nation’s largest private corrections firms. This official was a
former director of BOP.

Page 22

GAO/GGD-96-158 Private and Public Prisons

Appendix II

Description of Studies Comparing Private
and Public Prisons
We identified five studies completed since 1991 that compare private and
public correctional facilities in relation to operational costs and/or quality
of service. Table II.1 briefly describes each of these studies. Following the
table, separate sections respectively provide more details about each
study.

Table II.1: Citation, Evaluation Parameters, and Reported Results of Five Studies Comparing Private and Public Prisons
State
Study citation
Evaluation parameters
Reported results
Texas

“Information Report on Contracts for
Correction Facilities and Services,”
Recommendations to the Governor of
Texas and Members of the
Seventy-Second Legislature, Sunset
Advisory Commission, Final Report,
Texas Sunset Advisory Commission
(Austin: 1991).

Operational costs were studied.
Four private, prerelease, minimumsecurity prisons (500 beds each) for
males were compared with
hypothetical public facilities in Texas.

The private prisons’ operational costs
were 14 to 15 percent less than the
costs of the hypothetical state
facilities.

Fiscal year 1990 data were analyzed.
An empirical assessment of quality of The study noted that two of the four
service was not conducted.
private facilities had received ACA
accreditation, and the other two were
still involved in the accreditation
process. Also, the study noted that all
four facilities were in compliance with
11 of the 16 mandates of court
rulings applicable to Texas prisons.

New Mexico

Charles H. Logan, Well Kept:
A detailed analysis of operational
Comparing Quality of Confinement In costs was not conducted.
a Public and a Private Prison,
National Institute of Justice (1991).

Not applicable.

Quality of service was studied.
Three multicustody facilities
(minimum- to maximum-security) for
women were compared: a private
prison and a state-run prison in New
Mexico and a federal prison in West
Virginia.
The data analyzed for the private
facility covered June 1989 through
November 1989; data for the state
facility covered June 1988 through
November 1988; and the federal data
covered December 1987 through
May 1988.

The results of the study depended on
the data-collection instruments that
were employed. For example, data
from staff surveys and official records
showed that the private prison
“outperformed” the state and federal
prisons across nearly all dimensions.
However, inmate survey data showed
that one of the public facilities
“outperformed” the private facility on
every dimension except “activity”
(e.g., work and training programs).

(continued)

Page 23

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

State

Study citation

Evaluation parameters

California

Dale K. Sechrest and David Shichor,
Final Report: Exploratory Study of
California’s Community Corrections
Facilities, California State University
(San Bernardino: 1994).

Operational and construction costs
were studied.

Reported results

The private facility’s average annual
costs per inmate ($15,578) were
higher than comparable costs for one
Three for-profit community
of the government-run facilities
correctional facilities—one managed ($13,195) but were lower than such
privately (medium-security) and two
costs for the other government-run
managed by local governments (low- facility ($16,627).
to medium- and high-security)—for
males were compared with one
Due to methodological limitations,
another and with other state
conclusions could not be reached by
correctional facilities.
comparing the three for-profit
community correctional facilities with
Fiscal year 1991-1992 data were
other state correctional facilities. For
analyzed.
example, different cost components
were used for the two sets of facilities
in the comparison. In addition, it is
likely that the universe of the state’s
correctional facilities reflected
wide-ranging differences concerning
inmate populations and services.
Quality of service was studied.

Due to methodological limitations,
conclusions could not be reached by
The same three community
comparing the community
correctional facilities for males were
correctional facilities with two state
compared with two state correctional facilities. For example, the results
facilities.
could not be generalized to the
inmate or staff populations of the
Data were collected in summer 1992. facilities because small, nonrandom
samples were used.
Tennessee

Cost Comparison of Correctional
Centers, Tennessee Legislature
Fiscal Review Committee (Nashville:
1995).

Operational costs were studied.
Three multicustody prisons
(minimum- to maximum-security) for
males were compared: one private
and two state-run prisons.

There was little difference in the
average daily operational costs per
inmate across the three
facilities—$35.39 for the private
facility, versus $34.90 and $35.45,
respectively, for the two public
facilities.

Data from July 1993 through June
1994 were analyzed.
Comparative Evaluation of Privately
Managed CCA Prison and
State-Managed Prototypical Prison,
Tennessee Legislature Select
Oversight Committee on Corrections
(Nashville: 1995).
Washington

Quality of service was studied.
The same three multicustody prisons
for males were compared.

There was no difference in quality of
service between the private and
public facilities.

Data from March 1991 through
September 1994 were analyzed.

Department of Corrections
Operational costs of the same three
Privatization Feasibility Study, Report Tennessee facilities mentioned above
96-2, State of Washington Legislative were studied.
Budget Committee (Olympia: 1996).
Data from July 1993 through June
1994 were analyzed.

The average daily operational costs
per inmate for the private facility
($33.61) were slightly lower than
such costs for the two public facilities
($35.82 and $35.28, respectively).
(continued)

Page 24

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

State

Study citation

Evaluation parameters

Reported results

Operational costs of three
multicustody facilities in Louisiana
(two private and one state-run) for
males were studied.

The average daily operational costs
per inmate for the two private
facilities were $23.75 and $23.34,
respectively, compared with $23.55
for the public facility.

Projected data for July 1995 through
June 1996 were analyzed.
Operational costs of a Washington
state prison were compared with the
costs for one Tennessee state prison
and a Louisiana state prison.
Data analyzed for Washington state
covered calendar year 1995; time
frames for Tennessee and Louisiana
data are mentioned above.
Construction costs were studied.
The estimated costs for Washington
state to construct a planned
multicustody public prison for males
were compared with a private
company’s costs for constructing a
similar facility in Florida.

The average daily operational costs
per inmate for the Washington facility
($44.52) were higher than such costs
for the Tennessee and Louisiana
facilities ($37.07 and $24.04,
respectively).

The estimated cost per bed for the
Washington state facility ($60,400)
was approximately double the
estimated cost per bed for the Florida
facility ($33,900).

Data analyzed for Washington state
were based on projected July 1998
cost figures; and data analyzed for
the Florida facility were based on
projected July 1998 cost figures.
Quality of service was studied.
Three multicustody male facilities in
Tennessee (mentioned above), three
multicustody male facilities in
Louisiana (mentioned above), and
two multicustody male facilities in
Washington were compared.

Site visits showed that all prisons
were “clean and appeared to be
orderly.” Additional data indicated
that the prisons generally were
similar regarding quality of service.
However, the Washington facilities
had more counselors per inmate than
the Tennessee and Louisiana
facilities.

Data analyzed for Tennessee
covered 1994 (for review of
institutional records) and 1995 (for
on-site visits); data analyzed for
Louisiana covered 1995; data
analyzed for Washington covered
1995.
Source: GAO summary of information reported in the cited studies.

Page 25

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

Texas Study (1991)

Conducted by the Texas Sunset Advisory Commission, this study involved
a comparative assessment of operational costs; an empirical assessment of
quality of service was not conducted. The actual costs of operating four
privately managed prerelease minimum-security facilities (500 beds each)
for male prisoners were compared with the estimated costs of operating
similar but “hypothetical” public facilities in the state of Texas. Two of the
four prerelease facilities were managed by Corrections Corporation of
America, and the other two were managed by Wackenhut Corrections
Corporation.
The study considered direct and indirect costs to compute operational
costs for private and public sector management of the prerelease facilities.
Direct costs included items such as salaries and fringe benefits, food,
medical services, utilities, and supplies. Also, the study recognized that
another direct cost would be the expense of having state corrections
agency staff on-site to monitor the contractor’s performance. Indirect
costs included salaries and expenses for corrections department executive
personnel, an annual audit of facilities, and “other administration items”
attributable to the private or state facilities. Construction costs were
excluded because the state built the private facilities. Also, the study did
not include depreciation expenses and capital outlays, but debt service for
construction was included.
The study concluded that the state achieved savings from the privatized
facilities. Based on requirements specified in state statute, the state
estimated the costs of operating similar state-run prerelease facilities.
Specifically, because similar state-run facilities did not exist, the state
estimated the costs of operating hypothetical state-run prerelease
facilities. Contract provisions stipulated that contractors would receive at
least 10 percent less than the estimated costs for the state to operate each
facility.1 Therefore, in a sense, 10-percent “savings” to the state was
guaranteed. Cost data for fiscal year 1990 were analyzed. The state
estimated that the privatized facilities achieved 14- to 15-percent cost
savings (taking into consideration tax revenues paid to state and local

1

Texas state statutory provisions authorize the Texas Department of Corrections to contract with
private vendors and county commissioners to finance, construct, operate, maintain, or manage secure
correctional facilities. The provisions establish guidelines by which the state can enter into a contract
for such services. Among other things, the provisions specify that a contract proposal must provide the
state with a savings of at least 10 percent below the costs of similar state-operated facilities.

Page 26

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

authorities)2 compared with hypothetically equivalent state-run facilities.
The average daily operational costs per inmate for the private facilities
were $36.76. Because of staffing and construction differences between the
contractors, separate costs were estimated for the hypothetical state
operation of a facility for each contractor. The estimated average daily
operational costs per inmate for the hypothetical state-run facilities were
$42.70 for one contractor and $43.13 for the other. However, because the
state did not operate any prerelease facilities nor did any of its existing
facilities have prerelease components, the cost estimates for the state-run
facilities were not based on actual state experience. The method assumed
no unanticipated changes in components such as salary and other
expenses. Thus, an error in one or more assumption could have resulted in
different cost estimates, changing the size or even the direction of
estimated differences in private versus public management costs.
An empirical assessment of the quality of service was not conducted due
to the absence of comparable state facilities. However, the study noted
that two of the four private facilities had received ACA accreditation, and
the other two were still involved in the accreditation process. Additionally,
the study noted that all four facilities were in general compliance with 11
of the 16 mandates of court rulings applicable to Texas prisons.

New Mexico Study
(1991)

Funded by the Department of Justice’s National Institute of Justice,
Bureau of Prisons, and National Institute of Corrections, this study, of the
five we reviewed, made the most systematic effort to address quality of
service. However, a detailed cost analysis was not included. The study
compared three multicustody (minimum- to maximum-security) women’s
facilities—a privately run facility and a state-run facility in New Mexico
and a federal facility in West Virginia—across eight dimensions of quality.3
2

According to the Texas study, the private prisons have paid an estimated $400,000 per prison in state
and local sales taxes and payments in lieu of property taxes. The study noted that because the state
owned the prisons, property taxes were not assessed. Rather, the private contractor paid the local
taxing authorities an annual amount, which generally approximated the taxes that would have been
owed if the property were privately owned.
3

The New Mexico study used a “confinement model” of imprisonment to develop quality assessment
criteria. The model defined eight distinct dimensions or performance measures on which to evaluate
the quality of a correctional facility. According to the study, these dimensions are based on the
premise that confinement “carries with it an obligation to meet the basic needs of prisoners at a
reasonable standard of decency,” including standards to evaluate health care, safety, sanitation, and
nutrition, as well as constitutional standards to ensure due process and fairness. Examples of those
dimensions are (1) security (e.g., facility design and security procedures); (2) safety (e.g., personal
injury and harm); (3) order (e.g., discipline and control standards); (4) care (e.g., health care and
counseling); (5) activity (e.g., work and training programs); (6) justice (e.g., discipline and grievance
procedures); (7) conditions (e.g., food services); and (8) management (e.g., staff turnover and job
satisfaction).

Page 27

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

Data analyzed for the private facility covered June 1989 through November
1989, data for the state facility covered June 1988 through November 1988,
and data for the federal facility covered December 1987 through May 1988.
The study recognized, at least indirectly, that differences among the
facilities regarding age, architecture, and inmate programs made
comparisons somewhat difficult to interpret. For example, the private
facility was new, the state facility was 4 years old, and the federal facility
was about 60 years old. The respective inmate populations were 170
(private), 143 (state), and 814 (federal). The inmates at the New Mexico
facilities were nearly similar with respect to characteristics of age, race,
and offense type. However, they differed from the federal inmates in race
and offense type. The study indicated that one of the factors enhancing the
comparability of the two New Mexico institutions was that both were
applying for accreditation by the American Correctional Association (ACA).
However, ACA officials told us that ACA accreditation means that only
minimum standards are met, and since there can be wide variations among
facilities in exceeding minimum standards, accreditation should not be
used to assume that two or more facilities are comparable.
In assessing and comparing the quality of service at the three facilities, the
study derived multiple indicators for each of eight quality dimensions.
Data sources for all three facilities included various institutional records,4
such as incident and disciplinary reports as well as work and education
records. Also, staff surveys were conducted at all three facilities, and
inmate surveys were conducted at the private and the state facilities. In
total, the study made 595 comparisons among the institutions using 333
indicators. All of the indicators were available for the private and state
prisons, while 131 indicators were available for the federal prisons. Thus,
the study made three-way comparisons for 131 of the 333 indicators and
two-way comparisons (private/state) for 202 of the indicators.
The study concluded, generally, that “the private prison ‘outperformed’ the
state and federal prisons, often by quite substantial margins, across nearly
all dimensions.” It noted, however, that results varied by data source. For
example, contrary to other sources used in the study, inmate survey data
showed that the state facility “outperformed” the private facility in every
dimension except “activity” (e.g., work and training programs).
While the study did not include a detailed analysis of operational costs, it
suggested that the better performance of the private facility was
4

The study noted that fewer official records were collected for the federal prison.

Page 28

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

accomplished at lower cost. However, the study offered little evidence to
support this assertion. For instance, the full report consisted of 291 pages,
with only 2 pages devoted to costs. Without providing any detailed
analysis, the report noted that the average daily operational costs per
inmate for the private facility were $69.75 in fiscal year 1989-1990, the
average daily costs of housing an inmate in federal facilities (nationwide)
were $39.67 in 1988, the average daily costs for New Mexico state facilities
(statewide) were $68.00 in 1988, and the average daily costs for the
particular state facility studied were $80.00 in fiscal years 1988-1989.
Although no detailed cost analysis was attempted, the study appeared to
base the perception of lower costs of private facilities on the fact that
“financial analysts in the New Mexico Corrections Department believed
that the [private] contract was saving the state money.”

California Study
(1994)

Conducted by California State University, this study was neither
supportive nor critical of private facilities. To assess costs and quality of
service, the study compared three for-profit community correctional
facilities for males—one privately managed (medium-security) and two
publicly managed by local governments (one low-to medium-security and
one high-security).5 Specifically, the study compared the three facilities
with one another and with other state correctional facilities. Both
operational and construction costs were included in the comparison of the
three facilities with one another. The statewide comparison did not
include construction costs for the California Department of Corrections
facilities. Also, the study attempted to compare the quality of service of the
three facilities with two other California Department of Corrections
facilities.
The three community correctional facilities were generally comparable.
For instance, the inmates were nearly similar with respect to
characteristics of age, race, and offender status. However, one of the
public facilities had a greater percentage of drug offenders6 and Anglo- and
African-American inmates and a smaller percentage of Hispanics. The
private facility housed 400 inmates, compared with about 420 and about
450 housed at the public facilities. For the period studied, the number of
admissions was 1,498 for the private facility versus 392 and 1,073,
5

Correctional facilities that are operated for-profit by a private corporation under contract with the
California Department of Corrections are referred to as “private proprietary facilities.” Correctional
facilities that are operated for-profit by local government agencies under contract with the California
Department of Corrections are referred to as “public proprietary facilities.”
6

This public community correctional facility housed narcotics offenders who were sentenced under a
“civil addict program” and subsequently violated parole.

Page 29

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

respectively, for the two public facilities. The two public facilities were
relatively new (operational by mid-1991), while the study referred to the
private facility as “older.”
Using fiscal year 1991-1992 data, construction and operational cost
comparisons of the private and the two public facilities revealed some
differences. For example, the study found that the private facility’s average
annual costs per inmate ($15,578) were higher than comparable costs for
one of the government-run facilities ($13,195) but were lower than such
costs for the other government-run facility ($16,627). Construction and
operational costs, including overhead and capitalization costs, were
calculated for the three facilities. Costs were based on contracts
negotiated with the facilities and included capitalization, lease, renovation,
program development, and liability insurance. Costs that were not
included in the calculations were Parole Division overhead costs (for
community correctional facilities), state monitoring, medical costs
allocated to the California Department of Corrections, inmate clothing,
inmate pay, miscellaneous contracts, interest payments, and possible tax
breaks. Also, there were inexplicable inconsistencies in the cost data
obtained from two agencies within the California Department of
Corrections. These inconsistencies may have affected the reliability of the
cost estimates.
Attempts to compare the costs of the three community correctional
facilities and other state correctional facilities were not fully successful.
The cost calculations for the other state facilities used different
components than did the calculations for the community correctional
facilities. The latter costs included construction costs; however, the data
for the other state facilities did not. Therefore, these cost calculations
were not directly comparable. Further, given the unique characteristics of
community correctional facilities, the usefulness of comparing these
facilities to all other correctional facilities in California—many of which
are likely to be very different from the community-based facilities—is
questionable.
To assess quality of service, inmate and staff surveys were conducted at
the three community correctional facilities and at two state prisons.
However, due to small, nonrandom samples, the results could not be
generalized to the inmate or staff populations at any of the facilities.
The California study also attempted to compare the three community
correctional facilities and the state’s other correctional institutions in

Page 30

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

reference to recidivism rates. The study reported that, of the three
community correctional facilities, one of the publicly managed facilities
was “most impressive” in performance based on recidivism rates.
Sufficient data were not available to adequately complete the analysis
comparing the inmates released from the community correctional facilities
to inmates released from other correctional institutions in the state.
In summary, the California study’s methodological limitations prohibit
drawing any overall conclusions about quality of service. The study
acknowledged that any future comparative studies in California should
“incorporate more inclusive and better-selected survey samples.”

Tennessee Study
(1995)

The Tennessee state legislature conducted a two-part study. One part was
a cost assessment, and the other was an assessment of quality. Overall,
this effort was the most systematic attempt of all the studies we reviewed
to assess both the costs and quality of service. Three multicustody
(minimum- to maximum-security) prisons in Tennessee were
compared—one privately managed prison (Corrections Corporation of
America) and two state-run prisons.
The facilities were generally comparable. All three were new (e.g.,
operational by mid-1992), and all had been accredited by ACA also had met
other applicable professional standards. The inmates were similar on all
demographic characteristics mentioned, except race. No information was
provided on capacity level, but the institutions housed approximately the
same numbers of inmates—private (961) and state (929 and 1,029).7 The
study did not report any information regarding the inmate-to-staff ratios
for the facilities.
Similar criteria were used to compare the operational costs of the
facilities. The cost components and relevant adjustments—for direct and
indirect costs—were agreed to by all parties (private and public) prior to
data collection. Direct costs included salaries and fringe benefits, food,
professional services, equipment, maintenance, travel, utilities, and
supplies. Also, there was a cost provision for state employees to monitor
the private prison. Costs for medical and mental health services were
excluded. Indirect costs included salaries and expenses for corrections
department administration and overhead, and interest on working capital.

7

These data reflect the second year of operation.

Page 31

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

Using data that covered July 1993 through June 1994, the study concluded
that the costs of operating the private and both state facilities were
virtually identical. Specifically, the comparison showed that the average
daily operational costs per inmate for the private prison were $35.39,
versus $34.90 and $35.45, respectively, for the two public prisons.
The quality of service assessment consisted of three components—an
audit (given a weight of 60 percent), a security and safety index (a weight
of 25 percent), and a program and activity index (a weight of 15 percent).
The study period for the quality of service assessment was March 1991
through September 1994.
An operational audit was conducted at each of the facilities by an
inspection team, consisting of selected staff from the Tennessee
Department of Corrections and the Corrections Corporation of America.
The staff had varying degrees of expertise in major functional areas, such
as administration, safety and physical plant, health services, treatment,
and security. The team used a structured survey instrument to conduct a
detailed review of records, observe operations and practices, and conduct
interviews. By using the survey instrument, the team attempted to assess
compliance with the various programs and practices within each of the
functional areas. Examples of those programs and practices were
administration (e.g., fiscal management and affirmative action); safety and
physical plant (e.g., fire and occupational safety and sanitation); health
services (e.g., dental care and pharmacy services); treatment (e.g., inmate
orientation and social programs); and security (e.g., firearms and armory
control).
The security and safety index considered many factors, including
disciplinary reports, use-of-force incidents, assaults, deaths, injuries, and
escapes. These reports were counted over a 15-month period (from
July 1993 through September 1994) for each facility.
The program and activity index measured the percentage of inmates who
were eligible for a work or program assignment but remained inactive and
unassigned. The data used in this review were derived from monthly
reports that measured actual numbers of prisoners assigned to the
particular program or activity and the percent unassigned.
The results of the quality of service assessment stated that “all three
facilities were operated at essentially the same level of performance.” No
differences were found among the facilities on the security and safety

Page 32

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

index or on the program and activity index. All estimated variation across
the facilities was due to differences in audit scores. The overall
performance scores were 98.49 for the private facility and 97.17 and 98.34,
respectively, for the two public facilities.

Washington Study
(1996)

When the Washington State Legislative Budget Committee conducted this
study, the state had no privately run prisons but was considering the
feasibility of such. Therefore, using pertinent information available in
other states, this study made several intrastate and interstate comparisons
of correctional facilities. For example, the study compared the operational
costs of the three Tennessee facilities (mentioned above) as well as three
multicustody male facilities in Louisiana. Of the three facilities in
Louisiana, two were privately operated (Corrections Corporation of
America and Wackenhut, respectively), and the other was state operated.
All three Louisiana facilities were in full operation by the beginning of
1991.
Each of the Tennessee prisons had a rated capacity for 1,336 inmates, and
each of the Louisiana prisons had a rated capacity for 1,474 inmates. The
average daily inmate population at each of the Tennessee prisons was
slightly over 1,300, compared with a range of over 1,300 to more than 1,400
at the Louisiana prisons. Also, within each state, there was little difference
among the prisons’ inmates with respect to demographics such as
education, age, offense types, and sentence lengths.
Several cost comparisons were made between the private and public
facilities. First, the operational costs of the one private and the two public
prisons in Tennessee (actual data for July 1993 through June 1994) were
compared, as were the operational costs of the two private prisons and the
one public prison in Louisiana (estimated data for July 1995 through
June 1996). The unadjusted operational costs of the three Tennessee
facilities were similar. However, after adjustments to equalize the numbers
of inmates, the private facility’s average daily operational costs per inmate
($33.61) were slightly lower than the comparable costs for the two public
facilities ($35.82 and $35.28, respectively). For the Louisiana facilities, the
average daily operational costs per inmate for the two private prisons
were $23.75 and $23.34, respectively, versus $23.55 for the public facility.
The Washington study also compared the operational costs of one
Washington state prison with the operational costs of one Tennessee state
prison and the operational costs of one Louisiana state prison. The three

Page 33

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

facilities were similar on some characteristics; however, adjustments were
made to the costs and number of beds of the Tennessee and Louisiana
facilities to further equalize the comparison. The study showed that the
average daily operational costs per inmate for the Washington facility
($44.52) were higher than the costs for the Tennessee ($37.07) and the
Louisiana ($24.04) facilities.
The comparison of the Washington facility with the Tennessee and the
Louisiana facilities was problematic. While the facilities were similar in
capacity, there were differences in inmate demographics, such as race and
offense type. Other factors (e.g., cost-of-living differences) served to
complicate further the interstate comparisons. In any event, these
interstate cost comparisons involved state-run facilities only and did not
consider any private facilities.
Further, the Washington study looked at construction costs by comparing
the estimated costs for Washington state to construct a planned
multicustody public prison for males with a private company that was
constructing a similar facility in Florida.8 In making the interstate
comparison, the study noted that it focused on the “major elements
contributing to capital costs,” which included amounts and types of facility
space, actual construction costs, and ancillary construction costs such as
design and administration.9
Although the Washington study noted that the facilities were comparable
in terms of size and inmate mix (e.g., “large multicustody”), it made cost
and space programming adjustments to the facilities to further equalize the
comparison. For instance, land and site-related costs, taxes, and unique
local costs were excluded from the comparison. For the Florida facility,
the study made upward cost adjustments to account for differences
between the two facilities in labor and material costs, the later completion
(about 2 years of construction inflation) of the Washington facility, and
state oversight of the construction. For the Washington facility, the study
made downward cost adjustments to account for differences between the
two facilities, such as budget reductions of 20 percent, and space
reductions of 18 percent to account for differences in inmate security
levels and other space allocations. In addition, the study made downward
cost and space adjustments to reflect the Florida facility’s lower mix of
close custody beds.
8

When completed, the Florida facility is to be operated by Wackenhut.

9

The private prisons in Tennessee and Louisiana could not be used for capital costs comparisons
because the prisons were state built rather than privately built.

Page 34

GAO/GGD-96-158 Private and Public Prisons

Appendix II
Description of Studies Comparing Private
and Public Prisons

The study indicated that there were other differences in the space
programming between the two facilities, for which no adjustments were
made. For example, the Washington facility assumed single cells for “Close
Housing and Administrative Segregation,” while the Florida facility
assumed double cells for those beds. Also, the Washington facility’s
minimum-security beds had relatively high per-bed space allocations,
reflecting the incorporation of service and program space in the housing
calculation. The space allocations for the Florida facility, however,
reflected medium-security beds with centralized program and service
space.
The study showed that the adjusted estimated per-bed cost for the
Washington state facility ($60,400 per bed) was approximately double the
estimated cost for the Florida facility ($33,900 per bed). The cost
difference was explained largely as due to different operating and
programming approaches or philosophies between the two states.
The Washington study concluded that privatization per se would not result
in cost savings to the state. Rather, the report noted that savings could be
accomplished through privatization or through changes in the state’s
operational policies and practices. For example, savings directly related to
privatization would be due primarily to a private company’s flexibility to
operate outside state rules and procedures, collective bargaining
agreements, and the state’s employee compensation system.
Finally, the Washington study comparatively assessed the quality of
service at the selected private and public prisons in Tennessee and
Louisiana and two multicustody male facilities in Washington. While this
portion of the study was not as detailed or as comprehensive as the
portion involving costs, the quality assessment included visiting the
prisons and reviewing institutional records for several topics, such as
escapes, major disturbances, and inmate infractions. The study concluded
that the private and public prisons studied within the respective states
(Tennessee and Louisiana) generally were similar in quality of service.
However, the study noted that Washington’s two state-run facilities had
more counselors per inmate than the facilities in the other states.

Page 35

GAO/GGD-96-158 Private and Public Prisons

Appendix III

Factors That Could Affect Interstate
Comparisons of Prison Costs And/or Quality
of Service
Several factors could affect interstate comparisons of prison costs and/or
quality of service. In addition to cost-of-living and other economic
differences among the nation’s geographic regions and states, these
factors include the (1) extent of prison overcrowding, (2) history of court
interventions, (3) status of ACA accreditation of facilities, (4) rate of
incarceration (as an indicator of the punitiveness of the corrections
system), and (5) rights of correctional employees to organize and bargain
collectively.
Regarding the studies that we reviewed, the sponsoring states—Texas,
New Mexico, California, Tennessee, and Washington—are located in the
southern or western regions of the United States, the areas where most
privatized facilities are located (see app. IV). Regional differences, such as
cost of living, may affect comparisons with other states or regions. For
example, the Washington state study adjusted Tennessee’s operational
costs upward by 20 percent to account for regional cost-of-living
differences between the two states.
By definition, prison overcrowding1 occurs when the number of inmates
actually incarcerated exceeds the rated capacity2 of the correctional
facility. As table III.1 shows, the extent of overcrowding (if any) varied
among the five states studied—at the time of initial privatization of
selected correctional facilities within the respective state.3 For example,
on January 1, 1991, California’s prisons held 94,050 inmates, which was
41,352 inmates (or 78.5 percent) above the total rated capacity (52,698). In
comparison, at that time, the average state correctional system was
operating at 12.6 percent above rated capacity.
Texas was not reporting overcrowding on January 1, 1989. However, most
states, including Texas and the other states studied, have been involved, at
some point, in litigation challenging various conditions of confinement,
such as overcrowding, in their prisons. The following general descriptions
are examples of prison litigation that have occurred in the states studied:

1

Overcrowding, in and of itself, is not a violation of the Eighth Amendment’s prohibition of cruel and
unusual punishment unless specific effects flowing from that condition form the basis for a violation.
See Hoptowit v. Ray, 683 F.2d 1237, 1249 (9th Cir. 1982).
2

The term “rated capacity” refers to the maximum number of beds or inmates assigned by a rating
official to institutions within the jurisdiction.
3

At the time of its study, the state of Washington had no privatized prisons. Thus, for the purposes of
table III.1, we present data for Louisiana rather than for Washington. The Washington study included a
comparative evaluation of three multicustody prisons (two private and one state-run) in Louisiana, as
discussed in appendix II.

Page 36

GAO/GGD-96-158 Private and Public Prisons

Appendix III
Factors That Could Affect Interstate
Comparisons of Prison Costs And/or Quality
of Service

•

•

•

•

•

In Texas, the Ruiz v. Estelle4 line of decisions includes a 1980 ruling, which
found that various conditions (such as overcrowding and inadequate
sanitation, recreational facilities, and health care) within the Texas
Department of Corrections violated the U.S. Constitution. Thus, the court
appointed special masters and monitors to supervise the implementation
of and compliance with its decree.
In New Mexico, the Duran v. Apodaca5 line of cases includes a 1980
consent decree that contained mandatory and prohibitive injunctions
relating to conditions and practices at the state’s penitentiary. Among
other subjects, the consent decree addressed living conditions, medical
and mental health care, and inmate discipline.
In California, various court decisions in the 1980s addressed segregation
procedures, double-celling, and other conditions of confinement at several
prisons located in northern areas of the state.
In Tennessee, the 1982 Grubbs v. Bradley6 decision found that certain
practices and conditions of confinement at the state’s adult penal
institutions were unconstitutional.
In Louisiana, in the 1977 Williams v. Edwards7 decision, the court held that
conditions at the state penitentiary at Angola violated the U.S.
Constitution and certain state laws.
Another factor that could be considered in making interstate comparisons
of correctional facilities is the extent of ACA accreditation. Obtaining ACA
accreditation signifies that a facility has met minimum standards. ACA
officials told us that, at the time of the respective state’s initial
privatization efforts, Texas and Louisiana had no ACA-accredited facilities;
but New Mexico, California, and Tennessee had “some” accredited
facilities. For these latter three states, the ACA officials were unable to
specifically quantify the number of accredited facilities that existed when
the respective state began its privatization efforts. However, the officials
were able to tell us that, as of 1989, the state-run women’s facility in New
Mexico was not accredited by ACA.

4

503 F. Supp. 1265 (S.D. Tex. 1980), aff’d in part and vacated in part, 679 F.2d 1115, amended in part,
688 F.2d 266 (5th Cir. 1982), cert. denied, 460 U.S. 1042 (1983).
5

C.A. No. 77-721-C (D.N.M. July 14, 1980).

6

552 F. Supp. 1052 (M.D. Tenn. 1982).

7

547 F.2d 1206 (5th Cir. 1977).

Page 37

GAO/GGD-96-158 Private and Public Prisons

Appendix III
Factors That Could Affect Interstate
Comparisons of Prison Costs And/or Quality
of Service

The degree of punitiveness of a corrections system, as reflected, for
example, by the system’s incarceration rate,8 may affect operating and
programming approaches, and therefore, expectations of service. States
with higher incarceration rates tend to be found in the west and south,
although there is variation within regions. On December 31, 1988, Texas
and New Mexico had incarceration rates of 240 and 180, respectively,
while the national average rate for state institutions was 227. On
December 31, 1990, the incarceration rates in California (311) and
Louisiana (427) were higher than the national average rate for state
institutions (227); Tennessee’s incarceration rate, however, was lower
(207). Furthermore, the Washington study found differences in
programming (e.g., education and work programs) when comparing
Washington with Tennessee and Louisiana.
Both opponents and proponents of privatization have suggested that active
correctional employees’ unions can affect whether a state decides to
privatize corrections; for example, union agreements with the state may be
a disincentive to privatization. According to the results of an April 1994
ACA survey of state adult correctional departments, the public employees
of such organizations in the five states studied shown in table III.1 had the
right to organize. However, the percentage of such correctional employees
represented by unions in two of the states was 20 percent or below and
one state did not provide data; in three states, the correctional employees
could not bargain collectively; and in none of the states did the
correctional employees have the right to strike.
Table III.1: Factors That Could Affect
Comparisons With States Studied

States studied
Texas

New
Mexicoa

California

Tennessee

Louisiana

1989

1989

1991

1991

1991

South

West

West

South

South

Rated capacity

41,252

2,878

52,698

8,700

15,006

Inmate population

39,525

2,932

94,050

8,380

13,849

320

1,157

Factors
Year of initial
operation of
private facilitiesb
Geographic
regionc
All state facilitiesd

Under or (over)
capacity

1,727

(54)

(41,352)

(continued)

8

As defined earlier, the rate of incarceration is the number of sentenced prisoners (per 100,000 resident
population) in correctional facilities.

Page 38

GAO/GGD-96-158 Private and Public Prisons

Appendix III
Factors That Could Affect Interstate
Comparisons of Prison Costs And/or Quality
of Service

States studied
Factors
Status of ACA
accreditation of
state correctional
facilitiese
Incarceration ratef
Unions

New
Mexicoa

California

Tennessee

Louisiana

none
some
accredited accredited

some
accredited

some
accredited

none
accredited

Texas

240

180

311

207

427

g

Right to
organize

yes

yes

yes

yes

yes

Right to bargain

no

yes

yes

no

no

Right to strike

no

no

no

no

no

not
applicable 20 percent

83 percent

55 percent

12 percent

Percent of
correctional
employees
represented

a
The New Mexico data apply to all state facilities, while the New Mexico study (see app. II)
focused on women’s correctional facilities.
b

The year that inmates were first sent to privatized facilities in the respective state.

c
The Bureau of Justice Statistics groups all 50 states and the District of Columbia into four
regions—northeast (CT, ME, MA, NH, NJ, NY, PA, RI, VT), midwest (IL, IN, IA, KS, MI, MN, MO,
NE, ND, OH, SD, WI), south (AL, AR, DE, D.C., FL, GA, KY, LA, MD, MS, NC, OK, SC, TN, TX, VA,
WV), and west (AK, AZ, CA, CO, HI, ID, MT, NV, NM, OR, UT, WA, WY).
d

Texas and New Mexico data are for January 1, 1989. California, Tennessee, and Louisiana data
are for January 1, 1991.
e
ACA provided us information on the general accreditation status of state facilities in each state
studied at the time of privatization; specific numbers were not available.
f
Incarceration rates for Texas and New Mexico are for December 31, 1988. Incarceration rates for
California, Tennessee, and Louisiana are for December 31, 1990.
g

Responses shown are from an ACA-sponsored survey of state adult correctional departments
(April 1994) and refer only to public, state-level employees. Regarding the percentage of
correctional employees represented by unions in Texas, the term “not applicable” was not
defined. The data refer only to adult correctional departments.
Source: Data on geographic region and incarceration rates are from Kathleen Maguire and Ann L.
Pastore, eds., Sourcebook of Criminal Justice Statistics, 1993, Department of Justice, Bureau of
Justice Statistics, 1994. Data on overcrowding are from the Corrections Yearbook, 1989 and
1991, Criminal Justice Institute, Inc., 1989 and 1991. Data on accreditation status and union
rights and representation are from ACA.

Page 39

GAO/GGD-96-158 Private and Public Prisons

Appendix IV

Use of Private Corrections in States

As table IV.1 shows, as of March 1996, a total of 47 private correctional
facilities (secure facilities for adults) were being operated or being
planned for operations by private companies in various states. These 47
private correctional facilities are located in 12 states. However, the most
use (actual or planned) of privatized correctional facilities is in 3
states—Texas, with 21 facilities; Florida, with 7 facilities; and California,
with 5 facilities.
Table IV.1: Private Adult Secure
Correctional Facilities Operating or
Planned, as of March 1996
State

Number of private
facilitiesa

Security
classification of
facilities

Total rated
capacityb

Arizona

2

minimum to medium

Colorado

1

medium

California

5

minimum

1,446

7

minimum/medium to
medium

4,636

Kentucky

3

minimum

1,300

Louisiana

2

medium/maximum

2,948

Mississippi

2

medium

2,034

New Mexico

1c

all security levels

Tennessee

1

medium

Florida

Texas
21
Utah
Virginia
Total

minimum to
maximum

1

minimum/medium

1

medium

47

850
752

322
1,506
15,702
400
1,500
33,396

Note: According to the author of the data, the information presented is subject to change and
represents the number of actual or planned facilities at a particular point in time only. We did not
verify the accuracy or completeness of the information.
a
The information presented includes only state-sponsored private facilities that primarily house
inmates from the sponsoring state’s correctional system.
b

The term “rated capacity” refers to the maximum number of beds or inmates assigned by a rating
official to institutions within the jurisdiction.
c

This is an all-women’s facility.

Source: Charles W. Thomas, Private Corrections Project, Center for Studies in Criminology and
Law, University of Florida, Private Adult Correctional Facility Census, 9th ed. (Gainesville, FL:
Mar. 1996).

Page 40

GAO/GGD-96-158 Private and Public Prisons

Appendix V

Major Contributors to This Report

General Government
Division, Washington,
D.C.

Douglas M. Sloane, Assistant Director, Design, Methodology, and
Technical Assistance Group
David P. Alexander, Senior Social Science Analyst
Barbara A. Stolz, Evaluator-in-Charge
Brenda I. Rabinowitz, Evaluator
Pamela V. Williams, Communications Analyst

Office of the General
Counsel, Washington,
D.C.

Geoffrey R. Hamilton, Senior Attorney

Dallas Field Office

Danny R. Burton, Assistant Director
Steve D. Boyles, Evaluator
Donna B. Svoboda, Evaluator

(182827)

Page 41

GAO/GGD-96-158 Private and Public Prisons

Ordering Information
The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
following address, accompanied by a check or money order
made out to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.
Orders by mail:
U.S. General Accounting Office
P.O. Box 6015
Gaithersburg, MD 20884-6015
or visit:
Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC
Orders may also be placed by calling (202) 512-6000
or by using fax number (301) 258-4066, or TDD (301) 413-0006.
Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information on
how to obtain these lists.
For information on how to access GAO reports on the INTERNET,
send an e-mail message with "info" in the body to:
info@www.gao.gov
or visit GAO’s World Wide Web Home Page at:
http://www.gao.gov

PRINTED ON

RECYCLED PAPER

United States
General Accounting Office
Washington, D.C. 20548-0001
Official Business
Penalty for Private Use $300
Address Correction Requested

Bulk Rate
Postage & Fees Paid
GAO
Permit No. G100