Private Prisons in the
United States
An Assessment of Current
Practice

July 16, 1998

Douglas McDonald, Ph.D.
Elizabeth Fournier
Malcolm Russell-Einhourn, J.D.
Stephen Crawford
With the assistance of:
Julianne Nelson, Ph.D.
Gerald G. Gaes, Ph.D.
Scott D. Camp, Ph.D.
William G. Saylor

55 Wheeler Street  Cambridge Massachusetts
USA  02138-1168  617 492-7100 telephone 
617 492-5219 facsimile

Contents
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Does Contracting for Prison Operations Save Money? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
Do Privately Operated Facilities Provide Better Services? . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Legal Issues Relevant to Contracting for Imprisonment . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
Implications for Federal Prisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
1.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
The Structure of This Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2.

An Overview of the Private Imprisonment Industry, Past and Present . . . . . . . . . . . . . . 4
A Short History of Correctional Privatization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Why the Private Prisons Industry Emerged . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
State and Federal Experience with Private Prisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Contracting for Secure Confinement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Reasons for Contracting or Not Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Characteristics of Privately Operated Prisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Industry Concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Where Private Prisons are Located . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Facility Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Facility Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
The Functions of Private Prisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
The Private Sector’s Experience with Medium and High-Security Prisoners . . . . . . . . . 24
Comparing Security Classifications in Louisiana and the Bureau of Prisons . . . . . . . . . 27
Inmate Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Expected Demand for Prison Space and Plans for Contracting with
Privately Operated Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3.

Does Contracting for Prison Operations Save Money? . . . . . . . . . . . . . . . . . . . . . . . . . .
Privately and Publicly Operated Facilities May Not Be Sufficiently Comparable . . . . . . . .
Inconsistent Accounting Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Different Treatment of Capital Spending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dispersed Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Assigning Overhead Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Studies Comparing Privately and Publicly Operated Correctional Facilities . . . . . . . . . . . . .
State Prisons in Tennessee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adult Prisons in Louisiana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adult Prisons in Florida . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adult Prisons in Texas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Arizona’s Prison for Men and Women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Abt Associates Inc.

Private Prisons

33
34
35
35
36
37
37
38
40
41
43
45
46

4.

5.

How Well Do Privately Operated Prisons Perform? . . . . . . . . . . . . . . . . . . . . . . . . . . . .
What Does “Performance” Mean With Respect to Prisons, Public or Private? . . . . . . . . . .
What States Ask Contractors to Do . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monitoring Contractors’ Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Monitoring Performance as Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Well Do Privately Operated Facilities Perform Relative to
Government Operated Ones? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Systematic Research to Evaluate Performance of Privately Operated Prisons . . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

47
47
49
50
51

Legal Issues Relevant to Private Prisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Legality of Delegating Correctional Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Liability for Private Prison Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Employment and Labor Relations Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Inmate Labor Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Other Important Legal Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Access to Private Prison Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Private Contractor Access to Criminal History Records . . . . . . . . . . . . . . . . . . . . . . . . .
Bankruptcy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Use of Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Environmental Law Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Regulating Interjurisdictional Private Prison Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . .
Legal Dimensions of Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57
57
59
61
62
62
63
63
64
65
66
66
67

53
54
56

Appendix 1: Comparing Public and Private Prison Costs, by Julianne Nelson, Ph.D. . . . . . . . . 1
1 Preliminary Findings and a Basis for Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 The Tennessee Experience: The 1995 Report of the Fiscal Review
Committee (supplemented with data from the Washington State Legislative
Budget Committee) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Operating Cost Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Per Diem Estimates Reconciled . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Spending Patterns Compared . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 The Louisiana Experience: Archambeault and Deis (1996) . . . . . . . . . . . . . . . . . . . . . . . 8
Consistency and Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Appropriateness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Intertemporal Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Interpreting Regression Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
The Net Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4 The Florida Experience: (i) The Florida Department of Corrections ;
(ii) The Florida Correctional Privatization Commission; and (iii) The Florida
Office of Program, Policy Analysis and Government Accountability . . . . . . . . . . . 13
5 Experience with Privatization Elsewhere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
The Texas Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
The U. K. Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6 Recommendations for Further Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Abt Associates Inc.

Private Prisons

Appendix 2: The Performance of Privately Operated Prisons: A Review of Research,
by Gerald G. Gaes, Ph.D., Scott D. Camp, Ph.D., and William G. Saylor . . . . . . . . . . . . 1
Massachusetts and Kentucky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
California Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Tennessee Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Washington State Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Review of Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Qualitative Observations of Operations in Louisiana and Tennessee . . . . . . . . . . . . . . . . . . 11
Arizona Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Louisiana Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
New Mexico Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
The Probable Bias in Institution Performance Indicators Based on Staff Perceptions . . . . . 24
Interpreting Institution Performance Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Putting the Public-Private Comparison in Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Other Methodological Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Florida Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Summary of Existing Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Outline of How the Taft Evaluation Meets the Criteria of a Sound Evaluation . . . . . . . . . . . . . . 33
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Appendix 3: Legal Issues Relevant to Private Prisons, by Malcolm Russell-Einhorn, J.D.
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. The Legality of Delegating Correctional Services to Private Contractors Generally . . . . . . . 3
3. Liability for Private Prison Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4. Employment and Labor Relations Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5. Inmate Labor Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6. Other Important Legal Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7. Regulating Interjurisdictional Private Prison Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
8. Legal Dimensions of Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Appendix 4: State and Federal Prisoners in Government and
Privately Operated Facilities, December 31, 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Abt Associates Inc.

Private Prisons

Foreword
In the National Capital Revitalization and Self-Government Improvement Act of 1997 (Pub. L.
No.105-33 § 11201(c)(3)(A)), Congress mandated that the Attorney General "conduct a study of
correctional privatization, including a review of relevant research and related legal issues, and
comparative analysis of the cost effectiveness and feasibility of private sector and Federal, State, and
local governmental operation of prisons and corrections programs at all security levels...." In
response to this, the National Institute of Corrections issued a cooperative agreement (#98K38GIG5)
to Abt Associates Inc. to conduct the study.
We are grateful to Kathleen Hawk Sawyer, Director of the Federal Bureau of Prisons, who facilitated
our work by asking state correctional directors to take the time to respond to our survey. We are also
grateful for the support of the National Institute of Corrections, and especially to Larry Solomon,
Deputy Director, who facilitated the cooperative agreement, and to Brenda Maynor, the project
monitor for the Institute.
This project relied heavily upon many officials who responded to our survey requesting information
about contracting practices. These include the directors of correctional systems in 48 states, in Puerto
Rico, the District of Columbia, and the Virgin Islands; the administrators in these agencies who
manage 91 different contracts; and the monitors of these contracts. All devoted a substantial amount
of time to providing us much detailed information about these contracts, their procurement and
monitoring practices, their past experience with privatization, and their future plans. We were also
assisted by a number of people who responded to early drafts of the questionnaire, including staff
from the Bureau of Prisons; Dale Parent and Brad Snyder, colleagues at Abt Associates; and James
Austin and Darlene Grant, of the National Council on Crime and Delinquency.
The National Council on Crime and Delinquency provided us information from a survey its staff
conducted of privately operated facilities; we thank James Austin, Darlene Grant, and Owen Tulloch
for their assistance. Preliminary data were also obtained from the 1995 census of state and federal
correctional facilities, conducted by the Bureau of Justice Statistics (BJS) and the Bureau of Census.
Allen Beck at BJS facilitated this, as did the Inter-University Consortium for Political and Social
Research.
Tom Rolfs, retired Director, Division of Institutions in the State of Washington's Department of
Corrections, devoted a substantial amount of time interviewing by telephone a number of persons in
several states who responded to our survey. We are grateful to both Mr. Rolfs and to those he
interviewed, as our understanding of contracting at state government levels was enlarged by them. A
number of private correctional firms also responded to inquiries from us and provided us a substantial
amount of information about their operations.
Malcolm Russell-Einhorn's review of the legal issues associated with privatization benefitted from the
help and guidance of several individuals. These included Richard Crane, formerly Chief Legal
Counsel for the Louisiana Department of Corrections and later Vice President and General Counsel
for the Corrections Corporation of America. Mr. Crane provided his reactions and suggestions at
several stages in the work, including reviewing drafts of this document (included here in appendix 3).

Abt Associates Inc.

Private Prisons

i

Others who provided valuable assistance included William Collins, editor of the Correctional Law
Reporter, Harold Sklar of the Federal Bureau of Investigation's Criminal Justice Information
Division, and Michael McManus of the Texas Department of Criminal Justice. Mr. Russell-Einhorn
is especially grateful to four law student interns who helped with the legal research: Valerie Simons
and Charles Everage of Georgetown University Law Center, Russell Drazin of George Washington
Law School, and Steve Mercer of Washington College of Law, American University.
Four others made an especially valuable contribution to this study. Julianne Nelson, Ph.D., of NW
Partners, undertook an assessment of the best and most recent cost analyses of private and public
correctional facilities. Her report, included here as appendix 1, was supported by a separate
cooperative agreement from the National Institute of Corrections. We have relied heavily upon her
work in chapter 3. A similar review of studies comparing the performance of private and public
facilities was undertaken in connection with this report by Gerald Gaes, Ph.D., Scott Camp, Ph.D. and
William Saylor, all of the Federal Bureau of Prisons’ Office of Research and Evaluation. This review,
included here as appendix 2, provides such a detailed assessment of the most significant studies in the
field that we have not attempted to do more than summarize their findings in the main body of this
report.
Special thanks are owed to the staff at Abt Associates, who have worked extremely hard on this
project. This includes, most significantly, Joan Gilbert and Ruthel Watson, as well as Susan
Anderson, and Patricia Harmon. Carl Patten's reading of several sections was also valuable.
Douglas C. McDonald, Ph.D.

Abt Associates Inc.

Private Prisons

ii

Summary
This report examines the current state of practice, law, and research with respect to privately operated
prisons in the United States, at all levels of security. The body of current law relevant to privately
operated prisons is reviewed, including federal and state statutes, regulations, and case law. Current
knowledge regarding the cost and cost-effectiveness of privately operated prisons is assessed by
means of reviews and evaluations of existing research on cost and quality of performance.
The report’s focus is limited principally to contracting by state and federal governments for the
private management and operation of facilities that are analogous to secure state or federal prisons.
Although the private imprisonment industry also provides secure detention for defendants awaiting
trial, for illegal immigrants, for prisoners transitioning back into their community, and for juveniles,
such facilities are not examined here. Contracting for these latter facilities was excluded from this
study for several reasons. First, state and federal correctional agencies are responsible for sending
most of the prisoners held in privately operated facilitiesabout 80 percent of all privately held
prisoners at year-end 1997. Second, most of the studies comparing cost and performance of privately
and publicly operated facilities examine state prisons or their private equivalents, rather than local
jails and privately operated detention facilities. To survey more than 3,000 local governments to
identify their practices of contracting for privately operated detention would have required more time
and resources than were available for this study.
By 1997, the numbers of adults in privately operated secure facilities of all sortsprisons, jails, and
illegal immigrant detention centersreached about 64,000, which constituted less than three percent
of the entire United States population of confined adults. The number of facilities had grown to
approximately 140, and the industry’s revenues probably approached $1 billion.
Information about the subset of the industry that encompasses state and federal contracting for secure
imprisonment was obtained from a variety of sources, including a survey of contracting practices
conducted by Abt Associates Inc. of prison administrators in all state and federal governments, in the
District of Columbia, Puerto Rico, and the Virgin Islands. All but two jurisdictions provided
information. This survey determined that 28 of these jurisdictions had 91 active contracts at year-end
1997 with 84 different privately operated facilities. These facilities reportedly held 37,651 prisoners
at that time. The surveyed jurisdictions also reported having an additional 14,719 prisoners in still
other privately operated facilities. Many of these prisoners had been transferred to another
jurisdiction under the authority of an intergovernmental agreement, and the receiving government had
placed them in a privately operated facility with which it had a contract. Others were held in nonsecure facilities with which the governments contracted.
Ten correctional agencies were responsible for placing the vast majority of prisoners in privately
operated facilities. These were, in descending order, the Federal Bureau of Prisons, the States of
Texas, Florida, Oklahoma, Louisiana, Tennessee, California, Mississippi, and Colorado, and the
District of Columbia. The most commonly reported reason for contracting with private management
was not to reduce costs but to alleviate overcrowding in the public system and to acquire needed beds
quickly. It is not surprising, therefore, that most (sixty percent) of the privately operated facilities are
privately owned. In most places, contracting for imprisonment services was not taken at the initiative

Abt Associates Inc.

Private Prisons

iii

of the correctional agency, but was instead mandated by either the legislature or the chief executive of
the jurisdiction, typically the governor.

Does Contracting for Prison Operations Save Money?
Even though the quest for financial savings was not given by prison administrators as their primary
reason for contracting, it is no doubt a common hope among policymakers that governments will save
money by doing so. Some proponents argue that evidence exists of substantial savings as a result of
privatization. Indeed, one asserts that a typical American jurisdiction can obtain economies in the
range of 10 to 20 percent. Our analysis of the existing data does not support such an optimistic view.
Comparing public and private prisons’ costs is complicated for a variety of reasons. Comparable
public facilities may not exist in the same jurisdiction. Private facilities may differ substantially from
other government facilities in their functions (e.g., the private facility in Arizona houses men and
women, or some in Texas that are used for drug abuse treatment services or for pre-release
populations placed in halfway houses by other jurisdictions). Or they may differ in their age, design,
or the security needs of inmates housed, all of which affect the cost of staffing them. Cost
comparisons are also difficult because public and private accounting systems were designed for
different purposes; that is, public systems were not designed principally for cost accounting.
Spending to support imprisonment is often borne and reported by agencies other than the correctional
department, and computation of these costs is often difficult for lack of data. The annual costs of
"using up" the physical assets are not counted in the public sector, as capital expenditures are
generally valued only in the year that they are made, rather than being spread across the life of the
assets. Nor is the cost to the taxpayer of contracting readily apparent from tallies of payments to
contractors. Governments incur expenses for contract procurement, administration, and monitoring;
for medical costs above amounts capped by contracts; and for sentence computation, transportation,
and other activities performed by governments. Cost comparisons often fail to account for such
expenditures.
Of the approximately 140 secure confinement facilities currently in existence, or the 84 that held
active contracts with state or federal agencies at the end of 1997, only a handful have been studied to
learn if contracting is less costly to the taxpayer. Fewer still have employed reasonably strong
research designs and reported the data in sufficient detail to permit an assessment of the validity of the
findings. The results are mixed and subject to different interpretations.
Some studies report finding that contracting saved the taxpayers money; others report small
differences, if any. Some of these apparent differences may not reflect actual savings but may instead
be accounting artifacts, especially those associated with lower estimated costs of government
overhead activities. That is, signing lower overhead costs to privately operated facilities may not
result in actual savings because these costs are quite fixed. Savings will accrue to taxpayers only if
(a) overhead activities are cut back as a result of contracting (an unlikely event where only a few
facilities are contracted in a larger public system), or (b) if planned increases in overhead costs are
averted as a result of contracting. It is unclear in the studies reviewed if purported savings associated
with reduced use of government overhead functions actually resulted in less spending.

Abt Associates Inc.

Private Prisons

iv

Other apparent sources of savings in some states reflect lower spending for prisoner health care and,
perhaps, in other aspects of facility operations, including lower salaries for line staff in some
jurisdictions. In some states, payments to contractors may be offset somewhat by accounting
deductions or adjustments for contractors' tax payments to governments.
Our conclusion regarding costs and savings is that the few existing studies and other available data do
not provide strong evidence of any general pattern. Some states may be willing to pay high prices for
private imprisonment if they need the beds to solve short-term deficiencies. In other states,
expenditures for contracting may indeed be lower than for direct public provision. However, the
bottom line with respect to costs and savings is difficult to discern given the data and the assumptions
made by the analysts. Drawing conclusions about the inherent superiority of one or the other mode of
provision, based on a few studies, is premature.

Do Privately Operated Facilities Provide Better Services?
Interest in better imprisonment services was infrequently given as a reason for contracting, but the
issue of whether governments or private firms seek lower costs by sacrificing service quality is a
major concern, especially in an environment where public safety and staff and inmate lives are at risk,
and where lawsuits are common. Unfortunately, assessing the quality of imprisonment services is
controversial, in part because the various objectives government and the larger society establish for
prisons do not always suggest unambiguous measures. At a minimum, we expect prisons to meet the
standards established by the courts.
Most contracts require that privately operated facilities conform to the law, rules, and regulations that
prevail in the public correctional agency of that jurisdiction. Performance objectives are thereby
framed in procedural terms, rather than outcomes to be achieved by means of imprisonment. Contract
monitoring is highly variable from one jurisdiction to another, with about half of the contracts
involving daily on-site monitoring by government officials.
Few studies have been conducted to compare the relative performance of privately and publicly
operated prisons. Most are affected by a variety of methodological problems that severely limit the
conclusions that one can draw from them. These include insufficiently comparable public facilities
and insufficient attention to the possibility that observed differences resulted from dissimilar
populations of prisoners or dissimilar facility designs. Given these shortcomings and the paucity of
systematic comparisons, one cannot conclude whether the performance of privately managed prisons
is different from or similar to that of publicly operated ones.
With respect to public safety and inmate programming, the available data do not support definite
conclusions. Programs for inmates are especially important in prisons because they prepare offenders
to become reintegrated into society upon their release. Well designed and implemented programs
increase the likelihood that prisoners will become law-abiding citizens. The available surveys of
either privately or publicly operated facilities do not provide the information needed to compare the
quality of such programs or the extent of prisoners’ engagement with them.

Abt Associates Inc.

Private Prisons

v

Legal Issues Relevant to Contracting for Imprisonment
Whereas the legality of governments delegating correctional authority to private firms was much
debated in the 1980s, it now appears that objections to prison privatization on constitutional
delegation grounds have little force. Unless a government has absolutely no persuasive statutory
authority for entering into private prison contracts, courts will be reluctant to invalidate contacts on
delegation grounds. Only delegated rulemaking and adjudication functions that directly purport to
exercise a government power are deemed to require special constitutional due process safeguards and
to be subject to heightened judicial scrutiny. No clear case law has been developed to define with
precision how general due process standards will be applied to private prisons. These issues have
been handled in some states through legislation, and through contracts in others.
With respect to liability, private prisons are generally treated as "state actors" for purposes of civil
rights suits, so that all relevant constitutional requirements apply with equal force to private as well as
public facilities. Private prison employees do not have the advantage of "qualified immunity" shields,
nor are they protected by other governmental immunities that limit monetary damages available to
inmates suing over prison conditions. Governments cannot shield themselves entirely from liability
by means of contracting, but they can lessen their exposure by doing so. In the vast majority of suits
brought by prisoners, alleging individual rather than systemic group harms, governments will
generally not be deemed to have knowledge of the specific acts and injuries, and will not be held
responsible. Public correctional authority’s litigation costs can be kept as low as possible by requiring
that a private contractor indemnify them and have relevant public entities named as an insured on the
contractor’s comprehensive general liability insurance policy.
While the right to strike exists in privately operated prisons, the risk of such disruptions can be
minimized by private firms seeking to have employees agree, individually or collectively, to no-strike
pledges. Moreover, contractors can also insist on notification requirements that allow them to make
arrangements for assumption of certain essential responsibilities in the event of a labor action.
The possibility of bankruptcy has excited much concern regarding private imprisonment, but this has
not yet been a problem in the industry. The few exceptions involved some small firms that
speculated by building facilities in the absence of contracts with an agency. Public correctional
agencies should nonetheless seek to protect themselves against the untoward consequences of
bankruptcy by means of proper monitoring and careful contracting.
Regarding the use of force, including deadly force: the major legal issue is whether the use of force is
properly mandated by the relevant laws of the jurisdiction. Without proper enabling legislation or
contractual provisions authorizing the use of force by designated private prison officials, it is possible
that such persons and their firms could face civil and criminal liability.
Sending prisoners from one jurisdiction to privately operated prisons in another poses some special
legal questions. Several jurisdictions rely not on contracts but instead on intermediary governments
to place prisoners in private facilities. Several jurisdictions also lease bed space directly from private
contractors in foreign jurisdictions, sometimes in prisons that are built and operated on a speculative
basis. These interjurisdictional arrangements have raised questions about the sufficiency of legal
authority and regulation. For example, receiving facilities may lack legal authority to allow them to
Abt Associates Inc.

Private Prisons

vi

respond to a variety of critical situations, including escapes and other prison emergencies. Legislation
may be needed to govern these practices.
The success or failure of a private prison arrangement may depend upon the skill with which contracts
are designed and negotiated. Public authorities must give close attention to the purposes served by
contracting and the degree of specificity that they seek to build into the agreement.

Implications for Federal Prisons
Bureau of Prisons officials assert that the private sector’s experience in operating higher security
prisons or managing inmate populations with higher security needs is too limited to warrant the
privatization of such facilities in the federal system. (Correctional authorities distinguish between the
physical security of prisons and and the security needs of inmates.) To be sure, maximum security
facilities and prisoners remain almost entirely in the government-operated sector: only four percent
(1,661) of prisoners in privately operated facilities were classified as maximum security, whereas 20
percent (or 200,000) of all prisoners in government facilities were so classified.
The private sector’s experience with what the Federal Bureau of Prisons terms medium security
prisoners is more ambiguous, however. Surveyed correctional agencies reported that half of their
prisoners housed in private facilities under contract at year-end 1997 were classified as medium
security. But procedures for classifying prisoners according to their security needs vary from one
jurisdiction to the next. It is not clear that a prisoner classified as medium security in one will be
classified the same in all others. For example, a comparison of prisoners classified by Louisiana and
the Federal Bureau of Prisons suggests that the private sector’s experience in at least this one state
may not be comparable to the federal government’s practices at the medium security level.
Eighty percent of all prisoners in two privately operated facilities and in a third government operated
facility in Louisiana were classified by the state as being medium security; the remaining prisoners
were split evenly between maximum and minimum security. A random sample of prisoners from
these three facilities was picked and they were classified according to federal procedures, which uses
a four- rather than a three-tiered classification system: high, medium, low, and minimum security.
Approximately 33 percent of these prisoners would have been classified as medium-security in
federal prisons; 29 percent and 28 percent would have been classified as "low" and "minimum"
security, respectively. The proportion of high security prisoners was the same as classified by
Louisiana (10 percent). This distribution of prisoners so classified is not characteristic of mediumsecurity prisons in the federal prison system. Rather, it is closer to that found in low-security prisons.
Lacking similar data from other states, it is not possible to determine how the security needs of
inmates housed in other privately operated facilities correspond to those prisoners held in federal
facilities.

Abt Associates Inc.

Private Prisons

vii

1.

Introduction

This report examines the current state of practice, law, and research with respect to privately operated
prisons in the United States, at all levels of security. The body of current law relevant to privately
operated prisons is reviewed, including federal and state statutes, regulations, and case law. Current
knowledge regarding the cost and cost-effectiveness of privately operated prisons is assessed by
means of reviews and evaluations of existing research on cost or performance.
The focus of this report is limited principally to contracting by state and federal correctional agencies
for the private management and operation of facilities that are analogous to secure state or federal
prisons. Although several private firms operate correctional or pretrial detention facilities under
contracts with local governments that are functionally equivalent to local jails, this report does not
attempt to identify or study them. Nor does it attempt to study the practice of contracting for
detention centers designed to hold illegal immigrants pending deportation, for juvenile detention or
correctional facilities, or for privately operated non-secure institutions, such as halfway houses or
work-release facilities, in the adult or juvenile system. Reference is made to such facilities in the
historical overview of privatization in chapter two, but no sustained attention is given to them
elsewhere. Also beyond the study’s purview are contracts for services narrower than management
and operation of the entire facility, such as contracts for health care or food services.
Several sources of information were relied upon for this report.
Survey of state and federal correctional agencies. To obtain the information needed to describe

the breadth and variety of current practice, a mailed survey was conducted by Abt Associates, Inc., in
FebruaryMarch, 1998, of all correctional agencies in state governments, the federal government (the
Bureau of Prisons), the District of Columbia, the Commonwealth of Puerto Rico, and the U.S. Virgin
Islands. The questionnaire sought information about a wide range of issues. The response rate was
high: all questionnaires were returned with the exception of those from two states. The findings of
this survey are reported below.
Survey of private facility operators. A complementary survey was conducted by the National

Council on Crime and Delinquency (NCCD) during early 1998. This survey was sent by mail to the
firms that manage privately operated facilities in the United States; information was returned for
approximately half of those facilities in existence at the time of the survey. NCCD provided Abt
Associates with the applicable findings of the survey for this report.
Surveys by other private and federal agencies. Information about privately operated facilities was

also obtained from the Center for Studies in Criminology and Law at the University of Florida.
Professor Charles Thomas and his associates in Florida publish an annual census of private prisons
and jails in the United States and abroad.1

1

The most recent is Charles Thomas, Diane Bollinger, and John Badalamenti, Private Adult Correctional Facilities Census, Tenth
Edition (Gainesville, FL: Center for Studies in Criminology and Law, University of Florida, 1997). Preliminary results from the
1998 survey (which provides data for 1997) were obtained from the center’s website (http://web.crim.ufl.ed/pcp).

Abt Associates Inc.

Private Prisons

1

Research studies. During the past decade, a number of analysts have studied certain privately

operated facilities, comparing their cost and/or performance to certain publicly operated ones. These
studies, and their findings, are reviewed and assessed.
Review of contracts, requests for proposals, and other documents. In addition to information

from the surveys, documents pertaining to specific contracting arrangements in a selected number of
governments were examined. These include, variously, the requests for proposals that specify the
scope and nature of the privatization project, the resulting contracts, formal requirements for
monitoring contractors’ performance, audit and monitoring reports, and the like.
Interviews. A variety of interviews were conducted with officials in state governments to clarify

information reported to Abt Associates in the mail survey described above.
Review of relevant law. The body of law relevant to private prisons was reviewed, including federal

and state statutes and regulations, as well as case law. In addition, a wide range of pertinent legal
articles was consulted.

The Structure of This Report
A brief overview of the private corrections industry in the United States is provided in chapter two.
This includes: a short history of the industry’s development and the reasons for it; the industry’s
growth; the prevalence of privately operated facilities, their size, types of prisoners held, scope of
services provided, and various other characteristics; and a discussion of the prospects for future
growth in the industry.
It also examines the various arrangements by which private correctional services are obtained by
governments. This includes, among other arrangements: contracting with private firms to design,
build, and operate privately owned facilities; contracting with private firms to manage and operate
government-owned facilities; "renting" beds in privately operated and/or owned facilities on a per
diem or similar basis; and using intergovernmental agency agreements to transfer prisoners to other
jurisdictions, wherein they are then placed in private facilities. Chapter two also discusses the legal
authorization for these various relationships, some common differences in contractual structures,
payment provisions, and monitoring activities, and other topics.
State correctional administrators sometimes chose not to report non-contractual arrangements for
housing prisoners in private facilities, because they were not the contracting party. (Such situations
occur when a state sends prisoners to another government correctional agency, using the vehicle of
the intergovernmental agency agreement, and the receiving state then places the prisoners in a
contracted facility.) Although the survey attempted to identify such arrangements, we did not
generally obtain much information about them.2 The surveyed agencies did provide us with the
numbers of prisoners held in privately operated facilities under such agreements, however.

2

Correctional agencies were not reluctant to provide these data. Rather, the Abt Associates survey failed to request specific
information about them, as our questionnaire asked about contractual relationships.

Abt Associates Inc.

Private Prisons

2

Consequently, the analysis that follows is generally limited to facilities that operated under contract
with state or federal governments.
What is known about the costs and savings associated with privately operated prisons is examined in
chapter three. A number of studies are reviewed and their findings discussed. The data in a selected
number of studies are reanalyzed for the purpose of clarifying the differences in expenditures for
privately and publicly operated correctional facilities.
How the performance of privately operated facilities is measured and assessed by contracting
agencies is examined in chapter four. Contracts vary in the extent to which they require performance
achievement. The Abt Associates survey collected information about whether and how contracts
attempt to bolster performance. Few contracts have been evaluated for cost and/or performance
achievement. However, there is some research literature regarding measured performance of private
and public prisons. The methods and findings of eight such studies are discussed.
The principal legal issues associated with privatization are reviewed in chapter five. It addresses both
the legality of delegating the incarceration management function to private entities and the impact that
this delegation has on the liability exposure of both private prison contractors and public correctional
authorities.

Abt Associates Inc.

Private Prisons

3

2.

An Overview of the Private Imprisonment
Industry, Past and Present

This chapter offers a thumbnail portrait of correctional privatization. Its focus is generally on the
practice of contracting with state and federal correctional agencies to hold prisoners, mostly
convicted, who would otherwise be in government operated prisons. It begins with a brief history of
the industry since the mid-1960s and a discussion of why private prisons emerged. It then describes
several different constellations of public/private involvement in owning and operating prisons, stateto-state variation in experience with privately operated prisons, reasons why some state or federal
agencies have chosen to contract or not to contract, where privately operated facilities are located,
their size, their experience with housing higher-security prisoners and with operating programs for
prisoners. Finally, correctional agencies’ plans regarding growth and future privatization are
discussed.

A Short History of Correctional Privatization
The phenomenon of private prisons and jails came into the public eye in the mid-1980s, when the
fledgling Corrections Corporation of America (CCA) offered to take over the entire State of
Tennessee’s troubled prison system, with a 99-year lease from the state, for which it would pay $250
million dollars. CCA would, in return, house the state’s convicted prisoners for a negotiated per diem
payment. Moreover, it would guarantee that the prisons would meet the standards set by a federal
judge, who had earlier found the state’s correctional system to be in violation of the U.S. Constitution
because of the conditions of confinement in its prisons. The state refused, but the offer ignited
widespread press attention and public debate.
Despite the apparent novelty of the idea, privately operated correctional facilities were not new in this
country.3 Private imprisonment had been common in earlier centuries in both England and the United
States. By the beginning of this century, however, governments nearly everywhere had assumed
responsibility for imprisonment and most other criminal justice functions.4 By mid-century, the
notion that governments were responsible for the administration of justice, and especially
imprisonment, had become so well entrenched that many argued that imprisonment is an "intrinsic" or
"core" function of government.5 But even though state responsibility for imprisonment was well

3

A longer discussion of this history can be found in Douglas McDonald, "Private Penal Institutions," in Michael Tonry (ed.), Crime
and Justice: A Review of Research (Chicago, IL: University of Chicago Press, 1992); "Introduction," in Douglas McDonald (ed.),
Private Prisons and the Public Interest (New Brunswick, NJ: Rutgers University Press, 1990); and Aric Press, "The Good, the Bad,
and the Ugly: Private Prisons in the 1980s," in McDonald (ed.), Private Prisons, pp.19-41.

4

W.S Holdsworth, A History of English Law, Vol. 4, 3rd ed., (London: Cambridge University Press, 1922-24), p. 397; A. Crew,
London Prisons of Today and Yesterday (London: I. Nicholson & Watson, 1933), p. 50; Blake McKelvey, American Prisons: A
History of Good Intentions (Montclair, NJ: Patterson Smith, 1977), p. 197-216.

5

American Bar Association, Report to the House of Delegates (Chicago, IL: American Bar Association, February 13, 1989), p. 3;
John J. DiIulio, Jr., "The Duty to Govern: A Critical Perspective on the Private Management of Prisons and Jails," in McDonald
(ed.), Private Prisons, p. 172-177; Ira P. Robbins, The Legal Dimensions of Private Incarceration (Washington, DC: American Bar
Association, 1988), p. 44.

Abt Associates Inc.

Private Prisons

4

established, contracting continued for a variety of services associated with imprisonment, including
facility management and operation, albeit confined to various niches of the juvenile and adult
correctional systems. Private, mostly not-for-profit charities and organizations had played a long and
distinguished role in operating facilities for juvenile offenders. Indeed, the private Society for the
Reformation of Juvenile Delinquents established the first house of refuge in New York City in 1825.
By the mid-nineteenth century, private cottages had been established in several states.6 During the
1960s, the number of privately operated juvenile facilities began to grow rapidly. By the time a
national census of juvenile correctional facilities in the United States was conducted in 1989, the
number of privately operated facilities had grown to 2,167, compared to 1,100 public ones. 7 The
private facilities were very different from the public ones, however. Most were small communitybased group homes or halfway houses, whereas most of the government facilities were training
schools and detention centers.
In the adult correctional system, private firms had long been contracting with federal and state
governments to provide a variety of specific services to correctional facilities, such as food services,
maintenance, education, vocational training, health care, prison industries programs, and counseling.8
Such contracting received little attention because it did not seem to pose fundamental questions about
the state’s authority to incarcerate prisoners. Nor were questions raised when the Bureau of Prisons
began in the late 1960s to contract with private firms to operate community treatment centers,
halfway houses to which federal prisoners were transferred prior to being released or paroled. 9 These
were outside the mainstream of secure prisons.
The private sector began to approach that mainstream in 1979 when the U.S. Immigration and
Naturalization Service (INS) began contracting with private firms to detain illegal immigrants
pending hearings or deportation, some of whom had finished terms in state or federal prisons, in
secure confinement facilities. These contracts provided the seedbed for the contemporary private
imprisonment industry in the United States, as several of the now-significant players in the industry
started with them. This includes the Corrections Corporation of America (CCA), a Tennessee-based
firm that incorporated in 1983 and opened its first detention center in Houston, Texas, the following
year. Wackenhut, Inc., a long-established private security firm, entered the private imprisonment
business when it won a contract to build a detention facility outside Denver, Colorado, for the INS.
Similarly, the Correctional Services Corporation (formerly ESMOR) won a contract in 1989 to
operate a immigrant detention facility in Seattle, Washington.

6

Philip B. Taft, Jr., "The Fiscal Crisis in Private Corrections," (Part I), Corrections Magazine VIII, no. 6 (December 1982):27-32.

7

Biannual Census of Public and Private Juvenile Detention, Correctional, and Shelter Facilities (Washington, DC: U.S. Bureau of
the Census, 1989 census).

8

Camille G. Camp and George M. Camp, Private Sector Involvement in Prison Services and Operations (South Salem, NY: Criminal
Justice Institute, 1984).

9

Matthew J. Bronick, The Federal Bureau of Prisons’ Experience with Privatization (Washington, DC: Federal Bureau of Prisons,
March 1989).

Abt Associates Inc.

Private Prisons

5

These developments drew little attention, but this changed in 1985 and 1986 when governments began
to contract with private firms to operate secure facilities that functioned as county jails and state
prisons. In 1985, CCA contracted with Bay County, Florida, to operate its jail and with
Santa Fe County, New Mexico in 1986, to operate its jail. In January 1986, U.S. Corrections
Corporation opened a 350-bed prison in St. Mary’s, Kentucky to hold sentenced prisoners for the
state. At approximately the same time, a small privately operated facility was opened in rural
Cowansville, Pennsylvania, and the District of Columbia government arranged to transfer 55 inmates
from the District’s overcrowded jails to it. Their arrival created an uproar. Local residents came
together and patrolled the streets with shotguns, fearing escapes. A prison reform group in
Philadelphia learned of this and successfully petitioned the state legislature to declare a moratorium
on privately operated prisons in that state.10
These events set off a nationwide debate about the legality, propriety, and desirability of private
imprisonment. Congress held hearings in 1986; the National Institute of Justice convened a
conference; and many criminal justice professional associations took a stand. The latter included the
American Federation of State, County, and Municipal Workers (opposed), the National Sheriff’s
Association (opposed), the American Correctional Association (cautious support), and the American
Bar Association (which asked for a moratorium pending further study). 11 The American Bar
Association (ABA) study concluded that delegating operating authority to private entities posed
"grave constitutional and policy problems."12 This debate did not stop correctional privatization in its
infancy, however. By the end of 1989, there were 44 secure private facilities in this country, housing
about 15,000 prisoners.13 By the close of 1997, the total number of privately operated secure facilities
for adult prisoners in the United States had grown to 142, and the number of prisoners held in them to
64,086, according to the most recent estimates developed by Charles Thomas and his associates. 14
Comparisons between conditions prevailing in 1986 and 1996 give some measure of the rapid growth
experienced in this industry. (1996 is the most recent year for which a consistent data series is
available, developed by Professor Charles Thomas).15 The number of beds in privately operated
facilities in operation or under construction in the U.S. increased at an average annual rate of 45

10

Joint State Government Commission, Report of the Private Prison Task Force (Harrisburg, PA: General Assembly of the
Commonwealth of Pennsylvania, March 1982).

11

U.S. House of Representatives, Privatization of Corrections: Hearings before the Subcommittee on Courts, Civil Liberties, and the
Administration of Justice of the Committee of the Judiciary, House of Representatives, Ninety-ninth Congress, First and Second
Sessions on Privatization of corrections, November 13, 1985, and March 18, 1986. (Washington, DC: U.S. Government Printing
Office, 1986).; John Peterson, Corrections and the Private Sector: A National Forum (Washington, DC: U.S. Department of Justice,
National Institute of Justice, 1988), Conference Proceedings.

12

American Bar Association, Report to the House of Delegates (Chicago, IL: American Bar Association, February 13, 1989).

13

Douglas McDonald, "Private Penal Institutions," in Michael Tonry (ed.), Crime and Justice: A Review of Research (Chicago, IL:
University of Chicago Press, 1992).

14

Charles Thomas, Private Corrections Project (Gainesville, FL: Center for Studies in Criminology and Law, University of Florida,
June 29, 1998) http://web.crim.ufl.ed/pcp/.

15

Thomas, Bolinger, and Badalamenti, Private Adult Correctional Facility Census, Tenth Edition (Gainesville, FL: Center for Studies
in Criminology and Law, University of Florida, 1997). These annual census of privately operated facilities use similar rules for
including facilities in each. Generally, they include only "secure" facilities.

Abt Associates Inc.

Private Prisons

6

percent. Few of these beds were empty: the occupancy rate of all private adult facilities averaged 96
percent during 1996. In 1987, there were about 3,000 prisoners in such facilities. By 1996, the
number had soared to more than 85,000. During 1996 alone, the number of prisoners increased 30
percent.16
As a result of these trends, growth in the private correctional industry’s revenues has been explosive:
from about $650 million in 1996 to about $1 billion in 1997.17 Wall Street and individual investors
have been impressed with these growth statistics and with the apparently bright prospects for future
growth (private facilities have less than 3 percent of the "market share" of prisoners held in state and
federal prisons and in local jails in the United States).18 Stock prices of the four publicly traded firms
consequently have been bid up very high, providing these companies with substantial amounts of cash
to finance further expansion.

Why the Private Prisons Industry Emerged
The contemporary private imprisonment industry owes its emergence to several dynamics, other than
the obvious fact that entrepreneurs saw business opportunities and seized them. One was the desire of
many government correctional agencies to expand their capacity quickly. For example, the INS,
faced with the need for more beds to house illegal aliens, turned to private firms to design, build, and
operate detention facilities. This could be done quickly, using funds budgeted for detention
operations, rather than waiting months or years for a capital appropriation to be requested and
approved.
For cities and counties, contracting for private imprisonment resolved certain problems peculiar to
local governments. In Bay County, Florida, for example, the county commissioners were displeased
with the pace at which the independently elected sheriff was making improvements to the local jail,
which had been found by the state to be in violation of established standards. The commissioners
turned to the Corrections Corporation of America, which promised to make the necessary
improvements quickly and to get the state off the county’s back within a matter of months. In
Santa Fe, New Mexico, the county had built a larger jail than it needed, resulting in a higher per
prisoner expenditure than anticipated. County officials decided that it would be more economically
advantageous to contract with a private firm to operate the facility and to pay only for the space used,
leaving the rest of the facility to be used for housing prisoners in other "markets": for example,
prisoners from other counties that lacked sufficient bed space or federal prisoners being transported
by the U.S. Marshals Service.

16

C.W. Thomas, Private Adult Correctional Facility Census (1997).

17

1996 data from Nzong Xiong, "Private Prisons: A Question of Savings," The New York Times, July 13,1997; 1997 data from Alex
M. Singal and Raymond F. Reed, An Overview of the Private Corrections Industry: Industry Analysis, (Baltimore, Md: Legg Mason
Wood Walker, Inc., July 16, 1997), p. 12. Singal and Reed estimate the "revenue power" of the industry to have been about $1
billion in mid-1997; we have not attempted to determine the actual revenues received by year-end.

18

On June 30, 1997, a total of 1.2 million persons were held in state and federal correctional facilities, and 570,000 in local jails. As
discussed below, the private industry’s share of the state and federal market is slightly larger, about 4.2 percent.

Abt Associates Inc.

Private Prisons

7

Different dynamics created business opportunities for the private sector in state governments.
Beginning in 1973, the nation’s state and federal prison population began growing rapidly, and by
1986, it had almost tripled, growing more than fifteen times faster than the general population. To
accommodate the increase from 1985 to 1986 aloneabout 43,000 additional prisonersseven new
medium-sized (500-bed) prisons were needed each month.19 Governments did not build facilities
quickly enough to handle this flood of prisoners, and severe overcrowding became the norm. By
1986, all but seven states were operating their prisons in excess of 95 percent capacity; 38 were either
full or above capacity; and seven states exceeded capacity by more than 50 percent. The federal
prison system was also operating at somewhere between 27 and 59 percent above capacity.
Throughout the nation, prisoners were sleeping in hallways, day rooms, gymnasiums, sometimes even
in bathrooms, or were doubled up in small cells.
Overcrowding exacerbated another serious problem: a large proportion of the nation’s penal facilities
were outmoded and even obsolete by contemporary standards. A government survey conducted in
1983 found that half of all state and federal prisons in operation at the time were more than 35 years
old and a substantial number more than a hundred.20 Only about one-fifth of all state and federal
prisons had been accredited by the Commission on Accreditation for Corrections. The confluence of
these dynamics resulted in a spate of lawsuits challenging the constitutionality of the conditions under
which prisoners were being confined. By mid-1988, 39 states, the District of Columbia, Puerto Rico,
and the Virgin Islands were operating prisons and jails under court orders to remedy unconstitutional
conditions. Several states were even forced to release prisoners ahead of time to bring occupancy
levels down to mandated levels.
Correctional administrators found themselves in a difficult position, unable to stanch the flow of
prisoners and constrained in the ability to build more prisons quickly. Following the passing of
Proposition 13 in California in 1978, expenditure controls or revenue restrictions were placed on
many state and local governments. Federal aid to state and local governments had also been shrinking
since 1980. By 1986, the general revenue sharing program was dead, leaving many local
governments without any federal assistance. Many state governments were reaching their debt
ceilings and were unable to issue more bonds to finance prison construction. Even though voters were
supporting legislation to send more criminal offenders to prison and for longer times, they often voted
down prison construction debt proposals.
Turning to the private sector to provide new prison beds was an attractive solution to many
governments facing debt restrictions. If a private firm financed, constructed, and operated a new
prison, payments to the firm by governments for housing the state’s prisoners could be charged
against operations budgets, rather than capital budgets, thereby avoiding any need for increasing debt.
Some jurisdictions relied on a variant of this arrangement: private corporations were brought into
being, operating on behalf of a government, which would issue bonds to finance the construction of a
new prison that would then be leased to the state. Lease payments could be paid using government

19

McDonald (ed.), Private Prisons, p. 4.

20

Bureau of Justice Statistics, Prisoners in State and Federal Institutions on December 31, 1983 (Washington, DC: Department of
Justice, 1986), Table 16.

Abt Associates Inc.

Private Prisons

8

operating funds, and the state’s corrections department could operate the facility as if the state owned
it outright.21
To this day, the market for private imprisonment is largely confined to the provision of new beds in
state and federal prisons, rather than takeovers of existing government operated facilities. Although
privatization came onto the public stage with the Corrections Corporation of America’s bid to take
over Tennessee’s prisons in 1985, such a transfer of assets has occurred only once. Effective January
30, 1997, CCA acquired a twenty-year lease on the District of Columbia’s Correctional Treatment
Facility for which it pays $233,000 a month. CCA also has a twenty-year contract to operate the
facility.
The growth and development of the private imprisonment industry received important support from
broader political and ideological developments in the mid-1980s. On both sides of the Atlantic, in the
United States and in Great Britain, conservative governments held sway and launched concerted
attacks on the institutional structures and ideology of the welfare state. "Privatization" initiatives
gained influential proponents, although the opportunities in this country were fewer, largely because
governments here hold fewer assets that can be privatized and because a wide variety of services
funded by governments had long been delivered by private contractors. Despite this, the public
landscape in the United States was combed in search of targets for privatization or contracting, and
prisons were identified by some as promising opportunities for expanding private sector
involvement.22
For a variety of reasons, the belief emerged that contracting for services, including correctional
services, was superior to direct government provision.23 Private firms were said to be more efficient
as they are not mired in the "red tape" that encumbers public agencies, especially in procurement and
labor relations. Private managers can hire and fire without the constraints of civil services and
restrictions on creating budget lines for new employees; labor can be disciplined and reassigned with
far greater ease in the private sector, especially if labor is not unionized.
Another purported advantage of the private sector was its greater efficiency in the face of competition.
According to this line of argument, public agencies have monopolies on services, and few incentives
exist to discover and implement ways of improving efficiency. Shielded from the demands of the
marketplace, public managers may not strive for greater productivity but for maintaining their
positions by avoiding risks. Government positions and agency budgets, once established, are difficult

21

Jan Chaiken and Stephen Mennemeyer, Lease-Purchase Financing of Prison and Jail Construction (Washington, DC: U.S.
Department of Justice, National Institute of Justice, 1987).

22

President’s Commission on Privatization, Privatization: Toward More Effective Government, (Washington, DC: President’s
Commission on Privatization, 1988).

23

The arguments in support of this belief are offered in, for example: E.S. Savas, Privatizing the Public Sector: How to Shrink
Government (Chatham, NJ: Chatham House Publishers, 1982); Stuart Butler, Privatizing Federal Spending: A Strategy to
Eliminate the Deficit (New York, NY: Universe Books, 1986); James K. Stewart, "Costly Prisons: Should the Public Monopoly Be
Ended?" in Patrick B. McGuidan and Jon S. Pascale (eds.) Crime and Punishment in Modern America (Washington, DC: The
Institute for Government and Politics of the Free Congress Research and Education Foundation, 1986), pp. 365-388; President’s
Commission on Privatization, Privatization: Toward More Effective Government (Washington, DC: President’s Commission on
Privatization, 1988).

Abt Associates Inc.

Private Prisons

9

to reduce because constituencies are created inside and outside government that press legislatures for
continued funding. In contrast, competition in the private marketplace, and the risk of losing money
or going out of business, supposedly stimulates the search for increased efficiency.
Critics of privatization have challenged these beliefs, however. For example, Donahue argues that
there is little room for technological innovation in prisons because of their labor intensive nature.24
Others argue that the high priority given to maximizing profits creates incentives to minimize costs,
which may lead to reductions in service quality.25 Critics also point to examples of flagrant
overcharging by contractors to state and local governments for shoddy goods and poor services. 26
Some opponents of privatization in corrections point to the dismal experience with the convict leasing
arrangements that pervaded the South during the Civil War decades. The conditions of those
privately operated facilities were generally appalling, and the death rates in them were considerably
higher than in public prisons.27

State and Federal Experience with Private Prisons
There were a total of 142 privately operated secure facilities at the end of 1997, according to
Professor Charles Thomas’ census. The focus of this report is on a subset of these: privately operated
prisons, rather than jails or detention centers. That is, we examine the experience of privately
operated secure confinement facilities that are most equivalent to secure confinement facilities in state
or federal prison systems and which contract with the correctional agencies in the surveyed
jurisdictions to provide prison space. In contrast with jails, prisons are designed to hold inmates for
longer terms and have a variety of programs for the inmates. We exclude from our purview: all
privately operated facilities that function as jails; detention centers for illegal immigrants or others;
facilities operating under contract with the U.S. Marshals Service, the U.S. Immigration and
Naturalization Service, or local governments; all privately operated non-secure facilities; and all
juvenile facilities.
To identify the subset of privately operated facilities that most closely correspond to state or federal
prisons, we mailed questionnaires to the heads of all correctional agencies in all states, the Federal
Bureau of Prisons, the District of Columbia, the Commonwealth of Puerto Rico, and the U.S. Virgin
Islands. (Hereafter, these governmental entities are referred to as "jurisdictions.") This questionnaire
had two sections. The first inquired about the agencies’ practices and plans regarding privatization.
The second section was designed to obtain information about each contract, its administration, and
monitoring. Of the 55 surveyed government agencies, 28 all but two responded to the survey.

24

John D. Donahue, The Privatization Decision (New York, NY: Basic Books, 1990).

25

American Bar Association, Report to the House of Delegates, unpublished document (Chicago: American Bar Association), p. 4.

26

American Federal, State, County, and Municipal Employees (AFSME) AFL-CIO, Passing the Bucks: The Contracting Out of Public
Services. (American Federation of State, County and Municipal Employees, AFL-CIO, 1984).

27

George Cable, "The Convict Lease System," in The Silent South (Montclair, NJ: Patterson Smith, 1969).

28

Two agencies contract for private prisons in Florida: the state Department of Corrections and the Florida Privatization Commission.
Both responded to the survey.

Abt Associates Inc.

Private Prisons

10

Twenty-three states reported having contracts with private firms on December 31, 1997, to house
prisoners, as did the District of Columbia, the Federal Bureau of Prisons, and the Commonwealth of
Puerto Rico. In addition, two other states reported placing prisoners in private facilities located in
other states through an agreement between two public agencies (hereafter, an "intergovernmental
agency agreement" or "IGA"). These were reported by the two respondents as equivalent to contracts,
and we have included them in our analysis. (In the discussion that follows, we call all reported
agreements "contracts," even if this is not technically accurate in the case of these two jurisdictions.)
These 28 jurisdictions reported having a total of 91 active contracts on that date, with 84 different
private facilities. (The number of contracts exceeded the number of private facilities because some
facilities contracted with more than one political jurisdiction.) These 84 facilities held a total of
37,651 prisoners at year-end 1997.
In addition, surveyed agencies reported having active contracts on the same day with 529 communitybased facilities, such as work-release or educational-release facilities or half-way houses. For the
purposes of our survey, "community-based facilities" were defined as they are in the periodic Bureau
of Justice Statistics census of state and federal prisoners. That is, a community-based facility is one in
which 50 percent or more of the inmates are regularly permitted to depart unaccompanied for work or
study release or for other rehabilitation programming.
In addition to the 84 privately operated secure facilities that received prisoners under contract directly
from correctional agencies in the surveyed jurisdictions, others held these jurisdictions’ prisoners via
a more circuitous route. That is, some state or federal correctional agencies sent prisoners to a local
government agency using an intergovernmental agency agreement. These local governments in turn
contracted with private companies to house the prisoners. By requesting information about only those
facilities with which correctional agencies in the surveyed jurisdictions contracted, private facilities
receiving prisoners through the intermediary of a local government were not reported in our survey.29
The numbers of prisoners in these facilities can be estimated from the information provided in the
survey, however. All surveyed jurisdictions reported a total of 52,370 prisoners housed in privately
operated facilities on December 31, 1997, 14,719 more than they accounted for in the 91 contracts
they identified as being active on that date. We assume that those prisoners housed in privately
operated facilities through intergovernmental agency agreements were included in this number. We
cannot assume that all of these nearly 15,000 prisoners were held in secure facilities, however.
Because of some ambiguity in the wording of the question posed to correctional administrators, some
respondents may also have counted among these 14,719 inmates those held in non-secure communitybased private facilities.
In summary, the 53 responding agencies reported that privately operated facilities held a total of
52,370 prisoners on December 31, 1997.30 This included all prisoners under the correctional authority

29

The two exceptions to this were the two states that reported such IGA agreements as though they were contracts; their responses are
included in this report.

30

This number differs dramatically from that reported by the Bureau of Justice Statistics in its 1995 Census of State and Federal
Correctional Facilities (Washington, DC: U.S. Government Printing Office, 1997). This census identified 29 privately operated
secure confinement facilities, holding 12,534 inmates. This discrepancy does not reflect simply a two-year difference in data
collection points. Rather, it is a result of different inclusion rules in the Abt Associates and the Bureau of Justice Statistics censuses.
The former asked correctional officials to report on all facilities under contract to hold their prisoners, whereas the BJS census asked

Abt Associates Inc.

Private Prisons

11

of these surveyed jurisdictions who were held in privately operated secure facilities (whether by direct
contracts or through intermediary local governments), and some in non-secure facilities. This figure
represents 4.3 percent of the nation’s 1.2 million prisoners held by state and federal correctional
agencies.
Table 2.1 and Figure 2.1 report how each responding state’s prisoners were housedin government or
privately operated facilities, in-state or out-of-state. Most of these (36,441) were held in facilities
located within the political boundaries of the state. (This does not include the 9,951 federal prisoners
held in private facilities.) A small number (5,978) were held in facilities located in other states.
These included, most notably 1,725 prisoners from the District of Columbia held outside the district;
1,008 prisoners from Colorado held principally in Minnesota facilities; and 933 prisoners from
Oklahoma held in Texas facilities. Most of these prisoners "exported" by state or federal correction
agencies to other jurisdictions were being held in Texas, where private firms have established
contractual relationships with local county governments.
At the close of 1997, ten correctional agencies were responsible for the vast majority (41,965) of the
52,370 privately held prisoners. The Federal Bureau of Prisons reported having 9,951 prisoners, or
10 percent of its total population of 98,631 sentenced prisoners in privately operated facilities. Texas
had the second largest number of prisoners in private facilities7,223, or 5 percent of its state
prisoners.31 Other high-use states include: Oklahoma (4,588), Florida (3,877), Louisiana (3,581),
Tennessee (2,958), California (2,948), Colorado (2,390), Mississippi (2,522), and the District of
Columbia (1,927).
Although the Federal Bureau of Prisons, Florida, and Texas report having jurisdiction of the most
prisoners held in privately operated prisons, other states rely on private facilities to house larger
proportions of their state prison populations. Oklahoma, for example, housed 23 percent of its state
prisoners in such facilities on December 31, 1997, the highest percentage of any state or federal
correctional agency in the nation. The District of Columbia followed, with 22 percent of its 8,590
prisoners in privately operated facilities. Other heavy users included Colorado, with 21 percent of its
state prisoner population in private facilities, Tennessee (20 percent), Louisiana (19 percent), and
Mississippi (16 percent).

correctional officials to exclude private facilities in which fewer than 50 percent of the prisoners were from state or federal agencies.
Correctional officials may have been unable to make that determination effectively, especially because many facilities accept
prisoners from a number of different federal or state agencies and no single agency may know the total state/federal share.
31

The State of Texas contracted with private firms to house other prisoners as well but in facilities that functioned as local jails.

Abt Associates Inc.

Private Prisons

12

Figure 2.1
Number of State or Federal Prisoners in Government and Privately Operated Facilities on
December 31, 1997, By State

Source: Abt Associates survey of state and federal correctional administrators
Notes:

1. Alaska and Maine are not described in the figure; no survey response was received from these
jurisdictions.
2. Jurisdictions above the line report using private facilities to house prisoners; those below the line
do not.

Abt Associates Inc.

Private Prisons

13

Table 2.1
The Housing of State and Federal Prisoners, December 31, 1997
In-State
Facilities
Prisoners in Government
Facilities
Prisoners in Private Facilities
Total

Out-of-State
Facilities

1,162,357 (95.2%)

Total

5,840 (.5%)

1,168,197 (95.7%)

46,392

(3.8%)

5,978 (.5%)

52,370 (4.3%)

1,208,749

(99%)

11,818 (1%)

1,220,567 (100%)

Notes:

1. "Private facilities" include all those privately operated, regardless of facility’s ownership.
2. "In-state" facilities are located in the geographic boundaries of the relevant governmental unit, "out-of-state" facilities are not.
3. Excludes prisoners in privately operated facilities that function as jails in Texas, even though they are under contract with the State
of Texas.
4. Prisoners in out-of-state government facilities for the Federal Bureau of Prisons includes prisoners in a correctional facility operated
by any other governmental agency at state or local levels.
5. For this table, some responding agencies included prisoners housed in community-based facilities (such as halfway houses, home
confinement, etc.) in the reported figures, especially those describing private facilities. These facilities and prisoners are not
included elsewhere in this report. However, we were unable to disaggregate them from the figures reported here.

Source:

Computed from data supplied to Abt Associates Inc. in its survey of state and federal correctional administrators.

Contracting for Secure Confinement
The survey indicated that certain jurisdictions have far greater experience in contracting with private
firms for secure beds than others (see table 2.2). The Texas Department of Criminal Justice reported
having 14 active contracts on December 31, 1997 for secure imprisonment services (excluding
contracts with six privately operated jails). The State of Florida, either through the Department of
Corrections or the Florida Privatization Commission, reported five active contracts on that day;
Oklahoma reported seven; California reported ten.

Abt Associates Inc.

Private Prisons

14

Table 2.2
Number of Active Contracts for Secure Confinement on December 31, 1997, by Contracting
Jurisdiction
Arkansas
Arizona
California
Colorado
Florida
Georgia
Hawaii
Idaho
Indiana
Kentucky
Louisiana
Minnesota
Mississippi
Montana

2
3
10
5
5
2
3
3
2
3
2
1
3
2

Nevada
New Mexico
North Carolina
North Dakota
Oklahoma
Oregon
Tennessee
Texas
Utah
Virginia
Wyoming
Federal Bureau of Prisons
District of Columbia
Puerto Rico

1
3
2
1
9
1
4
14
1
1
1
2
2
3

Total Contracts

91

Source: Abt Associates Inc. survey of state and federal correctional administrators.

Reasons for Contracting or Not Contracting

Responding agencies that reported having contracts for secure facilities on the last day of 1997 were
asked to rank their reasons for turning to private firms. The most commonly cited reason was to
reduce overcrowding in their public prison system; fourteen of the 28 agency directors listed this as
their state’s primary objective. All but five surveyed correctional departments that were contracting
with private firms were operating their prisons in excess of 100 percent of reported capacity.32 Being
overcrowded was not a distinguishing characteristic of those jurisdictions that chose to contract with
private firms, however. Most other states also reported operating prison systems that were filled
beyond the design capacity of their prisons. The most commonly cited second reason was the need to
acquire new bed space quicklysuggesting that the private sector was able to create the needed space
faster than the public agency. The most commonly cited third reason was to gain "operational
flexibility" (see table 2.3).

32

This was computed by comparing actual prisoner population on December 31, 1997, with the reported "design" capacity in existence
on that same day.

Abt Associates Inc.

Private Prisons

15

Table 2.3
Reported Objectives, by Rank, for Contracting with Private Correctional Firms
Rank

Number (and Percent) of
States Citing this
Objective
5th-8th

1st

2nd

3rd

4th

14

2

3

3

2

24 (86%)

Speed of acquiring additional beds

2

9

4

1

4

21 (75%)

Gaining operational flexibility

1

0

8

2

5

17 (61%)

Operational cost savings

8

3

1

3

1

16 (57%)

Construction cost savings

0

6

3

4

3

16 (57%)

Improving caliber of services

1

0

0

1

9

12 (43%)

Reducing legal liability exposure

0

0

1

2

7

11 (39%)

Other

3

2

0

0

1

6 (21%)

Reducing overcrowding

Note:

One state listed several objectives without ranking them. Its responses are counted in the column on the right summing the
responses, but not in the columns indicating rank. The sum total reported here does not, therefore, always equal the total of
the ranked objectives.

Source:

Abt Associates survey of state and federal correctional administrators.

Among those not citing the need to reduce overcrowding as the principal reason for contracting, the
most common reason for contracting was the desire to save money by virtue of reducing operating
costs (8 of the 28 states). Two states reported that their primary objective was to acquire bed space
quickly: two others sought to gain operational flexibility, and two others sought to improve the
"caliber of services."
Table 2.3 also indicates the frequency with which various reasons were reported, regardless of their
rank order. These responses from correctional administrators indicate that the driving force behind
contracting for privately operated prisons is the desire to reduce crowding, to acquire additional beds
speedily, or to gain operational flexibility. Less frequently mentioned concerns were interests in
reducing the operating costs of imprisonment or saving money in construction. Reasons given least
often were improving the service quality in the prisons, and reducing exposure to legal liability.
Interestingly, contracting in most of the jurisdictions surveyed did not result from the correctional
agencies’ initiative. Of the 28 jurisdictions that were contracting with private firms in the closing
days of 1997, 11 reported that the initiatives were mandated by the legislature, five by the governor,
and three resulted from a court order or consent decree. Only seven reported taking the initiative to
contract for private prisons.33

33

This probably oversimplifies the dynamics by which privatization programs were initiated.

Abt Associates Inc.

Private Prisons

16

The Abt Associates survey asked the 25 jurisdictions that reported not having contracts with private
imprisonment firms on December 31, 1997 their reasons for not doing so; seventeen provided reasons.
Four reported not having contracts because their prison systems either already had sufficient space or
because planned construction was expected to be sufficient to handle the demand. Three others cited
legal prohibitions that prevented them from contracting; four reported that contracting was not under
consideration because of concerns about labor relations or because of labor opposition; two reported
that the issue of contracting was under study or that a decision was pending; two were not convinced
that cost savings would result; one reported that no funding was available for such contracting; and
one reported having concerns about accountability and the quality of private management.

Characteristics of Privately Operated Prisons
This section describes, in summary fashion, those privately operated prisons that were reported to be
contracting with state and federal correctional agencies at the end of 1997. This includes information
about their geographical location, ownership, size, type (e.g., general confinement, parole violator
confinement), reported levels of physical security, and numbers and proportions of prisoners
classified according to the risk they pose.
Unfortunately, we lack uniform information about all privately operated facilities. The Abt
Associates survey of state and federal correctional agencies was designed primarily to obtain
information about the correctional agencies’ contracting practices for privately operated
imprisonment. Much less information was obtained for their partners in these contractual
relationshipsthe private firms and their facilities. For this latter information, we have obtained data
from the National Council on Crime and Delinquency (NCCD), which undertook to survey private
correctional firms in early 1998, and data from the Bureau of Justice Statistics’ 1995 census.
Reliance on these two different sources of information about private facilities at year-end 1997
creates some difficulties because the Abt Associates and NCCD surveys obtained information for
different subsets of facilities. The NCCD survey collected information from six private corrections
management firms that reported information about only 43 of the facilities identified in the Abt
Associates survey. Figure 2.2 compares the numbers of facilities counted in both the Abt Associates
survey of government administrators and the NCCD survey of facility administrators, classified
according to the management firms that operated them.
For reasons not determined, information was not obtained in the NCCD survey for 41 others that Abt
Associates identified as having active contracts with state and federal agencies on the same date.
Information is missing in the NCCD survey for facilities operated by some of the smaller firms, but
also for several facilities operated by CCA and Wackenhutthe two industry leaders. Because of this
selective reporting and incomplete data collection, the data collected by NCCD cannot be considered
representative of the entire set of privately operated facilities. Readers must keep this limitation in
mind when considering these data characterizing privately operated facilities in the following pages.
These characterizations can only be considered tentative, at best.

Abt Associates Inc.

Private Prisons

17

Figure 2.2
Comparing Numbers of Privately Operated Facilities Reported in NCCD Survey and Abt
Associates Survey of State and Federal Correctional Agencies, by Management Firm

Sources: Abt Associates survey of state and federal correctional administrators; NCCD survey of private facility
administrators.

Abt Associates Inc.

Private Prisons

18

Table 2.4
Management Firms who Operate Facilities Reported in NCCD Survey and Abt
Associates Survey
Corrections Corporation of America
Wackenhut Corrections Corporation
U.S. Corrections Corporation
Cornell Corrections Inc.
Correctional Services Corporation
Management and Training Corporation
The Bobby Ross Group
Civigenics Incorporated/Fenton Securities
Alternative Programs, Inc.
Capital Corrections Resources, Inc.
GRW Corporation
CCA Prison Realty Trust
Marantha Production Company, LLC
Louisiana Corrections Services
Turning Point of Central California
Sources: Abt Associates survey of state and federal correctional administrators; NCCD survey of private facility
administrators.

Industry Concentration

Two firms—CCA and Wackenhut—dominate the market, holding 61 of all 91 reported contracts with
state and federal agencies that were active on December 31, 1997. CCA had 41 active contracts with
state or federal correctional agencies, Wackenhut 20. (This reflects the dominance of these two in
contracting with all governments generally, not just state and federal.) The third-ranking firm, U.S.
Corrections Corporation, lagged behind the two leaders with seven contracts.
Long dominated by a few big players, the industry appears to be experiencing still further
consolidation, as well as some diversification. Smaller firms are being acquired by larger ones, and
some are developing new capacities—such as drug treatment services—to augment their “core”
capabilities. Firms that have been focusing on the adult corrections market are also moving into
juvenile corrections.
Where Private Prisons are Located

Most of the privately operated prisons holding state or federal prisoners in this country are located in
the West and in the South (see table 2.5). Texas is home to the most such facilities by far—23, not
including six privately managed jails under contract with the state and some others not reported
because state or federal government agencies relied on intergovernmental agency agreements to place
prisoners in them. California is the state next most populated with such facilities, having 11 within its

Abt Associates Inc.

Private Prisons

19

boundaries. Three other states have five each (Arizona, Tennessee, and Florida). In the remaining 15
states, the District of Columbia, and in Puerto Rico, three or fewer facilities hold state or federal
prisoners.
Table 2.5
Number of Private Facilities Contracted by State and Federal Correctional Agencies,
by State of Location
Arkansas

2

Mississippi

3

Arizona

5

North Carolina

2

New Mexico

2

California

11

Colorado

4

Nevada

1

District of Columbia

1

Ohio

1

Florida

5

Oklahoma

3

Georgia

3

Puerto Rico

3

Indiana

1

Tennessee

5

Kentucky

3

Texas

Louisiana

3

Utah

1

Minnesota

1

Virginia

1

23

Total Facilities

84

Source: Abt Associates survey of state and federal correctional administrators.

Facility Ownership

Of the 84 facilities that contract with state or federal correctional agencies to house their prisoners, 34
are owned by governments (see table 2.6). Fifty others are owned by the management firms that
operate the facilities or by other private entities. Several of these latter facilities are only nominally
private, as they have been created by governments to own the facility on behalf of the government.34

34

For example, the Industrial Development Authority of Brunswick County owns Lawrenceville Correctional Center on behalf of
Brunswick County.

Abt Associates Inc.

Private Prisons

20

Table 2.6
Number of Privately Operated Facilities, by Ownership
Private Owners
Corrections Corporation of America (CCA)
Wackenhut Corrections Corporation
United States Corrections Corporation
Cornell Corrections, Inc. (CCI)
Corrections Services Corporation (CSC)
Management and Training Corporation
Civigenics Incorporated/Fenton Securities
Marantha Production Company, LLC
Dove Development Corporation
Industrial Development Authority of Brunswick Co.
Delta Correctional Authority
Appelton Prison Corporation
Louisiana Corrections Facility Corporation
Louisiana Corrections Services
Mooreland Corporation
CCA Prison Realty Trust
Turning Point of Central California
Wilkinson County Industrial Development Authority
Hardeman County Correctional Facility Corporation
Subtotal Private Owners

Facilities
16
5
6
4
3
2
1
1
1
1
1
1
2
1
1
1
1
1
1
50

Government Owners
Arkansas
Bear County, TX
Crystal City, TX
Brown field, TX
Mansfield, TX
Odessa, OK
Davidson County, TN
Florida
Frio County, TX
Hamilton County, TX
Karnes County, TX
Limestone County, TX
Mississippi
Marion County, IN
Newton County, TX
Puerto Rico
Teller County, CO
Tennessee
Texas Department of Criminal Justice
Utah
Federal Bureau of Prisons
Subtotal Government Owners

Abt Associates Inc.

2
1
1
1
1
1
1
4
1
1
1
1
1
1
1
3
1
1
8
1
1
34

Private Prisons

21

Table 2.6
Number of Privately Operated Facilities, by Ownership
Total
Note:

84
Includes only facilities holding prisoners under contracts with state or federal correctional agencies on December 31, 1997;
excludes privately operated state jails contracting with State of Texas.

Source: Abt Associates Inc. survey of state correctional administrators

Facility Size

Privately operated facilities that held state or federal prisoners under contract at the end of 1997 were
approximately the same size, on average, as all government operated prisons in this country.
Table 2.7 shows the average “rated capacity” for facilities that had active contracts with state or
federal governments on this date, as reported both by the facility administrators in the NCCD survey
and by correctional agencies administrators in the Abt Associates survey, and as well as the average
for all government operated prisons in the nation on June 30, 1995.35 (The rated capacity of a
correctional facility is assigned by a rating official and is meant to reflect the available space to house
prisoners and the ability to staff and operate the facility.)
Table 2.7
Comparing Capacities of Privately Operated and Government Operated
Prisons in the United States
Privately Operated
NCCD Survey

Abt Associates

Government
Operated

834

716

797

224 – 2,192

45 – 2,048

20 – 6,381

643

NA

811

Rated capacity
Average
Range
Average daily population

Sources: NCCD survey of private facility administrators, Abt Associates survey of state and federal correctional administrators, 1995,
Bureau of Justice Statistics census of state and federal correctional facilities.

Although the average capacity of the privately operated facilities was similar to those in the public
sector, the average daily populations differed from that reported by all government operated
confinement facilities on June 30, 1995. This suggests that privately operated facilities were less
crowded than government operated ones, but this is largely an artifact of the way questions were
asked in the survey. Respondents were asked for facility capacity at year-end, but average daily

35

1995 capacity figures computed from data provided by the Bureau of Justice Statistics, 1995 Census of State and Federal
Correctional Facilities (1997).

Abt Associates Inc.

Private Prisons

22

population throughout the year. Facilities that expanded their capacity late in the year only appear to
be underutilized, whereas they actually used their available capacity heavily. Most facilities operated
at high occupancy rates throughout the year.
In most cases (60 percent of all contracts reported in the Abt Associates survey), contractors are
allowed to determine whether single- or multiple-bed housing arrangements are used. Among the 36
contracts that specified occupancy arrangements, most allowed double bunking and dormitory
housing. Only three required single cells; two others did specify single cells for particular offenders.
One state specified (in 6 of its contracts) that its inmates may not be housed with another state’s
inmates nor with any federal prisoners unless a detailed plan for such housing is submitted and
approved in advance by the contract monitor.
The Functions of Private Prisons

General adult confinement is the usual function of state and federal prisons, whether privately or
publicly operated. According to the NCCD survey of privately operated facilities, 91 percent of those
that responded provided general confinement services of adult prisoners (Table 2.8). Eighty-seven
percent of all secure government operated facilities in operation in June 1995 were so characterized.
However, it appears the range of functions served by privately operated prisons is somewhat different
than that found in state and federal prisons operated by governments throughout the nation.
Somewhat larger proportions of privately operated prisons served as drug and alcohol treatment units
(26 percent of the privates, compared to 16 percent of government operated prisons), and as facilities
for parole violators returned to custody (16 percent compared to 7 percent ).36 In contrast,
proportionately fewer private prisons contracting with state and federal correctional agencies provide
“reception center” functions of diagnosing and classifying newly admitted prisoners, or confinement
of special inmate populations (such as those needing mental health services, those sentenced to
death, or geriatric patients).
Although this study did not address the question explicitly, it appears that governments may be
utilizing private corrections firms to fulfill particular needs in their correctional systems. Indeed,
‘gaining operational flexibility’ was the most common third-ranked objective of correctional
administrators choosing to contract with private correctional firms. It appears that correctional
systems may be achieving this flexibility by writing contracts that specify a particular inmate profile
for each facility. Thus, the type of prisoner housed in the contracted facility is restricted. Twentynine (29) surveyed contracts provide for the contractor’s right to refuse custody of individual
prisoners. When asked, contract administrators explained that the return of prisoners was usually
based on a decision that the prisoner did not meet the facility’s profile, either because of custody
classification or medical needs.
Prisoner assignments designed to select appropriate prisoners for a facility’s profile in turn affect the
typical profile of the public facilities in the system. Some state administrators and public facility
superintendents reported a higher concentration of badly behaved, unhealthy, or otherwise difficult

36

Facilities can serve more than a single function; the sum of facilities providing each of the discrete functions therefore exceeds the
total number of private and government facilities in operation on the date of the survey or census.

Abt Associates Inc.

Private Prisons

23

prisoners housed in the public facilities. For example, few private operators are being called upon to
provide facilities capable of hospital confinement. In fact, administrators for 63 contracts reported
that their contracts limited the contractor’s medical liability. The limit is set either by an annual dollar
amount ($500 -$10,000) or by specifying a length of hospital stay (48 hours - 5 days) beyond which
the agency takes over both medical and security responsibilities. (Some correctional agencies
reported paying for all offsite medical care.) This differentiation of privately and publicly operated
facilities may make sense from an administrative point of view, but it creates an obstacle to comparing
their performance.
Table 2.8
Comparing Functions Served by Privately Operated and Government Operated Facilities
Number of Private
Facilities (n=43)

Number of Public
Facilities (n= 1167)

General adult population confinement

39 (91%)

1015 (87%)

Alcohol/ Drug treatment confinement

11 (26%)

186 (16%)

Primarily for persons returned to custody (e.g.,
parole violations

7 (16%)

83 (7%)

Reception/ diagnosis/ classification

4 (9%)

155 (13%)

Work release/ prerelease

4 (9%)

137 (12%)

Medical treatment/ hospitalization confinement

2 (5%)

176 (15%)

Youthful offenders

2 (5%)

39 (3%)

0

53 (4%)

3 ( 7%)

246 (21%)

Boot Camp
Other

Figures sum to more than the total number of facilities because facilities may serve more than one function.
Source: NCCD survey of private corrections firms (1998), and Bureau of Justice Statistics, 1995 Census of State and Federal
Correctional Facilities (1997).

The Private Sector’s Experience with Medium and High-Security Prisoners

As discussed above, the private sector got its start outside of the correctional mainstream, in
community-based or less secure facilities. In recent years, however, increasing numbers of facilities
are designated as offering higher levels of security. This section examines the experience of the
private sector in managing criminals who pose high security risks. It also explores the relevance of
this experience with respect to managing federal prisoners.
Prisons differ according to their physical security. Minimum security institutions often lack secure
perimeters. Medium security prisons have secure perimeters—often two fences with a bank of razor
wire between them or, in older facilities, high concrete walls ringed with razor wire and fences.
Maximum security prisons typically have secure perimeters and guard towers, in which armed officers
are posted. There are many variations in these general configurations, and classification of the

Abt Associates Inc.

Private Prisons

24

physical security also depends on the architecture of the housing units and the procedures that are
followed inside the prison.
Some private facilities have the physical characteristics to meet ACA standards for medium and
higher security facility ratings. Among those surveyed by NCCD in early 1998, and which were
identified as contracting with state and federal correctional agencies, two-thirds (69 percent) reported
their facilities as offering medium or maximum physical security (Table 2.9). The one privately
operated maximum security prison (South Bay Correctional Facility in Florida, operated by
Wackenhut) held 866 maximum security prisoners at year-end 1997.37 It appears, then, that only one
of the 295 maximum/close/high security prisons in the United States in 1995 was operated by the
private sector.38

Table 2.9
Physical Security of Privately Operated Confinement Facilities
Number

Proportion

1

2%

Medium

29

67%

Minimum

13

30%

Total

43

100%

Maximum/Close/High

Source: NCCD survey of privately operated facilities.

Operating medium and high security prisons is not the same as managing medium and high security
prisoners, however. Almost all correctional agencies classify prisoners according to the risk that they
pose to staff, to other inmates and to the public. One common convention is to use a three-tiered
classification: minimum-security, medium-security, and maximum-security (sometimes called ‘highsecurity’ or ‘close supervision’). Many jurisdictions use procedures employing objective criteria
rather than subjective assessments to classify prisoners.
The Abt Associates survey of correctional agencies asked respondents to report the numbers of
prisoners in each of the privately operated facilities under contract with that agency, by the prisoners’
security classifications. Fifty percent of all prisoners held in privately operated prisons under contract
with these governments on December 31, 1997 were classified as medium-security; 4 percent were
maximum-security, and 45 percent were minimum/low security (Table 2.10). This compares with 39

37

Charles Thomas has also reported in his annual census that this facility is the only privately operated maximum security prison. in
the nation

38

The Bureau of Justice Statistics’ 1995 Census of State and Federal Correctional Facilities reported identifying this number in
existence on June 30, 1995. The number of government-operated high-security facilities has grown slightly since then.

Abt Associates Inc.

Private Prisons

25

percent medium-security prisoners housed in all state and federal prisons on June 30, 1995, 20 percent
maximum-security prisoners, and 33 percent minimum security.

Table 2.10
Classification of Prisoners’ Custody Levels in Confinement Facilities
Privately-Operated (1997)

All State and Federal (1995)

1,661 (4%)

201,996 (20%)

Medium

18,823 (50%)

404,256 (39%)

Minimum

16,993 (45%)

337,779 (33%)

174 (0.4%)

39,302 (4%)

Maximum/Close/High

Not Classified
Total

37,651

992,333

Notes:

The Abt Associates Inc. survey counted prisoners in custody in privately operated state and federal correctional agencies on
December 31, 1997.
The BJS census counted prisoners in custody on June 30, 1995 and includes unsentenced prisoners. Numbers do not sum to total
shown here; data computed from that reported as shown in BJS Census. The BJS census includes all prisoners in custody in
confinement facilities (i.e., not community-based) in state and federal correctional agencies. These include a small number of
privately-managed facilities.

Sources:

Abt Associates Inc. survey of state and federal correctional administrators; and Bureau of Justice Statistics, 1995 Census of State
and Federal Correctional Facilities (1997), Table 13.

To further characterize private facilities’ experience with more dangerous prisoners, surveyed
correctional administrators were also asked if they send violent offenders to the privately operated
facilities with which they contract. Administrators replied that this was done in 63 of the 91 total
contracts that were active on December 31, 1997. They did not report the number or security
classification of the violent offenders actually placed in private facilities.
The private sector’s experience with higher risk prisoners is not fully described in Table 2.10,
however, for several reasons. First, classification procedures vary from one jurisdiction to another,
and a prisoner classified as ‘medium security’ in one may be classified as minimum in another.
(Indeed, some respondents to the Abt Associates survey had difficulty conforming to our requested
categories of minimum, medium, and high security.) Second, prisoner classifications are not static.
Rather, each prisoner’s security rating may change throughout his institutional career, depending upon
his behavior, among other things. That is, prisoners may be classified as medium security, for
example, but may be housed in lower security facilities by virtue of their sustained good behavior.
Indeed, it is not uncommon for maximum security prisoners to be placed in lower security institutions
and then subsequently be reclassified to medium security (or they may simply keep their initial
designation). Prisoners classified as medium-security, therefore, may include persons who pose
substantial risk as well as those seen as posing relatively little risk. Some respondents reported that
the private facilities were used to house “better” inmates, although others reported that the most
troublesome prisoners were transferred to private facilities. It is difficult to know, therefore, whether
the medium-security prisoners held in private facilities were equivalent, as a whole, to the populations
of medium-security prisoners in the same jurisdiction’s government facilities.
Abt Associates Inc.

Private Prisons

26

Comparing Security Classifications in Louisiana and the Bureau of Prisons

The congruence between different jurisdictions’ classification procedures was explored in a
comparison of the Federal Bureau of Prison’s procedures with those used in Louisiana. State
correctional administrators reported, in the Abt Associates survey, that 80 percent of all prisoners in
three medium security facilities in Louisiana (the privately operated Allen and Winn Correctional
Centers and the government-operated facility, Avoyelles Correctional Center) were classified as
medium security. Ten percent were maximum-security prisoners, and another ten percent were
classified as minimum-security. (These were the prisoners’ current classifications, reflecting their
institutional behavior, rather than their classifications at time of entry to the prisons system.)
To determine how these prisoners would be classified in the Federal Bureau of Prisons, BOP staff
classified, using the bureau’s procedures, a random sample of prisoners from these institutions Unlike
the State of Louisiana, the bureau’s classification procedures produce a four-tiered rating: high,
medium, low, and minimum-security. Bureau staff determined that 10 percent of the Louisiana
prisoners would be classified in the federal system as high security, 33 percent as medium security, 29
percent as low security, and 28 percent as minimum security. Table 2.11 compares the numbers of
prisoners in each security level, as classified by the State of Louisiana and the facilities, and the
estimated proportions of prisoners at each level according to the Bureau of Prisons’ standards. (These
latter estimates are shown in ranges, because bureau staff estimated the classification of all prisoners
by means of a small random sample. This small sample indicates a projected proportion for the entire
population within a specified range of error.39 ) Although both the facility and the State of
Louisiana’s correctional department had classified the majority of these prisoners as medium-security,
it appears that about half of the Louisiana prisoners would not be considered medium security in a
bureau facility. Rather, they would be classified as minimum or low security.40

39

Ranges were defined as being within a 95 percent confidence level.

40

These comparisons may be distorted somewhat by the fact that the distribution of security levels reported by Louisiana referred to
prisoners’ current classification, whereas the bureau staff computed a classification that would have been made at the time of the
prisoners’ admission to the prisons. Bureau staff did not classify prisoners’ institutional behavior.

Abt Associates Inc.

Private Prisons

27

Table 2.11
Classification of Prisoners’ Security Levels in Two Privately Operated Facilities and One State
Prison, According to Louisiana Standards and to Federal Bureau of Prisons Standards
Prisoner Classification

Louisiana

Bureau of Prisons

Maximum/Close/High

10%

10% ± 4.9

Medium

80%

33% ± 7.7

----

29% ± 7.4

10%

28% ± 7.3

Low
Minimum
Notes:

Louisiana reported the current classification of all prisoners. The Bureau of Prisons used initial classification sources to classify
a sample (n=153) of prisoners from the same facilities. The range reported for the Bureau’s classification categories reflects a
95% confidence interval.

Sources:

Louisiana Department of Public Safety and Corrections; and Bureau of Prisons classification team.

This distribution of prisoners is not characteristic of medium-security prisons in the Federal Bureau of
Prisons. In all medium-security federal prisons, 63 percent of the prisoners are classified at
admission as medium custody (Table 2.12). These facilities hold very few minimum and low security
prisoners (18 percent). The security classification profile in the Louisiana prisons more closely
resembles the profile of BOP inmates in low physical security institutions (with 21 percent mediumsecurity inmates, 41 percent low-security, and 36 percent minimum-security). Thus, from the Bureau
of Prisons perspective, private sector experience, in Louisiana, at least, has been closer to what the
Bureau calls low security facilities than medium or high security ones. How Louisiana’s experience
corresponds with others states’ experience with ‘medium’ security prisoners was not determined.
Table 2.12
Distribution of Prisoners, by Their Security Classifications, in Federal Bureau of
Prisons Facilities
Physical Security Level
minimum

low

medium

high

Maximum/Close/High

.3%

2%

19%

66%

Medium

8%

21%

63%

27%

Low

22%

41%

13%

4%

Minimum

70%

36%

5%

3%

Prisoner Classification

Source:

Bureau of Prisons KI/SSS data for initial classification of prisoners in all BOP prisons of each type.

Abt Associates Inc.

Private Prisons

28

Table 2.13
Comparison of Louisiana Medium Security Profile to Bureau of Prisons Profiles
Prisoner Classification

Louisiana
Facilities

Bureau of Prisons
Facilities
low

medium

Maximum/Close/High

10.2% ± 4.9

2%

19%

Medium

33.1% ± 7.7

21%

63%

Low

28.7% ± 7.4

41%

13%

Minimum

28.0% ± 7.3

36%

5%

Notes:

Louisiana reported the current classification of all prisoners in 3 medium security facilities (2 private, 1
public).
The Bureau of Prisons used initial classification sources to classify a sample (153) of prisoners from the same
facilities. The range reported for the Bureau’s classification categories reflects a 95% confidence interval.

Sources:

Bureau of Prisons classification team and Bureau’s KI/SSS data.

In summary: the private sector has very limited experience in managing high security prisoners. Only
one maximum-security facility was being operated by a private company. Some other facilities, even
though classified as medium-security, held a number of high risk offenders. For example, one
medium-security prison—the Prairie Correctional Facility, operated by CCA, held 586 maximum
security prisoners. Twelve other privately operated medium-security facilities were reported holding
an average of 47 maximum security inmates. The private sector’s experience with managing mediumsecurity prisoners is much more substantial. Given the variability from state to state in classification
procedures, however, it is difficult to determine how many of these prisoners so classified in one state
would be similarly classified in another.
Inmate Programming

An important aspect of prison management is the extent and quality of programs designed to prepare
inmates for their release and their lawful reintegration into society. Unfortunately, the national
surveys of programming in prisons, whether public or private, do not reveal much. Both the NCCD
survey of private facilities, and the Bureau of Justice Statistics census which was used as a model,
simply ask whether programs were available and the proportions of prisoners enrolled. No measure is
used to assess quality, or even intensity of program involvement. Moreover, the NCCD survey was
returned by only half of the private facilities in existence at the end of 1997. The representativeness
of the survey’s findings is therefore unknown and we do not report them here.

Expected Demand for Prison Space and Plans for Contracting with
Privately Operated Facilities
Many jurisdictions project substantial increases in the demand for prisons beds. Several states plan to
rely upon privately operated prisons to meet at least some of this demand. If these private prison beds
are indeed contracted at the scale currently planned, it will result in significant growth for the private
prison industry.
Abt Associates Inc.

Private Prisons

29

The Abt Associates survey asked each correctional agency administrator to report the increase in
numbers of prisoners expected by the end of five years (i.e., by December 31, 2002), or by the end of
their forecasting horizon, whichever is sooner. Most (83 percent) reported projects for this five-year
period. Others reported projected increases within shorter planning horizons. For all responding
agencies, the total projected increase during this five-year period was expected to be 356,771, or
29 percent more prisoners than were in custody on December 31, 1997. Adjusting for the different
planning periods reported by agencies, this amounts to an increase of nearly 71,000 prisoners each
year.41
Anticipating these increases in demands, the responding agencies reported having definite plans to
expand the overall capacity of their systems by 240,967 beds (Table 2.14). Most of this planned
increase (85 percent of all beds)will be made in government operated facilities; agencies plan to
contract for the remaining 15 percent, or 36,752 beds.42
Correctional agencies differ widely in their plans regarding private prison beds. Twenty-six reported
planning not to procure beds in privately operated prisons (Table 2.14). Twenty-one others did have
definite plans to do so. California’s Department of Corrections plans the largest procurement of beds
in privately operated facilities: 15,000 by the year 2000. This represents 45 percent of all additional
beds to be acquired or built by the department during this period. Five other states plan to rely heavily
upon the private sector to meet the expected demand for prison beds. Four plan to rely entirely, or
nearly so, on private facilities for additional beds: New Mexico (1,800 new beds), Mississippi (500),
Connecticut (500), Idaho (96 percent of planned expansion, or 1,250), and Oklahoma (87 percent of
planned expansion, or 4,000 beds). Like California, Utah and Wisconsin plan to rely on the private
sector for about 40-45 percent of their capacity expansion needs.
Interestingly, those agencies that are currently contracting for the largest numbers of beds reportedly
plan little further utilization of privately operated prisons. Texas plans to rely entirely upon expanded
or new government operated facilities; Arizona has plans to contract for 1,000 more beds (which
represents only 18 percent of its planned expansion); Florida’s Department of Corrections reports no
plans for such contracting, and the Florida Privatization Commission reports not having any definite
plans regarding further privatization.
It is difficult to estimate how many state and federal prisoners will be held in privately operated
facilities at the end of this five-year period. However, if one assumes that the state and federal
agencies that now rely upon private facilities to house approximately 55,500 prisoners continue to do
so at this level, and that the planned expansion of nearly 37,000 beds does not replace the existing
stock of privately held prisoners, the total number of private beds for state and federal prisoners would
be nearly 93,000 by the end of 2002, or 40 percent more than were held at year-end 1997. This may
be an overestimate. The responding correctional administrators may have counted these
recompetitions or re-procurements of existing contracts in their answers to the Abt Associates survey.

41

This number may be an overestimate, as the two Florida agencies may have reported each other’s plans.

42

It is possible that agencies’ plans for relying upon privately operated facilities are even more ambitious. The survey asked about
definite plans for contracting with the private sector. Agencies that plan to rely upon intergovernmental agency agreements to
transfer prisoners ultimately to privately operated prisons may not have included these prisoners in their counts.

Abt Associates Inc.

Private Prisons

30

Table 2.14 (continued)
Table 2.14
Planned Increases in Capacities of Government and Privately Operated Facilities (as of
December 31, 1997), by State

Total Number
of Planned
Beds

Alaska

--

Government Facilities
By Year

--

Additional
Beds

Percent of
New Beds

Private Facilities
Additional
Beds

Percent of
New Beds

--

--

--

--

1,300

1999

1,300

100

0

0

683

2001

683

100

0

0

5,610

2002

4,610

82

1,000

18

California

33,593

2000

18,593

55

15,000

45

Colorado

4,078

1999

2,478

61

1,600

39

0

0

500

100

Alabama
Arkansas
Arizona

Connecticut

500

n/a

1,800

2000

1,500

83

300

17

Florida DOC

22,078

2002

22,078

100

0

0

Florida CPC*

27,025

2002

27,025

100

don’t know

0

Georgia

4,764

2000

3,264

69

1,500

31

Hawaii

2,980

--

2,980

100

0

0

Iowa

1,550

2002

1,550

100

0

0

Idaho

1,302

2005

52

4

1,250

96

Illinois

8,170

2001

8,170

100

0

0

Indiana

4,548

2002

4,448

98

100

2

Kansas

309

1999

309

100

0

0

Kentucky

2,409

2002

2,409

100

0

0

Louisiana

3,742

2002

3,026

81

716

19

--

--

--

--

Delaware

Maine

--

--

1,024

1998

1,024

100

0

0

Maryland

896

2003

896

100

0

0

Michigan

7,052

2000

6,572

93

480

7

742

2002

692

93

50

7

9,111

2002

9,111

100

0

0

Massachusetts

Minnesota
Missouri

500

2002

0

0

500

100

Montana

1,418

2002

918

65

500

35

North Carolina

4,104

2003

3,048

74

1,056

26

380

2001

380

100

0

0

1,336

2001

1,336

100

0

0

700

2001

700

100

0

0

New Jersey

1,625

1999

1,625

100

0

0

New Mexico

1,800

--

0

0

1,800

100

Mississippi

North Dakota
Nebraska
New Hampshire

Abt Associates Inc.

Private Prisons

31

Total Number
of Planned
Beds

Government Facilities
By Year

Additional
Beds

Percent of
New Beds

Private Facilities
Additional
Beds

Nevada

3,350

2002

3,350

100

New York

3,300

1999

3,300

100

0

0

Ohio

6,837

2002

4,837

71

2,000

29

Oklahoma

4,600

1998

600

13

4,000

87

Oregon

1,739

1999

1,739

100

0

0

Pennsylvania

4,321

2002

4,321

100

0

0

Rhode Island

--

--

--

0

--

--

don’t know

Percent of
New Beds

0

3,804

2002

3,804

100

0

0

624

2002

624

100

0

0

Tennessee

1,922

2001

1,922

100

N/A

0

Texas

4,200

1999

4,200

100

--

0

Utah

1,856

2003

1,056

57

800

43

Virginia

6,504

2002

6,504

100

0

0

466

2004

466

100

0

0

Washington

5,460

2006

5,460

100

0

0

Wisconsin

3,600

2003

2,100

58

1,500

42

West Virginia

5,140

2005

5,140

100

0

0

500

2000

400

80

100

20

23,553

2003

21,553

92

2,000

8

0

N/A

0

--

--

--

2,000

2002

2,000

100

0

0

62

1998

62

100

--

0

South Carolina
South Dakota

Vermont

Wyoming
Federal BOP*
District of
Columbia
Puerto Rico
Virgin Islands

240,967

Totals
Note:

204,215

84.7%

36,752

15.3%

Projected Federal BOP number does not include provision of section 11201(c)(1)(B) of the National Capital Revitalization and
Self-Government Improvement Act of 1997. Florida’s Correctional Privatization commission may be double-counting beds to be
added by the Department of Corrections.

Source: Abt Associates survey of state and federal correctional administrators

Abt Associates Inc.

Private Prisons

32

3. Does Contracting for Prison Operations Save
Money?
Even though cost containment is not the most frequently reported reason for contracting with private
firms to operate prisons (as discussed in chapter 2), it is no doubt a common hope among
policymakers that governments will indeed save money by doing so. Whether contracting for prison
operations actually saves the taxpayers’ money remains an open question, however. Professor Charles
Thomas of the University of Florida, commenting on the last decade’s experience with privatization,
states that a:
reasonable assessment of the available cost analyses lends at least qualified support to
the claims of privatization proponents that meaningful costs savings can be achieved by
contracting out. To be sure, there is little one can find in this body of evidence that
would support an expectation of massive cost savings. All other things being equal, for
example, a typical American jurisdiction could anticipate economies in the rough range
of 10 to 20 percent would be realized by privatization—perhaps at the low end of the
range for initiatives focusing on the privatization of existing facilities and at the high
end of the range for new design-finance-construct-manage projects. It would be quite
misleading to describe cost savings of this magnitude as trivial. 43
This is consistent with one assessment of the United Kingdom’s experience. Studies by Coopers &
Lybrand and Her Majesty’s Prison Service found that the operating costs of four privately operated
prisons in the U.K. were between 9 and 30 percent lower than comparable publicly managed prisons.44
Others draw more cautious conclusions from the research literature. The U.S. General Accounting
Office, reviewing the five studies deemed to have the strongest designs and methods among those
published between 1991 and mid-1996, concluded that “because the studies reported little cost
difference and/or mixed results in comparing private and public facilities, we could not conclude
whether privatization saved money.”45 Julianne Nelson’s assessment of several recent studies, in
appendix 1, generally concurs with the GAO’s.
At first glance, this controversy may seem surprising. It is reasonable to expect that one could easily
answer the question by comparing expenditures for imprisoning offenders in a public correctional
agency to the payments to private firms. Unfortunately, such comparisons are usually not so
straightforward and, indeed, are often misleading for three principal reasons. First, the facilities may
differ in ways that confound comparison of costs. Second, differences between public and private
sector accounting procedures make the very identification of comparable costs difficult. Although

43

Charles W. Thomas, Comparing the Cost and Performance of Public and Private Prisons in Arizona (Unpublished report to the
Arizona Department of Corrections, August 21, 1997), pp. 50-51.

44

Jo Woodbridge, Review of Comparative Costs and Performance of Privately and Publicly Operated Prisons, Prison Service
Research Report No. 3 (UK: HM Prison Service, Home Office, Offenders and Corrections Unit, Research and Statistics Directorate,
December 1997).

45

U.S. General Accounting Office, Private and Public PrisonsStudies Comparing Operational Costs and/or Quality of Service,
(Washington, DC: Report to the Subcommittee on Crime, Committee on the Judiciary, House of Representatives, August 1996), p.
7.

Abt Associates Inc.

Private Prisons

33

several studies comparing public and private correctional costs have sought to overcome these and
other difficulties, not all have succeeded fully. Analysis of the studies’ methods and assumptions, and
reanalysis of their findings, suggests in some instances that the purported cost advantages, if any, of
privately operated facilities may be less dramatic than reported.

Privately and Publicly Operated Facilities May Not Be Sufficiently
Comparable
Determining whether and how privatization saves the government money is difficult in many
jurisdictions because no publicly managed prisons exist that are similar enough to warrant direct
comparison with the private facilities. In New Mexico, for example, the one women’s prison in the
state’s correctional department is operated by a contractor. Finding a comparable publicly operated
facility in the state is impossible, as women’s prisons are operated quite differently than facilities for
men and generally cost more. In Arizona, the one facility for men and women is contractor-operated,
and no comparable prison exists in the Department of Corrections. Similar problems of comparability
exist for other privately managed facilities that have specialized missions, such as substance abuse
treatment programming or return-to-custody facilities for parole violators.
Nor is comparability assured if one compares non-specialized facilities having a common security
classification—two medium-security prisons for men, for example. The cost of imprisonment
generally varies according to the mix of different types of prisoners held by the facility. In general,
health care costs vary according to prisoners’ ages, and staffing costs vary according to the security
levels of the inmates. Medium-security prisons generally house a mix of differently classified
inmates, including some maximum-security ones and some classified as minimum-security, as
discussed in chapter 2. Publicly managed prisons may differ from privately managed ones in the
distribution of older and younger, healthy and unhealthy prisoners, as well as in the distribution of
high, medium, and low-risk prisoners.
The cost of prisons is also a function of their size, as larger facilities benefit from economies of scale
up to a certain point.46 How full—or overcrowded—they are also affects their cost. Other things
being equal, more heavily utilized prisons, and even more overcrowded ones, are less costly on a perprisoner basis because staffing levels in prisons are relatively fixed. Privately operated prisons are
probably less likely to be overcrowded than public ones, and thus, finding similarly utilized public
prisons in the same jurisdiction may be difficult.47
Another source of non-comparability stems from facility design. When states turn to private firms for
facility construction and operations contracts, these firms have the opportunity to build the most
labor-efficient buildings possible. Comparing the cost of operating these facilities to older and less

46

In their 1984 study, Schmidt and Witte analyzed facilities in the Federal Bureau of Prisons and concluded the "prisons are cheapest
to run when they are quite large but not behemoth." Peter Schmidt and Ann D. Witte, An Economic Analysis of Crime and Justice
(Orlando, FL: Academic Press, Inc., 1984), p. 355.

47

At a minimum, one would have to adjust cost estimates to account for such differences if utilization levels were dissimilar.

Abt Associates Inc.

Private Prisons

34

efficiently designed public prisons may lead to misleading conclusions about the reasons for an
apparent cost advantage attributed to privatization.
Finally, public and privately operated prisons may differ in the levels and quality of services provided.
In general, one would expect costs to be higher for more extensive and higher-quality programs and
services. The most telling comparison, therefore, would be between the cost of delivering an
equivalent service of equivalent quality by both a public and a privately managed facility. Measuring
the quality of imprisonment services is not a straightforward task, as chapter four discusses.

Inconsistent Accounting Procedures
The accounting procedures followed by public and private organizations differ, which makes
comparison of costs incurred by the two types of organizations difficult. Accounting methods
developed for private firms are designed to value all inputs used for producing goods and services; a
large proportion of costs incurred are thereby captured. In contrast, public accounting systems were
designed not to identify costs but to control the allocation of appropriated public funds through
agencies and to detect misuse of taxpayers’ money.48 Three characteristics of public accounting
procedures make cost identification especially difficult in government: the deficient treatment of
capital expenses, the focus on the agency rather than the service delivered, and methods for allocating
overhead costs.
Different Treatment of Capital Spending

Private firms have adopted accounting conventions to spread expenditures for capital assets
(e.g., buildings, land, equipment) throughout the period of the assets’ useful life, so that the year-toyear cost of a service includes some portion of the physical assets that are “consumed” in the
production of that service. Public accounting systems make no such attempt. In the public sector,
capital expenses are counted only in the year that they are made. Because no attempt is made to
spread these costs across subsequent years, it is impossible to determine how much it actually costs to
deliver a service without conducting an inventory and appraisal of existing capital assets. The
commonly cited costs of imprisonment by public agencies are, therefore, often lower than they would
be if physical assets were capitalized. In years where large capital construction projects are
undertaken, however, the operating costs will be overestimated as a result of “expensing” these costs
fully in that same year rather than spreading them across the useful life of the asset.
Among those facilities identified by the Abt Associates survey of correctional administrators as being
under contract with state or federal correctional agencies on December 31, 1997, 59 percent were
owned by the contracting firms. Because the fees charged by these firms for their services are
generally designed to recover capital investments in these facilities, the appropriate comparison is the
cost of ongoing operations in a similar public institution plus some annualized portion of the capital
investments by governments in this institution. Because the latter costs are, for all practical purposes,
uncounted, comparing the relative cost of private owner-operated facilities to public facilities is
subject to considerable uncertainty.

48

Herman B. Leonard, Checks Unbalanced: The Quiet Side of Public Spending ( New York, NY: Basic Books, 1986); Douglas C.
McDonald, The Cost of Corrections: In Search of the Bottom Line (Washington, DC: National Institute of Corrections, 1989).

Abt Associates Inc.

Private Prisons

35

Dispersed Costs

The practice of assuming that the cost of public imprisonment is equivalent to the correction
department’s expenditures often obscures the true cost to government for another reason as well. In
many jurisdictions, several costs of operating prisons and jails are borne not by the correctional
agency but by a number of different agencies or different government accounts. For example:
• employee benefits may be paid out of separate overhead government accounts rather than the
correctional department’s account;
• some medical care of inmates may be charged to the government’s health care agencies or may
be subsidized variously by state hospitals, mental health departments, and the like;
• utilities may be charged to the department of public works;
• legal work in support of prisons may be borne by the attorney general or equivalent;
• the true cost of insurance may be overlooked, as governments generally self-insure by paying
for unpredictable costs when they are incurred, rather than spreading these costs over all years;
• the true cost of supplies and equipment may be understated if they are purchased and stored by
a separate agency (such as the federal General Services Administration); and
• payments for contracted services (such as food or health care) may be made by central or
regional offices rather than by the facilities.
To identify all such dispersed costs of correctional operations, special studies are needed, rather than
simple reliance on measures of costs based on the correctional agency’s budget or reported
expenditures. Overlooking expenditures by other agencies or accounts may undercount actual costs
by as much as 30-40 percent. One study of New York State’s prisons estimated that only 77 percent
of the direct cost of imprisonment in the state’s prisons were covered by the department’s budget.
The remainder was paid by general government accounts or by other agencies.49
A parallel problem exists with respect to identifying all costs to government for contracting. Private
firms may not always bear all the costs of imprisonment, even when they own the buildings and all the
assets. Some services might be performed by the public correctional agency, and the cost of those
services should be counted along with the contractors’ costs. For example, surveyed government
officials who were charged with managing contracts with privately operated facilities reported that the
following types of activities were sometimes necessary to support private facilities but were paid by
government:
• contract monitoring and oversight;
• inspection and licensing;
• personnel training;

49

Douglas C. McDonald, The Price of Punishment: Public Spending for Corrections in New York. (Boulder, CO: Westview Press,
1980), p. 16.

Abt Associates Inc.

Private Prisons

36

• medical and/or dental care (contracts often limit the contractor’s liability for such care, placing
the risk for high-cost and catastrophic care on the government);
• inmate transportation;
• inmate case management;
• background checks for visitors and volunteers;
• reviews for pardon and/or parole boards;
• correctional industries programming;
• disciplinary and grievance appeals processes;
• accounting and banking of inmate funds;
• payment of inmate wages; and
• response teams for emergencies.
As with public correctional departments, agreements with private facilities must be scrutinized to
determine if these or similar costs are paid not by the facility itself but by the government agency with
which it contracts. All such costs, or estimates thereof, need to be included as costs to government of
contracting.
Assigning Overhead Costs

Some “overhead” expenditures incurred by various agencies of government in support of prisons and
prison administration must be assigned to both government operated and privately operated facilities
in a jurisdiction. The methods by which these costs are allocated affect the calculation of operating
costs of each facility and, consequently, the estimation of difference in public and private facility
costs. As Julianne Nelson writes in her paper (in appendix 1, at note 11), most analysts addressing the
question of public and private imprisonment costs have estimated the costs assigned to facilities
according to the use made of central office services by either public or private facility managers. She
argues that this tends to overstate the relative cost of public facilities. This is because the cost of
governmental overhead activities are quite fixed. That is, the marginal cost associated with increasing
a prison system by one or two more facilities is quite small, regardless of whether the additional
facilities are operated by government or a private firm. The appropriate method for allocating the cost
of these fixed overhead costs is to assign identical amounts to all facilities in the jurisdiction, whether
public or private. The only conditions in which privatization would lower overhead costs in
government would be (a) where the expansion of overhead activities were averted by relying upon
privatization or (b) where contracting for privately operated prisons resulted in cutting back
government overhead operations, which is not likely to occur.

Studies Comparing Privately and Publicly Operated Correctional
Facilities
A number of research studies have been reported since the early 1980s that aimed to compare the cost
of privately operated facilities with publicly operated ones. Most have been focused on facilities for
adult prisoners, although some have examined juvenile facilities. Only those pertaining to adult
facilities are examined here. Moreover, attention is restricted to those few studies that have employed

Abt Associates Inc.

Private Prisons

37

reasonably rigorous research designs, and which report their methods and data in sufficient detail to
assess the strength of their findings.50
State Prisons in Tennessee

In 1991, the State of Tennessee contracted with the Corrections Corporation of America (CCA) to
operate the South Central Correctional Center (SCCC), a 961-bed, multi-custody (minimum-tomaximum-security) facility. As required by state law, the Tennessee Fiscal Review Committee (FRC)
undertook to “compare the full costs of the contractor with the state’s full costs of operating similar
facilities.” 51 Two publicly operated prisons were selected for comparison: Northwest Correctional
Center (NWCC) and Northeast Correctional Center (NECC).
The FRC study is a model cost comparison. It allocated “line item” cost data for each prison to
specific management functions (administration, security, etc.). Adjustments were made to operating
costs (also reported by management function), which included netting out commissary revenues
against commissary expenses, allowing for changes in food services inventories and non-comparable
programs, eliminating depreciation expenses as a public facility expense (because there were
corresponding sets of expenses in the private facility), and adding monitoring costs to the expenses
reported for the private facility. After such adjustments, the FRC found that the daily operating cost
during FY 1994, exclusive of any costs allocated to central office, averaged $31.95 in the two public
facilities, compared to $33.78 in the CCA facility. By this measure, the private facility was 5.7
percent more costly than the two public facilities, on average. (Both public facilities were less costly:
$30.91 at NECC and $33.06 at NWCC.) When adjustments were made for differences in the size of
the facilities to equalize the comparison (more specifically, in the average numbers of prisoners under
custody in each one), the cost difference diminished to one percent. Adding in an allocated share of
the Department of Corrections’s central office costs to the three facilities changed the estimates
slightly. The CCA facility was then found to be one percent less costly, on average, than the public
facilities.
The FRC report did not directly address the question of whether or not privatization actually saves
money for the taxpayers of Tennessee. This question was addressed by the Washington State
Legislative Budget Committee (LBC) as it tackled its own version of the prison privatization

50

The studies not reviewed here include the following: A. Brown et al., Private Sector Operation of Correctional Institution: A Study
of the Jack and Ruth Eckert Youth Development Center, Okeechobee, Florida (Washington, DC: National Institute of Corrections,
1985); John D. Donahue, The Privatization Decision (New York, NY: Basic Books, 1990), Chapter 8; Keon S. Chi, "Private
Contractor Work Release Centers: The Illinois Experience," in Innovations (Lexington, KY: Council of State Governments, 1982).;
Joint State Government Commission, Report of the Private Prison Task Force (Harrisburg, PA: General Assembly of the
Commonwealth of Pennsylvania, 1987).; Douglas C. McDonald, “The Costs of Public and Private Correctional Facilities,” in
McDonald (ed.) Private Prisons and the Public Interest (New Brunswick, NJ: Rutgers University Press, 1990).; Douglas C.
McDonald, “The Costs of Public and Private Correctional Facilities,” in McDonald (ed.) Private Prisons and the Public Interest
(New Brunswick, NJ: Rutgers University Press, 1990); Charles H. Logan and Bill W. McGriff, “Comparing Costs of Public and
Private Prisons: A Case Study,” NIJ Reports (No. 216, 1989), pp. 2-8.; and Dale K. Sechrest and David Shichor, Final Report:
Exploratory Study of California’s Community Corrections Facilities (San Bernardino: California State University, 1994).

51

State of Tennessee Legislative Fiscal Review Committee, Cost Comparison of Correctional Centers ( Nashville, TN, 1995), p. 1; for
a detailed analysis of this study, see Julianne Nelson’s paper, in appendix 1 of this report.

Abt Associates Inc.

Private Prisons

38

question.52 The LBC conducted further analyses of the Tennessee data but followed a slightly
different tack. Instead of analyzing the cost incurred by CCA of operating SCCC, the LBC started
with the revenues received by CCA and adjusted these data for differences between public and private
management responsibilities. The FRC had excluded medical expenses for CCA from its analysis
because it lacked sufficient data. It was necessary to estimate these costs in order to compute CCA’s
operating surplus and to compare private per diem rates with those derived for the state-run facilities
NWCC and NECC. The LBC concluded that the estimated total daily cost to the taxpayers of the
privately operated CCA facility was $37.65, as compared to $40.16 at NECC and $40.19 at NWCC.
Julianne Nelson undertook further analysis of the Tennessee data, reported in both the FRC and the
LBC studies, to discern differences in how costs were allocated in the CCA facility compared to the
two public facilities. (This study was commissioned by the National Institute of Corrections in
conjunction with the Abt Associates study and is included in appendix 1.) She found that:
• Non-medical operating costs per inmate day were virtually identical in the public and private
prisons studied.
• At the CCA facility, more was spent on administration and less on security, relative to the
public facilities, including corporate overhead and monitoring by state officials.
• Contracting for SCCC may have saved money for Tennessee’s taxpayers by averting
expenditures for state government overhead activities. That is, the observed state-level
overhead expenses were lower than what they would have been had SCCC been publicly
managed.
• Labor costs were lower at the private facility: $16.89 per prisoner/day at the CCA facility, as
opposed to $19.63 at NECC and $20.96 at NWCC. Security staff costs were also lower: $9.96
per prisoner/day at the CCA facility, versus $12.57 at NECC and $14.55 at NWCC.
• Employee benefits were also lower at the CCA facility. The average employee benefit rates
were 28 and 29 percent at the two public facilities, and 22 percent at the CCA facility. Wide
variation existed among CCA staff: administrative employees enjoyed a benefit rate of almost
80 percent, while others received benefits at the rate of 13-14 percent of salary.
She summarized that the pattern of expenditures indicates that “the bulk of reported cost savings at the
privately managed prison studied in Tennessee can be attributed to differences in per-inmate medical
expenditures and allocated state overhead costs. (Labor cost savings were offset by the increased
administrative cost...).”
What remains unclear, however, is whether the apparently lower overhead costs are merely
accounting artifacts rather than reductions in expenditures. It is unlikely that overhead activities were

52

Washington Legislative Budget Committee, The Tennessee Experience: The 1995 Report of the Fiscal Review Committee (Seattle,
WA, 1996), appendix 3. In particular, LBC analysts both estimated the medical expenses incurred by CCA on behalf of prisoners at
SCCC and obtained an unofficial report of these expenses directly from CCA. The LBC analysts also collected data on the fees paid
by Tennessee to CCA for operating SCCC. This information was collected after the FRC report was published.

Abt Associates Inc.

Private Prisons

39

actually cut back, or that significant increases in expenditures for overhead were averted by
contracting for a single facilities. The actual difference in expenditures was therefore small, if any.
Adult Prisons in Louisiana

The State of Louisiana contracted with two different private correctional firms—the Corrections
Corporation of America and Wackenhut Corrections Corporation—to operate two prisons for men.
The former firm contracted to operate the Winn Correctional Center, the latter the Allen Correctional
Center; both opened in 1990. In 1996, William Archambeault and Donald Deis, professors at
Louisiana State University, reported the findings of their study which compared the cost and
performance of these two facilities with a third, the Avoyelles Correctional Center, operated by the
Louisiana Department of Public Safety and Corrections.53
The basic question addressed by Archambeault and Deis was how the full cost per inmate-day in a
publicly managed prison (i.e., Avoyelles Correctional Center) compared with the full per diem cost of
housing an inmate in either of two privately managed facilities. Their choice of comparison
institutions was appropriate. All three institutions were built by the state and shared a common
design. The capacities of all three were identical—1,474 beds. They are located in the same section
of rural Louisiana and draw from the same general labor pool. All three house maximum, medium,
and minimum security prisoners in approximately the same proportions. Moreover, the private
facilities appear to be well-integrated into the ongoing operations of the state’s public correctional
agency. They adopted the same rules and procedures that govern the state’s prisons, and the upper
management of these two privately operated prisons participate in agency-wide management
functions. A number of the potentially confounding factors are thereby held constant, permitting a
comparison of institutions that vary only in their being managed and operated by either the state
correctional agency or by private firms.
Archambeault and Deis compared expenditure data for all three prisons for five fiscal years, 1992
through 1996. As they were interested in assessing the total cost and savings to the taxpayers, they
based their analysis upon the state’s fees paid to the management firms and sought to identify all costs
incurred to support either the privately operated or the government operated facilities, including all
costs incurred by other agencies or other government accounts. They also sought to estimate a variety
of costs not accounted for on an annualized basis, such as self-insurance. Numerous other
adjustments were made, such as for costs associated with prison industry programs and services
provided by local vocational/technical schools. After such adjustments, Archambeault and Deis
concluded that the taxpayers of Louisiana saved an average of 12.75 percent for both private facilities
combined, throughout the five years studied, relative to what taxpayers spent for the Avoyelles facility
during the same period. The average five-year savings were greater for the Allen facility operated by
Wackenhut: 13.80 percent, as opposed to 11.69 percent at the CCA facility in Winnfield.
At approximately the same time as Archambeault and Deis were conducting their study, the State of
Washington’s Legislative Budget Committee examined expenditure data for the same three facilities

53

William G. Archambeault and Donald R. Deis, Jr., Cost Effectiveness Comparisons of Private Versus Public Prisons in Louisiana:
A Comprehensive Analysis of Allen, Avoyelles, and Winn Correctional Centers (Baton Rouge, LA: School of Social Work,
Louisiana State University, 1996).

Abt Associates Inc.

Private Prisons

40

and drew different conclusions about the savings.54 The LBC’s report was released before
Archambeault and Deis released theirs. Comparing the two reveals that they took different
approaches to estimating some costs or made different assumptions. In the end, the LBC concluded
that the state could expect to break even on its two contracts during FY 1996 when all facilities were
operating at full capacity. Whereas the CCA facility was costing about one percent more than the
state facility during that year, according to the LBC's analysts, the Wackenhut facility was saving the
state an equivalent amount. (The LBC suggested that costs of the three private and public facilities
had recently converged and that the private facilities had been costing the state approximately four
percent less during FY 1994.)
Julianne Nelson reviewed both the Archambeault and Deis study and the LBC report and identified a
number of difficulties in the former. (See appendix 1) She concluded that the savings associated with
private operation in the Winn and Allen facilities are more in the range estimated by the LBC
analysts. That is, during fiscal year 1996, the savings were more likely to be below 5 percent than
above 12 percent. Savings to the taxpayers could be greater, she argues, if contracting out operations
of facilities results in cutbacks in state-level overhead expenses.
Adult Prisons in Florida

The controversy over the appropriate measure of operating costs in Florida prisons shows how hard it
is to get all sides to agree on whether or not privatization saves money. Current state law helps
identify the participants in this debate. Chapter 89-526 of the Florida state statutes, enacted in 1989,
first authorized the state Department of Corrections (DOC) to enter into contracts with private prison
management companies. The subsequent Chapter 93-406, adopted in 1994, created the Correctional
Privatization Commission (CPC), to expedite prison privatization; it is housed in the state Department
of Management Services and operates independently of the Department of Corrections. Chapter
957.05 qualifies the authority of the CPC, requiring it to demonstrate—before any contract is
signed—that private prison management will result in at least a 7 percent savings (relative to the cost
of allowing the state Department of Corrections (DOC) to run the prison). Chapter 957.07 also
empowers the state Auditor General to decide whether or not this condition has been satisfied before
any contract with a vendor has been signed.
In the summer of 1995, two prisons opened under contract with the Correctional Privatization
Commission. One, the Bay Correctional Facility, is operated by CCA; the second, the Moore Haven
Correctional Facility, is operated by Wackenhut Corrections Corporation. Both are 750-bed prisons
that were designed for medium-security inmates, although the department has been assigning a large
percentage of minimum-security prisoners to them (about half, during FY 1997).55
To provide for ongoing independent oversight, Florida state law requires the Office of Program Policy
Analysis and Government Accountability (OPPAGA) 56 to review the performance of all management

54

State of Washington Legislative Budget Committee. Department of Corrections Privatization Feasibility Study, Report 96-2
(Olympia, WA, 1996).

55

Florida OPPAGA, Review of Bay Correctional Facility and Moore Haven Correctional Facility, Report No. 97-68 (Florida
OPPAGA, April 1998).

56

Although the OPPAGA is a unit of the Office of the Auditor General, it reports directly to the state legislature.

Abt Associates Inc.

Private Prisons

41

companies that operate private prison facilities in Florida (this may include both vendors under
contract to the CPC and those under contract to the DOC). A separate oversight board, the Florida
Corrections Commission, was established in 1994 and is “charged with reviewing the effectiveness
and efficiency of [Florida’s] correctional efforts, recommending policies, and evaluating the
implementation of approved policies.”57
Not surprisingly, these different parties do not always favor the same measure of privatization
benefits. In its Annual Report for the fiscal year 1996-97, the DOC argued that all three privately
managed adult prison facilities were more expensive than similar publicly run facilities would have
been.58 Using a somewhat different approach, the OPPAGA reached a similar conclusion.
Nevertheless, the OPPAGA report included responses from the Correctional Privatization
Commission, the Corrections Corporation of America, and Wackenhut proposing alternative
methodologies—ones that led to conclusions directly opposed to those reported by the DOC and the
OPPAGA.
The Department of Corrections’ estimates are somewhat biased against the privately managed firms.
The per diem rate for the men’s public prison is derived as the average of nine other facilities; all but
one of these public facilities is substantially larger than any of the private facilities being analyzed. 59
Given the economies of scale in prison operation, this approach understates the public sector cost of
running a prison comparable in size to Bay or Moore Haven.
Nevertheless, the OPPAGA reached a similar conclusion from a more appropriate starting place. It
compared two private men’s facilities (Bay and Moore Haven) with a single comparably-sized public
facility, the Lawtey Correctional Institution. In its analysis, the OPPAGA started with the
management fee due to CCA and Wackenhut (for running Bay and Moore Haven, respectively); it
then deducted monitoring costs paid by the management companies, property taxes not paid, and
medical insurance co-payments collected from prisoners and retained by the prison operators. For the
Lawtey facility, the OPPAGA used the annual cost of operations, health services, and education
programs reported for this facility by the Department of Corrections in its Annual Report. All three
cost estimates were then adjusted for differences in scale economies, the scope of management
responsibilities and the quality of services provided.
The OPPAGA estimated that Moore Haven was more expensive to operate than its public sector
equivalent; it also argued that the Bay facility was only 0.2 percent cheaper. Neither private facility
achieved the 7 percent cost savings required by law.

57

Commissioners are appointed by the Governor and with the approval of the state Senate. The Commission’s mission statement
appears on its web site, located at http://www.dos.state.fl.us/fgils/agencies/fcc/.

58

These estimates are reproduced and analyzed in Florida Department of Corrections, Privatization in the Florida Department of
Corrections (1998). Available at web site (time stamped April 28) www.dc.state.fl.us/administrative/reports/privatize/index.html.

59

Apart from the one public facility with an ADP of 733, the remaining public prisons housed between 1214 and 1719 inmates on
average. In contrast, the observed average daily populations for the privately operated Bay and Moore Haven facilities were 708
and 706 respectively. If these two prisons were assumed to operate at 95 percent of capacity, they would house approximately 713
inmates on average. Each of these private facilities has a contract maximum of 750 beds.

Abt Associates Inc.

Private Prisons

42

In response to this finding, CCA proposed some adjustments that substantially change the
interpretation of the data: they argue that private management at Bay and Moore Haven saved
taxpayers, respectively, 8.3 and 6.0 percent of state costs. This difference arises primarily from three
simple changes: an increase in the CCA’s credit for taxes paid to the state, an increase in CCA’s
credit for the “Inmate Welfare Trust Fund Net Revenue;” 60 and a decrease in the state’s adjustment
for “medical costs for higher medical grade inmates at Lawtey.” 61 The OPPAGA itself rejected these
three CCA amendments, observing that CCA sought to claim credit for more tax revenue than was
properly attributable to its Bay facility, 62 that it retained custody over Inmate Welfare Trust Fund
balances, 63 and that Lawtey inmates had considerably more medical needs than those at the Bay
facility.
This debate is further complicated by an adjustment that the OPPAGA could have made but ultimately
rejected. As in Louisiana, the state retirement system was not fully funded by past payroll deductions.
To eliminate this shortfall, a surcharge of 5.78 percent was included in the the payroll costs reported
for Lawtey. The OPPAGA justified counting this surcharge, stating that “it is still a cost to the
taxpayers. Privates do not pay this and thus should be able to reduce costs in comparison to state by at
least 5.78% of payroll.”64
This accounting decision biases the comparison against the public facilities, however. The unfunded
liability was incurred in the past; the surcharge was imposed to collect the funds needed to cover that
liability. To compare the current costs of operating public and private facilities, the surcharge should
not be counted. In other words, payroll costs at the Lawtey facility should be 5.78 percent lower than
what is reported by OPPAGA. Omitting this surcharge brings the cost of the public and private
facilities yet closer together, although the end result does not change the basic conclusion: there is no
strong evidence indicating that contracting for these facilities rather than direct government operations
saved taxpayers' money in Florida.
Adult Prisons in Texas

In 1987, the Texas State Legislature enacted a law authorizing the Texas Department of Corrections to
contract with private vendors and county commissioners to finance, construct, operate, maintain or
manage correctional facilities. In June of that year, a request for proposals was issued offering the
opportunity to operate four 500-bed prisons, to be structured as pre-release centers, with programming
designed to reintegrate prisoners back into society upon their release. Such programming would

60

This trust fund is an account containing the revenue (in excess of expenses) from the prison commissary, vending concessions and
inmate telephone service.

61

A fourth small adjustment by CCA remains unexplained: the "unadjusted" per diem reported by CCA for Lawtey exceeds that
found in the OPPAGA report by $.36, OPPAGA, p. 38.

62

According to the OPPAGA report (p. 59), CCA sought to claim credit for taxes paid on income from other facilities.

63

According to the OPPAGA report (p. 13), this trust account is established in the name of the private contractor operating the prison.
However, current law requires the contractor to obtain the approval of the CPC commissioner before using these funds for anything
other than the purchase of items for resale. In contrast, state profits from the prison commissary and inmate phone usage are used to
offset prison operating costs.

64

OPPAGA, p. 36.

Abt Associates Inc.

Private Prisons

43

include educational and vocational programs as well as counseling and treatment for drug and alcohol
abuse, domestic difficulties, and other problem areas. Contracts were ultimately awarded to the
Corrections Corporation of America (CCA) for two facilities, and to the Wackenhut Corrections
Corporation for two other 500-bed facilities.65 The two facilities operated by CCA were located in
Cleveland and at Venus, Texas. The two operated by Wackenhut were in Bridgeport, and Kyle,
Texas. These facilities began receiving prisoners during the summer of 1989.
The law specified that the state could not contract with a private vendor without an assurance that
services would cost at least 10% less than the state would have to spend for an equivalent facility.
The task of determining if the private firms were delivering the service at least 10% below the state’s
cost was given to the Sunset Advisory Commission. The challenge was to find comparison
institutions operated by the state. The statute specified that the commission should analyze the cost
and quality of services in the private prisons as compared to the cost and quality of any similar state
services. However, as the commission noted in its 1991 report, “The development of a cost estimate
was complicated by the fact that the TDCJ did not, and still does not, operate a comparable facility.”66
The commission therefore requested that the Texas Department of Criminal Justice develop two cost
estimates: (1) the total cost to the state of contracting for the operation of the privately operated
prisons, and (2) the total cost to the state if TDCJ operated the facilities directly. The estimate of the
former cost included actual contract costs plus a number of other direct and overhead costs borne by
the state in support of this contract. The second estimate included all direct and indirect costs that
would be associated with direct government operation of the CCA facilities and, separately, the
Wackenhut facilities. The commission concluded, from a comparison of the actual costs and the
estimated costs, that contracting saved the state 14-15 percent—more than the required 10 percent.
However, some of the savings to the state was in the form of payments by each of the firms to local
governments that approximated the taxes that the firms would have paid if the property were privately
owned. Because the state owned the facilities, with the firms contracted to operate them, it makes
little sense to require the firms to be liable for any such taxes. Moreover, the cost to the state of the
capital asset is the same, regardless of whether the facilities are managed by the private firms or the
TDCJ. Not counting these payments in lieu of taxes, private operation appears to cost the state 8.8 or
9.7 percent less than direct TDCJ operation, depending upon the facility—or slightly less than the
savings required by statute. The payments in lieu of taxes appear to be simply a fee paid by the
private firms to governments so that the net cost of contracting to Texas exceeds the 10 percent
threshold.
An examination of the commission’s report indicates that the principal sources of savings associated
with contracting, apart from this payment, were lower operating costs and a lower overhead
expenditure by the state. Operating costs were $1.53 to $1.96 lower on a daily per prisoner basis,
state overhead costs were $1.36 less for the private facilities. The precise reason for the lower
operating costs cannot be identified because the report provides only aggregate-level expenditure data.

65

The RFP initially requested that contractors construct the prisons as well as operate them, but the state subsequently decided that it
was less expensive for the state to construct the facilities and to own them. Consequently, the contract was written as an operations
contract only.

66

The State of Texas, Sunset Advisory Commission, Recommendations to the Governor of Texas and Members of the 72nd
Legislature: Final Report (March 1991).

Abt Associates Inc.

Private Prisons

44

In short: the evidence from Texas suggests that the private firms are delivering a service that would
cost the government approximately 9-10 percent more if the state’s corrections department operated
the facilities directly. This assumes that the estimates of the department’s costs of direct operation are
accurate, of course. Lacking more information about how these costs were estimated, it is not
possible to evaluate them.
Arizona’s Prison for Men and Women

In 1993, the Arizona Department of Corrections awarded a contract to the Management and Training
Corporation to design, construct, operate, and own a 450-bed minimum security state prison for men
and women. This prison, the Marana Community Correctional Treatment Facility, was evaluated by
Charles Thomas to assess how the facility’s cost and performance compared with other comparable
institutions in Arizona.67 Unfortunately, no equivalent facility operated by the Department of
Correction existed in the state. The Marana Facility is the only one in the state that houses both men
and women. Other things being equal, that fact alone would affect any cost comparison significantly
because women prisoners are generally more costly to house than men. Thomas acknowledged a
variety of differences between the Marana facility and others. For example, many of the publicly
operated prisons used for comparison housed prisoners in tents, which no doubt affected the level of
per inmate expenditure in these facilities. Moreover, the Marana Facility, by design and legislative
wish, provides a higher level of programming—both in the types of programs and the level of inmate
involvement in them—than other Arizona facilities generally. Thomas’s solution to this was to
compare costs at Marana to the average costs of all 15 minimum-security prisons operated directly by
the state’s Department of Corrections.
During fiscal year 1996, the operating costs of the Marana Facility, including adjustments for
payments by the management for ad valorem property taxes, were $13,140 per prisoner per year, or
$35.90 per day. This compared to an average operating cost for the 15 state-operated minimum
security prisons that year of $15,766 per prisoner per year, or $43.08 per day, a difference of almost
17 percent.
However, the Marana Facility was not the least expensive of the minimum security prisons operating
during FY 1996. Three publicly operated facilities were less costly (before crediting the private
facility for payment of property tax). Indeed, seven of the fifteen publicly operated facilities were
within five percent of the cost of the Marana Facility. Even after adjusting for payment of property
taxes by the management firm, Marana still remained more expensive to operate than two of the
public facilities. In light of this, Thomas concluded “Given the limitations of the available data and
the heterogeneity of the facilities, it would be speculative to conclude that a precise cost savings had
been achieved.”68

67

C.W. Thomas, Comparing the Cost and Performance of Public and Private Prisons in Arizona (Gainesville, FL: Center for Studies
in Criminology and Law, University of Florida, 1997).

68

C.W. Thomas, Comparing the Cost and Performance of Public and Private Prisons in Arizona (1997), p. 93.

Abt Associates Inc.

Private Prisons

45

Summary
Perhaps the main finding of this review is that only a very small percentage of those facilities operated
by private firms have been evaluated systematically to determine how much more or less the relevant
government would spend in the absence of contracting for operations. Ninety-one contracts existed
on December 31, 1997 with privately operated secure confinement facilities that held prisoners on that
day, and some undetermined number of facilities also held prisoners sent to them through noncontractual means (that is, via an intergovernmental agency agreement). The number of contracts that
have been issued during the past decade is much higher. But only a handful have been subjected to
close scrutiny with respect to their costs, relative to the cost of directly government provision.
Moreover, some of these studies examined contracts that were negotiated in the early days of this
industry. It is difficult to have much confidence in conclusions about the relative costs of public
versus private provision when we have systematic cost comparisons of such a small and dated subset.
In Florida and Arizona no strong evidence of significant savings exists. In Louisiana, Tennessee, and
Texas, the studies find that contracting did reduce costs to government. Some of the savings
reportedly associated with contracting are seen to flow from lower spending by state governments for
“overhead” administrative functions. These savings may only be apparent, an artifact of how
government expenditures are allocated to all correctional facilities in the state. Governments will not
actually realize these savings unless they actually reduce overhead spending in conjunction with
contracting, rather than merely shifting the allocated share of an unchanged expenditure to the
government-operated facilities.
Generally unstudied in these few evaluations is the potential for either governments or private firms to
control specific types of costs, such as spending for prisoner health care. Expenditures for prisoner
health care has been rising at a much faster rate than for other non-medical correctional activities, in
part because prisons are subject to the same inflationary pressures experienced in the broader health
care marketplace. However, prisons have generally lagged behind other sectors in developing
procedures for managing these costs. Private firms that operate prisons are certain to implement
“managed care” strategies to contain costs, all of which could also be adopted by public correctional
agencies.69

69

See Douglas C. McDonald, Managing Health Care and Costs (Washington, DC: National Institute of Justice, 1995).

Abt Associates Inc.

Private Prisons

46

4. How Well Do Privately Operated Prisons
Perform?
If the principal objective of contracting out the management and operations of prisons is, as surveyed
correctional administrators most often reported, to expand capacity in order to alleviate overcrowding,
then privatization no doubt achieves that purpose. But advocates of privatization often proclaim other
goals. For example, the President’s Commission on Privatization, established by President Reagan,
described contracting—and broader private participation in government generally—as a means not
only of increasing the efficiency but also the effectiveness of public administration. Thus: “Contracting the administration of jails and prisons at the federal, state and local levels could lead to
improved, more efficient operation.” 70 More recently, the National Performance Review has also
supported privatization as one means of improving effectiveness as well as efficiency. Is there any
evidence that contracting has, indeed, improved prison operations or, more broadly, the performance
of prisons? Or, does it appear that the increased attention to cost containment has resulted in reduced
service quality?
Some read the studies to date as providing favorable support for privatization on this question.
Charles Thomas and his associates write that studies of private prison operations in Florida,
Louisiana, and the United Kingdom “demonstrate that these cost savings are often matched with
performance improvements (e.g., fewer disturbances, fewer escapes, increased prisoner involvement
in work programs, and more programs aimed at reducing recidivism).” 71 Others are more cautious.
Analysts at the U.S. Government Accounting Office (GAO) reviewed what they thought were the
strongest studies and concluded that they “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.” 72
This chapter discusses different approaches to defining and monitoring performance, both with
respect to government services generally and to prisons specifically. The current practice of
monitoring and evaluating the performance of privately operated prisons is described, reporting results
of the Abt Associates survey of correctional administrators. Finally, it reviews and assesses several
studies that have sought to evaluate the relative performance of privately and government operated
facilities.

What Does “Performance” Mean With Respect to Prisons, Public or
Private?
Much of current thinking about performance in government has been framed, at least at the federal
level, by the Government Performance and Results Act of 1993 (GPRA), the Clinger-Cohen Act of

70

President’s Commission on Privatization, Privatization: Toward More Effective Government (Washington, DC: President’s
Commission on Privatization, 1988), p. xviii.

71

Thomas, Bolinger, and Badalamenti, Private Census, 1997, p. vi.

72

General Accounting Office, Private and Public Prisons: Studies Comparing Operational Costs and/or Quality f Service
(Washington, DC: U.S. General Accounting Office, 1996), p. 3.

Abt Associates Inc.

Private Prisons

47

1996, the revised handbook to the Office of Management and Budget Circular A-76, and the Clinton
Administration’s major management reform initiative—the National Performance Review (NPR). All
have stimulated greater attention to improving government performance. More importantly, GPRA
aims to improve the operations of federal government programs by changing management emphasis
from inputs and processes, to performance and results. It mandates that each federal agency develop
a strategic plan describing the agency’s goals, objectives, and operations. Each agency is, in turn, to
derive from its strategic plan an annual performance plan for each of its program activities, defining
program-level goals, operations, and performance indicators.
The Senate Report accompanying GPRA emphasizes that measures of program performance should,
to the extent possible, concentrate on outcomes, rather than outputs. Outcomes refer to “the actual
results, effects, or impact of a program activity compared to its intended purpose,” whereas outputs
refer to actual levels of activity or effort that are realized. The report provides an example of the
distinction: eligible clients completing a job training program are outputs, but an increase in their rate
of long-term employment is an outcome. “While recognizing that outcome measurement is often
difficult, and is infeasible for some program activities,” notes the report, “the Committee views
outcome measures as the most important and desirable measures, because they gauge the ultimate
success of government activities.”
But how does one gauge the success of prison operations? What are the outcomes of prisons intended
to be, whether publicly or privately operated? Reaching agreement on those questions has been
difficult.
Throughout the two-hundred-year history of prisons in America, penal administrators, theorists,
reformers, and lawmakers generally have sought to advance different goals through imprisonment, or,
at least, to give certain ones more salience than others. Establishment of the nation’s first
penitentiary, in Pennsylvania, was animated in great measure by a desire to civilize punishment
(depriving persons of newly-valued liberty rather than more “barbaric” corporeal punishments and
executions), to save souls, and to teach self-discipline. Later reformers, and then penal administrators,
embraced prisoner rehabilitation, which could be accomplished more by the application of scientific
principles. “Penitentiaries” became “reformatories” and later “correctional institutions.”
Conceiving of imprisonment’s purpose as rehabilitation or reformation of criminal offenders suggests
an obvious outcomes-based performance measure: offenders’ criminality following release from
prison. But recidivism rates of prisoners have been depressingly high, and advocates of imprisonment
have generally turned away from justifying their existence and assessing their performance according
to such measures. Many have embraced, in recent years, a more minimalist conception of the
purposes of imprisonment: the containment, or “incapacitation,” of criminals. If nothing else, prisons
do remove prisoners from civil society and prevent their committing crimes outside the walls.
Simple incapacitation cannot be the only purpose, however. Rather, containment must also be
constitutional. Faced with claims of appalling conditions in prisons, the federal courts began in the
early 1970s to intervene in prison administration. Entire state prison systems were found to be in
violation of the Eighth Amendment’s prohibition of “cruel and unusual punishment,” and nearly all
states were forced to spend large amounts of money to institute the required improvements.

Abt Associates Inc.

Private Prisons

48

In response to the perceived absence of national standards, correctional professional associations
began to promulgate such standards for prison administration. Most prominent are those established
by the American Correctional Association (ACA), first in the mid 1970s. Since then, the ACA has
published standards for many different types of correctional facilities, including detention centers,
boot camps, juvenile facilities, and adult confinements. The ACA also established an auditing system
to accredit facilities.73 Other standards for health care services were established by the National
Commission for Correctional Health Care. Both sets of standards primarily prescribe procedures.
The great majority are written in this form: “The facility shall have written policies and procedures
on . . . .” They emphasize the important benefits of procedural regularity and effective administrative
control, which flow from written procedures, and the careful documentation of practices and events.
For the most part, the prevailing professional standards prescribe neither the goals that ought to be
achieved nor the indicators that would let officials know if they are making progress toward those
goals over time. Two facilities could conform equally to an ACA standard by having a written policy
on a particular issue, yet they could have diametrically opposite practices and outcomes on that issue.
For example, both could have a written policy on the use of isolation, each of which covers topics like
when isolation may be used, how long inmates can be isolated, and how reviews should be conducted
to extend periods of isolation. One facility, however, might use isolation rarely, if at all, while the
other could use it prodigiously.
The effect of these various trends has generally been to conceive of prison performance quite
narrowly as conformance to law, state rules and regulations, and professional standards.74 That is,
performance tends to be measured according to procedural compliance, or outputs, in the language of
the above-mentioned U.S. Senate report. The dominant focus is not on outcomes, other than achieving
compliance or accreditation—which, arguably, is not a true outcomes-oriented measure of
institutional performance. However, procedural compliance is what most state correctional agencies
contract with private facilities to do.

What States Ask Contractors to Do
In general, state and federal governments demand in their contracts that privately operated facilities
perform like their public sector counterparts. Statements of work in the contracts generally specify
that compliance will be required with the same procedural rules, regulations, and standards that are in
force in the public facilities. Although the Abt Associates survey did not explicitly ask if compliance
with departments of corrections’ rules and regulations was required of contractors, monitors for
eleven contracts volunteered that such compliance was indeed specified as a performance objective in
the contract. An examination of a selected number of contracts also found this to be true.
Correctional administrators also reported that 57 of the contracts in force at the end of 1997 required
that facilities achieve ACA accreditation within a specified time. In addition, administrators reported
that 61 contracts explicitly required compliance with conditions established in consent decrees or
other court-mandated standards.

73

ACA accreditation requires a facility to meet all mandatory standards during a three-day inspection by an ACA audit team. The
facility receives a score based on the proportion of optional standards the facility meets. To maintain accreditation, the facility must
pass inspection every three years.

74

There are some exceptions to this, as discussed below.

Abt Associates Inc.

Private Prisons

49

There is no doubt that inclusion of such procedural requirements in contracts is necessary. As
discussed in chapter five, governments may remain liable for the performance of contractors.
Compliance with national standards and laws provides governments a measure of protection against
lawsuits. However, examples of contracts that go beyond these to specify other performance
outcomes are rare.

Monitoring Contractors’ Compliance
Governments’ practices of monitoring privately operated facilities vary widely. All contracts receive
some oversight from the contracting agency. This ranges from minimal attention from a centrally
located contract administrator to a combination of a contract administrator and one or more on-site
monitors. Of the 28 state and federal government agencies surveyed by Abt Associates that reported
having active contracts with privately operated facilities on December 31, 1997, twenty reported using
monitors in addition to contract administrators. Several states required a strong background and
knowledge of a government agency or corrections as prerequisites for monitoring correctional
contracts. Others required experience as an internal auditor or investigator. Most governments
required at least some level of formal training (up to 40 hours annually) in areas such as auditing,
establishing performance standards and measures, ethics, and reporting disputes and conflicts.
Contract monitors surveyed by Abt Associates often reported receiving more training than was
required. One reportedly received more than 120 hours of training as a contract monitor. However,
only 12 states reported providing training specifically geared towards the monitoring of privately
operated correctional facilities.
The amount of time and labor spent on monitoring privately operated facilities also varied widely:
from full-time monitors who spend a substantial proportion of their time on-site to virtually no
monitoring at all. Of the 91 contracts for which we have information, 46 (52 percent) reported having
monitors on-site on a daily basis (see table 4.1). These monitors generally worked full-time on these
assignments, although some had responsibilities for more than one contract. A smaller number of
contracts (16) had part-time monitors, who averaged about one day a week, visiting the facilities on a
monthly basis. Ten contracts had monitors who devoted about the same level of attention to the
contract, but visited only quarterly.

Abt Associates Inc.

Private Prisons

50

Table 4.1
Contract Monitoring Practices, by Frequency of Visits to Facility

Daily Visits

Weekly
Visits

Monthly
Visits

Quarterly
Visits

No On-site
Monitoring

Number of monitors

35

3

7

6

2

Number of contracts

46

5

16

10

3

Average hours per month
spent on site, per
contract

145

57.6

19

Average hours per visit,
per contract

7.25

14.4

19

20

Total monitoring time
(% FTE), per contract

84%

43.3%

18.6%

23.1%

Note:

6.7

0

0
0%

Assumes a 40 hour work week. Nine contracts are not included in this table. Monitors of these contracts reported other
arrangements for site visits, including visiting on an “as needed” basis or on an annual or semi-annual basis.

Source: Abt Associates survey of state and federal correctional administrators.

Monitoring Performance as Outcomes
An alternative to conceiving of prisons’ performance as procedurally based is to frame goals as
outcomes and to identify and measure indicators that assess the facility’s progress toward
accomplishing these goals. This conforms to how many correctional administrators see their job.
Even though legislators, governors, correctional theorists, and other parties may differ in how they
define the mission of the prison, correctional administrators in public and private facilities generally
agree with one another on both the general and specific goals of their assignment. Prisons should be
operated to maximize public safety, either by keeping prisoners behind walls or maintaining strong
controls over those who are released for short periods of work or education or for short furloughs and
the like. Prisons should also be safe internally, for both prisoners and staff alike. Prisoners should be
occupied rather than idle. Health care should be sufficient, more or less equivalent to the standards
prevailing in the free community (if only because the courts have so required). Each of these
objectives could be translated into specific outcomes that are amenable to measurement. Public safety
could be measured by numbers of escapes, for example, or institutional safety by numbers of assaults
by prisoners on other prisoners or on staff.

Abt Associates Inc.

Private Prisons

51

Professor Charles Logan attempted to formalize such an approach in his definition of what he calls a
“confinement model” of corrections and associated performance indicators.75 This model does not
assume the success of prisons is determined by their effects on criminality in the larger society.
Rather, goals are limited to aspects that prison officials have the ability and resources to affect. Thus,
the mission of corrections is to “keep prisoners in, keep them safe, keep them in line, keep them
healthy, and keep them busy—and to do it with fairness, without undue suffering, and as efficiently as
possible.” The goals associated with this mission statement are security (“keep them in”), safety
(“keep them safe”), order (“keep them in line”), care (“keep them healthy”), activity (“keep them
busy”), justice (“do it fairly”), conditions (“without undue suffering”), and management (“as
efficiently as possible”). These goals are then given operational objectives and detailed performance
measures, some of which rely on objective data (e.g., escape rates), some on subjective data (e.g.,
inmates’ perception of their risk of being assaulted). Logan has used this approach, and his model, to
compare the performance of a privately managed facility in New Mexico to a federal facility. 76 (See
below for a discussion of these studies.)
Even though most contracts for private operations of facilities do not incorporate outcomes-oriented
performance objectives in a systematic way, monitors do report measuring such indicators in many
instances. Monitors for 71 contracts reported in the Abt Associates survey that they track variously
specified incidents occurring in the privately operated facilities. Sixty-nine reported monitoring
inmate misconduct, 68 monitor escapes, 73 monitor inmate programming, and 49 staff turnover. How
these measures were used to assess the contractors’ performance was not reported in the surveys. It
does not appear that thresholds were contractually established to define acceptable or unacceptable
ranges.
Nor do most contracts tie performance on such indicators to financial rewards. One exception is the
Federal Bureau of Prisons’ contract with Wackenhut for the operation of the Taft Correctional
Institution in California, which contains provisions for an award-fee incentive worth up to 5 percent of
paid invoices. The contract also has provisions for reducing payment for poor performance. The
other exception is a contract between the District of Columbia and the Corrections Corporation of
America for the operation of the Correctional Treatment Facility, which permits financial rewards for
meeting targets based on performance indicators.
In summary: even though most correctional administrators could agree upon a number of outcomes
that could be used to assess their success, in addition to compliance with various procedural rules and
regulations and recognized standards, most contracts have not taken such an approach. Rather, good
performance is seen as complying with the terms of the contract—the rules, regulations, and standards
included explicitly in the statement of work or by reference.

75

Charles Logan, "Criminal Justice Performance Measures for Prison," in John J. DiIulio et al., Performance Measures for the
Criminal Justice System (Washington, DC: Bureau of Justice Statistics, 1993), pp. 19-59).

76

Charles H. Logan, Well Kept: Comparing Quality of Confinement in a Public and a Private Prison (Washington, DC: U.S.
Department of Justice, National Institute of Justice, 1991.

Abt Associates Inc.

Private Prisons

52

How Well Do Privately Operated Facilities Perform Relative to
Government Operated Ones?
The Abt Associates survey of state administrators asked contract monitors to assess the performance
of each of the facilities that were operated under contract with their agencies. More specifically, the
monitors of these contracts were asked, “In your professional opinion, does the quality of service
provided in this facility fall below, equal, or exceed (a) that of the contract requirements? (b) the
quality of service in other reasonably comparable facilities in your system operated directly by public
correctional authorities?” Most (68 of the 80 responses) reported that the contractors’ performance
met the contractual requirements; only three exceeded requirements; ten were seen as failing to meet
requirements. Similarly, most (58) were assessed to be the equals in terms of performance to similar
facilities operated by the government correctional agency; ten were thought to exceed the performance
of these government facilities; and twelve were deemed inferior. Thus, about three-quarters of those
facilities under contract with state or federal agencies in the closing days of 1997 were assessed as
delivering performance equivalent to that found in comparable government operated facilities in the
same jurisdictions.
Another indicator of contractor performance is whether contracts are terminated or whether
contractors are found to be in violation of the terms of contract. The Abt Associates survey of state
and federal correctional administrators asked for information about contract terminations and contract
violations. However, these inquiries focused on contracts that were active as of December 31, 1997.
Terminations occurring before this date were reported summarily, with little characterizing
information. Based on the data reported in the survey, it is not possible to determine whether
contracts terminated before this date were terminated as a result of contract violations. Of the sixteen
terminations reported, nine agencies reported that their sole reason for terminating their contracts was
reduced need for beds. In probably only five instances, terminations occurred as a result of vendor
violations.







North Carolina terminated its contract with the Limestone County Detention Center in Texas
citing operational problems and a reduced need for contractor capacity.
Oklahoma also terminated its contract with Limestone County Detention Center in Texas for use
of excessive force, idleness, and lack of attention to need. Reduced need for beds also contributed
to the decision to terminate this contract.
Colorado terminated its contract with Bowie County Correctional Facility in Texas citing
“conditions.”
Utah terminated its contract with Frio County Detention Center and Crystal City Detention in
Texas citing several escapes with legal inability to prosecute.77
Montana terminated its contract with Dickens County Correctional Facility in Texas citing
unsuitability of facility for long term inmate housing.

In each of these cases, the agency and the contracting facility were located in different states. These
facilities were typically engaged to house prisoners in what amounts to a temporary “bed renting”
arrangement to alleviate crowding in the state’s prisons.

77

See chapter 5 for further discussion on prosecutorial jurisdiction and statutory authorization.

Abt Associates Inc.

Private Prisons

53

Eight agencies reported issuing violations to eleven facilities. For example, the Federal Bureau of
Prisons reported several escapes occurring at Eloy Detention Center, which is operated under contract.
Other agencies reported four other security-related violations. Oklahoma reported violations of
inmate rights which included excessive use of force, disciplinary measures, idleness, and lack of
attention to need.78 Other common violations involved case management and medical issues.
That some private facilities have been found in violation of their contracts reveals little, if anything,
about their relative performance vis-a-vis their public counterparts. Sub-par performance exists in
publicly operated prisons, but no equivalent mechanism exists for finding the facilities in violation of
written agreements (other than court orders or consent decrees). Superintendents of public facilities
do not contract with their superiors to deliver institutional performance specified in legally binding
terms. Moreover, many public facilities fail to achieve American Correctional Association
accreditation, and federal courts have found the performance of many public prisons deficient.

Systematic Research to Evaluate Performance of Privately Operated
Prisons
Eight research studies have been conducted and reported during the past decade of privately operated
secure confinement facilities in the United States. All but one were described in chapter three, as each
sought to evaluate both cost and performance of private and public facilities.
Perhaps the most striking aspect of this research literature is that it is so sparse and that so few
government agencies have chosen to evaluate the performance of their contractors formally. Even
though there exist over a hundred privately operated secure confinement facilities, there have been
very few systematic attempts to compare their performance to that of public facilities. Most
government agencies have been satisfied with monitoring compliance with the terms of the contracts.
An examination of the available literature indicates that only a few states have sought formal
evaluations of a contractor facility’s performance. In a small number of other states, researchers have
received grants to conduct research and have gotten agreements from private and public facilities to be
studied.
The authors of these studies measured institutional performance using a variety of different indicators.
These included outcomes-oriented measures such as post-release recidivism, numbers of escapes, uses
of force, disciplinary problems, and major disturbances. Some studies also measured the extent to
which prisoners were engaged in productive activities, including educational programming. Some
assessed compliance with standards; some conducted audits, especially of security procedures. Others
surveyed staff and inmates to learn their subjective assessments of the facilities. For an assessment of
these studies, see the paper by Gerald Gaes, Scott Camp, and William Saylor, in appendix 2.
Some of the researchers argued that their findings indicated superior performance in the private
facility studied. Charles Thomas, for example, concluded that the performance of the privately

78

Oklahoma was the only state to report violations which caused termination. Strict adherence to our questionnaire, as noted, would
not permit such a report.

Abt Associates Inc.

Private Prisons

54

operated Marana Community Correctional Facility in Arizona was comparable and in some respect
superior to the average performance of all other 15 state-operated minimum-security facilities.79
However, as discussed in chapter three, this facility was a specialized one for which no comparable
government-run facility existed. Comparing its performance to an average of all other minimumsecurity prisons obscures the fact that it did better than some and worse than other public facilities.
The more reasonable conclusion to draw from the data is that the Marana facility appears to have
delivered a level of service within the same range found in the other fifteen facilities.
In his ten-year old study of New Mexico’s only facility for women, Logan reported finding superior
performance under private management, compared to the direct government management of the same
facility one year earlier and compared to the performance of a federal prison in West Virginia.80
Although a large number of dimensions were measured, using both objective data and subjective
assessments of staff, the small size of the sample limits the generalizability of the findings. Moreover,
the claim of superiority relies mostly upon the subjective data obtained from staff surveys, which were
possibly biased in favor of private management, if only because contractors were invited into New
Mexico to remedy an unsatisfactory system for housing women offenders.
Only a few studies had the advantage of comparing similar public and private facilities in the same
jurisdiction. The State of Tennessee compared three facilities, two publicly operated and one
privately operated, and concluded that all three were operating at essentially the same level of
performance.81 In Louisiana, William Archambeault and Donald Deis compared two privately
operated facilities with one government facility that was similar in design and mission. Performance
was measured by a number of different indicators. On some measures, the authors interpreted the data
as evidence of the private facilities’ superiority; on others, the reverse was thought to be the case. In
general, however, they concluded that the two private facilities outperformed the public one used as a
comparison. We read the study and the data differently. We see no compelling evidence of the
private facilities’ superiority with respect to performance.
In their review of several studies comparing public and private prisons’ performance, Gaes, Camp and
Saylor reach the same general conclusion that GAO reported: that all of the existing evaluations are
flawed in fundamental ways, and there is little information that is widely applicable to various
correctional settings (see appendix 2). Accordingly, they believe that it is not possible to make any
general statements about the ability of private contractors to ensure quality correctional services in
comparison to public prison management.
The authors argue that the existing evaluations fail to coherently examine factors other than
management differences that might influence correctional outcomes. That is, the existing evaluations
assume that any observed differences in performance of public and private prisons can be attributed to
management practices, when these differences may be caused by a variety of other factors that often

79

Thomas, Comparing Cost and Performance, 1997.

80

Logan, Well Kept, 1992.

81

State of Tennessee Select Oversight Committee on Corrections, Comparative Evaluation, 1995, p. 68.

Abt Associates Inc.

Private Prisons

55

are not under the direct control of management. Such factors include dissimilarities in the prisoner
populations and the architecture of the facilities.
Gaes, Camp and Saylor note also that none of the existing evaluations tried to determine the extent to
which management practices, ideas, and personnel were shared among public and private sectors.
Little research attention has been given to the impacts that contracting may have had upon the public
prison systems that have relied upon contracting for private imprisonment. Moreover, few researchers
have attempted to document innovations developed by private contractors to improve services and/or
reduce costs.

Summary
In general, it appears that state and federal governments are getting what they ask for in privately
operated prisons, with some notable exceptions. Where contractors are sought to operate facilities
owned by the contracting agencies, the contract procurement and monitoring procedures appear to be
designed to obtain a facility that is a close equivalent to the public facility, differing mainly in the
legal status of the operator (i.e., private rather than government). Rules and regulations to be followed
are consistent with, if not identical to, those promulgated in the public corrections agency.
Governments monitor contractors to ensure that they comply with standards, rules and regulations,
and court orders, if for no other reason than to reduce their exposure to lawsuits. Great variability
exists in the time and attention government monitors spend on their tasks, with some devoting
virtually no time at all and others spending all of their time at a single facility. Where monitoring is
so limited, it is unlikely that contracting agencies are able to provide more than a cursory assessment
of the contractor’s performance.
When state governments house prisoners in out-of-state facilities, they are looking more for beds on a
temporary basis than for another facility that will be integrated into the states’ public correctional
agency. Although these contractual relationships were not studied closely, it is likely that the
threshold for acceptable performance differs from that applied to management contracts in stateowned or perhaps privately-owned, in-state facilities. How well these out-of-state facilities are
integrated into the public correctional agency is an issue, and expectations may be lower. Whether or
not private prisons have gone beyond contract compliance and actually perform better on other
measures of outcome is not well documented.
Only a few of the more than a hundred privately operated facilities in existence have been studied, and
these studies do not offer compelling evidence of superiority.

Abt Associates Inc.

Private Prisons

56

5. Legal Issues Relevant to Private Prisons
The legal issues surrounding the privatization of prisons have elicited considerable interest and
concern from public correctional officials and the general public since private prisons became a reality
in the mid- to late 1980s. Over the past decade, however, the thrust of this interest and concern has
changed as practical experience and legal precedent have answered some questions and raised others.
A host of subsidiary legal and contractual matters continue to receive attention and provoke
discussion. These range from questions about the use of force by private prison guards to highlypublicized problems resulting from certain states’ lack of legislation governing out-of-state prisoners
in private prisons.
With these and other legal matters increasingly well illuminated, it becomes possible to discern how
relatively few legal topics turn on public-private distinctions. To a considerable extent, the legal
requirements applicable to public correctional agencies—most notably the constitutional standards
governing incarceration—are the same for their private counterparts. Meanwhile, only a handful of
the legal issues peculiar to private correctional facilities have elicited legislative treatment. For the
most part, the contracting process has become the real focus of legal innovation. In the coming
decade, more of the resulting contractual arrangements will be carefully scrutinized and modified by
public correctional officials and the courts.
This chapter briefly focuses on the major legal issues that have emerged in prison privatization over
the past decade. The broad conclusions contained herein are treated at greater length in appendix 3.

The Legality of Delegating Correctional Services
While the legality of governments delegating the incarceration function to private entities generated
considerable controversy in the 1980s,82 it would now appear that objections to prison privatization
per se on constitutional delegation grounds have relatively little potency. To begin with, the federal
constitutional delegation doctrine, grounded in separation-of-powers principles, is rarely invoked and
has little direct application to private delegations. 83 Unless a government has absolutely no persuasive
statutory authority for entering into private prison contracts, courts will be extremely reluctant to
invalidate privatization arrangements strictly on delegation grounds. More generally, privatization is
usually viewed merely as a delegation of certain administrative functions related to incarceration.
Accordingly, only delegated rulemaking and adjudication functions that directly purport to exercise a
government power are deemed to require special constitutional due process safeguards and to be
subject to heightened judicial scrutiny.84

82

See, e.g., Field , Note, Making Prisons Private: An Improper Delegation of a Government Power, 15 Hofstra L. Rev. 649 (1987); I.
Robbins, The Legal Dimensions of Private Incarceration 396-413 (1988).

83

The Supreme Court has not invalidated a private delegation since the New Deal-era case of Carter v. Carter Coal, 298 U.S. 238
(1936).

84

See generally Lawrence, Private Exercise of Governmental Power, 61 Ind. L.J. 647, 650-653.

Abt Associates Inc.

Private Prisons

57

These constitutional due process mandates—often subsumed under a broader “delegation”
heading—appear to require that certain types of rulemaking and adjudication be kept in public
correctional agency hands.85 As a practical matter, these mandates have generally meant that private
delegations are permissible where the private entity’s role is subordinated to adequate public oversight
and/or approval.86 In this context, a due process or delegation challenge might be sustained if, for
example, private correctional authorities had the final say in functions like writing rules on release or
making disciplinary decisions.
The courts have yet to determine whether a private contractors’ potentially strong interest (it need not
actually be shown to exist) in maximizing their inmate populations87 would disqualify them from
making rules or decisions directly affecting inmates’ liberty interests, including rules and decisions
involving sentencing, parole, release dates, furloughs, discipline, and other matters affecting “good
time” credits.88 Still, numerous states have enacted legislation directed at retaining ultimate releaserelated decision-making and rulemaking in the public sector.89 Other states have sought to retain such
powers by contracts. In most cases, these statutory or contractual provisions have mandated that
initial decisions or recommendations, even where formulated by private contractors, must be subject
to final approval or ratification by public authorities.

85

The Due Process Clause has given use to an influential line of Supreme Court cases invalidating the delegation of discretionary
rulemaking or adjudication powers to financially interestedor potentially interestedparties. In the rulemaking context, see, e.g.,
Eubank v. City of Richmond, 226 U.S. 137 (1912) (striking down local ordinance authorizing majority of landowners to impose setback requirements on neighbors); Carter v. Carter Coal Co., supra. In the adjudication context, see, e.g., Sniadach v. Family
Finance Corp. of Bay View, 395 U.S. 337 (1969) (invalidating law authorizing private parties to garnish employee wages before
employees can respond to garnishment in court); Gibson v. Berryhill, 411 U.S. 564, 579 (1973) (striking down state optometry
board’s adjudication of conduct by corporate salaried optometrists where board was composed of individual optometrists with
financial interest in reducing corporate competition). Note that these standards are applicable to the federal and state governments
alike, since the constitutional restrictions on private delegations apply to the fifty states through the Fourteenth Amendment. See
Liebman, Delegation to Private Parties in American Constitutional Law, 50 Ind. L. J. 650, 652-54 (1975).

86

See, e.g., Currin v. Wallace, 306 U.S. 1, 15-16 (1939) (creation of market controls by producers judged constitutionally acceptable
based on government-imposed requirement that 2/3 of producers waive antitrust restrictions); Sunshine Anthracite Coal Co. v.
Adkins, 310 U.S. 381, 399 (1940) (industry officials allowed to propose non-binding minimum prices where officials functioned
"subordinately" to a public commission); Fuentes v. Shevin, 407 U.S. 67, 80-83 (1972) (state-sanctioned prejudgment repossession
of property by private companies permissible only if property owners have opportunity to resist seizure before a judge).

87

In some cases, the nature of the contract between the responsible agency and the private provider eliminates any financial interest on
the part of the provider to maximize the inmate populationfor example where the contract sets a fee that is independent of the
number of inmates housed.

88

The most persuasive guidance, albeit in a property rights, rather than liberty interest, context, are the so-called "Todd Standards"
first articulated in the case of Todd & Co. v. SEC, 557 F.2d 1008 (3d Cir. 1973). Under Todd, three factors were presented that
enabled rulemaking and adjudicatory functions by private securities associations to accord with due process. First, the private
entity’s rules had to be approved by a governmental authority before becoming effective. Second, all private decisions regarding
rule violations and penalties had to be subject to governmental review. Finally, all such private adjudications had to be governed by
a de novo standard of review giving no particular deference to the earlier decision. While the third factor has not been accorded
much attention, many states have interpreted the first two factors to require that rulemaking for private prisons and adjudications by
private prison contractors be subject to final government review and approval. See, e.g., the state statutes mentioned in fn. 7 below.

89

Taking just a few examples of the many statutory restrictions of this nature, see, e.g., Va. Code Ann. § 53.1-265 (barring private
contractors from developing and implementing procedures for calculating inmate release and parole eligibility dates); Ariz. Rev.
Stat. Ann. § 1609.01 (prohibiting contractors from taking any disciplinary action against an inmate); Colo. Rev. Stat. Ann. § 17-1203 (preventing private contractors from making conclusive recommendations about parole of particular inmates); Fla. Stat. Ann. §
957.06 (preventing private prison contractors from granting, denying or revoking inmates’ good-time credits).

Abt Associates Inc.

Private Prisons

58

Liability for Private Prison Conditions
Of central importance to governments contemplating correctional privatization is the question of how
privatization affects liability for the conditions in private prisons. For some time, relatively
extravagant claims were made by both advocates and opponents of privatization that the advent of
private prisons would insulate governments from liability exposure.90 Claims were also made that
privatization would substantially shield private contractors from inmate civil rights suits alleging
constitutional harms.91 The following questions hung heavily over the debate: Can prisoners’
rights— including those established under the Constitution—be adequately protected in a private
correctional context? Does the delegation of day-to-day responsibility for facility management to a
private contractor yield lower potential liability exposure for government correctional authorities?
And does correctional privatization result in a lower litigation price tag for the government?
The answer appears to be a qualified “yes” to the first and second questions and a “maybe” to the
third. As to the matter of safeguarding inmate rights, it is generally accepted that private prisons will
be treated as “state actors” for purposes of civil rights suits, 92 so that all relevant constitutional
requirements will apply with equal force to private as well as public correctional facilities. Moreover,
private prison employees will not be covered by the “qualified immunity” that shields from liability
public correctional authorities who reasonably believe that their discretionary actions are lawful.93
Finally, private prisons and officials will not be protected by other governmental immunities that may
otherwise limit the monetary damages available to inmates suing over prison conditions.94
As for liability exposure, a government’s exposure will generally be lower if a private contractor is
running a private facility. A contractor will be the primary defendant in inmate litigation, and
government authorities generally will not have direct responsibility for the actions of contractor
employees. In the vast majority of inmate cases—including most prisoner civil rights litigation where
individual, rather than systemic, group harms are at issue—governments will not be deemed to have
specific knowledge of, and therefore bear responsibility for, the specific acts and injuries alleged.95

90

See, e.g., Wash. Post, March 23, 1986, at F6, col. 2 (describing how some politicians believed that government could escape liability
through use of private prisons); M. Walzer, "Hold the Justice," New Republic, April 1985, at 12 (privatization may result in an
"escape from the enforcement of constitutional norms").

91

See, e.g., Sullivan, Privatization of Corrections: A Threat to Prisoners’ Rights, in G. Bowman, et al., eds., Privatizing Correctional
Institutions 139 (1994).

92

See West v. Atkins, 487 U.S. 42 (1988). See also, e.g., Street v. Corrections Corp. of America, 102 F. 2d 810, 814 (6th Cir. 1996);
Payne v. Monroe County, 779 F. Supp. 1330, 1335 (M.D. Fla. 1991).

93

Richardson v. McKnight, 117 S. Ct. 2100 (1997).

94

The Eleventh Amendment and state sovereign immunity statutes may prevent inmates from suing federal or state correctional
agencies and supervisory officials in their official capacities. In these cases, plaintiffs may be limited to suits for monetary damages
against officials in their personal capacity.

95

Under "Section 1983" civil rights litigationwhich represents the overwhelming majority of inmate claimsa firm and its
supervisory employees must be shown to have been directly involved in an alleged violation, have known about the violation or its
likelihood of occurring, and been "deliberately indifferent" toward the risk, or have generated or validated a policy or custom that
led to the violation. See, e.g., Street v. Corrections Corp. of America, supra at 818. See also generally Monell v. Dep’t of Social
Services, 436 U.S. 658, 694 (1978). Since public correctional authorities will have entrusted day-to-day management of prisons to
private contractors, they will be less likely to have notice or knowledge of specific harms alleged to have caused injury to individual
inmates.

Abt Associates Inc.

Private Prisons

59

While reliance on a private contractor will not prevent government authorities from being named in
lawsuits or being exposed to liability for widespread or obvious problems relating to facility
conditions, private contracting will greatly lessen the liability of government supervisory officials for
most inmate claims alleging individual harm. These claims represent the most common type of
prisoner lawsuit and assume a significant proportion of a correctional agency’s litigation budget.
Government litigation costs at a particular facility may or may not be lower with prison management
in the hands of a private contractor. A great deal depends on the nature of the facilities involved, the
profile of the inmate population, and the risk management approaches taken by particular public
versus private authorities. The latter approaches can themselves vary greatly according to the overall
management plan, quality of staff, and claims handling procedures employed by those administering a
specific prison facility or system. While it may be contended that private authorities can exercise
greater flexibility than their public sector counterparts in given situations, this assertion cannot be
readily substantiated.96 Thus, even though a prudent government correctional agency will insist as a
matter of course that a contractor indemnify and hold it harmless against all acts and omissions of the
contractor arising under a management contract—and even though such an agency will similarly insist
that it be named as an insured on any private comprehensive general liability insurance policy
covering particular prison facilities—there is no way to tell for sure whether its litigation expenditures
under privatization will be lower.97
While some degree of liability exposure will still attach to governments that have privatized certain
prison facilities, such exposure can be further reduced through the sensible use of monitoring
personnel. Having on-site monitoring provides a mechanism by which government supervisory
officials can take remedial steps upon learning of certain problems, thereby limiting the potential for a
negligence lawsuit.98

96

Some privatization proponents have asserted that private firms may prove more responsive to inmates and the general public in
remedying problems that result in prison litigation. They content that a private contractor can more more expeditiously to correct a
problem or settle a lawsuit when political pressures are absent, fewer approvals are needed, and funding is more readily accessible.
Also, a private firm’s concern with market reputation and profitability is alleged to create incentives for flexibility and holding costs
down. There remains no solid evidence or study to support these contentions, however plausible they may be.

97

Clear answers to this question are likely to be exceedingly difficult to obtain. To begin with, public entities do not carry third party
liability insurance, so direct cost comparisons are impossible. Also, public correctional authorities and contractors alike would find
it very hard to isolate with any precision all of their relevant litigation costs (staffing costs, particularly in the public sector, are hard
to attribute exclusively to litigation-related functions), and may be reluctant to disclose their damage awards paid out. Ascertaining
what portion of the contractor’s per diem is attributable to litigation defense would probably be quite speculative. Finally, it would
be extremely difficult to control for potentially wide differences in prison populations, facility conditions, and the past history of
litigation at one or more facilities. Still, one could surmise that public authorities have a number of factors that could aid them in
keeping certain litigation defense costs relatively low. These include the judicious use of salaried government attorneys to handle
virtually all prison litigation, possibly less expensive claims handling, and the lack of a profit mark-up. Private contractors, by
contrast, might feature relatively more expensive claims handling and legal assistance (e.g., though contract attorneys billing at
market rates), as well as that portion of their fee allocated to such work.

98

Government correctional authorities may be liable for their own negligence in oversight even though management of a correctional
facility has been turned over to an independent contractor. See Logue v. United States, 412 U.S. 521, 532-33 (1973). A monitor can
prevent such negligence from occurring. Moreover, in Section 1983 cases, a monitor can keep supervisory officials accurately
informed about situations that, if left unattended to, could give rise to charges of "deliberate indifference."

Abt Associates Inc.

Private Prisons

60

Employment and Labor Relations Issues
Employment and labor relations questions pose a number of important issues to government
correctional authorities as well as private contractors, including concerns about private employees’
right to strike. As a preliminary matter, it is important for a government correctional agency to ensure
that a private prison management company is indeed treated as an independent contractor, with its
own responsibility for compliance with all legal requirements imposed on private industry, including
the Fair Labor Standards Act and relevant federal and state nondiscrimination laws.99 It is also
important for government officials to ensure that they have adequate statutory authorization to
privatize corrections from the standpoint of civil service laws.100
Private contractors will be subject to the National Labor Relations Act (NLRA), rather than labor
legislation governing public employees. At an entirely new facility, this will mean that a contractor
will not be permitted to interfere with any concerted activity taken by its employees. At an existing
facility, however, it is extremely unlikely that a contractor will have to bargain with an existing union
or adopt an existing union contract. Under the NLRA, an incumbent public employees’ union will not
be able to represent prison guards in a private employment setting.101 Only a special guards
union—such as the United Plant Guard Workers of America—can represent prison guards under the
NLRA. Even in the extremely rare case where guards at a public facility are already represented by
such a union, a private contractor would not be bound by the substantive provisions of the existing
labor contract unless the contractor were to indicate affirmatively that it would hire most of the named
correctional officers at a facility or would explicitly consent to the terms of the existing contract.
While the right to strike clearly exists at private prisons, the potential for work interruptions also
exists in the public sector, i.e., work slowdowns and “blue flu” outbreaks. Still, the potential for more
serious disruption of prison operations can be blunted in two important ways. First, a private
contractor can seek to have employees agree, individually or collectively, to a no-strike pledge.102
Contractors can also seek notification requirements that allow contractors to make arrangements for
assumption of certain correctional responsibilities in the event of a work interruption.

99

If a management company were not an independent contractor, it could be deemed a "joint employer" with the government, which
would prevent the contractor from having unfettered discretion over a range of personnel issues. This might emerge most critically
where a government, as joint employer, was deemed to retain civil service control over employees, thereby negating the purported
benefits of personnel flexibility that are supposed to attend privatization.

100 For example, in Washington and Colorado, courts have ruled that those states’ civil service laws bar privatization. See Colo. Ass’n
of Public Employees v. Dep’t of Highways, 809 P.2d 988, 995 (Colo. 1991)(en banc); Wash. Fed’n of State Employees v. Spokane
Community College, 585 P. 2d 474, 476 (Wash. 1978)(en banc).
101 Under the NLRA, "no labor organization shall be certified as the representative of employees in a bargaining unit of guards if such
organization admits to membership. . . . employees other than guards." NLRA § 9(b)(3), 29 U.S.C. §159(b)(3).
102 Unions often agree to the inclusion of such clauses in labor contracts in return for a management agreement to seek peaceful
settlement of contract disputes through an agreed-upon grievance procedure. The U.S. Bureau of Prisons contract for its privatized
Taft, California facility provides that any collective bargaining agreement entered into during the period of the contract "should
provide that grievances and disputes involving the interpretation or application of the agreement will be settled without resorting to
strike, lockout or other interruption of normal operations." Taft Contract at 14.

Abt Associates Inc.

Private Prisons

61

Inmate Labor Questions
The Thirteenth Amendment to the U.S. Constitution excepts convicted prisoners from its bar against
“involuntary servitude.”103 Courts, accordingly, acknowledge this “convict labor exception” and
uphold laws mandating prison labor while finding no prisoner right to remuneration for prison
work.104 Against this backdrop, there is little reason to believe that private prisons, merely based on
their for-profit nature, would create a persuasive basis for a Thirteenth Amendment objection.
In fact, federal and state regulatory schemes governing prison industry labor are designed to prevent
exploitative situations from developing. In the federal system, a government corporation known as
Federal Prison Industries (“Unicor”) manages all prison industry labor (roughly 20% of all federal
inmates) and cannot sell its products and services in regular interstate commerce. Moreover, because
they do not work voluntarily, prisoners working for Unicor are not “employees” under the Fair Labor
Standards Act.105 In the case of state prisons, the federal government has largely preempted state
legislation by generally restricting the flow of prison products and services in interstate commerce.106
While the Prison Industry Enhancement (PIE) Act of 1994 does allow prison-made goods from 50
pilot industry programs in various states to be sold in interstate commerce, prisons employing inmates
in these programs must pay them prevailing or minimum wages and provide them with workers’
compensation coverage.107 Questions remain as to whether private prisons will be authorized to
operate prison industry programs pursuant to the PIE Act.
The biggest worry about inmate labor does not relate to such prison industries, but to increasing use of
inmate labor to perform ordinary prison maintenance or service work. There is understandable
concern that inmate labor may increasingly be used by for-profit contractors to handle more aspects of
prison operation—including tasks traditionally performed by government employees. One way to
draw boundaries in such situations is to use the same inmate staffing plans for private prisons that are
used in public facilities. This creates a level playing field and allows for reasonable cost comparisons.
Public correctional authorities should also attempt to ensure that, wherever possible, prison labor
staffing innovations are accompanied by commensurate reductions in the per diem rates charged to the
government by private contractors.

Other Important Legal Issues
A number of other legal issues have important implications for correctional privatization. These
include public access to private prison records; private contractor access to criminal history records;

103 Only pretrial detainees are protected from involuntary labor. See, e.g., Channer v. Hall, 112 F.3d 214 (5th Cir. 1997).
104 See, e.g., Berry v. Bunnell, 39 F.3d 1056 (9th Cir. 1994).
105 See, e.g., Vanskike v. Peters. 974 F. 2d 806, 810 (7th Cir. 1992), cert. denied, 507 U.S. 928 (1993); Nicastro v. Reno, 882 F. Supp.
1128 (D.D.C. 1995), aff’d, 84 F.3d 1446 (D.C. Cir. 1996).
106 See the Sumners-Ashurst Act, 18 U.S.C. § 1761.
107 18 U.S.C. § 1761(c).

Abt Associates Inc.

Private Prisons

62

the impact of bankruptcy laws on possible private contractor insolvency; private contractor use of
force; and environmental law responsibilities. These topics merit brief discussion in turn.
Access to Private Prison Records

The shift toward private prisons has prompted concern that the public may be cut off from access to
information necessary for proper accountability. A few key principles have emerged from the federal
FOIA framework; state analogues, though too numerous and various to catalogue here, are likely to
follow the same pattern. First, records relating to privatized facilities that are generated by public
correctional authorities will be subject to public disclosure. Second, only private contractor records
obtained and held by public agencies will be similarly available for public review.108 Even then,
however, exemptions for confidential “financial or commercial information” may limit disclosure,
particularly if public authorities determine that disclosure might cause substantial competitive harm to
a contractor.109 Such is often the case with private prison proposals in competitive bidding processes.
While the systematic collection of private contractor documents by public authorities might ensure
adequate public access, other, less burdensome solutions include contractual requirements that simply
require contractor compliance with FOIA.110
Private Contractor Access to Criminal History Records

To perform its prison management duties properly, private correctional contractors need access to
criminal history records for two purposes: (1) to classify inmates properly and take other steps for the
proper safety and care of prisoners; and (2) to screen potential private correctional employees. Of
necessity, both of these needs affect the privacy rights and expectations of inmates and private citizens
seeking employment in the field of law enforcement and corrections. As it turns out, there exist two
related, but distinct, legal regimes at the federal level that safeguard these privacy rights and that
would be accessible to private contractors.
The federal mechanisms established to permit transfer of selected federal criminal history records
information (CHRI) to private firms represent a compromise between full disclosure and redacted or
edited data which could hamper contractors’ ability to perform their contractual obligations.111 For
prisoner data, private contractors may rely on the Federal Bureau of Investigation’s National Crime
Information Center (NCIC), which is able to link local, state, and federal criminal justice agencies for
the purpose of exchanging CHRI. Public correctional authorities must first obtain such information

108 See Dep’t of Justice v. Tax Analysts, 429 U.S. 136 (1989).
109 See, e.g., Critical Mass Energy Project v. Nuclear Regulatory Commission, 975 F.2d 871, 878 (D.C. Cir. 1992)(en banc), cert.
denied, 507 U.S. 984 (1993).
110 In the future, another solution might be special legislation that mandates such a result for all privatized facilities. See N. Casarez,
Furthering the Accountability Principle in Privatized Federal Corrections: The Need for Access to Private Prison Records, 28 U.
Mich. J. L. Reform 249, 301 (1995).
111 At the state level, the policies are governed by state law, which reflects varied approaches too numerous to summarize within the
scope of this report. Most of the mechanisms at the state level parallel and intersect with those employed by federal authorities.

Abt Associates Inc.

Private Prisons

63

and then summarize or characterize it for use by private firms.112 Access thus far has not appeared to
be a problem.
A separate federal regulatory regime, which embraces employment screening, governs the use of
CHRI for purposes unrelated to the “administration of criminal justice.” CHRI can only be used for
screening and licensing purposes pursuant to state and local statutes or regulations that are certified by
the FBI’s Criminal Justice Information Division, Access Integrity Unit, and that are administered by
designated state and local agencies. Here too, contractor access has not appeared to pose a problem.
It is worth noting, however, that the Federal Bureau of Prisons has decided to conduct criminal
background and records checks itself in the case of the Taft facility. This would obviously maintain
consistent screening procedures.
Bankruptcy

Few topics surrounding privatization seem to generate as much concern as bankruptcy, yet generally
speaking, such concern seems exaggerated. Bankruptcies involving private prisons have been
virtually nonexistent over the decade-and-a half of privatization and have been confined to those firms
concentrating on the building of private prisons on a speculative basis. More important, public
correctional authorities should be able to protect themselves against the untoward consequences of a
potential contractor bankruptcy through proper monitoring and contracting.
A properly monitored contract should provide public authorities with some warning that financial
problems exist. Annual financial investigations provide a baseline, while interim signals, such as
complaints by certain vendors about missed payments, provide additional guidance that may militate
in favor of prompt contract termination and the selection of another contractor (or resumption of
public control). Public correctional agencies can also protect themselves by insisting in the contract
that a contractor purchase business interruption insurance that names the relevant public agency as an
insured.
Careful contracting and the competitive nature of the private prison industry are the best safeguards
against serious problems developing from a bankruptcy. A general “termination for convenience”
clause with a ninety-day phase-out or transition period can keep the public authorities outside the
bankruptcy process (removing a facility from the bankruptcy process’ automatic stay and control by a
bankruptcy trustee) 113 and permit the public authority time to resume management of a facility or find
another private operator. Moreover, a 90-day transition period also provides a government facing an
impending private contractor bankruptcy with a significant financial resource with which to engineer a
successful transition and cover most expected or unexpected damages. Due to the lag time in
contractor invoicing—usually 30 days, followed by another 30 to 60 days for payments to be
due—public authorities can withhold payment and use these sums for purposes of covering current
operations and securing a new contractor.

112 See generally 20 C.F.R. §§ 20.21(b)(3), (c)(3).
113 The pre-petition termination of the contract can be ensured by a contractual requirement that notice of a filing of bankruptcy be
given to the contractor several days or weeks in advance. See Ruben, Legislative and Judicial Confusion Concerning Executory
Contracts in Bankruptcy, 89 Dickinson L. Rev. 1029, 1058-59 (1995).

Abt Associates Inc.

Private Prisons

64

Use of Force

Legal questions about the use of force often revolve around subtle matters of proper statutory
authorization. The major issue for private prisons is whether the use of force generally, or specific
variants thereof, including the use of deadly force, is properly mandated by the relevant law of the
jurisdiction. Without proper enabling legislation or contractual provisions authorizing the use of force
by designated private prison officials, it is possible that such personnel and the private firm could face
criminal and civil liability.
The applicable legal standards for the use of force vary tremendously from place to place. Some laws
may adequately treat the use of force generally, but insufficiently address the use of deadly force in a
variety of situations. Other states, for example, recognize a common law right of any citizen to use
deadly force in certain circumstances, including the apprehension of one previously arrested for a
felony.114 For states that have adopted the Model Penal Code, which permits correctional officers to
use deadly force under certain conditions, private correctional firms may be so authorized if their
employees are brought statutorily within the definition of “correctional officers.” 115 A still different
approach is taken by some states like Louisiana, which, in its correctional privatization statute,
specifically authorizes private firms to use force and further prescribes standards and procedures for
licensing private correctional employees to use weapons.116
Even if a contractual solution is adopted on a case-by-case basis—delegating certain use of force and
weapons use powers to private correctional management firms—care must be taken to explore all of
the nuances involved in various scenarios, including escapes and actions taken outside of a
correctional facility. American Correctional Association (ACA) standards address many dimensions
of the use of force in generic ways,117 but specific policies and procedures are necessary to comply
with state statutes and case law that have evolved under the Eighth Amendment. So-called rent-a-cop
statutes delegating certain use of force powers to private security companies may not be sufficient to
cover many of the circumstances encountered by private contractor guards. Also, correctional
officials must be sensitive to interjurisdictional issues about the use of force, particularly in the case
of escapes where use of certain types of force may not be authorized beyond a facility’s perimeter.
All of this requires government correctional officials and their legal staffs to think through very
carefully all of the possible scenarios requiring different kinds of force and to ensure that there is
proper authorization for their use.

114 This is ostensibly the case, for example, in Michigan. See, e.g., People v. Whitty, 96 Mich. App. 403, 292 N.W. 2d 214 (1980).
115 See Model Penal Code §§ 3.04-3.07, cited in Wooley, Prisons for Profit: Policy Considerations for Government Officials, 90
Dickinson L. Rev. 307, 323 (1985).
116 See La. Rev. Stat. Ann. § 39:1800.4(D)(5), (6). Note, for example, that the State of Ohio similarly requires a private correctional
officer working at a correctional facility housing out-of-state prisoners in Ohio to carry and use firearms only if the officer is
certified as having satisfactorily completed an approved training program designed to qualify persons for positions as special
policemen, security guards or persons otherwise privately employed in a police capacity. Ohio Rev. Code § 9.07(E).
117 See, e.g., ACA Standard 2-4206, for example, which simply requires a written policy and procedures to restrict the use of force as a
last resort in cases of justifiable self-defense, protection of others, protection of property, and prevention of escapes. Elaboration is
left to prison officials.

Abt Associates Inc.

Private Prisons

65

Environmental Law Concerns

Environmental concerns unique to a privatized correctional facility are few. The only major questions
concern responsibility for environmental impact statement (EIS) preparation in the case of new prison
construction and proper allocation of the risk of hazardous waste cleanup liability.
The former issue will usually be the responsibility of the public correctional agency. Although a
project proponent could be public or private, the government agency most involved with the project is
the party responsible for the preparation of an EIS.118 The latter issue revolves around the way in
which public correctional agencies and private contractors choose to allocate their risk through
carefully drawn indemnification agreements. Due to the transfer of land and property rights that may
occur between governments and private contractors in a particular construction, sale, or lease
arrangement, hazardous waste liability presents real concerns. Experience has shown that contracting
parties need to articulate indemnification arrangements expressly to minimize the risk of nonindemnification.119

Regulating Interjurisdictional Private Prison Contracting
Prison privatization, as commonly conceived, involves a government correctional agency contracting
out correctional management work to a private contractor that operates within the agency’s political
jurisdiction and is subject to statutes and regulations promulgated by that jurisdiction. Correctional
privatization can, however, involve other arrangements between public and private correctional
entities that span multiple jurisdictions. These arrangements can encompass the transfer of prisoners
to private contractors in distant localities and the interplay of several jurisdictions’ laws that affect
such a transfer. Two basic patterns characterize this phenomenon of interjurisdictional contracting.
The first involves indirect arrangements where public correctional authorities send inmates to another
jurisdiction via an intergovernmental agreement or contract, and the receiving jurisdiction in turn
subcontracts to a private contractor to run the facility where the inmates are housed. The second
arrangement involves the leasing of bed space directly from a private contractor in a foreign
jurisdiction. Many of these prisons are built and operated on a speculative basis.
Both of these interjurisdictional arrangements have generated considerable controversy and legal
problems. Correctional specialists have criticized the fact that the receiving facility may not have the
legal authorities to allow it to cope with a variety of critical situations, including inmate escapes and

118 See Seattle Audobon Society v. Lyons, 871 F. Supp. 1291, 1318 (W.D. Wash. 1994) ("Agencies may not delegate the responsibility
of preparing an EIS to private parties...."). This is the standard enunciated with respect to the federal National Environmental Policy
Act (NEPA). State environmental protection acts (SEPAs) may have different requirements. See, e.g., G. McGregor, Environmental
Law and Enforcement 10 (1994). Moreover, some SEPAs require different EIS preparations depending on whether a project is
deemed public or private in nature. See, e.g., Wash. Rev. Code § 43.21C.030. In most cases, however, a private prison should be
deemed a public project based on its governmental function. Cf. Weyerhauser v. Pierce County, 873 P.2d 498, 504-07 (1994) (en
banc) (holding that a private contractor’s construction and operation of a landfill is a public project because the handling and
disposal of waste is a governmental function).
119 See Hill, Contractual Indemnification for Environmental Liabilities, 21 Colo. Law. 943, 944 (1992).

Abt Associates Inc.

Private Prisons

66

prison emergencies. The law in most cases has lagged behind these correctional realities.120 A good
example is the case of two Oregon prisoners who escaped a few years ago from a private prison in
Texas and yet were unable to be charged with a crime, since Texas law at the time did not recognize
escape from a private prison as a felony.121 More recently, a lack of legal oversight mechanisms was
highlighted in the case of the privately-operated Northeast Ohio Correctional Center (NOCC) in
Youngstown, Ohio, which accepted out-of-state prisoners from the District of Columbia. The NOCC
experienced two fatal prison stabbings within a year of its opening, allegedly due in part to the
acceptance of out-of-jurisdiction inmates without an adequate classification system and without
standards requiring sufficient criminal history information about the prisoners.122
In response, many states have recently crafted detailed legislative solutions. Two notable examples
are Ohio (in direct response to the NOCC difficulties) and Texas.123 These new statutes attempt to
give state correctional agencies greater authority to prescribe minimal operating standards for such
facilities and to eliminate gaps and inconsistencies that have emerged in the legal treatment of private
facilities housing in-state versus out-of-state inmates. They address in particular the need for private
prisons with out-of-state inmates to be inspected by state authorities, maintain proper staffing,
coordinate closely with local law enforcement officials, reimburse them for the costs associated with
prison escapes and disturbances, and reject foreign prisoners if necessary for a variety of reasons,
including patterns of institutional violence in their correctional record. More such legislation can be
expected in coming years.

Legal Dimensions of Contracting
Today, the success or failure of a private prison arrangement may critically depend on the skill with
which contracts are negotiated. Public authorities must give close attention to the purposes served by
contracting and the degree of specificity which they seek to build into the agreement. In the absence
of an adequate statutory or regulatory framework, a contract may have to import from the outside and
treat in detail a number of issues that would otherwise govern various aspects of prison operation.
Correctional authorities should also be sensitive to the need to create, wherever possible, a level
playing field between public and private prisons to promote fairness, increase competition, and allow
for meaningful cost comparisons.124 Requests for Proposals (RFPs) can be structured in such a way as
to elicit many kinds of contractor information and plans which public officials can scrutinize and
evaluate in putting together a tight and effective contract.

120 See Associated Press, Private Prisons Shackle Texas with Confusion: State Laws Haven’t Caught Up with New Phenomenon,
Chicago Tribune, Nov. 7, 1996 at 34.
121 See Prison Brake: Texas Needs More Scrutiny of Private Prisons, Houston Chronicle, Sept. 3, 1996 at 18A.
122 See Washington Post, February 23, 1998 at B1.
123 Ohio House Bill 293, effective March 17, 1998; Texas Senate Bill 367, May 7, 1997.
124 A good example of this is ensuring that private contractors do not make money from things like inmate phone service, commissary
sales, and the like. Income from these, if it exists, should be reflected in lower per diem rates or should be paid into the general fund.

Abt Associates Inc.

Private Prisons

67

Contracts themselves can ensure better performance and cost effectiveness through well-drafted
provisions, the inclusion of objective standards, relatively short contract periods (e.g., 3-4 years) and
renewal opportunities, and broad termination clauses and penalty provisions. Contracts can even
specify cost savings requirements, which can be phrased in terms of a provider delivering correctional
services at significantly less cost or with significantly better services at the same or less cost than that
which public correctional authorities would need to pay if they did the job themselves. 125 Drafters
must be careful, however, not to inadvertently bestow third-party contract beneficiary rights on
inmates or the public.126
Public correctional authorities should also be wary of excessive use of alternative dispute resolution
provisions in a contract—particularly those relating to arbitration—to handle disputes with
contractors. Although arbitration is appropriate for handling issues surrounding payment and possibly
certain issues regarding the satisfaction of certain technical standards (including certain American
Correctional Association standards), its use in other contexts can deprive public officials of the
effective contractual and legal tools they need to ensure prompt contractor compliance. In cases
where mandatory ACA standards are at issue, or where resort to injunctive relief may be necessary
(which covers a good deal of contractual content), maintaining a public agency’s flexibility to seek
immediate recourse in the courts is essential.

125 The State of Tennessee actually mandates by law that private prison management contracts be at least 5% less expensive than the
likely full cost of the state providing correctional services itself. See Tenn. Code Ann. § 41-24-1-4(c)(1)(E).
126 Prisoners do not automatically possess such rights. A contract must intend to grant them legally enforceable rights to sue under the
contract, which can occur through conscious adoption of contractual language or through a court reading such an intent into a
contract and its surrounding circumstances. There is some precedent for courts reading an implicit intent to bestow third-party
beneficiary status on prison inmates, so drafters must explicitly disavow such an intent if they do not wish to create such rights in
prisoners. See, e.g., Owens v. Haas, 601 F.2d 1242 (2d Cir.), cert. denied, 444 U.S. 980 (1979) (court finds intent to bestow thirdparty beneficiary status on federal inmate housed in county jail based on U.S. Bureau of Prisons statutory policy to safeguard federal
prisoners while in the care of county governments).

Abt Associates Inc.

Private Prisons

68

Appendix 1
Comparing Public and Private Prison Costs

by
Julianne Nelson, Ph.D.
N. W. Partners
4200 Cathedral Ave., N.W.
Suite 412
Washington, D. C. 20016
(202) 363-8163
jnelson@nwpartners.com
May 15, 1998

Appendix 1

Private Prisons

1

How much money do taxpayers save when a prison is privatized? What are the sources of these
savings? These questions have received increasing attention from elected officials, civil servants,
academics and policy analysts as more and more prisons are transferred to private management.
Today, roughly 5 percent of adult jail and prison inmates in the United States are housed in privatelyrun facilities.1 This proportion is likely to increase as states come to rely more heavily on private
companies to supervise expanding prison populations. For example, the Governor of Tennessee
announced his support early in 1998 for proposed legislation that would “privatize the management of
up to 70 percent of the state’s prison system.”2 Some analysts predict that the private sector’s share of
the prison “market” will more than double in the next five years.3
This privatization movement has stirred considerable controversy, on both philosophical and
financial grounds.4 Among those invoking financial data, supporters of prison privatization cite
potential cost savings as high as 50 percent; others claim that the evidence to date in support of these
claims is ambiguous at best.5 The strength of these claims ultimately depends on the extent to which
they can be replicated in carefully-designed studies that compare cost reports for a number of years
from similar privately- and publicly-managed facilities. Although studies published to date do not yet
attain this level of rigor, it is still possible to draw several conclusions and to formulate a number of
hypotheses that merit further testing.

1

Preliminary Findings and a Basis for Comparison
Not surprisingly, reported costs per inmate-day are highly sensitive to
i.
ii.

the assumptions made about public sector overhead and indirect costs; and
the definition of variable costs used to adjust for differences in inmate populations across
facilities in a particular jurisdiction.

Some of the more extravagant claims made on behalf of prison privatization can be traced to
inappropriate handling of these issues. This chapter reviews the evidence from a number of recent
privatization studies, with particular attention paid to data reported for Tennessee, Louisiana and
Florida. These three reports seem to indicate that the cost savings attributable to privatization have
been (and will continue to be) far less than 10 percent of state spending on comparable facilities.

1

Johnston (1998), p. 35. Once CCA merges with the REIT that it created in July 1997 (i.e., CCA Prison Realty Trust), then
slightly more than 2.5 percent of all U. S. adult inmates will be housed in facilities managed by a single company.

2

Sundquist (1998), Wade (1998). Death-row inmates would remain in a state-operated facility. Although the sponsors of the bill
later decided that they lacked the support needed to gain passage this year, they will "probably" try again in 1999. Locker (1998),
Humphrey (1998).

3

Bates (1998).

4

See Murrell (1997) for a statement of the arguments against privatization and Logan (undated) for a refutation of the most common
objections.

5

The US-GAO (1996) report concluded that existing evidence was inconclusive. This report has been strongly criticized by those
generally favorable toward privatization initiatives. See, for example, Logan (1996) and Thomas (1996). See Florida Corrections
Commission (1996) and Zupan (1996) for more extensive annotated bibliographies.

Appendix 1

Private Prisons

2

Reports for California, Arizona, Texas, and the United Kingdom paint a far rosier picture of the
gains to be realized from privatization, but the aggregate nature of the data published in these
reports makes it difficult to evaluate these claims. Nevertheless, in every study that itemized
expenditures and adjustments, much of the reported difference between public and private sector
cost estimates can be traced to differences in the allocated burden of state-level overhead costs. If
this is reported difference is to reflect actual cost savings, then privatization must induce cutbacks
in state spending on central office operations before taxpayers realize this benefit.6
The review of two public and one private facility undertaken by the Tennessee Fiscal Review
Committee (1995, hereafter cited as the FRC report) stands out as the most detailed of the studies now
available. To a considerable extent, this report could serve as a model for future privatization studies.
Supplemented with data gathered for the same three facilities by the Washington State Legislative
Budget Committee, the Tennessee privatization experience suggests some systematic differences
between privately- and publicly-managed prisons. In particular, audited prison cost data for the fiscal
year ending June 30, 1994 indicate that
i.
ii.

iii.

iv.

v.

non-medical operating costs per inmate-day were virtually identical in the public and
private facilities studied in Tennessee;
non-medical operating costs were allocated differently in public and private facilities
studied in Tennessee: more was spent on corporate overhead and administration
(including monitoring by state officials) in the private facility (and correspondingly less
was spent on security);
the level of spending on prison employees differed substantially between the public and
private facilities studied in Tennessee: the amount spent on wages and salaries per
inmate-day was higher in the publicly-managed prison;
the nature of labor costs also differed between the public and private facilities studied in
Tennessee: benefits paid to administrative staff at privately-managed prisons averaged
almost 80% of salary; benefits paid to other employees at these facilities averaged 14%;
in contrast, the ratio of benefits to salary paid to staff at public facilities ranged from just
under 22% to almost 32%;
the bulk of reported cost savings at the privately-managed prison studied in Tennessee can
be attributed to differences in per-inmate medical expenditures and allocated state
overhead costs. (Labor cost savings were offset by the increased administrative cost cited
above.)

These findings have yet to be confirmed for other years and other jurisdictions. (They were, in
fact, not directly stated in the FRC report itself; they were merely implied in the data.) These findings
do, nevertheless, indicate a number of hypotheses that merit further study. In particular, the
supplemented FRC report suggests that
i.

6

privatization changes the way taxpayer dollars are spent on prison inmates: less is spent
on employees having direct daily contact with prisoners; more is spent on prison-level

These "cutbacks" in overhead cost are defined relative to the central office expenses that would have been incurred had the prisons
not been privatized. The only justification for assigning different overhead rates to private and public facilities is that the central
office expenses are smaller in a system with one or more private prisons.

Appendix 1

Private Prisons

3

ii.

administration, monitoring activities by state officials, corporate overhead and corporate
profits;
the lower labor cost found in private sector firms does not alone produce significant
savings from privatization; it is offset by increased administrative expenses (including
monitoring); significant cost savings do occur when private sector managers cut spending
on prisoner health care and induce the state to scale back its own administrative overhead.

How could these hypotheses be tested rigorously? An analyst undertaking such a project would
start by selecting roughly comparable public and private prisons and then collecting several years’
worth of data on the cost to the state of each facility.7 Ideally, these costs would be recorded as line
item amounts (wages, salaries, supplies, etc.) allocated to functional areas (administration, security,
food service, etc.)8 Next, the amounts budgeted for publicly-managed prisons and the amounts paid
by the state to private prison management companies would be adjusted to ensure the direct
comparability of estimated per-inmate costs. In particular, the analyst would need to identify the costs
that depend directly on the number and/or type of inmates and would need to adjust estimated
operating costs for any differences found in inmate populations.9 Allowances would also have to be
made for differences in public and private sector operating responsibilities and in the quality of
services received by inmates.10 Other adjustments to on-site operating cost include the amount spent
by the state when monitoring the privately-run facility, the taxes paid to the state by the private
management company and allowances for services provided by state employees not working for the
local department of corrections.
A final adjustment for state “overhead” cost often proves the most logically complex (and hence
the most controversial). In principle, the state overhead allocated to a privately-run prison should be
less than that imputed to a publicly-managed one only if the amount spent by the state department of
corrections on central office activity shrinks as more prisons come under private management. If the
budget for the state department of corrections central office does not change with the extent of
privatization, then per diem overhead allocations should be identical in publicly- and privatelymanaged facilities.11

7

The task of selecting fully comparable prisons is difficult, if not impossible. Differences in the age, health, gender, and security risk
of inmates would imply differences in the reported cost per inmate-day even among equally efficient prison facilities. The same
may be said for differences in the size of the inmate population and regional differences in the cost of labor and materials. One
could use regression analysis to "correct" for these differences if there were a large enough number of roughly comparable prisons in
a given system. In the absence of such data, the only available alternative is to adjust on a case-by-case basis. Not surprisingly, this
approach leaves room for considerable disagreement over the appropriate scale of the adjustment factors.

8

This program/performance budgeting approach has become an essential element of the campaign to "reinvent" government.

9

These costs would typically include food service, health care, security staff, and other "variable costs". The amount spent per
inmate on these items would, of course, depend on the health, gender, and security risk of the inmate population.

10

For example, in some jurisdictions the state would be responsible for installing and maintaining certain equipment at privately-run
facilities. These costs would have to be either (a) added to the amounts paid directly by the state to the private management
company or (b) subtracted from the amount budgeted for the state facility before private sector per diem rates are computed.

11

This "incremental" (or marginal) approach to overhead allocation is not generally the one followed in the literature on the cost of
corrections. Instead, analysts have generally sought to estimate the use of central office services by private and public prison
managers. This approach tends to overstate the relative cost of public prisons to the extent that central office expenses are truly

Appendix 1

Private Prisons

4

This idealized “model” study provides a standard against which existing research can be
measured. The remainder of this chapter provides an in-depth analysis of the methodology used in
two recent studies of prison privatization in Tennessee and Louisiana. It also provides briefer
discussions of privatization studies for prisons in Florida, California, Arizona, Texas and the United
Kingdom.

2

The Tennessee Experience: The 1995 Report of the Fiscal
Review Committee (supplemented with data from the
Washington State Legislative Budget Committee)

As required by state law, the Tennessee Fiscal Review Committee (FRC) undertook to
“compare the full costs of the contractor [authorized to operate a prison to house Tennessee inmates]
with the state’s full costs of operating similar facilities.” 12 Three prisons were selected: Northwest
Correctional Center (NWCC), Northeast Correctional Center (NECC), and South Central Correctional
Center (SCCC). The first two are operated by the state; the last is operated by Corrections
Corporation of America.
This study is unique in its attention to detail: it follows the procedure described in the previous
section for the “ideal” study, allocating “line item” cost data for each prison to specific management
functions (administration, security, etc.). There are, nevertheless, several limitations to the FRC
report. It covers a single year, it excludes medical expenses incurred by the private contractor, and it
does not directly address the question of whether or not privatization actually saves money for the
taxpayers of Tennessee. The first of these limitations is not surprising: the FRC report was intended
to serve as a template for future cost studies. Happily, the second and third issues were addressed
with considerable success by the Washington State Legislative Budget Committee (LBC) as it tackled
its own version of the prison privatization question.13
Operating Cost Estimates:

By combining the data found in the FRC report with that available from the Washington State
LBC, it is possible to compare the pattern of spending at public and private facilities and to gain some
insight into the profitability of the latter. 14 For the sake of completeness (and to illustrate the nature of
adjustments needed to ensure the comparability of data), Tables 1-3 reproduce the cost data found in
FRC report (Appendices A, B, C and the CCA Statement of Department Expenditures). Adjustments

fixed. In the tables that accompany this chapter, we use the overhead allocations as reported by the jurisdictions in question and
discuss the extent to which these reported allocations reflect an incremental approach.
12

Fiscal Review Committee (1995), p. 1.

13

Washington Legislative Budget Committee (1996), Appendix 3. In particular, LBC analysts both estimated the medical expenses
incurred by CCA on behalf of prisoners at SCCC and obtained an unofficial report of these expenses directly from CCA. The LBC
analysts also collected data on the fees paid by Tennessee to CCA for operating SCCC. Bob Thomas of the LBC graciously
provided spreadsheets containing the supplemental information (as promised in the LBC report itself). This information was
collected after the FRC report was published.

14

It is also possible -- as suggested by Logan (1996) -- to reconcile the apparent differences between the conclusions found in the
Tennessee and Washington studies.

Appendix 1

Private Prisons

5

to operating costs (reported by management function) include: netting out commissary revenues
against commissary expenses, allowing for changes in food services inventories and non-comparable
programs, eliminating depreciation expenses as a public facility expense (because there is no
corresponding set of expenses in the private facility) and adding monitoring costs to the expenses
reported for the private facility. These three tables indicate how sensitive the reported cost per inmate
is to the scale of operations at a particular facility. Although the privately-managed SCCC appears to
have the highest per diem operating cost ($33.78 as opposed to $33.06 at NWCC and $30.91 at
NECC), it also has the smallest number of prisoners (1053 as opposed to 1070 and 1149 at NWCC
and NECC respectively). Adjustments for scale and for amounts actually paid to the private
management company lead to somewhat different conclusions.
Table 4 reflects a different approach to the task of comparing public and private facilities.
Instead of analyzing the cost incurred by CCA of operating SCCC, this approach starts with the
revenues received by CCA and adjusts this data for differences between public and private
management responsibilities.15 Unfortunately, the FRC did not report medical expenses for CCA; it is
therefore necessary to estimate these costs in order to (a) compute CCAs operating surplus and (b)
compare private per diem rates with those derived for the state-run facilities NWCC and NECC.
The first half of Table 4 reports the operating surplus earned by CCA using two different
estimates of medical costs: these are deducted — along with “official” reports of non-medical
operating costs — from the revenues received by CCA to derive the benefit (before taxes) that the
company realizes from operating the prison facility.16 It is clear from these estimates that operations
at SCCC make a significant contribution to corporate overhead.17
The second half of Table 4 provides two estimates of the cost to the State of Tennessee of the
privately-operated facility. To estimate this cost, revenue received by CCA was reduced by estimated
medical expenses and augmented by the relevant adjustments for state expenses contained in the FRC
report. These expense estimates were further reduced by an estimate of the tax revenue received by
the state from CCA operations.18
Per Diem Estimates Reconciled:

Table 5 illustrates the impact of these various approaches on per diem rate estimates — and
hence on the estimated cost savings to be realized from privatization. Once cost reports are adjusted

15

This is the approach followed by the Washington State Legislative Budget Committee as it reviewed the research available at the
time. It is also the approach used in most other privatization studies.

16

The first surplus estimate assumes that CCA medical expenses are equal (on a percentage basis) to the average of the amount spent
at state-run facilities. (This is the basis for the per diem rates found in the LBC report.) The second surplus estimate uses an
unofficial report of CCA medical expenses (a fax from D. Massengale at CCA to Bob Thomas at LBC).

17

It is not clear from these data whether or not CCA is earning a profit over and above the facility’s allocated share of corporate
overhead of $970,417. Using one estimate of medical expenses, CCAs operating surplus (i.e., facility revenue net of facility
expenses) is below the share of corporate overhead allocated to SCCC. Under an alternative estimate of medical expenses, the
reverse holds true.

18

The LBC supplemental data contained tax revenue estimates.

Appendix 1

Private Prisons

6

for differences in the average daily population of inmates (ADP)19, we find the surprising result that
the non-medical per diem operating cost of the privately-managed SCCC is virtually identical to the
cost of the two public facilities.20 This conclusion obtains whether (a) one starts with individual line
item costs and aggregates to find total cost or (b) one starts directly with management fees received by
the private prison operator. It is also independent of the way in which one handles estimates for taxes
paid by CCA to the State of Tennessee.21 It follows that the savings to be realized from the
privatization of SCCC must arise elsewhere.
When official state overhead allocations and medical expense estimates are included in the
estimated cost of operating the three facilities, the picture changes yet again: the estimated total cost
per inmate day at the privately-run SCCC is as low as $37.65, while the estimated total cost of the
public facilities ranges from $40.16 (at NECC) to $40.19 (at NWCC). (See Table 5.) In other words,
the differences among these estimated per diem rates can be attributed to a lower overhead allocation
for SCCC (a $.78 difference); a credit for taxes paid by SCCC (a $.83 difference); and lower medical
costs ($3.07 at SCCC as compared with $4.37 at NWCC and $4.89 at NECC).
These tables indicate that (for at least the fiscal year ending June 30, 1994) privatizing SCCC
saved money for Tennessee taxpayers, if at all, by reducing spending on prisoner health care, by
inducing “cuts” in the central office (or state overhead) expenses of the Tennessee Department of
Corrections, and generating tax revenues for other parts of the state government22 Other savings
from restructuring employee compensation appear to be offset by the increased cost of monitoring by
state employees. Obviously, this estimate of cost savings understates (overstates) the benefit of
privatization to the extent that savings in DOC state-level overhead expenses amount to more (less)
than $.78 (i.e., $2.38-$1.60) per prisoner.23 It also understates (overstates) the benefit of privatization
to the extent that health care is qualitatively better (worse) in the private facility.
Spending Patterns Compared:

There are more insights to be gleaned by returning briefly to the operating cost data reported in
the FRC study. Even though public and private operating costs appear to be roughly comparable,

19

These adjustments reflect the assumption that not all operating costs vary directly with the number of inmates. A sufficient, but not
necessary, justification for this assumption would be "economies of scale" in prison operations.

20

Slight differences in the "equalized" operating cost estimates reported by the FRC and the LBC can be attributed to differences in
the way that size adjustments were handled in the two reports. The Washington State study allocated FRC adjustments (for noncomparable programs, supervision of industrial work crews, etc.) to specific line items and recomputed variable cost estimates
accordingly.

21

It is not clear from the FRC report whether or not the cost data given for SCCC include taxes paid by CCA. The per diem rates
given in Table 5 reflect the assumption that these data do include taxes paid (since the report was supposed to indicate the full cost
to CCA of its operations).

22

Here "cuts" refers to the differences between observed state-level overhead expenses and the overhead expense that the state would
have incurred had SCCC been publicly managed.

23

The per diem rates reported in Table 5 use the LBC estimates for overhead allocations. The LBC allocations for the publicly-run
prisons are lower than those found in the FRC report, as the LBC analysts sought to estimate an average annual cost for major
maintenance at all three facilities. The FRC report uses observed major maintenance expenses for SCCC, an approach that -- for the
year in question -- appears to understate the expected annual cost of maintenance.

Appendix 1

Private Prisons

7

Tables 6-8 provide evidence of some systematic differences between public and private prisons. From
Table 6, it appears that CCA is spending considerably more per prisoner on administration and less on
security. This result obtains even when monitoring costs are excluded from the per diem estimates.
Table 7 provides evidence of the substantially lower labor costs at the private facility: $16.89 per
prisoner/day at SCCC as opposed to $19.63 at NECC and $20.96 at NWCC. Much of this savings
comes from the smaller amount spent on security staff: $9.96 per prisoner at SCCC as opposed to
$12.57 at NECC and $14.55 at NWCC.24 The average employee benefit rates (i.e., the ratio of fringe
benefits to salary) found in Table 8 indicate that the nature of employee compensation also differs
between the public and private facilities in Tennessee. CCA administrative employees enjoyed a
benefit rate of almost 80 percent, while the rates received by other CCA employees ranged from just
under 13 percent to just over 14 percent. The average rate for CCA employees was 21.84 percent. In
contrast, there was much less variance in benefit rates across functional areas in the two public
facilities: the rates ranged from a low of just under 22 percent to almost 32 percent. The average
benefit rate was 29.19 percent at NECC and 28.38 percent at NWCC.
These differences in the structure of employee compensation at public and private prisons have
been strongly criticized. Bates (1998) cites testimony given before the Tennessee state legislature
indicating that SCCC has an employee turnover rate that is more than double the one found at
Tennessee public facilities. According to Bates (ibid.), many staff members who continue to work at
SCCC blame this high turnover on low wages and poor treatment by supervisors. The long-run
consequences remain to be seen. Nevertheless, Bates (ibid.) notes that members of the state legislature
who oppose privatization are quite ready to blame last year’s higher reported incidence of violence
and drug usage among inmates on the inexperience of the SCCC staff.
In conclusion, the FRC report provides a detailed look at the structure of prison costs and
suggests a number of hypotheses whose testing must await the availability of additional data.

3

The Louisiana Experience: Archambeault and Deis (1996)

The basic question addressed by William Archambeault and Donald Deis (A&D 1996) was
quite similar to the one underlying the Tennessee FRC report: “How does the full cost per inmate-day
in a publicly-managed prison (i.e., Avoyelles Correctional Center) compare with the full per diem cost
of housing an inmate in either of two privately managed facilities (i.e., Wackenhut’s Allen and CCA’s
Winn Correctional Centers)?” On one level, A&D appear to adhere more closely to the model for the
“ideal” study proposed in section 1 of this chapter: they use data from five years (instead of just the
single year found in the FRC report); and they use management fees paid to private prison operators to
address directly the question of taxpayer savings. They find that the publicly-managed prisons are 12
to 14 percent more expensive to operate than the privately-managed alternatives. (A&D, p. 439)
Nevertheless, the aggregate nature of the data presented and a variety of analytical quirks call into
question the strength of this conclusion.

24

These per prisoner data were derived from the "observed cost" information found in the FRC report. As a result, they do not reflect
adjustments for the scale of operations or the LBC allocation of general adjustments to specific line item categories (like salaries,
wages, materials, etc.)

Appendix 1

Private Prisons

8

A review of the methodology used by Archambeault and Deis helps identify the range of
difficulties that arise when using data that cover a span of years. Table 9 reproduces the data used in
the A&D study, along with the adjustments made in the interests of comparability. An analyst
wishing to use this data must make a number of assumptions about its consistency, accuracy, and
appropriateness, as well as choose a basis for making intertemporal comparisons. These assumptions
will, in part, determine the insights to be gleaned from regression results. Consider each of these
issues in turn.
Consistency and Accuracy:

Archambeault and Deis’ discussion of the annual state budget allocation for Avoyelles (see the top of
Table 9 in this report) provides an example of a classic data consistency problem. They observe (pp.
428-9) that this facility generally tends to be “under budget” and that the warden therefore ends up
returning unused funds to the state at the end of each fiscal year. Unfortunately, final data were not
available for FY 1995-96 when Archambeault and Deis finalized their report. As a result, they used
the full annual allocation for Avoyelles thereby (probably) overstating public sector costs . Table 10
illustrates an alternative approach: the gross appropriation for FY 1995-96 is reduced by same
proportion (1.7 percent) as the gross appropriation for FY 1994-95.
Louisiana budget data collected by the Washington State Legislative Budget Committee also
illustrate the difficulty of obtaining accurate estimates. The Washington State LBC estimates for
insurance costs at Allen and Winn are respectively $83,732 and $83,061; the data reported by
Archambeault and Deis are $66,534 and $66,001, a difference of almost 26 percent.25
The aggregate nature of the Avoyelles budget reported by A&D makes it difficult to evaluate
the scope of the costs covered. However, the data reported by Washington State researchers suggest
that some relevant costs may have been omitted. The WA-LBC cost estimates for FY 1995-96 include
allowances for a state-funded “record system analyst” at both Allen and Winn (at a cost of $26,255
and $29,193 respectively). Table 10 reflects this adjustment.
Appropriateness:

The fact that data are reported precisely does not guarantee that they provide relevant
information. The discussion of vocational education costs found in A&D illustrates this point. The
cost concept of interest is the full burden borne by the state for each prison. It is therefore appropriate
that A&D add grant money awarded by the state to Allen (for its vocational education programs) to
the other costs reported for this institution. It is also appropriate that A&D augment the cost reported
for Avoyelles by the value of services obtained from the local technical school. Ideally, this
adjustment would reflect the opportunity cost to the state of the time and materials used. In other
words, the operating cost reported for Avoyelles should be adjusted upwards only to the extent that

25

Bob Thomas graciously provided us with the spreadsheets used to generate the tables found in the WA-LBC (1996) analysis of
Louisiana prison costs. Oddly enough, the ratio of the A&D insurance cost estimate to the corresponding LBC estimate is identical
out to five decimal places. This suggests that the DOC Risk Management Division inadvertently used a different algorithm when
deriving the two sets of cost estimates from previously reported data. This may be due to the fact that A&D reported the cost of
"fire and property" insurance, whereas the LBC reported the cost of "property and boiler" insurance. One would suppose that the
appropriate figure would cover the cost of fire, property and boiler insurance.

Appendix 1

Private Prisons

9

other divisions of state government spent money to acquire these goods and services. Time and
materials donated by volunteers have no opportunity cost to the state and hence would not be
reportable as a prison operating cost.
A similar problem arises when adjustments are made for the “prison industries” operated at the
Allen and Winn facilities. Before computing private sector per diem rates, Archambeault and Deis
deducted the costs incurred by CCA and Wackenhut when inmates are assigned to work in prison
industries. They reasoned that since the state received the profits from these ventures it should also
bear the costs. This is not the relevant frame of reference: the adjustment should reflect the net
benefit to the state of these activities. The adjustment should also be made in a way that renders the
private sector data comparable with those reported for the publicly-operated facility. Ideally, an
estimate of the value of services rendered should be subtracted from the cost of each of the three
prisons. However, since no value is assigned to the work done by inmates at Avoyelles, the same
procedure should be followed for Allen and Winn. The cost to CCA and Wackenhut of running their
respective prison industries simply becomes one of the services they provide under the terms of their
contracts. This is the approach followed in Table 10: the benefits imputed to prison industries (and
originally subtracted from the state’s cost of operating Allen and Winn) are added back before per
diem rates are computed.
The perennial problem of allocating overhead costs provides yet another version of this
opportunity cost problem. In principle, the share of state overhead cost allocated to privatelymanaged facilities should be lower than that allocated to public facilities (all other things being equal)
only if the fact of privatization reduces the amount spent on the “central office” of the state
department of corrections.26 In all other cases, the overhead amounts allocated on the same basis to all
institutions, no matter who manages them. It is unlikely that overhead cost allocations based on the
methodology found in OMB Circular A-27 follow this rule. In Table 10, “indirect cost allocations”
used by Archambeault and Deis are subtracted from the other costs reported for each facility so that
per diem operating costs of facilities (excluding state corrections overhead) can be computed.27
Intertemporal Comparisons:

The Archambeault and Deis study illustrates a number of problems that arise when comparing
data reported at different points in time. These include (i) adjustments for inflation, (ii) the treatment
of “pre-paid” costs, and (iii) the treatment of accrued liabilities. Consider each of these issues in
turn.
The data reported for a given fiscal year should reflect consistent assumptions about prevailing
prices. Archambeault and Deis failed to allow for this when they constructed the time series used to
adjust private sector costs for the effects of prison industries and the taxes paid by CCA and
Wackenhut. As discussed in the previous section, the credit for prison industries should be deleted
altogether. A further problem with these adjustments is the fact that they are all expressed in prices

26

Again, the relevant point of comparison is the amount that would have been spent on state overhead if all prisons had remained
under state supervision. Bob Thomas (of the Washington State LBC) also makes this point in a letter dated March 7, 1997 and
addressed to William Archambeault.

27

Analysts at the Washington State LBC also followed this approach.

Appendix 1

Private Prisons

10

appropriate to the fiscal year FY95-96. Although Archambeault and Deis say that the prison
industries adjustments reflect a 4 percent rate of inflation, the cost savings actually reported for Allen
and Winn do not do so.28 It also appears from the text of A&D (1996) that no adjustment for inflation
was even contemplated when past observations for tax benefits to the state were constructed. As a
consequence, past estimates for these state benefits are likely to be too high — and past estimates for
per diem rates at privately-managed prisons are likely to be too low.
The treatment of ACA accreditation costs is curious for several reasons. Since the ACA is
chartered as a non-profit organization (not as a government agency), it is not clear why Archambeault
and Deis treat the accreditation fees paid to the ACA by CCA and Wackenhut as a benefit to the state
government.29 Even if it were appropriate to reduce the estimated cost to the state by the amount of
these accreditation fees, the full amount should not be allocated to a single year: the accreditation is
good for three years. It would be more appropriate to treat these fees as prepaid expenses, i.e., an
asset to be consumed over time.
The treatment of insurance costs poses a somewhat different intertemporal “smoothing”
problem. The cost of risk management insurance at Avoyelles appears to reflect the actual number of
claims made in a given year, since the costs reported do not correspond to differences inmate
populations.30 A more realistic approach would be to compute the expected (or average) annual cost
of such claims and use this as the basis for computing annual per diem rates.
The treatment of accrued pension and compensated leave liabilities poses yet another
intertemporal smoothing problem. Archambeault and Deis (1996, p. 434) chastise the authors of the
Washington State LBC study for deducting from the Avoyelles budget a contribution of $410,000
used to help amortize the unfunded pension liability accrued for employees working at the facility
prior to 1988. Archambeault and Deis claim that privatizing Louisiana prisons has the potential to
reduce this accrued liability. In response, Bob Thomas (1997) cites the Legislative Actuary for the
State of Louisiana as his authority for assuming that this liability would be invariant to the extent of
privatization. The revised operating cost estimate in Table 10 follows the approach found in the LBC
study.
Archambeault and Deis augment the cost budgeted for Avoyelles by the amount of unused sick
and annual leave accrued by current employees.31 It would be more appropriate to base this
adjustment on the expected (or average) annual leave accumulation — this would more accurately
reflect the average annual cost of operating a state facility.

28

See Table 9. The reported year-to-year differences in cost savings are directly proportional to annual changes in the prison
population.

29

The ACA web site indicates that publicly-managed prisons are not exempt from inspection fees.

30

Specifically, the reported cost of insurance fell by 40 percent between FY 93-94 and FY 94-96 even though the number of prisoners
housed at Avoyelles rose slightly. There is also no cost reported whatsoever for risk management insurance at Avoyelles in FY 9596.

31

These accrued leave hours represent a future liability to the state, as departing employees may be compensated for up to 300 hours
of annual. Hours in excess of this limit are used to compute retirement benefits.

Appendix 1

Private Prisons

11

Interpreting Regression Results:

These data problems combine to make it difficult to interpret the regression results found in
Table XI-8 of Archambeault and Deis (1996, p. 449). The simplest specification (Model A) reflects
the assumption that per-inmate variable costs are independent of the size of the facility and tests to see
whether or not the fixed cost of operating a prison is significantly different in the private sector. 32
This is a reasonable starting point, but the regression coefficients should not be computed until the
data are adjusted for inflation.33 Since this adjustment was not made, the coefficient reported for
“average inmates per day” does not provide an estimate of the impact on total cost of an increase in
the average daily population. Nor does the coefficient for the public/private dummy variable indicate
the impact of privatization.
The specification of Model B has the same “inflation problem” as Model A, with a further
wrinkle. Archambeault and Deis assume that the number of “Vo-Ed Slots Filled” helps explain the
total cost of operating a prison facility. When they find that the coefficient for this variable is
negative and statistically significant, they conclude that “lower operating costs are associated with
increases in the use of vocational education” (p. 439) and that “placing inmates in vocational
education programs is a less costly strategy than alternative work assignments” (p. 440).
Unfortunately the reported coefficients do not support this conclusion. Ignoring (for the
moment) the inflation adjustment problems mentioned above, the negative coefficient for inmates
placed in vo-ed may simply reflect economies of scale in prison operation. To see why, note that the
“average number of inmates per day” will generally be positively correlated with the number of
inmates placed in vocational education in a given facility.34 Note also the specification of Model A is
not quite right: over at least part of the relevant range of inmate populations, the marginal cost of an
additional inmate is likely to be falling as the inmate population grows.35 It is quite likely that the
“prediction errors” of the simplistic Model A are correlated in some way with the differences between
the growth rates in inmate population and vo-ed placement. This correlation implies nothing about the
relative cost effectiveness of vocational education, prison industries or roadside trash pickup. It is
merely the result of the “linearity” restriction imposed by the specification of Model A.36
Model C presents another version of this specification problem: in this model, the privatization
dummy is replaced by the number of security staff positions at each facility. The inflation and

32

I.e., the cost of operating a prison is assumed to take the form C(x) = F i + vx, where the index i is either "private" or "public."

33

In other words, the data should be adjusted for inflation so that all observations reflect the same assumptions about prevailing
prices. For example, if FY95-96 were chosen as the "base year" and if the relevant rate of inflation were assumed to be 4 percent per
annum, then observations for FY91-92 would be increased by the factor (1.04)4, observations for FY92-93 would be increased by
the factor (1.04)3, etc.

34

The larger the inmate population, the more inmates there are to place.

35

In other words, the prison "technology" exhibits economies of scale.

36

An alternative specification would have been to use the proportion of inmates placed in vocational programs as an explanatory
variable. Nevertheless, this would still only establish a correlation between prison costs and vo-ed.; it would not prove that
vocational education actually caused cost savings.

Appendix 1

Private Prisons

12

correlation problems described above again make it difficult to interpret these coefficients.37 The
approach used in the Tennessee FRC report provides a much more useful method of uncovering the
sources of cost savings.
The Net Effect:

It is clear from Table 10 that the cumulative impact of these additional adjustments may
drastically reduce the estimated cost savings from privatization: for the fiscal year 1995-96, it appears
that the actual savings in operating costs is more likely to be below 5 percent than above 12 percent.
However, as in the Tennessee case, the total cost savings from privatization will increase if it induces
cutbacks in state-level overhead.

4

The Florida Experience: (i) The Florida Department of
Corrections ; (ii) The Florida Correctional Privatization
Commission; and (iii) The Florida Office of Program, Policy
Analysis and Government Accountability

The controversy over the appropriate measure of operating costs in Florida prisons provides
yet another indication of just how hard it is to get all sides to agree on whether or not privatization
saves money. Current state law helps identify the participants in this debate. Chapter 89-526 of the
Florida state statutes, enacted in 1989, first authorized the state Department of Corrections (DOC) to
enter into contracts with private prison management companies. The subsequent Chapter 93-406,
adopted in 1994, created the Correctional Privatization Commission (CPC), to expedite prison
privatization; it is housed in the state Department of Management Services and operates
independently of the Department of Corrections. Chapter 957.05 qualifies the authority of the CPC,
requiring it to demonstrate — before any contract is signed — that private prison management will
result in at least a 7 percent savings (relative to the cost of allowing the state Department of
Corrections (DOC) to run the prison).38 Chapter 957.07 also empowers the state Auditor General to
decide whether or not this condition has been satisfied before any contract with a vendor has been
signed.
To provide for on-going independent oversight, Florida state law requires the Office of
Program Policy Analysis and Government Accountability (OPPAGA)39 to review the performance of
all management companies that operate private prison facilities in Florida (this may include both
vendors under contract to the CPC and those under contract to the DOC). A separate oversight board,
the Florida Corrections Commission, was established in 1994 and is “charged with reviewing the

37

For example, if we continue to the regression interpret coefficient as a measure of fixed cost, then the results reported for Model C
indicate that the annual fixed cost for operating a prison is -$5,000,000.

38

The CPC is state housed in the state Department of Management Services; it operates independently of the Department of
Corrections.

39

Although the OPPAGA is a unit of the Office of the Auditor General, it reports directly to the state legislature.

Appendix 1

Private Prisons

13

effectiveness and efficiency of [Florida’s] correctional efforts, recommending policies, and evaluating
the implementation of approved policies.”40
Not surprisingly, these distinct entities do not always favor the same measure of privatization
benefits. In its Annual Report for the fiscal year 1996-97, the DOC argued that all three privatelymanaged adult prison facilities were more expensive than similar publicly-run facilities would have
been.41 Using a somewhat different approach, the OPPAGA (1998) reached a similar conclusion.
Nevertheless, the OPPAGA report included responses from the Correctional Privatization
Commission, the Corrections Corporation of America, and Wackenhut proposing alternative
methodologies — ones that led to conclusions directly opposed to those reported by the DOC and the
OPPAGA.
Tables 11-13 (reproduced from the DOC and OPPAGA reports) help sort out this difference
of opinion. The Department of Corrections estimates, found in Table 11, are somewhat suspect: the
per diem rate for the men’s public prison is derived as the average of nine other facilities; all but one
of these public facilities is substantially larger than any of the private facilities being analyzed. 42
Given the economies of scale in prison operation, this approach understates the public sector cost of
running a prison comparable in size to Bay or Moore Haven.
Nevertheless, Table 12 shows how the OPPAGA (1998) reached a similar conclusion with
from a more appropriate starting place: it compares two private men’s facilities (Bay and Moore
Haven) with a single comparably-sized public facility. In its analysis, the OPPAGA started with the
management fee due to CCA and Wackenhut (for running Bay and Moore Haven respectively); it then
deducted monitoring costs paid by the management companies, property taxes not paid, and medical
insurance co-payments collected from prisoners and retained by the prison operators. For the
publicly-managed Lawtey prison, the OPPAGA used the annual cost of operations, health services,
and education programs reported for this facility by the Department of Corrections in its Annual
Report. All three cost estimates were then adjusted for differences in scale economies, the scope of
management responsibilities and the quality of services provided.
The OPPAGA inferred from its calculations that Moore Haven was more expensive to operate
than its public sector equivalent; it also argued that the Bay facility was only 0.2 percent cheaper. The
OPPAGA nevertheless rated the performance of both CCA (at Bay) and Wackenhut (at Moore Haven)
as “satisfactory,” while noting in passing that neither private facility achieved the 7 percent cost
savings required by law.
Table 13 reproduces the response by CCA to this finding. The adjustments proposed by this
management company substantially change the interpretation of the data: they indicate that the private

40

Commissioners are appointed by the Governor and with the approval of the state Senate. The Commission’s mission statement
appears on its web site, located at http://www.dos.state.fl.us/fgils/agencies/fcc/ .

41

These estimates are reproduced and analyzed in Florida Department of Corrections (1998).

42

Apart from the one public facility with an ADP of 733, the remaining public prisons housed between 1214 and 1719 inmates on
average. In contrast, the observed average daily populations for the privately-operated Bay and Moore Haven facilities were 708
and 706 respectively. If these two prisons were assumed to operate at 95 percent of capacity, they would house approximately 713
inmates on average. Each of these private facilities has a contract maximum of 750 beds.

Appendix 1

Private Prisons

14

management at Bay and Moore Haven saved taxpayers respectively 8.3 and 6.0 percent of state cost of
running the facility. This difference arises primarily from three simple changes: an increase in the
CCA’s credit for taxes paid to the state, an increase in CCA’s credit for the “Inmate Welfare Trust
Fund Net Revenue”43; and a decrease in the state’s adjustment for “medical costs for higher medical
grade inmates at Lawtey” (OPPAGA (1998), p. 38).44 The OPPAGA itself rejected these three CCA
amendments, observing (i) that CCA sought to claim credit for more tax revenue than was properly
attributable to its Bay facility,45 (ii) that it retained custody over Inmate Welfare Trust Fund
balances,46 and (iii) that Lawtey inmates had considerably more medical needs than those at the Bay
facility.
This debate is further complicated by an adjustment that the OPPAGA could have made but
ultimately rejected. As in Louisiana, the state retirement system was not fully funded by past payroll
deductions. To eliminate this shortfall, a surcharge of 5.78 percent was included in the the payroll
costs reported for Lawtey. The OPPAGA justified leaving this surcharge in place by stating that
it is still a cost to the taxpayers. Privates do not pay this and thus should be able to reduce
costs in comparison to state by at least 5.78% of payroll. (OPPAGA, p. 36).
This reasoning is not generally consistent with the basic question at issue, to wit, “How much money
will privatization save for Florida taxpayers?” If the unfunded portion of pension liabilities was
incurred in the past, then prison privatization does not help taxpayers avoid its consequences. Any
decease in DOC contributions toward this existing liability would necessarily be offset by increases
elsewhere in the Florida state budget.
Although there are many differences of opinion over particular line items, the OPPAGA
report tends to support the hypotheses suggested by the Tennessee data. The public and private
facilities reviewed appear to have quite similar non-medical operating costs; the potential cost savings
can be traced to medical expenses and to overhead allocation policies. If this conclusion is confirmed
over time, it then follows that prison privatization in Florida saves money through cutbacks in
spending on prisoner health care and state-level DOC staff.

5

Experience with Privatization Elsewhere

The cost comparisons reported by other states are generally less elaborate than those found in
the three cases reviewed thus far; they also tend to predict higher cost savings from privatization. For

43

This trust fund is an account containing the revenue (in excess of expenses) from the prison commissary, vending concessions and
inmate telephone service.

44

A fourth small adjustment by CCA remains unexplained: the "unadjusted" per diem reported by CCA for Lawtey exceeds that
found in the OPPAGA report by $.36.

45

According to the OPPAGA report (p. 59), CCA sought to claim credit for taxes paid on income from other facilities.

46

According to the OPPAGA report (p. 13), this trust account is established in the name of the private contractor operating the prison.
However, current law requires the contractor to obtain the approval of the CPC commissioner before using these funds for anything
other than the purchase of items for resale. In contrast, state profits from the prison commissary and inmate phone usage are used to
offset prison operating costs.

Appendix 1

Private Prisons

15

example, Sechrest and Shichor (1994) report savings on the order of 25 percent from the privatization
of community correctional facilities in California, and Thomas (1997, p. 93) concludes that “operating
cost savings [in Arizona] quite probably falls in the range of 13-17 percent.” Unfortunately, these
studies essentially report only aggregate costs, making it impossible to evaluate or interpret their
findings.
Two other analyses of privatization — one for Texas and one for the U.K. — have at least one
theme in common: much of the difference in estimates of public and private sector costs per prisoner
can be traced directly to differences in state overhead allocations.
The Texas Case:

A Texas law passed in 1987 authorized prison privatization only if the state can demonstrate
that it would lead to “a savings of not less than 10 percent over an equivalent state-run program.”
(Texas Sunset Commission (1991, p. 7). In its review of the state’s prison system, the Texas Sunset
Commission compared public and private sector costs to “construct and operate a 500-man pre-release
center, and found that “taking into account the tax revenues paid to local governments organized by
the state, the private prisons were operating at 14 percent below the state.” (Ibid.)
Despite the aggregate nature of the reported data, it is clear that the rule for allocating state
overhead allocation contributes substantially to this cost difference. Table 14 reproduces the cost
analysis found in Exhibit C of the Sunset Commission report. Once again, day-to-day operating
operations account for a relatively small part of the difference between public and private sector costs.
The combined contract payment and monitoring cost at a hypothetical facility designed by CCA is
reported as $29.71 per inmate day; the cost at a comparable facility designed by Wackenhut is
reported as $29.54. The estimated state cost per inmate day of operating these same facilities is
estimated to be $31.24 and $31.50 respectively. The remainder of the 14 percent cost savings lies in
the presumed tax benefit to the state from privatization ($2.19 per inmate day) and the estimated
decrease in administrative costs at the Texas Department of Criminal Justice (a difference of $1.36
per inmate day).
The U. K. Experience:

Woodbridge (1997, p. 29) predicts “an operational cost saving of 8 to 15% in 1996-97” from prison
privatization. Table 15 provides the aggregate data used in the U.K. study, and reports per diem rates
for adjusted operating costs. Two things are immediately obvious: (i) the average cost per inmate day
is much higher than in the studies from the United States; and (ii) the difference in cost between the
public and private sector prisons is much larger than in the United States. Without more detailed cost
data it is difficult to identify the reason for this difference. Nevertheless, it is clear that the allocation
of central office costs contributes substantially to the reported differences between in public and
private sector per diem estimates.

6

Recommendations for Further Study

This survey of recent cost studies does not resolve the question of whether privately-managed
prisons are cheaper than publicly-managed ones. The evidence is mixed, with the more detailed
studies indicating the smallest cost savings from privatization.
Appendix 1

Private Prisons

16

Nevertheless, it is possible to draw some preliminary conclusions. There do appear to be
some consistent differences between the public and private facilities, particularly if the Tennessee
experience can be applied to other jurisdictions. It appears likely that in privately-managed facilities,
the wage bill for non-administrative staff will be lower and prison-level administrative expenses will
be higher; that health care costs will be lower and that the imputed cost of state overhead will be
lower.
These generalizations suggest a number of questions for further study. These include:
1. Do the higher administrative costs found in the privately-managed prisons offset the
savings from different staffing patterns and compensation packages?
2. How can health care costs be cut without sacrificing quality?
3. Do the differences in reported public and private overhead rates reflect the overhead costs
actually avoided through privatization?
A yet broader question lurks in the background of this debate: “Is privatization necessary for
cost savings, or does the mere threat of privatization suffice?” In other words, will the risk of being
“reinvented” induce an appropriate set of changes in public sector practices? This question is familiar
from other public sector contexts. Experiments with “market testing” in Indianapolis indicate that
public employees can successfully compete against private firms and win city contracts. For the last
two decades, the rationale for deregulating telephones, airlines, trucking, and electric utilities has also
depended, in part, upon this sort of threat from potential (not actual) competitors.
It remains to be seen whether or not these market analogies — and their policy implications
— readily translate to the corrections context. Nevertheless, these analogies do suggest an alternative
the question of when and how privatization saves money. A more appropriate form of this question is
likely to be “Does competition make it possible to provide an appropriate level of service at a lower
price?”

Appendix 1

Private Prisons

17

Insert table 1 — Tables 1 through 15 are in the file app3_tab.pdf. The file can be downloaded from
the same site as the text of the report.

Appendix 1

Private Prisons

18

Insert table 2

Appendix 1

Private Prisons

19

Insert table 3

Appendix 1

Private Prisons

20

Insert table 4

Appendix 1

Private Prisons

21

Insert table 5

Appendix 1

Private Prisons

22

Insert table 6

Appendix 1

Private Prisons

23

Insert table 7

Appendix 1

Private Prisons

24

Insert table 8

Appendix 1

Private Prisons

25

Insert table 9

Appendix 1

Private Prisons

26

Insert table 10

Appendix 1

Private Prisons

27

Insert table 11

Appendix 1

Private Prisons

28

Insert table 12

Appendix 1

Private Prisons

29

Insert table 13

Appendix 1

Private Prisons

30

Insert table 14

Appendix 1

Private Prisons

31

Insert table 15

Appendix 1

Private Prisons

32

Archambeault and Deis. 1996. “Cost Effectiveness Comparisons of Private Versus Public Prisons in
Louisiana: A Comprehensive Analysis of Allen, Avoyelles, and Winn Correctional Centers.”
Bates, Eric. 1998. “Over the Next 5 Years Analysts Expect the Private Share of the Prison “Market”
to More than Double.” The Nation, January 5, p. 11.
Florida Corrections Commission. 1996. Annual Report, Appendix 5-4.. Available at web site
www.dos.state.fl.us/fgils .
Florida Department of Corrections. 1998. “Privatization in the Florida Department of Corrections.”
Available as a web page (time-stamped April 28) at site
www.dc.state.fl.us/administrative/reports/privatize/index.html .
Florida OPPAGA. 1998. Review of Bay correctional Facility and Moore Haven Correctional
Facility, Report No. 97-68. April.
Humphrey, Tom. 1998. “Private Prisons Bill bites Dust for this Year.” Knoxville News-Sentinel.
April 15, p. A1.
Johnston, David Cay. 1998. “Merger of Prison Operator Into REIT Sends Both Stocks Down.” New
York Times, April 21.
Locker, Richard. 1998. “Vote-Starved Prison bill Dies.” The Commercial Appeal. April 15., p. A1.
Logan, Charles. 1996. “Letter to the Honorable Bill McCollum, Chairman, Subcommittee on Crime,
Committee on the Judiciary, U. S. House of Representatives,” dated September 2. Available
at web site www.ucc.uconn.edu/~logan/ .
----------. Undated. “Prison Privatization: Objections and Refutations.” Available at web site
www.ucc.uconn.edu/~logan/ .
Murrell, David. 1997. “An Investigation into Privatization.” Roll Call, Florida PBA, December.
Available at web site www.flpba.org/publications/dec97.html .
Sechrest, Dale and D. Sichor. 1994. “Final Report: Exploratory Study of California’s Community
Correctional Facilities.” July 30.
Sundquist, Don. 1998. “State of the State/Budget Address.” February 2. Available at web site
www.state.tn.us/governor/sos1998.htm .
Tennessee, State of. 1995. Fiscal Review Committee. Cost Comparison of Correction Centers.
January 30.
Texas Sunset Advisory Commission. Final Report. “Chapter 4: Information Report on Contracts for
Correction Facilities and Services.” March 1991.

Appendix 1

Private Prisons

33

Thomas, Bob. 1997. “Letter to Dr. William Archambeault,” dated March 7. Copy obtained directly
from B. Thomas.
Thomas, Charles. 1996. “Letter to the Honorable Bill McCollum, Chairman, Subcommittee on
Crime, Committee on the Judiciary, U. S. House of Representatives,” dated September 1.
Available at web site www.ucc.uconn.edu/~logan/ .
Thomas, Charles. 1997. “Chapter 4: Comparing the Operating costs of Correctional Services for
Public and Private Prisons in Arizona.” Comparing the Cost and Performance of Public and
Private Prisons in Arizona. Available at web site web.crim.ufl.edu/pcp/html/test_iii.html .
United States Government Accounting Office. 1996. Private and Public Prisons: Studies
Comparing Costs and/or Quality of Service (GAO/GD-96-158). August.
Wade, Paula. 1998. “Sundquist Hits Hard for Bill to Privatize Prisons.” The Commercial Appeal.
January 21, p. A1.
Woodbridge, Jo. 1997. “Review of Comparative Costs and Performance of Privately and Publicly
Operated Prisons 1996-97.” Prison Service Research Report, No 3. (December).
Zupan, Linda. 1996. Privatization of Prison and Jail Operations: An Annotated Bibliography.
Available at web site www-cj.nmu.edu/cf/fin/forward.htm .

Appendix 1

Private Prisons

34

Appendix 2
The Performance of Privately Operated Prisons:
A Review of Research

Gerald G. Gaes, Ph.D.
Scott D. Camp, Ph.D.
William G. Saylor
U.S. Bureau of Prisons
Office of Research and Evaluation
320 First Street, N.W.
Washington, D.C. 20534
(202) 307-3198

June 10, 1998

Appendix 2

Private Prisons

1

There are good and bad public sector prisons, just as there are good and bad private sector prisons. As
Thomas and others have noted (Casile 1994; Thomas 1997a), whether improvements in quality can be
expected depends upon how well or poorly public-sector institutions are run by the respective
government agencies. In a public system with good management, good labor relations, and adequate
funding, the potential for improving quality by contracting prisons to private contractors is less than
where these conditions do not exist in the public sector.
Despite the paucity of evidence, those who argue for the private operation of prisons do so based on
one or two premises. They claim privatization introduces competition into an otherwise public
monopoly and this enhances services throughout the system while lowering the overall costs.
Secondly, they claim that private prisons can deliver the same or better services at a lower cost than
the public sector because “the marketplace” compels efficiency, ingenuity, and innovation.
Moore (1998) has published a recent review of the privatization research. He argues that private
companies save money through “new management approaches, new monitoring techniques, and
administrative efficiencies (p. 15).” Since labor is about two-thirds of correctional operating budgets,
labor cost savings, according to Moore, have been achieved through efficient facility design, reduction
in administrative personnel, minimization of overtime, and greater freedom to manage personnel.
Moore offers little evidence of these innovations and does not indicate how such savings translate into
affective inmate supervision and management. He cites the Archambeault and Deis study (which we
review in great detail), claiming the private sector can reduce significant incidents, such as prison
disturbances, relative to the public sector prisons. We found that evidence to be misleading, and, in
some cases, quite inaccurate. Moore makes the assumption that “... incidents lead to lawsuits, which
increase personnel costs.” He argues that market pressures and the competition for contracts result in
better direct services to inmates. This is typical of the argument-without-proof that is often found in
this literature. Anecdote is combined with “glittering generalities” to produce a conclusion having
little or no foundation. Rather than critically evaluating each study on its own merits, Moore’s review
of the prison quality literature merely cites those conclusions reached by the individual authors.
In this paper, we take a more systematic and critical approach to reviewing the research literature on
privatization. We examine the relative performance of publicly and privately operated prisons.
However, we also look beyond that comparison to see if there is evidence that privatization has an
effect on the entire public prison system and whether there is evidence that privately operated prisons
introduce ingenuity and innovation into the management of correctional institutions.
We critically analyze the literature that has accumulated which compares the quality of publicly and
privately managed prisons focusing on studies done in the United States. There have also been
evaluations conducted on prisons in the United Kingdom and Australia. Since we are unfamiliar with
the way these systems function, and we are unsure of the applicability of these studies to privately
operated prisons in the United States, we have excluded those studies from our analysis.
We review evaluations done in Massachusetts, Kentucky, California, Tennessee, Arizona, Louisiana,
New Mexico, Florida, as well as the Washington State review of the literature. We systematically
analyze these evaluations in terms of the methodology employed, in particular, whether the
evaluations compared institutions on the basis of performance measures and/or an audit/compliance
approach. We examine the evaluations with respect to how well they meet other methodological
criteria. These criteria were also identified by the Government Accounting Office (1996: 13) and
Appendix 2

Private Prisons

2

include whether equivalent facilities (and inmates) were compared; whether multiple indicators or
data sources were utilized for cross-validation; and whether the assessments were based on one-shot
or multi-year comparisons. Finally, we review the reports to assess the types of innovations (if any)
employed in the private (or public) sector that are intended to produce improvements in the quality of
services provided to inmates.
Although we are very critical of most of the studies that have been conducted, in our conclusion of
this paper, we try to build upon these criticisms and propose an optimal design for assessing
performance among institutions. This design can be used to evaluate aggregate measures of institution
performance regardless of whether one is interested in the private/public comparison or in an
understanding of those aspects of institution operations that produce positive or negative outcomes.

Massachusetts and Kentucky
The Urban Institute undertook a study between 1987 and 1988 to fill the void of empirical findings
available to aid states and local governments in making choices about private corrections (Urban
Institute 1989). While prison population growth and associated costs had led some to advocate the
privatization of corrections, opponents of privatization questioned the propriety, legality, and
constitutionality of private prisons. Advocates had argued that competition and less red tape would
enable private contractors to achieve lower costs and faster procurement of facilities and equipment
than was possible for government agencies.
Nevertheless, at the time of this study, there had been little empirical data used to test the assumption
that the cost and quality of private sector correctional facilities were superior to that of public sector
facilities. The Institute’s objective in this study was to assess and identify any differences in cost,
service quality, and effectiveness between publicly and privately run facilities and to identify reasons
for any differences that were found. Legal, propriety, and philosophical issues of private corrections
were intentionally not addressed by the study.
The study compared three pairs of facilities, one pair of minimum security adult facilities in Kentucky
and two pairs of facilities which housed violent juveniles in Massachusetts. Each pair consisted of
one private and one public facility.
Common methodological procedures were employed during the collection of data in both states. The
procedures included:
1
Extraction of data from agency records reflecting the number of escapes or
attempted escapes, returns to prison after release, results of facility inspections,
and cost data.
2
Surveys of inmates and staff at each institution using a modified version of the
Prison Social Climate Survey (PSCS) questionnaire designed by the Office of
Research and Evaluation of the Federal Bureau of Prisons (Saylor 1984).
3
Interviews with operations and oversight personnel at each facility.
4
A physical inspection of each facility by Urban Institute project staff using a
visual inspection rating form designed for the inspections.

Appendix 2

Private Prisons

3

The data collection took place between January 1987 and September 1988. It appears that
considerable effort was made to select pairs of facilities that were similar in mission and in the types
of inmates they housed. The authors acknowledge some major differences in the physical
characteristics of the facilities. For example, the Kentucky facilities housed minimum security adults
while the Massachusetts facilities housed violent juveniles. The public sector adult facility housed
more inmates than did the private sector adult facility (the public facility had an average daily
population of 353 while the private facility had an average daily population of 206). And, the juvenile
facilities all housed small numbers of inmates. Three of the facilities had 15-bed capacities and one of
the facilities had a 16-bed capacity. Presumably each facility was operating at capacity throughout the
duration of the study.
Some concern was expressed about the comparability of the inmate populations in the matched
facilities, particularly the adult facilities in Kentucky. However, after an examination of the inmate
characteristics in each pair of facilities, the evaluation team believed the comparison populations were
reasonably equivalent. It appears to us, however, that the initial concerns expressed by the researchers
were well founded. The differences appear greatest in the adult population, with the public sector
facility housing the more difficult population. Conversely, the private sector facilities appear to have
the more difficult juvenile population, with a larger segment of more serious criminal offenders.
In the adult population, the public sector facility had 14 percent more violent offenders. The private
facility had 11 percent more new offenders while the public facility had 10 percent more returnees
with new offenses. The public facility inmates had a median number of years to serve that was 3
years greater than the inmates in the private facility, with 38 percent of the public inmates serving
more than 10 years as compared to 22 percent of the private facility inmates serving sentences of more
than 10 years. Additionally, the median age of the public sector inmates was 5 years greater than the
median age of the inmates in the private sector facility, with 15 percent of the public facility inmates
over the age of 45 compared to 2 percent of the inmates in the private sector facility.
For the juvenile population, the primary differences were in race and offense. The public facilities
were composed of 50 percent black and 27 percent white inmates, while the private facilities were
composed of 30 percent black and 48 percent white inmates. While 52 percent of the juveniles in the
public facilities were committed for offenses against the person, 79 percent of the juvenile population
in the private facilities had committed an offense against the person. Furthermore, public facility
populations were composed of inmates who committed more property (27 percent) and miscellaneous
offenses (18 percent) than their private facility counterparts (12 percent property and 6 percent
miscellaneous offenses).
Although the study employed common data collection methods in both states, the analysis and
reporting were produced independently for each state. Consequently, there was considerable loss of
comparability in the application and interpretation of the measures and in the construction of the
tables that were used to summarize the findings. It seems that this lack of integration defeats any
benefit that might have derived from a common methodology. Admittedly, the differences in the
nature of the facilities and their populations might have diminished the comparability anyway,
particularly given the univariate nature of their analysis.
The report describes the sampling method employed for obtaining the adult inmate observations. The
warden initiated the process by drawing several numbers from a hat. These numbers were used to
Appendix 2

Private Prisons

4

select inmates from a list based on whether the 2 or 3 numbers drawn by the warden appeared in the
last digit of the inmates DOC identification number. A stronger sampling design for the Kentucky
inmates would have increased the comparability between the sample and the population. The
remainder of the staff and inmate observations from both states were intended to be a census of the
population.
The authors used a chi-square test of statistical significance throughout the report, although it is only
relevant to the tables of figures for the Kentucky inmates where a sample was drawn. The inferential
test was inappropriate for the remaining data since these were population characteristics of staff and
inmates at the facilities. Their use of the chi-square statistic as a measure of importance is
inappropriate both because (except for the adult sample) the study is an analysis of the populations
and also because it confuses statistical significance and substantive significance. The evaluators
attribute substantive significance to differences in group means simply because the chi-square was
statistically significant, while the metric and substance of the measure suggest that there is little
importance in the observed difference.
The principal findings of the study were that the quality of services and programs were superior at the
privately run facilities. The method for determining superior performance in the provision of services
was based on counting the number of measures (from each of the four types of data: agency records,
inmate and staff survey questionnaires, interviews with facility officials, and visual inspection ratings)
on which each private or public facility exceeded, or performed better, than its comparison facility.
Based on this method, the private facilities uniformly had a larger number of positive evaluations on
the set of measures. Many of the differences that favored the private facilities were obtained from the
staff questionnaire data. Ironically, the authors admit that juveniles in both the public and private
facilities had virtually indistinguishable responses to questionnaire items about service delivery .
The greatest deficiency of the study was its reliance on univariate analyses. The statistical analysis
consisted of univariate group mean comparisons. This method of statistical analysis created
comparability problems in spite of the researchers’ efforts to select comparable pairs of facilities. The
reality is that without an experimental design, it is virtually impossible to obtain two facilities that are
similar enough to provide meaningful comparisons without statistically adjusting for potentially
confounding aspects of each facility. A multivariate analysis would have been more appropriate.
There were no coherent or systematic models specifying desirable performance outcomes and
structural or operational processes that would be expected to accelerate or inhibit those levels of
performance. The absence of theories or models to guide the analysis resulted in a much more
voluminous and unwieldy report. Performance models would have allowed the evaluators to test for
institutional differences in a more systematic, precise manner, explicitly acknowledge preexisting
differences, and adjust the expected outcomes accordingly. The methods employed resulted in
arbitrary decisions and attributions about degrees of comparability and levels of performance, and in
general obfuscated the meaning of any public and private sector differences in the measurement set.
Statistical models of the sort proposed by Saylor (1996) and Camp et. al. (1997; 1998) could have
minimized the need for presumptions about comparability and would have made the determination of
performance differences, and the interpretation of those differences, straight- forward. Such
performance models would specify certain outcome measures and the process measures that are
believed to influence or control those selected outcome measures. The complexity of these models

Appendix 2

Private Prisons

5

would necessarily be limited by the small number of observations available. However, with
thoughtful preparation, meaningful models that fell within the limits of the number of staff and inmate
observations could have been identified. This same criticism can be applied to virtually every study
we reviewed. Rather than repeat that criticism throughout the report, we highlight the problem here
and in our summary of this research literature.
The study methods did meet the subsequent GAO (1996) criteria for evaluating the quality of service
delivery in correctional facilities; however, there were still quite a few deficiencies as we have noted.
There is no discussion of the types of innovations employed to achieve better quality of services and
programs.

California Evaluation
In adult corrections, California entered into contracting out for prison services around 1991 by
allowing cities, counties, and private companies to operate Community Corrections Facilities (CCFs).
Originally, these facilities were intended to house only inmates who had been returned to custody for
parole violations. But given the crowding pressures in the state of California, the decision was soon
made to allow new admissions into the facilities. Originally, there were 12 of these facilities, but by
the time of the Sechrest and Shichor study (1994), the number had declined to 11.
Sechrest and Shichor (1994) conducted an exploratory study of 3 of the 11 facilities. Two of the
facilities were run by public entities, one by the police department of a small community in the San
Joaquin valley and the other directly by the city administration of a small city in the Mojave Desert.
The privately operated facility was run by Management and Training Corporation (MTC) of Utah. It is
important to note that none of the comparison facilities were actually operated by the California
Department of Corrections (CDC).
Two types of data were collected for the quality comparisons. First, surveys were administered to staff
and inmates at each of the three study facilities as well as staff and inmates at two CDC facilities: the
California Institution for Men at Chino and the California Rehabilitation Center at Norco.
Additionally, interviews were conducted with the wardens at the three study facilities. The survey data
and on-site visits were used to assess conditions of confinement. The second source of data came from
official inmate data as provided by the Offender Information Services Branch of the CDC. In addition
to providing background information about the types of inmates at the respective institutions, the
official records also allowed the inmates to be tracked for recidivism (parole violation for a new
offense, technical violation of a condition of parole, or no violation).
The survey instruments used were taken from the surveys used in the study of private facilities in
Massachusetts and Kentucky undertaken by the Urban Institute (1989). Unfortunately, because of
study constraints, the surveys were administered to a fairly small number of inmates, and these
inmates were not chosen randomly. As such, it is not clear what confidence can be placed on the
results from the inmate surveys. The survey of staff suffered from the same problems, although the
number of surveys is even smaller (68 total surveys from all 5 facilities). Since the results cannot be
generalized to the staff and inmate populations from which they were drawn, we see no reason to
review the results. We simply do not know what they mean.

Appendix 2

Private Prisons

6

There are also methodological problems in terms of using the data on recidivism. We concur with the
summarization of the California study drawn by the General Accounting Office (1991: 31):
... 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.

Tennessee Evaluation
The Tennessee prison system, like prison systems in many jurisdictions, came under intense capacity
pressures in the 1980s that resulted in litigation (Grubbs v. the State of Tennessee, 1985). As a result,
the state legislature approved a substantial building program that started in 1985 and resulted in the
building of six prisons along similar architectural lines as well as a special needs facility. In 1991, the
state adopted legislation enabling the contracting out of correctional services to private contractors. A
decision was made to turn over one multi-custody facility to a private contractor (Corrections
Corporation of America as it turned out) to see what could be learned about best practice. The
enabling legislation required a research component to assess quality and cost (Tennessee Select
Oversight Committee on Corrections 1995). As stated:
TCA 41-24-105 (d) The contract may be renewed only if the contractor is providing at
least the same quality of services as the state at a lower cost, or if the contractor is
providing services superior in quality to those provided by the state at essentially the
same cost.
As a result, a bi-partisan committee from both houses of the General Assembly, the Select Oversight
Committee on Corrections (SOCC), brought together staff from the Tennessee Department of
Correction (TDOC) with executives from CCA to formulate a methodology for conducting the quality
and performance assessment. With the assistance of the Vanderbilt Institute of Public Policy Services,
formal meetings between TDOC and CCA produced a comparative methodology that was admittedly
not an academic research project, but it fulfilled the requirements of the legislation. Importantly, both
the public and private sectors agreed to the essence of the comparative methodology. Essentially, the
process entailed that an audit/compliance check would provide the basic methodology for comparing
the South Central Correctional Center (SCCC) operated by Corrections Corporation of America
(CCA) with the two state-operated facilities, the Northeast Correctional Center (NECC) and the
Northwest Correctional Center (NWCC). The three facilities chosen for comparison were all based on
the same general architectural design as discussed above. SOCC wanted to insure a “level playing
field” for all three facilities, consequently, in addition to similar physical design, all facilities came on
line at approximately the same time. By mid-1992, all three facilities were operational.

Appendix 2

Private Prisons

7

There were six elements in the comparative methodology used, although only three of the elements
actually received weight in computing the final aggregate score.47 The audit portion counted for 60
percent of the total score. A list of 200 elements was compiled. Joint teams comprising staff from both
TDOC and CCA conducted the audits. Ratings of compliance in the areas of Administration, Safety
and Conditions, Health Services, Mental Health, Treatment, and Security were compiled as well as an
overall rating of compliance. Two inspections were held at each of the three facilities. In general, the
results showed comparable levels of performance at each of the institutions. On the first inspection,
the overall compliance rates for the two public-operated facilities (NECC and NWCC) were
respectively 90.67 percent and 90.08 percent. For the SCCC facility operated by CCA, the overall
compliance rate was 84.53 percent. Both of the public institutions scored slightly better than the
private facility, although the differences are modest. On the second inspection, the three facilities
were virtually identical with 95.28 percent and 97.23 percent compliance at NECC and NWCC,
respectively, and 97.48 percent compliance at SCCC.
A security and safety index was the second element that received weight, and accounted for 25
percent of the final aggregate score. The parties agreed not to assign an objective score on this
dimension, even though the factors considered included disciplinary reports, use of force, assaults
(both inmate-on-inmate and inmate-on-staff), deaths, injuries, escapes, and a residual category for
other security and safety concerns. SOCC felt that scoring this area relied too heavily upon
professional judgment. The working assumption used in the evaluation was that all institutions were
in full compliance with safety and security standards, and the review would only note deficiencies in
safety and security practices.
As might be expected, even though the evaluation reports on differences in the factors of safety and
security, the conclusion is pretty mild. “Each of the institutions met the security and safety
requirements of the two annual inspections and an ACA audit. Their respective scores were
exceptionally high and almost identical. The administrative choices of how and when to use force,
how to dispose of disciplinary charges, or how many disciplinary tickets to write is really the
prerogative of management. However, in reviewing the entire period, in our judgment there was very
little difference in security and safety among the three facilities” (Tennessee Select Oversight
Committee on Corrections 1995: 56).
The final element that received weight (15 percent) was an index of programs and activities.
Generally, this was a review of the numbers of inmates in education programs and work status. The
indicator that received the most attention in the report was inmates in job waiting status. Because of a
lack of an operational industry program at SCCC (operated by CCA) and NWCC by the second year

47

The elements not used were the nature of inmates, professional standards, and a survey of staff and inmates. As noted in the report,
there is a need to make level-playing field comparisons, and as such, a need to control for the nature of inmates. However, this was
not done, and the report showed that there was substantial variation in inmate characteristics at the three facilities. For example, 47.5
percent of the inmates at SCCC (operated by CCA) were black, as compared to 22.6 percent at NECC and 78.2 percent at NWCC.
While these percentages may reflect the racial backgrounds of the regions of Tennessee where the prisons are located, the prisons
themselves are hardly comparable on this item. Similar differences were noted for custody classification of inmates.

The professional standards, those set by the American Correctional Association (ACA), State Fire Marshal reports, State Education
Department, and local and state health and sanitation standards, were considered minimum standards. However, the audit items created as
discussed above closely mirror items of concern in CCA accreditation inspections.
The surveys were intended only to provide subjective measures of satisfaction from staff and inmates and to provide insights into
operational issues.

Appendix 2

Private Prisons

8

of operation, these two institutions had higher percentages of inmates in job waiting status. At the
time of the review, SCCC was not in compliance with the policy that inmates job structures comprise
6 hours. CCA responded, though, that they changed their practice to be in compliance.
In determining a final weighted rating score to compare the institutions, the scores on security and
safety as well as program and activity were meaningless as all three institutions received the
maximum number of points on these scores. The only scores that differed among the three institutions
were for the percentage of compliance captured in the second audit. The scores for the first audit
(where CCA scored lower than the two public facilities) were not used. Since the CCA facility had a
slightly higher compliance rate (97.48 percent as compared to 97.23 at NWCC and 95.28 at NECC), it
came out slightly higher on the final weighted score (finals scores: SCCC, 98.49; NWCC, 98.34;
NECC, 97.17). But as SOCC (1995: 68) noted, “In reviewing the ratings we considered the range of
difference of up to 3 percent among the three facilities as essentially comparable. Therefore, our
conclusion was that all three facilities were operated at essentially the same level of performance.”
Despite this conclusion, the New York Times ran an article at the time the Tennessee evaluation was
released that concluded that there was strong evidence that CCA ran a better facility than the two
public comparisons in terms of quality and cost (Butterfield 1995).
The Tennessee evaluation is often cited as one of the more sound methodological attempts at
comparing private and public prisons. For one, it compares institutions that were of a similar
architectural design, were opened at about the same time, and were designed to house inmates of
similar custody levels. Nonetheless, there are some serious shortcomings to the Tennessee evaluation.
First, the review is based solely on operational audits of the three facilities. This means that no
performance measures were used in actually comparing the prisons, even though some attempt was
made to gather performance data related to inmate misconduct, programs provided to inmates
(primarily education), and the like. Generally, the position taken was that on the measures that could
be developed as performance indicators (measures of safety, security, and program activity), there
were no differences among the institutions. All institutions were comparable and received the
maximum number of points in these areas. However, even a cursory examination of the actual tables
presented on misconduct, education, and the like makes this conclusion suspect.
Second, it is not clear that the facilities provide an apples to apples and oranges to oranges
comparison. Even though all three facilities were multi-custody, they housed quite different types of
inmates in terms of the socio-demographic characteristics reported, age and race, criminal history, and
custody classification. Of course, these differences did not have as much impact upon the Tennessee
evaluation as they would have if factors other than operational compliance had been used to calculate
the final weighted scores of performance.
Third, the evaluation is a one-shot comparison even though data were collected over two years. For
whatever reason, the compliance data from the first audit at each institution is reported but not used in
determining the final comparative scores.
Fourth, only a single source of data, the compliance audit, was used to construct the final scores for
the three institutions. As mentioned previously, this flows post hoc from the decision to award all
three institutions the maximum number of points on safety and security as well as programs and
activities. It is not that multiple data sources were not compiled, it is that they were ineffectively used
(if used at all) in drawing the final comparisons between the private prison and the two public prisons.
Appendix 2

Private Prisons

9

Fifth, there was no attempt to document how the private sector had employed innovations in
maintaining or improving quality while at the same time holding down or maintaining costs. There is
not even any reference as to how the respective institutions were staffed. While not stated directly in
the report, it even appears that private sector innovation was deliberately thwarted by making the
private sector provider, CCA, abide by TDOC policy in running SCCC. In other words, it appears that
Tennessee took the position that SCCC was simply another TDOC facility, to be run by TDOC policy,
and that CCA would simply be given the opportunity to see if they could out-TDOC the TDOC. If this
were the case, then obviously this limits the knowledge that can be gleaned about the benefits of
privatization. Additionally, the design of the facility was set by the state, and the private contractor
(CCA) did not have the opportunity to incorporate potential design efficiencies into the facility.
Presumably, though, design considerations would impact more upon cost than quality.
Finally, there is no mention in the evaluation about the consequences of privatization for the TDOC
and how TDOC operations may have changed.

Washington State Review
As part of a wider inquiry into the privatization of government services, the Legislative Budget
Committee (LBC) of the state of Washington was asked to submit a report by January of 1996 on the
feasibility of privatizing Washington State Department of Corrections facilities. Part of the feasibility
study included an examination of quality issues.
The researchers did not collect original data for their assessment of the impact upon quality created by
contracting for correctional services. Instead, they reviewed studies conducted by Logan (1991) and
the Tennessee Select Oversight Committee (1995). Both of these studies are reviewed at length
elsewhere in this report.
The LBC also performed cost analyses of public and private prisons in Tennessee and Louisiana.
They collected published data for conducting this component along with additional data from the
respective state agencies and private contractors. The LBC researchers also went onsite to observe
operations. While we are not concerned with the cost analyses here, they did make some qualitative
observations about operations in Tennessee and Louisiana. It is these observations that are of interest
here.
Review of Quality
The LBC researchers concluded that the Logan (1991) and Tennessee (Tennessee Select Oversight
Committee on Corrections 1995) studies demonstrated no significant differences in quality between
the publicly and privately operated prisons. They actually do not provide any information about how
they reached this conclusion. This is surprising since their conclusion is at odds with Logan’s claim
that the private contractor provided better quality than the state and federal prisons in his study.

Appendix 2

Private Prisons

10

Qualitative Observations of Operations in Louisiana and Tennessee
The LBC researchers examined several questions to address the issue of whether similarities exist
between the inmate populations and behaviors both within and between states. The LBC sought to
examine whether the experiences with privatization in Louisiana and Tennessee could be generalized
to the state of Washington, and this led to the inter-state comparisons. For our purposes, the within
state comparisons are of more interest as they address how the private and public prisons compared in
Louisiana and Tennessee.
The LBC researchers concluded that inmates in the public and private prisons within each state
behaved about the same. They based this conclusion in part upon an examination of the number of
escapes, the major infraction rate, the minor infraction rate, and the percentage of inmates in school.
In part, though, it appears that the conclusion was based upon subjective evaluations. The LBC
researchers note that “(t)here were comments made to us in both Louisiana and Tennessee about a
belief of under reporting of infractions and incidents at the private prisons, but headquarters
administrators said they thought all of the prisons were safe and secure” (Thomas, Gookin, Keating,
Whitener, Williams, Crane, and Broom 1996: A4-4). In fact, in Louisiana, both the major and minor
infraction rates were higher at the state-operated prison, although little emphasis is given to this fact.
Other areas briefly covered in the Washington study are demographic and criminal history
characteristics of inmates, classification policies, inmate idleness, and program opportunities. Most of
the relevant comparisons for these factors were between states. However, the LBC researchers did
note that the private prison in Louisiana run by CCA had a lower percentage of inmates enrolled in
education as a result of CCA losing federal grant monies upon which they were dependent. Otherwise,
the public and private prisons were seen as fairly comparable. As the LBC researchers note: “The
prisons we visited appeared clean and orderly. ... Staff were professional both in appearance and
performance. There did not appear to be major differences in operations.” (Thomas et al. 1996: A4-9)
There are several shortcomings to the LBC report with regards to quality assessment. First, the LBC
review depends upon reviewing performance data (such as escapes, infractions, etc.), but it is not clear
how the individual data elements were pulled together to reach a general conclusion. Likewise, even
though there is recognition that reported incidents depend upon the nature of inmates and reporting
procedures, there was no attempt to adjust rates for these factors.
Second, as discussed in the reviews of the studies by the Tennessee Select Oversight Committee
(1995) in Tennessee and Archambeault and Deis (1996) in Louisiana, the evaluations do seem to
provide more apples-to-apples comparisons than are generally found in other evaluations. However,
unanswered questions remain about how much the institutional averages for inmate behavior are
influenced by race differences (e.g., Tennessee) or classification differences (e.g., Louisiana).
Third, the analysis does not address the time dimension. Generally, even though they review data from
two separate states, the data are for a single point in time. This does not allow for an assessment of
how quality in the public and private sectors may vary over time.
Fourth, there was no systematic attempt to obtain information on sources of innovation available to
private sector operators to improve quality. This was not the case for private-sector costs. The LBC
does attempt to disaggregate sources of cost savings for private-sector operators into savings from
using different staffing patterns, savings from providing different employee benefits, and savings from
Appendix 2

Private Prisons

11

salaries. In general, they claim that the private-sector operators use staff in more than one area, and
they have more flexibility in using staff. In addition to providing little detail about these claims, the
LBC researchers do not go on to discuss how these changes affected quality in the respective prisons.
Finally, there is only limited information on the impact of privatization on public-sector operations in
Tennessee and Louisiana, and most of those insights pertain to costs. In noting that the cost savings in
Louisiana have become smaller over time, the LBC offered this possibility (but no direct evidence):
One explanation for the convergence of costs over time may be the effect of
competition. This is an argument made by the private companies that was also
mentioned by some state correctional officials. Lean budget years may also have
made a difference. For some years the inflationary increases built into the private
contracts has been greater than the increases in the corrections budget. So while the
per diem cost for the private prisons has inflated, it has not inflated for the public
facility (Thomas et al. 1996: 12).
It is worth emphasizing that the Washington State study is a presentation and analysis of data that can
also be found elsewhere (Archambeault and Deis 1996; Tennessee Select Oversight Committee on
Corrections 1995). While the results of the LBC study generally mirror those of the Tennessee Select
Oversight Committee about the experiences with privatization in Tennessee, the LBC conclusions
about the experiences in Louisiana conflict with the data analysis presented by Archambeault and
Deis (1996).

Arizona Evaluation
The Utah-based firm Management and Training Corporation (MTC) won a contract from the state of
Arizona in 1993 to operate a 450-bed minimum security, mixed gender institution. The institution is
now known as the Marana Community Correctional Treatment Facility. MTC began receiving inmates
in October 1994. Generally speaking, the contract stipulates that MTC run the Marana facility in a
manner similar to that in which the state would have operated the prison. MTC must abide by all
applicable policy stipulating how Arizona state prisons are run (Nink 1998). By law, the contract with
MTC could not be renewed unless there was evidence of either 1) cost savings and comparable quality
or 2) comparable costs and superior quality. Dr. Charles Thomas was selected to conduct the
corresponding evaluation.
Thomas was faced at the outset with a number of serious methodological problems. First, the Marana
facility is a dual gender facility, but there are no dual gender facilities operated by the state of Arizona
to serve as a point of comparison. Twenty-two percent of the inmates incarcerated at Maranna are
female. Second, the prisoner population profile is different from those of the publicly operated
minimum security prisons in Arizona. In particular, Marana houses a much higher percentage of DWI
inmates (24.8 percent) than is true of all other Arizona facilities (with the exception of Papago, 98.9
percent, which is almost exclusively a DWI center). Also, the Arizona Department of Corrections
contractually agreed not to send to Marana prisoners who “have serious or chronic medical problems,
serious psychiatric problems, or are deemed unlikely to benefit from the substance abuse program”
(Thomas 1997a: 73). Third, the classification of the inmate population “tilts” toward being less
serious at Marana than at the other Arizona facilities (Thomas 1997a: 106). In particular, all of the
inmates at Marana are public risk 1 or 2 inmates (with 1 being the lowest risk), as they are, for the
most part, at the other Arizona prisons. However, whereas the other Arizona prisons have inmates
Appendix 2

Private Prisons

12

with internal risks greater than 2 (i.e, inmates who require more supervision), Marana has practically
no inmates with an internal risk factor greater than 2 (Thomas 1997a: Appendix A, Table 3: A9). This
is important as higher risk classification scores are generally predictive of misconduct. In other words,
14.11 percent of public risk level 1 or 2 inmates at publicly operated Arizona prisons have an internal
risk greater than 2. Only 0.24 percent of the inmates at Marana have a similar classification. Fourth,
the contract with MTC calls for providing a “heavier” load of programming to Marana inmates than is
the case in publicly-operated minimums. And, finally, the physical design of Marana is unlike that of
other Arizona facilities, primarily because it is newer.
Within these constraints, Thomas decided to compare cost and quality performance measures at
Marana against the average for all minimum security institutions under public operations. Arguably,
Thomas did not have many good options, but it is not clear that the choice he made is completely
satisfactory. Very different types of institutions make up the full contingent of minimum security
facilities in Arizona. We reserve this discussion for later. As Thomas notes in his chapter on
comparing quality, “Any or all differences could be caused by nothing more or less than the
fundamental dissimilarities between the Marana Community Correctional Facility and the fifteen
state-operated facilities” (Thomas 1997a: 108).
Even recognizing the serious limitations to the study, Thomas drew 13 conclusions from his study, 7
of which pertain directly to the issues associated with quality as determined in the Arizona evaluation.
We largely ignore Thomas’ review of existing literature on quality comparisons as the relevant studies
that Thomas reviewed are also reviewed as part of this report. We also exclude his conclusions about
cost. Thomas’ seven conclusions about quality of operations in Arizona are worth reporting in full,
however. They provide the basis for our examination of the methodological issues raised in the
Thomas study. The conclusions appear in full in both the Executive Summary and body of the report
(Thomas 1997a: ii-iv, 157-159):
Conclusion #4: There is a high risk that operating cost and performance comparisons
of the Marana Community Correctional Treatment Facility could yield misleading
results because there is no state-operated prison in Arizona that houses inmates
similar to Marana or runs similar programs.
Conclusion #7: The performance comparison on the dimension of protecting the
public safety interest as measured by the frequency of escapes, major disturbances,
and injuries caused to visitors revealed that the record for the Marana Community
Correctional Treatment Facility was superior to that of the state-operated Level Two
(minimum) prisons.
Conclusion #8: The performance comparison on the dimension of protecting staff and
prisoners from the risk of personal injury or death caused by homicide, battery,
assault, and arson revealed that the record of the Marana Community Correctional
Treatment Facility was superior to that of the state-operated Level Two prisons.
Conclusion #9: The performance comparison on the dimension of educational,
treatment, and work programs resulted in a best professional judgment that the
dissimilarities between the programs offered at the Marana Community Correctional
Treatment Facility and those found at the state-operated Level Two prisons were so
great that no fair comparative conclusions could or should be reached.
Appendix 2

Private Prisons

13

Conclusion #10: The performance on the dimension of compliance with professional
standards as measured by routine Department performance audits, litigation initiated
by either prisoners or staff members, inmate grievances, and compliance with inservice training requirements for staff members revealed that the overall record of the
Marana Community Correctional Treatment Facility was superior to that of the stateoperated Level Two prisons.
Conclusion #11: A balanced consideration of the entire set of individual performance
indicators revealed that the overall performance record of the Marana Community
Correctional Treatment Facility was superior to that of the state-operated Level Two
prisons.
Conclusion #13: Notwithstanding the conclusion that, when compared with all stateoperated Level Two prisons, the quality of performance at Marana was superior to
that of the state-operated prisons, it was found that one or more individual stateoperated prisons had performance records that were equivalent or superior to that of
Marana.
Thomas justified drawing these conclusions by pointing to the necessity to have the information for
policy purposes (Thomas 1997a: 103). Also, he claimed that the limitations cause the findings only to
have “fuzzy” rather than “crisp” edges (Thomas 1997a: 104).

Nonetheless, Thomas recognized the risk associated with the comparisons he makes and that while
Marana fared the best in his comparisons against the average for Level Two prisons in Arizona, some
individual prisons had better performance measures than Marana. To Thomas’ credit, he did report the
performance measures individually for the 15 publicly-operated prisons.
Of the conclusions about quality listed above, Conclusions 7, 8, and 10 are based on direct
examination of data presented by Thomas. As such, we review each of these conclusions in more
detail. Thomas’ other conclusions regarding quality are more subjective or global, and we do not
discuss them specifically, although these quality conclusions (Conclusions 4, 9, 11, and 13) are
reviewed in general.
Conclusion 7, the conclusion about public safety, is based on comparing rates of major incidences,
escapes, and injuries to visitors. Thomas concluded that Marana’s performance was “superior.”
However, as Thomas noted, none of the facilities (including Marana) had a single major disturbance,
only 4 escapes occurred at the 15 state facilities (3 from Papago where the DWI cases are
incarcerated) while there were none at Marana, and no injuries to visitors were reported at any
minimum security prison (including Marana). None of these findings are surprising given that the
facilities house minimum security inmates. What is surprising is that Thomas concluded, even with
the cautionary remark in the body of the report that “readers must refrain from making more of this
difference than is fair and reasonable” (Thomas 1997a: 111), that Marana was superior to the state
facilities because 2 of the 15 state-run facilities had escapes.

Appendix 2

Private Prisons

14

It seems that reasonable commentators could conclude that Marana was exactly equivalent to 13 of the
15 state-operated facilities on the dimension of public safety, and that Marana -- and the 13 publiclyoperated prisons -- were marginally better than two of the publicly operated facilities. The term
marginal seems appropriate because there was a difference on only one of the three indicators, and the
frequencies for the one indicator were not unusual. While escapes are rare, they do nonetheless occur
(from private as well as public prisons).
Similar problems exist for Conclusion 8 about protecting staff and inmates. Thomas again gave ample
warning about the tenuous nature of his conclusion, nonetheless concluding that Marana was superior.
The rates of the types of serious misconduct considered by Thomas are very low at any minimum
security prison. In essence, most of the forms of misconduct (like inmate-on-inmate assault), occurred
only a few times -- if at all -- and at a limited number of institutions. Thus, comparisons of Marana to
15 other facilities is fraught with difficulties, especially when one considers that 22 percent of the
inmates at Marana were female, a group with low rates of the types of misconduct considered.
In addition to Marana, three publicly operated facilities had perfect records on offenses related to
protecting staff and inmates from serious injury. Two of the three publicly operated facilities (Globe
and Maricopa) housed male inmates.
Thomas reported the rates of all Type A offenses (serious) charged and found guilty at each of the 16
prisons in his study. (He also reported comparable information for the less serious Type B and Type C
offenses, but we ignore these for our purposes, since less serious offenses typically involve a great
deal of discretion and, therefore, are often unreliable indicators of the “true” underlying pattern of
misconduct). What is interesting about the Type A offenses is that they occur often enough to provide
some confidence in the respective rates, although collapsing all serious offenses together means
information is lost about what types of offenses are being reported at the respective institutions.
The rate at which Type A offenses are charged at Marana (0.33) compared very favorably to the rate
(0.55) for the average of the 13 publicly operated facilities listed in Table 6 (Thomas 1997a:
Appendix A, page 12). Nonetheless, in addition to the two female minimum-security institutions, two
other male institutions (Globe and Piacho) had lower or comparable rates of Type A offenses (0.20
and 0.33 respectively). When we factor in that 22 percent of Marana inmates are female and females
commit almost no Type A offenses (the rates for charging inmates at the female prisons in Arizona
were 0.04 and 0.00), the rate of 0.33 at Marana is not as impressive.
Assuming that the 22 percent of the females at Marana were charged with Type A offenses at the rate
of 0.04 (the highest rate noted in an Arizona female, minimum-security prison), then males at Marana
were charged for Type A offenses at the rate of 0.41. At this adjusted rate, we see that Globe and
Piacho compared favorably to Marana, as did San Pedro where the rate was 0.37. This still means that
Marana was doing well, with a lower rate than 8 of the male, publicly operated minimum-security
facilities. Nevertheless, Marana was higher than 3 of the publicly operated males facilities, and it was
obviously higher than the rate at the two publicly operated female facilities. At best, Marana is
comparable to the best public facilities, but it is hard to argue that Marana is superior if we trust the
data on Type A offenses.
Regarding Conclusion 10, the factors considered included the comprehensive annual quality audits
conducted by Arizona DOC (including an audit of Marana), litigation filed against the prison, inmate

Appendix 2

Private Prisons

15

grievances, and compliance with staff training requirements. Marana was the only minimum security
institution that did not receive an overall rating of excellent on the audit. Thomas implied, however,
that the internal audits may have been stacked against Marana. Regarding litigation, there were only
14 lawsuits initiated by prisoners. This involved inmates at 8 of the 15 publicly operated facilities and
none at Marana. There were two lawsuits initiated by staff, both at publicly operated prisons. For the
rate of inmate grievances, the rate at Marana (0.26 per 100 inmates) was second only to Papago where
the rate was 0.20. On the dimension of training, there was no difference between the publicly operated
prisons and the Marana facility operated by MTC.
Despite the mix of evidence on performance -- including the fact that Marana did not score as highly
as some state-run facilities, that some kinds of infrequently occurring events transpired at some of the
fifteen state facilities but not at Marana (e.g., litigation), and the problematic audit review -- Thomas
nonetheless concluded that, on balance, a superior ranking “must” be assigned to Marana.
Thomas was faced with a host of methodological problems, some of which may have been
insurmountable. Nonetheless, an evaluation was mandated before the contract could be renewed, and
as Thomas correctly noted, the conditions of policy reviews seldom meet the pristine conditions of
laboratory research. In this context, let us review then how this research effort fulfills the conditions
of sound comparative research.
First, the review seemed to use both performance and audit data, relying most heavily upon
performance data. The audit data came from the normal audit cycle of the Arizona Department of
Corrections. As such, the audit data are suspect as Thomas implied in his analysis. Unlike the audits
performed as part of the Tennessee evaluation (Thomas 1997a: 73), where the protocols were
developed by both public and private officials, and where audit teams were comprised of members
from both the public and private sectors, the Arizona auditors were all Arizona Department of
Corrections employees conducting a standard Department of Corrections audit. At the very least, this
creates the appearance that bias may have been built into the audit of Marana. In our opinion, Thomas
was correct to not emphasize the audit results in his analysis. However, we are not so sanguine about
his uses of the performance data, as noted in the review above.
Second, it is abundantly clear that this evaluation does not provide an apples-to-apples comparison.
Thomas himself was aware of this fact, as was the Arizona Department of Corrections when they
contracted for the evaluation. It seems there was no comparable facility to Marana in the entire
Arizona prison system. Still, this does not justify the strategy Thomas followed of comparing the
Marana facility to the average of the other publicly operated facilities. If an average had been used, it
should have been an average of facilities that most closely resembled Marana. Using Thomas’
approach, the comparisons are not that informative. More informative are the comparisons made in
the body of the text of Marana to individual state facilities, but these comparisons do not provide the
basis for the findings presented in his 13 conclusions. For the most part, the 13 conclusions presented
by Thomas are based upon comparisons of Marana to the averages for the 15 publicly operated
facilities with the exception that Thomas does point to the facility-by-facility comparisons as
conditioning the other conclusions (see Conclusions 4 and 13).
Third, the study suffers from being a one-point-in-time comparison. A problem not noted above, but
which may be relevant, is the timing of the study. Generally, the Marana facility was in an activation
phase for most of the course of the study, which probably had some influence upon the types of

Appendix 2

Private Prisons

16

behaviors observed there. Whether patterns noted for this time period will hold as the facility matures
is unknown.
Fourth, there was no attempt to document how MTC employed innovations to facilitate changes in
quality (or cost for that matter). In part, this inattention may have resulted from the model of
privatization followed in Arizona. Basically, the State of Arizona has taken the position that a private
contractor should be given the opportunity to demonstrate it can out perform the state in running an
Arizona prison according to Arizona Department of Corrections policy. Some states take the position
that a private contractor should be given a great deal of flexibility and freedom in developing its own
foundation of policy and procedures, while holding the contractor to some minimum standards such as
those developed by ACA. While their approach may make some things easier for the Arizona
Department of Corrections (conducting audits for one thing), it certainly provides little room for the
private contractor to innovate. This means that differences in quality must arise from better
performance by the private contractor’s line staff and management.
Fifth, there was no attempt to assess whether the operation of the Marana facility by MTC has brought
about system changes to the Arizona Department of Corrections.

Louisiana Evaluation
The State of Louisiana has funded an evaluation of its prison privatization efforts. Archambeault and
Deis (1996) have produced a report comparing 3 institutions which have the same architectural
design, accommodate approximately the same number of inmates, were built and activated at
approximately the same time, are located in rural areas of the state, and house inmates of comparable
security levels. Allen Correctional Center is operated by Wackenhut Corrections Corporation (WCC).
Avoyelles Correctional Center is operated by the Louisiana Department of Public Safety and
Corrections, and Winn Correctional Center is operated by the Corrections Corporation of America
(CCA).
The primary purpose of the Archambeault and Deis evaluation was to compare the two privately
operated and one publicly operated facilities on measures of cost and performance. The performance
data used in this report came primarily from official records normally reported to the Secretary of
Public Safety and Corrections by all adult correctional institutions within the state. The researchers
made one field visit to each of the three sites. Based on the site visits and supporting documentation,
Archambeault and Deis characterized the management philosophies existing at the time of the study.
They rated the Wackenhut operated facility as most authoritarian, the publicly operated facility as
intermediate, and the CCA operated facility as the least authoritarian of the three. At the time of the
study, WCC employed 335 staff, the public facility 384 staff, and the CCA facility 340 staff. The
CCA facility employed more women and minorities than the other 2 prisons, while the public facility
employed predominantly white and male staff.
According to Archambeault and Deis, the Louisiana Department of Public Safety and Corrections
Secretary, Richard Stalder, standardized the policies and procedures of the three prisons in 1992 and
instituted a common reporting procedure that was used by the evaluators as their primary source of
data. In a letter to Archambeault, written by Robert C. Thomas, the Principal Management
Auditor/Supervisor for the State of Washington’s Joint Legislative Audit and Review Committee,

Appendix 2

Private Prisons

17

Thomas expressed his concerns about the validity of the reporting mechanism (Thomas 1997b). In his
letter, Thomas raised the possibility that “...different incentives can be at work in the private facilities
than in the public facilities (p. 3).” Although we agree with this criticism, without further
information, it is difficult to sort out the relationship between reporting incentives and the actual
behavior underlying these reports. While private prisons may have an incentive to under report
unfavorable incidents to bolster the impression that they are performing well, there is a
counterbalancing incentive for private prisons to report all significant incidents as failure to report
could provide grounds for contract termination. On the other hand, a public facility may want to over
report incidents as an argument to receive more resources, unless the public facility is in
“competition” with private sector providers of prison services.
Because it is difficult to disentangle accurate reporting from this subtle incentive structure, we assume
for the present purposes that the reporting mechanism is a true reflection of underlying conduct.
Nonetheless, we prefer that reporting mechanisms occur more as a by-product of institution operations
than as an additional requirement. As an example of what we mean, consider two separate reporting
mechanisms used by the Bureau of Prisons for representing serious inmate misconduct. One
mechanism is similar to the one adopted by Louisiana. On a daily basis, institutions report any serious
incident (e.g., fire, serious assault, disturbance) to both the Regional and Central Offices. There is a
certain amount of discretion in what people choose to report. The second mechanism of reporting
these incidents occurs as a by-product of the normal adjudication process that occurs in the Bureau of
Prisons institutions. A serious assault will typically result in an administrative action culminating in
an adjudication of the misconduct by a disciplinary hearing officer (assuming that the assailant is
known). Whether or not there is a sustained finding (conviction), all of the procedures and actions will
be recorded in an automated database of disciplinary hearings. The Bureau is able to cull aggregated
information on disciplinary actions for any and all of its institutions from this individual level
database. We have data demonstrating that this latter mechanism has much less reporting discretion
than the former.
To establish comparability of the three institutions, Archambeault and Deis assessed information on
the physical plant, its history of expansions, State-imposed policies and standards, comparability of
ACA standards, and the number and ethnicity of inmates during the evaluation period. While
acknowledging that a limitation of their study was that they lacked information on the characteristics
of the inmate populations, the evaluators minimized this limitation by stating that an “ideal evaluation
design” would include such data. We would characterize information on the characteristics of the
inmate population as fundamental in a comparison of institution performance measures. Almost every
institution performance measure used by Archambeault and Deis was a summary or aggregate
indicator of individual performance data. Since it is well known that inmate behavior whether
prosocial or antinormative is strongly correlated to criminal history, demographics (age, race,
ethnicity), and other social and actuarial predictors, these background data are a sine qua non of valid
institution performance comparisons. What may appear to be differences in institution performance
may be nothing more than differences in the background characteristics of inmates housed in the
different institutions that predispose them toward more or less favorable behavior.
Over 200 measures were collected and analyzed by Archambeault and Deis. These measures covered
issues about safety, risk, effectiveness, performance, and cost. The data covered several years (fiscal
years 1992 through part of fiscal year 1995) and represented monthly occurrences of the phenomenon.
Typically, the researchers would compute an average of the monthly data and use that average as the
Appendix 2

Private Prisons

18

performance indicator. In some cases, the researchers compared averages for different periods of time
when they thought changes in reporting or other artifactual problems might influence the data.
Unfortunately, the researchers did not capitalize on the trend data to evaluate performance over time.
The performance measures included data on escapes, assaults on staff and inmates, assault outcomes
(i.e., whether they involved serious injury), sex offenses, aggravated sex offenses, institutional
disturbances, deaths due to violence, suicide, or illness, disciplinary actions, gunshots, grievances,
drug tests, communicable diseases, inmate education and vocational training participation, GED’s
earned, basic literacy, college participation, and various indicia of medical care. Staffing data
included use of sick and maternity leave, resignations, and grievances. Each performance measure was
analyzed separately; however, scales of these measures were also constructed and analyzed.
As an example of how Archambeault and Deis evaluated the three institutions, consider the first
performance measure—escapes from the institution. Over the study period, the public institution
reported no escapes, the CCA-operated facility reported 5 escapes and the WCC-operated facility
reported 3 escapes. Archambeault and Deis submitted the month-to- month data to an analysis of
variance and contrasted the three average escape rates using a Scheffe post hoc comparison technique.
Using these methods, there was no statistical difference among the average of the monthly escapes
over the study period. The F- and Scheffe tests used by these researchers were probably inappropriate
techniques to compare such rarely occurring events and since they were measuring population
differences, not samples, the use of any statistical test to compare the averages may have been
inappropriate. Even more important is the interpretation of the absolute differences in the number of
escapes for the 3 institutions. Protecting the public is one of the most significant missions of a prison
system and clearly the public institution was doing a much better job than either of the two private
institutions. The researchers do acknowledge the public institution’s superiority but also claimed that
“each of the three prisons are fully meeting the obligation of protecting the public (p. 121).” From our
point of view the two private prisons were deficient in this obligation.
In chapters V through X of their report, Archambeault and Deis analyzed the remainder of nearly 200
performance measures. This was clearly an ambitious assessment of the comparative quality of the
three institutions. On some of the more important dimensions, data were also collected and reported
on the four other adult correctional institutions in Louisiana: Hunt, Dixon, Angola, and Wade.
Comparisons involving all 7 institutions were made primarily on serious misconduct such as escapes,
serious assaults on staff, inmate on inmate sex offenses, other serious inmate offenses, and
institutional disturbances. While the three primary comparison sites had almost identical inmate
population counts over the period of this study, the other four institutions used in some of their
comparisons had widely discrepant population counts. It appears to us that Archambeault and Deis
reported these performance measures without accounting for the large differences in the average
monthly inmate populations of the other four comparison institutions. Angola had an average monthly
inmate population almost 4 times as high as the 3 comparison institutions. Another institution, Wade,
had an average monthly population almost 20 percent lower than the 3 comparison sites. The
comparison of raw averages is meaningless given the huge disparity in the monthly inmate population
counts.
Since these serious misconduct measures were important indicators and the raw data were available,
we re-computed the performance measures by taking the data reported in Archambeault and Deis and
dividing the average monthly counts by the average monthly inmate population then multiplying the
Appendix 2

Private Prisons

19

result by 1,000. This results in a performance measure that is an average monthly rate per 1,000
inmates. A completely different picture emerges when you compare the raw counts, i.e., the average
number of incidents versus the rates adjusted by the average monthly population. Table 1 represents
the monthly counts, Table 2 the rates. Whereas Archambeault and Deis reported that the risk of
assaults by inmates on staff resulting in serious injury is marginally higher at the public facility than
the two privately operated facilities, Table 2 indicates that at least one other public facility had an
equally low rate of such assaults. The data in Tables 1 and 2 represent information on staff and inmate
safety. We are not confident that a statistical test is warranted, thus, for each performance indicator,
we have color coded the best rate in blue, the second best rate in green, and the worst rate in red. It is
clear that when the data are represented as counts, Dixon (public), and perhaps Allen (WCC) and
Avoyelles (public), were the best performing institutions on these performance indicators. Angola
(public) was the worst by far. Looking at Table 2, using the more appropriate rates, Dixon (public)
was still the best, Hunt (public) perhaps the second best, and Angola (public) -- while still the worst -did not look nearly as bad. It is not clear why the penitentiary at Angola should be included in any of
these analyses, since by reputation, this facility houses the most hardened prisoners in the Louisiana
system. Data on the background characteristics of these inmates would go a long way toward
answering these issues of comparability.
We have also included other counts and rates in Tables 1 and 2, for which there is no information
reported by Archambeault and Deis on the other public institutions, as well as information on total
disciplinary actions and gunshots. We did not color code disciplinary actions and gunshots because,
unlike Archambeault and Deis, we are not convinced that lower rates indicate better performance.
Archambeault and Deis cite Clear and Cole (1997) who argue that high levels of disciplinary
infractions indicate staff and inmates are intolerant of each other. We can think of other scenarios in
which high levels of misconduct can be interpreted as a sign that management is strict about enforcing
rules but fair in the disciplinary process. Alternatively, low levels of misconduct may indicate a lax
environment, and tolerance of nuisance behaviors that invite more serious misconduct. Misconduct
data can be very misleading unless we are knowledgeable of the institutional context. Looking at
Table 2, we can see that Angola, which had the worst record with respect to serious incidents, had the
lowest overall misconduct rate in that time period. Dixon, which had the best record with respect to
serious incidents, had a total disciplinary rate intermediate to the lowest and highest rates among the 7
institutions. If misconduct were related to serious incidents, we would expect institutions with the
highest misconduct rate to have the highest serious incident rate. The data argue against any
conclusion that low rates of misconduct indicate better performance.
The urinalysis “hit” rate was another important indicator of institution safety and security although
Archambeault and Deis treat the indicator as a measure of health risk. The rate at which inmates are
randomly detected using drugs is an indication of the extent to which an institution is able to control
the importation and use of drugs. These data were only available on the 3 comparison institutions. The
data showed that the public institution was much more likely to use urinalysis to test for drugs and
that the random tests revealed that the public institution had far fewer inmates testing positive for drug
use than either of the privately operated facilities. Overall, these data seem to indicate that each
institution had certain safety and security deficiencies and each had certain strengths; however, the
best performing institution was a publicly operated one during the period of this study while the worst
was the notorious penitentiary at Angola. All of these conclusions should be tempered by the fact we
have no basis for determining whether these differences resulted from institution management or

Appendix 2

Private Prisons

20

merely from housing inmates with different propensities toward violence and other serious
misconduct.
Other data were collected by Archambeault and Deis to indicate health risk. Information on
communicable diseases, including HIV infection and tuberculosis, were analyzed. Archambeault and
Deis used information on the incidence and prevalence of HIV and tuberculosis. They used various
indicators of HIV and AIDS which depended on the level of seriousness of the infection and the year
in which the data were recorded. For example, they separately assessed group I, group II, group III,
group IV-A, group IV-B, group IV-C, group IV-D, group IV-E, and the number of inmates receiving
AZT for fiscal years 1992-1993. In fiscal years 1993-1994, they categorized the infection into separate
subgroups that depended on the T-cell count and whether the cases were characterized as category A,
B, or C and levels 1,2, or 3. Similar measures were compiled on tuberculosis cases. The researchers
also created a composite health risk of these measures and included the drug testing measures. This
resulted in an index composed of 60 indicators.
They found that the WCC-operated facility had the highest medical risk score, the CCA-operated
facility the second highest, and the public facility the lowest. We question the validity of this
composite health risk predictor. Unless they could measure the quality of care delivered by the
different institutions, it is difficult to credit or fault an institution for its level of health problems.
Health problems can be related to the demographic characteristics of the inmate population and totally
unrelated to the health care delivery system at the institution. The drug assessment did indicate the
public institution had the lowest drug utilization rate and the highest use of random and total number
of drug tests and the lowest percentage of inmates testing positive among the random tests. This is
evidence that the public institution was actively combating drug use at least partly through the use of
testing.
In Chapter IX of their report, Archambeault and Deis assess the work environment of correctional
personnel. Their first set of indicators examines the relationship between the number of positions
allocated and the actual number of staff working on a daily basis. They found that the public
institution had more positions allocated to custody but a lower proportion of those positions were used
on a daily basis. In their comments, Archmbeault and Deis provide the explanation for this
discrepancy. When the public facility opened, many of the personnel transferred from other state
institutions. Employees with more tenure accrue more leave and therefore, more staff were needed to
cover the same shifts. Archambeault and Deis did not have sufficient data to investigate this
possibility.
Chapter X of their report examined the use of administrative remedy procedures by the inmates,
indicators of education and vocational opportunities, inmates’ medical duty status, and the number of
inmates evaluated and transferred to pre-release programs and community correctional centers. The
WCC-operated facility had the highest monthly total of administrative remedy complaints. The public
and CCA-operated facilities were almost identical in those rates. The public facility was the least
likely to accept a complaint.
The CCA-operated facility had the most complaints regarding medical care and quality of life. The
WCC-operated facility had the most complaints regarding property, legal issues, rules, threats,
communication, records, finances, institution programs, discipline, and classification. The public
facility had the most complaints about discrimination and protection.
Appendix 2

Private Prisons

21

Data on program enrollment and completions in education and vocational training programs indicated
that the CCA-operated facility was doing a better job in providing educational and vocational training
to inmates than either the public or the WCC-operated facility. However, the authors noted that the
public institution involved more inmates and volunteers in the education process, had more inmates
enrolled who completed the basic literacy program, and had more inmates enrolled in college courses.
Archambeault and Deis concluded that the CCA-operated facility was far more effective in its
education performance than the WCC-operated facility or the public facility. This conclusion was
partly based on efficiency ratings in which the CCA-operated facility had higher rates of completions
on some indicators.
Finally, Archambeault and Deis collected data on the duty statuses for inmates and whether the
institutions screened inmates for transfer and community rehabilitation and work centers. The
CCA-operated facility had the most occurrences of screening and highest ratio of transfers to the
community centers. The CCA-operated facility also had the highest ratio and most inmates on some
form of limited duty status. In the absence of any additional contextual information, it is almost
impossible to make any sense of the limited duty status data.
Archambeault and Deis concluded that the two private prisons outperformed the public institution on
most of the performance measures, including critical incidents, safer working environment for staff, a
safer environment for inmates, judicious use of disciplinary actions, more efficient use of security
personnel, higher inmate program participation, and higher use of inmate transfer to community
settings. The public institution was credited with outperforming the private institution on measures of
escape, aggravated sex offenses, substance abuse, the breadth of education and vocational training in
the number of inmates served, and the breadth of treatment, recreation, social services, and
habilitative services to inmates.
As we have already noted, a re-analysis of the safety and security data using the information on all of
the Louisiana institutions shows that the public facilities were the best and the worst. However, this
conclusion and any other conclusion should be strongly tempered since Archambeault and Deis were
unable to collect information on characteristics of the inmate population that could have been used to
control for potential inmate population differences among the institutions. We suspect that the
penitentiary at Angola probably houses the most hardened and violent prisoners incarcerated by the
State and that there are structural reasons (both architectural and sociological) that make it a difficult
prison to manage.
In Thomas’ (1997b) letter (mentioned earlier), he was also critical of Archambeault and Deis‘
conclusion on the role of privatization for other jurisdictions, although as far as we can tell, that
conclusion was misstated in the Thomas letter. Thomas wrote that Archambeault and Deis claimed,
“Private prisons have a definite place in any state’s total prison system (p. 3 of the Thomas letter).”
The closest comment we can find in the Archambeault and Deis report is the statement “Privately
operated prisons have a definite place in the planning of any state’s total prison system (p. 573).”
Thomas argued that the conclusion was unwarranted by the study and that a private alternative may
not be in the best interests of the state. In fact, the Archambeault and Deis conclusion was more
guarded and those authors went on to state that privatization of prison beds should be limited to
ensure that the incentive to compete is not lost.

Appendix 2

Private Prisons

22

The intent of the Archambeault and Deis evaluation was strictly to compare the public versus the
private operation of the facilities. If there is an innovation to be found, it might be the staffing patterns
used by Corrections Corporation of America. The public facility used predominantly white (81.5
percent) males (76 percent). The CCA-operated prison used many more female staff (42 percent) and
many more minority staff (50.3 percent). The WCC-operated facility was somewhere in between,
using mostly white male staff (63 percent). Since staffing is one of the most important components of
correctional service, the use of more women and minorities might be considered an innovation if it
could be shown to be related to the successful management of the institution. If Archambeault and
Deis did point to an innovation it would have been the management philosophy of the CCA-operated
facility. They considered CCA’s approach to be the least authoritative, involving staff in the
organizational decision making and giving “employees a vested interest in the overall success of the
prison organization ( p. 66).”
It was difficult to conclude how or whether the Louisiana public correctional system had changed in
any way in response to the privatization of its two facilities. The Secretary of Public Safety and
Corrections did introduce a reporting system to monitor all of the institutions; however, this system
might have been implemented in the absence of privatization.

New Mexico Evaluation
The discussion of the New Mexico evaluation is taken primarily from the report by Harer, Karacki,
and Gaes (1995).
Charles Logan has published, "Well kept: Comparing quality of confinement in private and public
prisons," as both a National Institute of Justice monograph and a condensed version in the Journal of
Criminal Law and Criminology (Vol. 83, pp 577-613, 1992). Logan's comparison of private and
public institutions is based on contrasting the operations of two women's prison in New Mexico and a
Federal institution. Women prisoners in New Mexico were originally housed in a state facility which
operated primarily as a diagnostic and orientation facility for men. To meet the needs of the female
inmates, a private facility was built and operated by the Corrections Corporation of America (CCA)
under contract to the state. Logan later collected a limited amount of data from a Federal women's
facility to do a comparison of all three kinds of operations.
In both his monograph and paper, Logan first outlined a theoretical model, the Confinement model,
which he used to form the basis for comparisons among the three institutions. He then proceeded to
develop institution performance indicators (IPI's) which represented the various dimensions of his
Confinement model. Logan's Confinement model is succinctly represented in his own words:
The mission of a prison is to keep prisoners--to keep them in, keep them safe, keep
them in line, keep them healthy, and keep them busy--and to do it with fairness,
without undue suffering and as efficiently as possible. (Logan 1992: 580)
Thus, the dimensions which Logan used to compare institutions were security, safety, order, care,
activity, justice, conditions, and management. We have little quarrel with Logan's conceptual
framework, although we believe there are critical goals not adequately addressed by this model. In
Appendix 2

Private Prisons

23

Logan's approach to corrections, inmate programs are merely activities to "keep them busy." But, as
the mission statements of most modern correctional systems imply, rather than merely "to keep them
busy," prison programs are intended to provide inmates with a positive influence in their lives, afford
them the opportunity to improve their skills, and provide a socializing agent so that they come to
accept the moral and legal norms of society.
One of the fundamental problems with Logan's study is that with only three institutions to compare, he
was limited in the kinds of analyses that could reasonably be used. Furthermore, he was really
comparing the management of essentially the same New Mexico female inmate population over time
by two different management groups (public versus private) housed in two different facilities. Almost
as an afterthought, he contrasted this before/after longitudinal assessment to a population composed of
imprisoned federal women. The appendices to Logan's NIJ monograph also include reports, written by
two consultants, which described the state and private institutions and the events culminating in the
transfer of female inmates from the New Mexico State prison to the privately run institution.
Logan used a large number of performance indicators which were quantitatively combined; however,
the large number of measures used does not compensate for the very small sample under analysis.
Other problems with this study, however, arise from the ways in which Logan analyzed and
interpreted the institution performance indicators that he used to compare the operations of the three
facilities.
The Probable Bias in Institution Performance Indicators Based on Staff Perceptions
Almost all of the differences which favor the private facility over the two public facilities in the
Logan study were based on staff perceptions as measured by the Prison Social Climate Survey (Saylor
1984) administered to staff at the three institutions. There are many reasons why staff surveys could
have biased the results in favor of the private facility.
Clearly, staff at the private facility would be keenly aware that the success of their employment could
depend, in part, on the responses they provided to the researcher. This most obvious, and potentially
fatal flaw of the research, was discussed only superficially and then dismissed in the Logan report.
A second biasing influence was the fact that many of the staff selected for the private facility had
worked in the public facility previously. It is likely that they would want to justify their decision to
leave the state government to work in the private facility by favorably responding to the survey
questions.
A third biasing influence in the measurement of these perceptions was the fact that the private facility
was brand new, with many new staff as well as some experienced former state corrections staff. The
facility built by the private corrections company was considered to be very well designed and staff
were excited by working in this new environment. Yet, one wonders how their attitudes would have
changed over time as the challenge of working in a new and exciting environment gives way to the
daily routine of operating a correctional facility. We were so curious about the possible change in
attitudes of staff over time that we asked the Corrections Department of New Mexico if we could
re-administer the Prison Social Climate Survey at the private facility two years after the private
operation began. However, management at the private facility would not allow this.

Appendix 2

Private Prisons

24

Another important biasing influence on staff perceptions was the low response rate among state and
Federal employees. The private facility response rate was 72 percent while the two public facility
response rates were less than 50 percent. Data from the Federal facility were collected in the very first
year the Prison Social Climate Surveys were administered for the entire Bureau of Prisons. In
subsequent administrations, the Bureau of Prisons has achieved no lower than a 72 percent response
rate and as high as an 88 percent response rate from 1989 through 1994. Low response rates could be
construed as indicating that only staff who had negative perceptions of the institutions were motivated
to complete the Social Climate survey at the two public institutions.
The last important biasing influence in Logan's comparison across the three institutions and, perhaps
the most damaging, is that Logan did not have a full complement of measures for each institution for
computing an overall index of "quality." In particular, approximately 30 percent of the measures that
were used in the comparison were missing for the Federal facility.
To exemplify this problem, suppose we are asked to rank order three employees on their overall
performance based upon 10 performance indicators. For two employees, data are available on all ten
measures; however, for the third employee, we only have information on 7 measures. Knowing this,
should we still try to rank order all three employees on the overall performance index? Obviously not,
yet this is exactly what Logan proceeded to do.
Interpreting Institution Performance Indicators
Another problem with the Logan study was what we believe to be the often questionable way in which
Logan interpreted the institution performance indicators (IPI’s). In all, Logan used 333 IPI's, most of
which were based on staff perceptions measured by the Prison Social Climate Survey. For each of the
333 IPI's, Logan assigned a ++, +, =,-, or -- value to indicate whether an institution was much better,
better, equal, or worse, or much worse than the comparison institution(s). It is noteworthy that a set of
indicators and measures of correctional effectiveness derived from the professional judgment of
multiple correctional experts and practitioners was available to Logan, namely the American
Correctional Association's (ACA) standards. Logan chose to ignore the ACA measurement scheme,
preferring his own instead. Logan himself admitted that "Interpreting each measurement item was
often difficult." He also pointed out that many items could quite legitimately be interpreted and
scored in many different ways. Yet despite his initial cautious remarks, he proceeded to make specific
judgments, a priori, about whether a measurement item was positive or negative without consulting
correctional experts. We found some of his judgments somewhat naive. A few examples should
demonstrate this. We will focus here on only the relatively objective measures he culled from
institution records for the institution security dimension.
In his section on security, Logan examined a number of objective indicators related to inmate
contraband, drug use, misconduct, staffing patterns, and furloughs. Other than his inclusion of
furloughs, we think these are valid issues related to institution security. Unfortunately, the way these
measures were interpreted by Logan makes little or no sense from any sound correctional management
point of view.
Logan compared the state and privately run prisons by looking at the rate shakedowns (i.e.,
contraband searches) were occurring and the proportion of shakedowns in which contraband was
found. He gave the state facility credit for performing shakedowns more often; however, he gave the
privately run facility more credit for finding contraband a lower proportion of time. This latter
Appendix 2

Private Prisons

25

judgment makes little sense. It may very well be that the state officials were doing a better job of
finding contraband, were less tolerant of contraband, or both. These explanations would have led to an
opposite conclusion from Logan's. Whereas Logan saw merit in finding less contraband, it is also
reasonable to assume that finding more contraband is a reflection of thorough and well-trained
security staff..
A second example involved the privately run facility, as well as the state and federal facility. Logan
found that among inmates suspected of using drugs, fewer tested positive for opiates in the Federal
facility than in either the state or privately run institution. Yet, Logan gave the Federal institution a
demerit for conducting fewer tests among inmates suspected of drug use, while giving the privately
run institution the highest rating for conducting the most urinalysis tests. The important issue here is
whether inmates were using drugs in the institution. Conducting more tests might be construed as
wasting money. How could the Federal institution be given credit for having the lowest opiate usage
rate and simultaneously be judged inadequate in its drug testing policy while achieving a better result?
Another objective dimension reported by Logan was the rate at which inmates committed significant
misconduct incidents during the 6-month period of the study. This is an example where Logan used
his judgment, rather than conduct a statistical test. He found that the privately run prison which had a
0 rate of significant incidents was judged better than the federal prison which had a rate of .01
significant incidents in a 6-month period. These rates were not statistically different from each other,
yet in Logan's judgment, the privately run facility performed better on this measurement item. Since
there is so much reporting discretion in these kinds of incidents, we are not sure it is a fair assessment
to conclude that the privately operated facility had out-performed the publicly operated facility when
the rates for both facilities were so low.
In his analysis of institution security, Logan evaluated furlough rates as well. He claimed that
furloughs indicated the extent to which prison administrators exposed the community to dangerous
criminals. He reasoned that the higher the exposure (i.e., more furloughs), the lower the rating for the
institution. Many correctional administrators look favorably at furloughs as a way of easing an
inmate's transition back into society. Most correctional systems use furloughs on a very limited basis
for inmates that are already close to their release date and who are thoroughly screened to minimize
the risk to a community. A study conducted by Harer (1994) demonstrated that prison furloughs are
one of the best predictors of an inmate's post-release success, when controlling for other risk
predictors. Inmates who receive furloughs for the purpose of establishing community ties before
release are less likely to be re-arrested or have their supervision revoked for the first three years after
their release.
The final objective measure Logan used in the security section of his paper was the custody staffing
levels of the different institutions. The Federal prison received 2 demerits (--), compared to the state
and privately run institutions. The private, state, and Federal prisons had inmate-to-staff ratios of 3.1,
2.3, and 8.1 to one, respectively. How is it that the Federal institution, which had the lowest
urinalysis rate, and an extremely low incident rate, could be judged unfavorably because it used fewer
custody staff, by far, than the other two institutions, yet still managed to operate a safe and humane
environment? A related problem with this interpretation is Logan’s failure to understand the
management context. The Bureau of Prisons considers all of its staff to be correctional officers first.
Thus, teachers and vocational training instructors have both responsibilities as instructors and
custodial staff. Under this model, if safety and security is still maintained, why would you debit an
institution? The real issue here is whether a performance indicator is an outcome or a process
measure. In the case of security, the primary outcomes are those that measure whether inmates and
Appendix 2

Private Prisons

26

staff are at risk to be assaulted or harmed. Other measures, such as staff to inmate ratios, shakedowns,
and even furloughs are process measures that should be regarded as related to outcomes, but not as
outcomes themselves.
The bottom line on the comparison between the private and the Federal facilities was that on objective
dimensions, the Federal facility seemed to perform better. Since we regard the perceptual measures as
likely biased, comparisons based on these measures strike us as highly questionable.
Although Logan has done a remarkable job conceptualizing the dimensions of confinement, we
strongly believe that his lack of institutional experience limited his ability to compare the correctional
institutions on specific performance indicators, especially since he relied on his own judgment about
those performance indicators. His lack of a fundamental understanding about how prisons operate and
how they are managed limits the usefulness of "Well Kept."
Putting the Public-Private Comparison in Perspective
In considering the history that led to the transfer of all female inmates from WNMAC, which served
as the publicly-operated facility in this analysis, to a new CCA facility, which stood as the
privately-operated facility, our impression is that we would have been surprised had the new facility
not been an perceived as an improvement over the past by those staff who were surveyed for the
purpose of this study.
First of all, as DiIulio commented in an appended section of the report, before the CCA facility
opened, "New Mexico's women prisoners could be described as correctional 'orphans' who were
housed in 'a make-shift' wing of Western New Mexico Correctional Facility, a large high-security
institution for males." He further added that, "Before that, the women prisoners experienced frequent
moves between different facilities, none of which was equipped to meet the needs of female inmates."
DiIulio's comments seem to indicate that the conditions of confinement for the female offenders were
not particularly favorable.
Moreover, when it is considered that the State correctional system was under court order to improve
conditions (WNMAC in June 1988 was found to be in non-compliance in over one-third (35.7
percent) of the 42 compliance provisions audited), it certainly appears that problems existed in the
operation of WNMAC. Our sense is that while State of New Mexico prison authorities attempted to
provide for the female offender population at WNMAC, there were serious limits as to how successful
this effort could be. This is suggested in the comments of Charles W. Thomas, a correctional
consultant, who, in his assessment of WNMAC in October 1988, was very positive about institutional
staff but was highly critical about institutional design features and the security problems these design
flaws created. Indeed, our assumption is that the contract with CCA to operate a new female facility
was intended, in part at least, to overcome problems which existed at WNMAC.
The point is that WNMAC was not just any State-operated facility for female offenders, but was
instead a facility with major inadequacies, and CCA, far from being just another privately-operated
female facility, was intended to replace WNMAC and presumably in the process to overcome the
inadequacies of WNMAC. Under these circumstances, could CCA be anything but better?

Appendix 2

Private Prisons

27

Other Methodological Problems
In this last section, we list other methodological problems with the Logan study.












Logan gave equal weight to each of his quality dimensions and sub-dimensions. Thus, a rating
on security was considered just as important as a rating on activity. Most correctional experts
consider security and safety to be the primary objectives of sound prison practices and would
not rate all dimensions equivalently.
Logan gave equal weight to each of his empirical measures, for example, the serious assault
rate had no more weight as an index of "OVERALL" prison quality than how often inmates
used the recreational facilities.
He made no attempt to show if, and by how much, the subjective measures predicted, or were
associated with, the objective measures.
He ignored the magnitude of differences in item scores between institutions, forcing
quantitative (interval scale) differences between institutions into a tripartite (+, -, and =) scale,
while often making conceptually questionable decisions about whether high or low on the
item score means "+". He then combined and re-quantified the tripartite difference measures
using a conceptual method of scoring similar to ice hockey. We question the adequacy of this
measurement process. A quantitative method could have been used.
Although he purported to be looking only at three all-female institutions, Logan, without
acknowledging it, introduced survey response bias by including responses by staff at the New
Mexico State prison who worked with male inmates when making comparisons of staff
Climate survey responses across institutions.
Inmate survey data showed that the public New Mexico State facility outperformed the private
facility in every dimension except “activity” (GAO report p. 28). This was contrary to the
results based on staff perceptions. Why was there this discrepancy between inmate and staff
perceptions? Are we to discount inmate perceptions based upon some underlying vested
interest? Or perhaps the greater vested interest lies with the staff who had a financial stake in
the success of the evaluation.

Although we find Logan's conceptualization of a theoretical framework from which to compare
institutions appealing, for the reasons given above his attempt to evaluate the relative merits of public
and private facilities falls far short of a rigorous or conclusive analysis. There was no indication in the
report whether the State of New Mexico had changed its policy or procedures in response to the
privatization of its female population. Furthermore, there is no indication in the report that there had
been any innovation on the part of CCA in its management of the women prisoners.

Florida Evaluation
A recent study by Lanza-Kaduce and Parker (1998) compared the recidivism rates of inmates released
from prisons operated by the Florida Department of Corrections to those inmates released from a
Appendix 2

Private Prisons

28

prison operated by the Corrections Corporation of America and a prison operated by Wackenhut
Corrections Corporation. The at-risk release period for the offenders was 12 months and recidivism
was defined either as a rearrest, a new offense, a new commitment, a technical violation, or a
summary measure based on the first four indicators. Inmates from public facilities were matched with
inmates from private facilities on the basis of classification (minimum, medium), offense category,
race, prior incarcerations (0, 1, 2, or more), and age ( 25 or less, age 26 to 30, age 31 to 35, age 36 to
40, and age 41 or older). Of 300 releases from private prisons, only 196 matched pairs could be
found. In fact, the researchers had to relax their categories to find even 196 matched pairs. The
researchers also identified whether an inmate had participated in educational, vocational, substance
abuse, behavioral education, or pre-release training. A seriousness of recidivism score was
constructed based on the nature of the recidivism. The score ranged from 0 to 5. Zero indicated no
recidivism, 1 indicated a technical violation, 2, a misdemeanor, 3, a drug or weapon possession
offense, 4, a property offense, and 5, a violent or personal offense.
Lanza-Kaduce and Parker (1998) reported the following results: (1) private releasees had a lower
recidivism percentage on every one of the 5 indicators except technical violations. The overall
measure indicated a recidivism percentage of 17 percent for the inmates released from private prisons
and 24 percent for inmates released from the public institutions. For the overall measure, this
translated into a recidivism rate of 172 per 1,000 released inmates for the private institutions and 237
for the public facilities. The recidivism scale measuring seriousness indicated that public sector
releasees were more likely to commit drug/weapon possession offenses, property offenses, and violent
offenses. The mean level of seriousness on this 0 to 5 point scale was 3.43 for the public sector
releasees and 2.32 for the private sector releasees.
Lanza-Kaduce and Parker (1998) also reported on the relative recidivism rates among releasees from
the privately-operated facilities who successfully completed one or more of the programs previously
listed against those who failed, dropped out, refused to participate, or were removed. Among the
successful completers, 15 percent recidivated, while 40 percent of the noncompleters recidivated. We
find this final result fraught with issues regarding selection bias. One can draw no conclusions about
program effectiveness when you compare dropouts with those who complete a program. Under these
circumstances, it is impossible to disentangle program effects from the inmate’s underlying motivation
to succeed (Pelissier, Rhodes, Gaes, Camp, O'Neil, Wallace, and Saylor 1998).
A critical assessment of this study has been written by the Florida Department of Corrections, Bureau
of Research and Data Analysis (1998). The analysts writing this report noted four significant
problems with the Lanza-Kaduce and Parker study. The first problem focuses on the putative
equivalency of the private and matched public inmate releasees. The Bureau of Research and Data
Analysis (BRDA) paper notes that while Lanza-Kaduce and Parker selected inmates who were
minimum or medium custody, this is not the same as using the level of custody which was apparently
higher, on average, for the inmates released from the public facility. Furthermore, publicly released
inmates were more likely to have longer sentences, another indicator of the seriousness of the instant
offense. Publicly imprisoned inmates served, on average, a much longer time than the privately held
prisoners and inmates released from a public facility were more likely to have a term of supervision.
This latter difference is quite important. As the BRDA researchers point out, the differences in
supervision indicate that the inmates released from public institutions were more serious offenders,
who were more closely monitored, thus magnifying re-arrest or other recidivism measures. This makes

Appendix 2

Private Prisons

29

the use of technical violations inappropriate. If an offender is not under supervision, he or she cannot
be technically violated.
The BRDA researchers also found that offenders in the private sample had, on average, a less serious
previous record. This was based on the number of prior incarcerations. Finally, the BRDA researchers
were critical of the broad age categories used by Lanza-Kaduce and Parker. As the BRDA group
noted, age is the most significant determinant of recidivism and either an exact match should have
been used or a multivariate analysis controlling for age and some of these other variables should have
been conducted.
A second, even more serious problem in the Lanza-Kaduce and Parker analysis is that 35 percent of
the inmates included in the private sample had also spent a significant amount of time in a public
facility. It would be impossible to disentangle the effect of the private “dose” from the public “dose.”
This might suggest a research design in which one compared exclusively private, exclusively public,
and mixed public-private incarceration. But even this might be meaningless unless we can have
assurances that the decision to select or place inmates into these facilities is not somehow entangled
with their propensity to recidivate -- the problem of selection bias at a broader level.
As a third criticism, the BRDA researchers also criticized Lanza-Kaduce and Parker for using
relatively small sample sizes. However, we would argue, if all of the other problems of this study
could be addressed, the sample size was probably sufficient. A fourth, and more serious error, was the
way in which Lanza-Kaduce and Parker procedurally measured recidivism. According to the BRDA
group which had the original data, Lanza-Kaduce and Parker evaluated the recidivism of inmates
released from private facilities in the period June 1, 1996, to September 30, 1996. The period for the
inmates released from public facilities was January 1, 1996, to September 30, 1996. As we noted
before, recidivism was assessed from the day of release until 12 months had been reached. Because
there is a lag between the occurrence of a recidivism event and the recording of that event into the
automated records, the recidivism differences reported by Lanza-Kaduce and Parker may be an
artifact of the recording process. The recidivism data were gathered in November 1997, thus the
publicly released inmates had, at most, a 23-month release period, while the privately released inmates
had, at most, an 18-month release period. Depending on how long the lag is between the event and the
recording of the event, this discrepancy between public and private release periods could seriously
bias the results in favor of the private facility.
While this particular study of the relative differences in recidivism among privately and publicly
incarcerated inmates had serious errors, we are skeptical that any such study can circumvent the
problems associated with matching inmates and precluding selection bias. Aside from these
methodological problems, what are the theoretical implications of such a contrast? On what
theoretical basis would we expect privately and publicly operated prisons to be different in ways that
would affect recidivism? If the issues revolve around programming, then it is certainly possible to
evaluate inmate programs recognizing the same methodological problems with matching and selection
bias. However, the public sector can also develop and deliver programs. In fact, many programs
delivered in public prisons are the result of a contractual arrangement with a private provider. Finally,
even if we could develop a satisfactory design to compare the recidivism rates of publicly and
privately operated prisons, we would want to know what it is about the nature of operations at either
the public or private facility that reduces criminality. Knowing this, we would export that knowledge
to all of our prisons.
Appendix 2

Private Prisons

30

Summary of Existing Studies
For the most part, those who have evaluated private corrections in comparison to public corrections
have concluded that the private correctional facilities performed as well or better than the public
institutions. However, in our assessment of these evaluations, we find that most of these studies are
fundamentally flawed, and we generally agree with the 1996 GAO report that there is “little
information that is widely applicable to various correctional settings (p. 11).” We think this
conclusion is still warranted despite the two recent evaluations conducted in Louisiana (Archambeault
and Deis 1996) and Arizona (Thomas 1997a), and the recidivism study conducted by Lanza-Kaduce
and Parker (1998).
In our opinion, the strongest of the existing studies are the Tennessee evaluation and the Washington
State review. The Tennessee evaluation found that Corrections Corporation of America was running a
prison on par with the two new facilities operated by the Tennessee Department of Corrections. The
review conducted by Washington State noted that while it would be hard to generalize the findings to
Washington (and, by implication, other states), it does appear that the experiences with privatization
in Tennessee, Louisiana, and Florida have been positive. Even in these studies, though, there are
serious shortcomings in the analyses (Thomas et al., 1996).
The most significant problem with all of these studies is that they fail to develop a coherent model of
institution performance in terms of cost and quality of operations. Such a model would include the
structure of the relationship between process and outcome measures. The model would also make
explicit those factors that must be controlled to make institutional comparisons meaningful (Camp,
Saylor and Harer 1997; Camp, Saylor and Wright 1998; Office of Research and Evaluation 1998;
Saylor 1996). In this review, we have criticized researchers either for not using statistical adjustments
or for incorrectly using univariate statistics. In most cases, it seems to us that researchers are
inappropriately using univariate statistics to infer population differences when, in fact, they are using
population data. We also were critical of Logan for not using statistics to assess the degree to which
the private and public institutions were different. We think descriptive statistics could help in
clarifying the strength of the differences between institutions, but this must be done in conjunction
with a model that allows us to understand the relationship between outcome and process variables
while simultaneously controlling for the important substantive differences among the institutions.
In Table 3, we summarize some of the other characteristics and failures of these studies. Most of the
studies do not use a variety of different measurement approaches; fail to study equivalent inmate
populations or have insufficient information on the comparability of the offenders; use inappropriate
or no statistics; and, use a single point in time, rather than a longitudinal assessment. Most also fail to
explain the nature of private innovation, the impact of privatization on the entire system, or how
innovation affects performance in terms of cost and quality of operations.
Practically all of the evaluation research literature that has been produced on privatization has been
designed to compare the relative performance of the public or private operation of a prison. The issue
has been framed as a competition and the scorecard has been based on cost and quality. Although this
will continue to be an important issue in the future, there are other fundamental ways of framing the
research questions. A second, yet complementary approach, is to ask the following two questions:
How does privatization change the provision of correctional services, especially in terms of changing
Appendix 2

Private Prisons

31

operations in public sector prisons, even in public systems that are generally considered to be well run
and accountable to the public interest? How does public sector management of prisons influence the
operations of privately managed prisons within its jurisdiction? The ways in which these systemic
changes take place depend on the approach to privatization each jurisdiction takes.
Since the majority of experience in managing inmates comes from the public sector, the private sector
must begin from that knowledge base. Indeed, there has been a great deal of public sector innovation
in corrections, some of it borrowed from private sector operations in other industries. Unit
management, objective classification of inmates, strategic planning, and the concept of direct
supervision all preceded the introduction of private sector management. There is no evidence to
contradict the assertion that basic management philosophy, technology, and correctional expertise
have been transferred to the private sector from the public sector. Often, there is a direct transfer of
personnel and policies. The question, then, is how does the private sector add value to prison
operations. One can even take a broad view of this issue and consider the increased flexibility of a
system that might use privatization to alleviate crowding or handle special needs populations.
The usual response is that market pressures force the private sector into a more efficient use of
resources. In particular, the market supposedly creates incentive for efficiencies in two general areas.
First, labor is more efficiently utilized in the private sector. This is important because labor costs
typically account for 65-70 percent of the operational costs of a prison. Second, there are efficiencies
realized from more flexible purchasing practices.
However, despite the claims about cost savings and increased value, in reality there have been no
empirical studies documenting innovations in the private sector in the use of labor or the purchasing
of goods and services. What is needed are case studies that document the innovations developed by
the private sector that produce added value in the use of labor or in purchasing practices (Camp 1998).
We also need to document how the labor use and purchasing practices of public sector prisons change
as a result of the dynamic interplay between public and private sectors.
It appears to us that the private sector’s approach to corrections has been to build upon correctional
practices that already exist in well-run public prisons. The private sector does not appear to argue that
they run prisons in a dramatically different way based on different philosophies of managing inmates.
However, there has been little attention given to documenting the private sector approach to
innovation or to the impact of competition from the private sector on the practices of the public sector.
A corollary to these systemic questions is the issue of whether the private sector is delivering too little
or the public sector is delivering too much. Does the private sector save on costs by providing fewer
essential services to inmates than the public sector? If so, what are the short- and long-term
consequences of this lack of services? Is there a possibility that public sector prisons provide “too
much” quality to inmates? That is, do they provide services to inmates that are not as readily available
to some law-abiding members of society (usually indigent citizens)? While most correctional
administrators (public and private) agree that U.S. prisons should meet American Correctional
Association (ACA) accreditation standards at a minimum, there is probably much less agreement as to
how far above the bar set by ACA standards those prisons should operate to perform effectively. How
much education is excessive? Is a community standard the appropriate standard for medical care?
Should we provide inmates with job training when other poor, law-abiding civilians may have less
access to such training? Privatization brings these paradoxical public policy issues into sharper focus.
Appendix 2

Private Prisons

32

Quite similar to these arguments is Harding’s (1997) contention that the next generation of
privatization research should look for evidence of system-wide changes. Some evidence of such
changes can be found in Vagg’s (1994) description of the introduction of privatization into the
English penal system. Vagg, who reserves judgment on the merits of privatization, notes that the
introduction of the first private remand48 prison in England—The Wolds, which opened in 1992—may
have improved accountability for how inmates are treated in public sector prisons as well. As he notes
(Vagg 1994: 307), “... private prisons have the potential to offer improved prison regimes; and
ironically, in England, they were a key factor in persuading the administration that standards were
necessary, if only for the purpose of monitoring contractual compliance ...” According to Vagg, the
English government had been reluctant to establish prison standards such as ACA standards which are
used in the U.S. With the need to oversee the Wolds contract, performance indicators were specified
that the contractor, Group 4 Remand Services, Ltd., had to meet. Standards were established for
security; health, safety, and hygiene; reception, registration, and discharge; regime activities (such as
the grievance procedure); inmate services; and other prison functions. The standards went so far as to
specify the amount of time that prisoners should spend out of cell, a key point as there was public
concern about unsentenced inmates being locked in their cells for extended periods in older remand
centers under public control (see James, Bottomley, Liebling, and Clare 1997). This explication of a
system change introduced by adding privately operated beds is not typical in the evaluation literature
on privatization. Similar efforts must be undertaken in the United States.
There are well over 100 adult correctional institutions currently being operated by private corrections
companies. Yet we have analyses on only a handful. Even if these studies were rigorous and
methodologically sound, the private institutions captured in these analyses may, or may not, represent
the performance of the industry as a whole.
In the next section of this critique, we propose a model designed to address many of the problems we
have noted throughout this paper.

Outline of How the Taft Evaluation Meets the Criteria of a Sound
Evaluation
The proposed evaluation of the experiences of the Federal Bureau of Prisons with the private Federal
prison operated by Wackenhut Corrections Corporation in Taft, California addresses many of the
criticisms of existing evaluations of privatization. A complete discussion of the proposed evaluation
can be found elsewhere (Office of Research and Evaluation, 1998). The discussion proceeds by
addressing each of the areas covered in the summary table of existing quality evaluations.
System Impact: The system impact of the private operation of the Taft prison upon other operations in
the Federal Bureau of Prisons is addressed by the modeling approach that will be used in the
evaluation. At about the same time that the Taft prison was built, three other low security institutions
were built by the Federal Bureau of Prisons in Yazoo City, Mississippi, Elkton, Ohio, and Forrest
City, Arkansas. These three institutions are the institutions against which Taft will be compared. Taft
and two of the comparison institutions also have minimum-security camps as part of the facility. In

48

A remand prison in the United Kingdom functions the same as a jail in the United States criminal justice system.

Appendix 2

Private Prisons

33

this discussion, though, we focus primarily upon the main facilities that house the low-security
inmates.
In order to ensure that the institutions are compared on measures that are adjusted for features that are
unrelated to institutional performance, it is necessary to develop models of the outcome measures that
follow the procedures suggested by Saylor (1996) and Camp and colleagues (Camp, Saylor and Harer
1997; Camp, Saylor and Wright 1998). In essence, this means that both Taft and the three BOP
comparison institutions will be evaluated against the performance of all of the other low security
institutions in the BOP. As such, it will be possible to see how outcome measures at the Taft
institution, the three comparison institutions, and the other low security BOP institutions change over
the 5 years of the evaluation. This should provide information about how BOP operations change in
response to the experience of having a private-sector institution as part of the BOP.
Innovation: As has been noted previously, innovation is touted as one of the reasons that privatesector companies can operate prisons more efficiently than the public sector without sacrificing the
quality of services provided to inmates and the public. However, descriptions of how private-sector
operators actually “do” corrections differently are usually missing (Camp 1998). To capture this
component in the Taft evaluation, a full-time, on-site researcher has been placed at Taft. It is the role
of this researcher to document the more qualitative aspects of how Wackenhut operates differently
than the BOP. In addition, by examining organizational charts and work patterns, a close examination
of the use of staff by Wackenhut and the BOP is planned.
Measures: The evaluation of Taft will use many different sources of information. There will be an
audit/compliance component. All of the institutions, Taft as well as the BOP comparison institutions,
are subject to ACA accreditation during the study period. In addition, there are plans to make use of
periodic reviews conducted by the BOP. The Office of Research and Evaluation also plans to gather
information through the annual survey of staff. Taft will participate in the staff survey, the Prison
Social Climate Survey, as do all other BOP correctional facilities. A corresponding inmate survey of
the social climate will also be administered at Taft and the comparison institutions. The inmate survey
of the social climate is conducted on an “as needed” basis unlike the staff survey which has been
conducted every year since 1988. Official records will comprise an integral part of the evaluation data.
For example, Wackenhut uses the same disciplinary process as the Federal Bureau of Prisons and
enters the disciplinary data into the same centralized database used by all BOP correctional facilities.
Finally, numerous and periodic site visits to Taft and the comparison institutions are already, and will
continue to be, part of the data collection effort for the evaluation.
Points in Time: As has been mentioned, the time frame for the evaluation is a 5 year time period.
Therefore, the study is less likely to be influenced by atypical performance by any of the institutions
over a relatively short period of time.
Equivalent Facilities: The Taft facility, as well as the three formal comparison institutions, were all
built on an almost identical architectural plan at about the same time. While the wardens at all four
facilities have been free to enhance the physical plant, there have not been any major renovations such
as the construction of new buildings or the relocation of security fences at any of the four study
facilities.
Equivalent Inmates: All of the study institutions house low-security inmates as defined by the Federal
Bureau of Prisons classification system. None of the institutions have units for special needs inmates,
such as units for those with severe medical or psychological problems or residential drug-treatment
Appendix 2

Private Prisons

34

programs. Each of the institutions receives inmates as proscribed by BOP policy. There is no attempt
to send specific types of inmates to any of the institutions. Nonetheless, assignment of inmates to the
institutions is not random. For example, Taft is located near Los Angeles. Taft receives a large
number of Hispanic inmates because of the BOP policy of locating inmates near their point of release
when this does not conflict with other BOP management concerns. Any discrepancy between the
institutions in the characteristics of the inmates, though, will be addressed in statistical models used to
generate the comparison measures.
Model Approach: Statistical models will be utilized that are appropriate for the different types of data
collected as part of the Taft evaluation. For example, Saylor (1996) has demonstrated how residuals
from regression models can be utilized to assess whether institutions are performing better, worse, or
the same as expected on measures of inmate per capita cost and staff perceptions of crowding. Camp
and colleagues (Camp, Saylor and Harer 1997; Camp, Saylor and Wright 1998) have demonstrated
how hierarchical linear models can be used to assess institution influences on measures obtained from
staff surveys, such as organizational commitment or evaluations of institutional operations. While the
details of the models employed in the respective studies are beyond the scope of this discussion, the
use of the measures derived is quite straight-forward. With all of the models used in the analyses, the
end result is the ability to evaluate how institutions perform relative to other institutions, after
controlling for factors that are known to be unrelated to management performance. As an example,
consider evaluations of institutional operations. Experienced and inexperienced staff at the same
institution differ in the evaluations that they provide of institutional operations. If we compare two
institutions for “typical” staff perceptions of institutional operations, we would certainly want to
control for any differences in the institutions in the proportions of experienced and inexperienced staff
providing the evaluations.
Statistical Approach: Most of the statistics employed in the Taft evaluation will be used to
statistically adjust comparison measures. For the most part, the emphasis is not upon making
inferences to some population of prisons. For the most part, we are dealing with the population of
low-security, Federal prisons. There will be some use of inferential statistics when the data come from
samples of staff or inmates at the respective institutions.
Type of Facility: All of the institutions in the Taft evaluation house only adult offenders.
Security Level: All of the institutions are designed to house low-security inmates. For the most part,
that is the type of inmate that is housed at the facilities.
Gender of Inmates: All of the inmates are male.

Concluding Remarks
Most evaluations of the respective strengths and weaknesses of public and private prisons have not
relied upon strong theory to guide the evaluations. In the Taft evaluation, the theory of what
constitutes a “good” prison as outlined by academics such as Charles Logan (Logan 1990) is
supplemented with the collective practical knowledge of practitioners in the Federal Bureau of
Prisons. Management at the Bureau have developed a performance measurement system called the
Executive Staff Management Indicators (ESMI) module. ESMI is organized around six goals. Within
Appendix 2

Private Prisons

35

each goal, vital functions that support the achievement of those goals are outlined. The organization of
these performance measures is based upon the collective wisdom and years of corrections experience
reflected in the current and past compositions of the executive management staff at the BOP.
This theoretical approach guides the choice of measures to evaluate the performance of Taft and the
three comparison BOP institutions. The data used in the evaluation will come from multiple sources.
Therefore it will be possible in some instances to test for consistency in conclusions suggested by
reliance upon different sources of information. The data itself will be adjusted by statistical models to
control for differences between the institutions that are not related to performance. This information,
in turn, will be supplemented with an examination of organizational innovations in correctional
practices at Taft.
The proposed evaluation of Taft avoids some of the more obvious problems that have plagued existing
evaluations of privatized prisons. The proposed evaluation does not, however, circumvent all
problems. For example, the evaluation is still for only one jurisdiction, the U.S. Federal Government,
for one private company, Wackenhut, and for only low- and minimum-security inmates. Likewise, all
of the inmates are male. The study does, we think, point the way for collecting systematic evidence
about the types of changes privatization produces in the operations of public systems, how private
firms innovate in providing services to inmates, and whether public and private firms can provide
comparable services to inmates given the constraints faced by the Federal Bureau of Prisons. To date,
we have not generally found that this type of information is readily available in existing evaluations.

Appendix 2

Private Prisons

36

References
Archambeault, William G. and Donald R. Deis. 1996. “Cost Effectiveness Comparisons of Private
Versus Public Prisons in Louisiana: A Comprehensive Analysis of Allen, Avoyelles, and
Winn Correctional Centers.” Louisiana State University, Baton Rouge, LA.
Bureau of Research and Data Analysis. 1998. “Preliminary Assessment of a Study Entitled: ‘A
Comparative Recidivism Analysis of Releasees from Private and Public Prisons in Florida’.”
Florida Department of Corrections, Tallahassee, FL.
Butterfield, Fox. 1995. “For Privately Run Prisons, New Evidence of Success.” Pp. 7 in New York
Times. New York.
Camp, Scott D. 1998. “Private Adult Prisons: What Do We Know and Why Don't We Know More?”
Federal Bureau of Prisons, Washington, D.C.
Camp, Scott D., William G. Saylor, and Miles D. Harer. 1997. “Aggregating Individual-Level
Evaluations of the Organizational Social Climate: A Multilevel Investigation of the Work
Environment at the Federal Bureau of Prisons.” Justice Quarterly 14:739-761.
Camp, Scott D., William G. Saylor, and Kevin N. Wright. 1998. “Creating Performance Measures
from Survey Data: A Practical Discussion.” Corrections Management Quarterly
(forthcoming).
Casile, Maureen. 1994. “The New Boss: How Privatization Changes the Strategy and Structure of
Formerly Public Organizations.” SOC 396M, Professor Ekland-Olson.
General Accounting Office. 1991. “Private Prisons: Report to the Chairman, Subcommittee on
Regulation, Business Opportunities and Energy, Committee on Small Business, United States
House of Representatives.” U.S. General Accounting Office, Washington, D.C.
General Accounting Office. 1996. “Private and Public Prisons: Studies Comparing Operational Costs
and/or Quality of Service.” U.S. General Accounting Office, Washington, D.C.
Harding, Richard W. 1997. Private Prisons and Public Accountability. New Brunswick, NJ:
Transaction Publishers.
Harer, Miles D. 1994. “Recidivism Among Federal Prison Releasees in 1987: A Preliminary Report.”
Federal Bureau of Prisons, Washington, D.C.
Harer, Miles D., Loren Karacki, and Gerald G. Gaes. 1995. “A Critical Analysis of Charles Logan's
"Well Kept".” Federal Bureau of Prisons, Washington, D.C.
James, Adrian L., A. Keith Bottomley, Alison Liebling, and Emma Clare. 1997. Privatizing Prisons:
Rhetoric and Reality. Thousand Oaks, CA: Sage.
Lanza-Kaduce, L and K.F. Parker. 1998. “A Comparative Recidivism Analysis of Releasees from
Public and Private Prisons in Florida.” Private Corrections Project, Center for Studies in
Criminology and Law, University of Florida, Gainesville, FL.
Logan, Charles H. 1990. Private Prisons: Cons and Pros. New York: Oxford University Press.
Logan, Charles H. 1991. “Well Kept: Comparing Quality of Confinement in a Public and a Private
Prison.” National Institute of Justice, Washington, D.C.
Logan, Charles H. 1992. “Well Kept: Comparing Quality of Confinement in Private and Public
Prisons.” The Journal of Criminal Law and Criminology 83:577-613.
Moore, Adrian T. 1998. “Private Prisons: Quality Corrections at Lower Cost.” Reason Public Policy
Institute, Los Angeles.
Nink, Carl. 1998. “Comments at National Workshop on Privatization.” Program sponsored by the
Corrections Program Office, Office of Justice Programs, U.S. Department of Justice.
Office of Research and Evaluation. 1998. “Proposal to Evaluate the Performance of a Privately
Operated Bureau of Prisons' Facility.” Federal Bureau of Prisons, Washington, D.C.
Appendix 2

Private Prisons

37

Pelissier, Bernadette, William Rhodes, Gerald G. Gaes, Scott D. Camp, Joyce O'Neil, Susan Wallace,
and William G. Saylor. 1998. “Alternative Solutions to the Problem of Selection Bias in an
Analysis of Federal Residential Drug Treatment Programs.” Federal Bureau of Prisons, Office
of Research and Evaluation, Washington, D. C.
Saylor, William G. 1984. “Surveying Prison Environments.” Federal Bureau of Prisons, Washington,
D.C.
Saylor, William G. 1996. “Modeling and Graphing Organizational Processes in Pursuit of
Performance Benchmarks: Methods for Establishing and Evaluating Performance Measures.”
Federal Bureau of Prisons, Washington, D.C.
Sechrest, Dale K. and David Shichor. 1994. “Final Report: Exploratory Study of California’s
Community Correctional Facilities.” Parole and Community Services Division, California
Department of Corrections.
Tennessee Select Oversight Committee on Corrections. 1995. “Comparative Evaluation of PrivatelyManaged Corrections Corporation of America Prison (South Central Correctional Center) and
State-Managed Prototypical Prisons (Northeast Correctional Center, Northwest Correctional
Center).” Tennessee Select Oversight Committee on Corrections. Nashville, TN.
Thomas, Charles W. 1997a. “Comparing the Cost and Performance of Public and Private Prisons in
Arizona.” Center for Studies in Criminology & Law, University of Florida (for the Arizona
Department of Corrections).
Thomas, Robert C. 1997b. “Letter to William G. Archambeault, P.D., School of Social Work,
Louisiana State University from Robert C. Thomas, Principal Management
Auditor/Supervisor.” Joint Legislative Audit and Review Committee, Washington State.
Thomas, Robert C., Kathy Gookin, Beth Keating, Valerie Whitener, Robert M. Williams, Richard
Crane, and Cheryle Broom. 1996. “Department of Corrections Privatization Feasibility
Study.” Legislative Budget Committee, State of Washington.
Urban Institute. 1989. Comparison of Privately and Publicly Operated Correctional Facilities in
Kentucky and Massachusetts. Washington, D.C.: NIJ.
Vagg, Jon. 1994. Prison Systems: A Comparative Study of Accountability in England, France,
Germany, and the Netherlands, Edited by R. Hood. New York: Oxford University Press.

Appendix 2

Private Prisons

38

Table 1. Average Monthly Counts

Institution

Institution Operated By: Wackenhut Corrections
Corporation (WCC), Corrections Corporation of America
(CCA), State of Louisiana (Public)
WCC Public CCA
Public Public Public Public
Avoyell
Allen es
Winn Hunt Dixon Angola Wade

Mean Inmate Population

1206.9 1227.2 1214.9 1775.5 1367.4 4741.8 1021.9

Performance Measure
escapes

0

5

5

4

6

1

assaults on staff with injury
assaults on staff/serious injury
assaults on staff/non-serious injury

1.78
1.06
0.0455 0.1389
1.73 0.917

2.26
0
2.261

1.64
0
1.639

1.31
0.139
1.167

24.11
0.278
24.08

1.33
0.556
1.278

Cat. I,II,III incidents
aggravated sex offenses
aggravated cat I sex offenses
aggravated class II sex offenses
institutional disturbances
major inst. disturbances
minor inst disturbances

28
11.1
2.023
9.114
0.9555
0.114
0.841

38.5
30
55.8
26.4 154.1
22.3
4.3
8.5
6.1
2
27.9
3.9
2.028 1.217 2.861 1.056 7.222 1.611
2.222 7.283 3.251 0.9444 20.667 2.778
4.5 4.457 17.556 1.444 26.694 1.694
0.194 0.0455 0.0278
0 1.0556 0.0833
4.306 4.413 17.556 1.444 24.694 1.694

assaults/inmate on inmate
assaults/serious injuries
assaults/non-serious injuries
assaults/weapon
assaults/no weapon
assaults/weapon and serious injury
assaults/weapon and no serious injury
assaults/no weapon/serious injury
assaults/no weapon/no serious injury

13.455
0.136
13.318
2.523
10.932
0.068
2.455
0.068
10.864

26.278 17.421 28.083 18.417 67.778 14.083
1.056 0.304 0.1389 0.3043 0.5556 0.111
25.222 13.978 27.944 18.028 67.222 13.318
3.861 2.652 0.0833 0.1111 0.0556 0.0556
22.417 11.63 19.778 16.417 38.639 11.778
0.528 0.261
3.333 2.391
0.528 0.043
21.889 11.587

total disciplinary actions
schedule A disciplinary actions
schedule B disciplinary actions
monthly gunshots

3

374.1
42.3
325.9
0.25

859.8
282.3
515.4
2.25

459.6 567.1 687.2
153.9 103.4 102.2
291.7 399.1 558.9
0.152 0.9722 2.5833

1009
1779
258.9 396.7
697.8 1274.1
7.222
1

Table 2. Average Monthly Rates / 1,000
Institution Operated By: Wackenhut Corrections
Corporation (WCC), Corrections Corporation of America
(CCA), State of Louisiana (Public)
WCC Public CCA
Public Public Public Public
Appendix 2

Private Prisons

39

Avoyell
es
Winn

Institution

Allen

Hunt

Dixon

Angola Wade

Mean Inmate Population

1206.9 1227.2 1214.9 1775.5 1367.4 4741.8 1021.9

Performance Measure
escapes

2.4857

assaults on staff with injury
assaults on staff/serious injury
assaults on staff/non-serious injury

1.4749 0.8638 1.8602 0.9237 0.958 5.0846 1.3015
0.0377 0.1132
0
0 0.1017 0.0586 0.5441
1.4334 0.7472 1.8611 0.9231 0.8534 5.0782 1.2506

Cat. I,II,III incidents
aggravated sex offenses
aggravated cat I sex offenses
aggravated class II sex offenses
institutional disturbances
major inst. disturbances
minor inst disturbances

23.2
9.1971
1.6762
7.5516
0.7917
0.0945
0.6968

31.372
3.5039
1.6525
1.8106
3.6669
0.1581
3.5088

24.693
6.9965
1.0017
5.9947
3.6686
0.0375
3.6324

31.428
3.4357
1.6114
1.831
9.8879
0.0157
9.8879

19.307
1.4626
0.7723
0.6907
1.056
0
1.056

32.498
5.8838
1.5231
4.3585
5.6295
0.2226
5.2077

21.822
3.8164
1.5765
2.7185
1.6577
0.0815
1.6577

assaults/inmate on inmate
assaults/serious injuries
assaults/non-serious injuries
assaults/weapon
assaults/no weapon
assaults/weapon and serious injury
assaults/weapon and no serious injury
assaults/no weapon/serious injury
assaults/no weapon/no serious injury

11.148
0.1127
11.035
2.0905
9.0579
0.0563
2.0341
0.0563
9.0016

21.413
0.8605
20.552
3.1462
18.267
0.4302
2.7159
0.4302
17.837

14.339
0.2502
11.505
2.1829
9.5728
0.2148
1.9681
0.0354
9.5374

15.817
0.0782
15.739
0.0469
11.139
0
0
0
0

13.469
0.2225
13.184
0.0812
12.006
0
0
0
0

14.294
0.1172
14.176
0.0117
8.1486
0
0
0
0

13.781
0.1086
13.033
0.0544
11.526
0
0
0
0

total disciplinary actions
schedule A disciplinary actions
schedule B disciplinary actions
monthly gunshots

309.97
35.048
270.03
0.2071

700.62 378.3 319.4 502.56 212.79 1740.9
230.04 126.68 58.237 74.74
54.6 388.2
419.98 240.1 224.78 408.73 147.16 1246.8
1.8334 0.1251 0.5476 1.8892 1.5231 0.9786

Appendix 2

0 4.1156 2.8161 2.9253 1.2653 0.9786

Private Prisons

40

Appendix 3
Legal Issues Relevant to Private Prisons

by
Malcolm Russell-Einhorn, J.D.
Abt Associates Inc.

Appendix 3

Private Prisons

TABLE OF CONTENTS
Appendix 3
Legal Issues Relevant to Private Prisons
1.

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

2.

The Legality of Delegating Correctional Services to Private Contractors Generally . . .
A. The Current Limits of Federal and State Delegation Principles . . . . . . . . . . . . . . . . . . . .
B. Addressing Possible Due Process Concerns in Correctional Privatization . . . . . . . . . . . .
1. Basic Due Process Principles Governing Delegation Scenarios . . . . . . . . . . . . . . . .
2. Specific Ways of Addressing Due Process Concerns in Correctional
Privatization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3
4
7
7
9

3.

Liability for Private Prison Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
A. The Impact of Privatization on Liability Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1. Inmates’ Overwhelming Reliance on Section 1983 Cases to Redress Grievances . . 14
2. Are Private Prison Contractors Amenable to Suit under Section 1983? . . . . . . . . . . 15
3. Do Private Prison Officials Have a ‘Qualified Immunity’ Under Section
1983? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4. Private Contractor and Governmental Liability Exposure in Section 1983 Litigation 19
B. The Role of Indemnification, Insurance and Monitoring in Reducing Liability and
Litigation Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1. Indemnification and Insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2. Monitoring to Safeguard Against Liability Resulting from Government
Duties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4.

Employment and Labor Relations Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. Potential Civil Service Impediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. The Significance of Private Correctional Firms’ Independent Contractor
Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C. Effects of Privatizing an Existing Unionized Facility . . . . . . . . . . . . . . . . . . . . . . . . . . .
D. The Right to Strike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. No-Strike Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2. Strike Notification/Mandatory Delays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

28
30
31
32
32

Inmate Labor Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. Constitutional Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. Federal prisons and inmate labor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C. State Prisons and Inmate Labor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

34
34
35
35

5.

Appendix 3

Private Prisons

27
27

6.

Other Important Legal Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. Access to Private Prison Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. Private Contractor Access to Criminal History Records . . . . . . . . . . . . . . . . . . . . . . . . .
C. Bankruptcy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
D. Use of Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
E. Environmental Law Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7.

Regulating Interjurisdictional Private Prison Contracting . . . . . . . . . . . . . . . . . . . . . . . 47
A. Interjurisdictional Contracting Out and Legal Problems . . . . . . . . . . . . . . . . . . . . . . . . . 48
B. Some Regulatory Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

8.

Legal Dimensions of Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. Regulatory Purposes and Contract Specificity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. The Importance of Objective Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Special Attention to Staffing and Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C. Effective Contract Performance and Enforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Contract Term and Renewal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2. Bolstering Good Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3. Contract Enforcement through Termination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4. Contract Enforcement Through Fines and Penalties . . . . . . . . . . . . . . . . . . . . . . . . .
5. Dispute Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appendix 3

Private Prisons

37
37
39
40
43
45

56
56
58
59
60
60
61
61
62
63

1.

Introduction

The legal issues affecting private prisons have attracted nearly as much attention as the moral,
cost, and performance-related issues surrounding the correctional privatization movement. From
questions about the constitutionality of delegating the incarceration function in the first place, to the
legality of private prison guards going on strike, the legal dimensions of correctional privatization
have often elicited great concern from public correctional officials and the general public. Specific
concerns have, however, changed over time, as practical experience and legal precedent have provided
answers to many previously unresolved questions. For example, a fundamental issue that arose in the
mid-1980s along with the birth of the private prison industry was whether private prisons would be
considered “state actors” amenable to inmate civil rights suits and subject to all of the constitutional
standards applicable to public prisons. Over the past several years, courts have affirmatively
answered that question. Likewise, the Supreme Court last year ruled on another significant matter,
deciding that private prison guards may not share public correctional officers’ qualified immunity
from suit in federal civil rights actions.1
While a host of subsidiary legal and contractual matters continue to receive attention and
provoke discussion, governments are becoming more active in passing legislation, issuing regulations,
and insisting on contractual provisions that limit contractor discretion in important areas of public
policy. Public correctional authorities and their lawyers are becoming more sophisticated about the
pitfalls of private prison contracting and how governments can protect themselves from both a cost
and performance standpoint. One highly-publicized example of this phenomenon involves states
passing legislation to address the growing number of private prisons that accept prisoners from
outside the state.
Few attempts have been made to address the entire breadth of legal issues surrounding private
prisons in a relatively succinct form.2 This appendix makes such an attempt, employing a
straightforward presentation that should be useful to correctional specialists and laypersons alike. It
relies on a variety of sources for its information, including private prison contracts and requests for
proposals (RFPs), law review and other journal articles, federal and state statutory and case law,
survey information gleaned from federal and state officials summarized in other parts of this report,
and interviews with public and private correctional authorities and specialists.
This appendix traces the legal system’s accommodation of prison privatization as the latter
has evolved over the past 15 years. The report begins with the broadest constitutional questions,
raised at the outset of the privatization debate, and concludes with a discussion of the latest efforts of
state legislatures to gain greater control over the incarceration of out-of-state inmates in private
prisons within their borders. Only legal issues that significantly turn on private entities as prison
operators are examined.

1

Richardson v. McKnight, 117 S. Ct. 2100 (1997).

2

One such early effort, which was quite detailed, nevertheless failed to treat some issues in depth, e.g. public access to private prison
records. Moreover, this study is now a decade old and has been superseded in many subjects by recent legal developments. See I.
Robbins, The Legal Dimensions of Private Incarceration 396-413 (1988).

Appendix 3

Private Prisons

1

The focus of the report is on practical issues of a kind most likely to confront public and
private corrections officials. It also focuses on contracts for the management of private prisons, rather
than on contracts for the design and construction and management of future correctional facilities:
except for detailed public finance issues that vary from jurisdiction to jurisdiction, there are few
issues of any significance peculiar to private prison design construction. Finally, the report addresses
both publicly- and privately-owned facilities managed by private contractors, although there is a
slightly greater emphasis on publicly-owned prisons. Where relevant, legal distinctions that turn on
ownership are noted (e.g., in the case of potential contractor bankruptcies).
This report does not make any recommendations per se, but occasionally notes where certain
statutory or contractual mechanisms have yielded positive and negative results. The major issues are
arranged by section.

Appendix 3

Private Prisons

2

2.

The Legality of Delegating Correctional Services
to Private Contractors Generally

When the whole enterprise of correctional privatization has been challenged, the legal basis
for such an action has often been couched in terms of constitutional ‘delegation’ issues. At the heart
of these constitutional considerations is the worry that political accountability and disinterested
decision making may be lost when administration and operation of places of incarceration are turned
over to the private sector -- something that, prior to the mid-1980's, had not been encountered. Critics
have also raised particular concerns about private correctional contractors’ financial interests leading
them to minimize releases and maximize sentences to keep their facilities filled. These core concerns
have seemed to reinforce doubts about the overall propriety of delegating the state’s power to
incarcerate to private, for-profit corporations.
Despite these worries, no serious delegation challenge has been mounted against a statute or
contract allowing the private sector to incarcerate inmates. Although delegation concerns were
prominently raised in the 1980s as a potential roadblock to correctional privatization,3 few observers
today believe that, with certain safeguards in place, a government may not legally contract out its
power to incarcerate.4 There now appears to be broad agreement that delegation concerns, if any, do
not hinge on the overall policy choice of entrusting the day-to-day management of prisons to private
parties, but rather on the details of how such transfers of responsibility are structured (including the
rules by which such prisons are intended to operate), and who has ultimate authority for a number of
decisions affecting inmate rights and welfare.
As discussed below, the often diffuse “delegation” concerns expressed about prison
privatization tend to subsume or confuse several discrete legal issues, including the federal
constitutional “delegation doctrine,” state constitutional delegation standards, the due process
requirements of the U.S. Constitution (made applicable to states and local governments through the
Fourteenth Amendment), and specific questions about statutory enabling authorization that may vary
from jurisdiction to jurisdiction. It is apparent, however, that while modern challenges to correctional
privatization may be couched in terms of “delegation,” the analytical basis for these challenges will
usually turn on due process concerns. A due process analysis is particularly sensitive to the need to
have state correctional officials retain responsibility for promulgating many kinds of prison rules and
making final decisions on release-related matters such as sentencing, parole, discipline, ‘good time’
awards, work assignments, and possibly classification. A second type of delegation challenge is often
based on inadequate privatization enabling legislation.

3

See, e.g., Field, Note, Making Prisons Private: An Improper Delegation of a Government Power, 15 Hofstra L. Rev. 649 (1987);
Robbins, The Legal Dimensions of Private Incarceration, supra note 2 at 396-413.

4

Perhaps acknowledging the contemporary importance of privatization generally, as well as the history of private prisons in this
country, no one appears to have convincingly argued that the running of prisons, as opposed to the decisional control over the
incarceration of inmates in the first place (a police power), is so much a ‘core’ governmental function that it may never, even with
proper safeguards, be delegated to private parties as a legal (as opposed to moral or ethical) matter. While at least one article has
made that argument, it tends to ignore the details of delegation jurisprudence and the degree to which government oversight and
accountability can be built into a prison privatization arrangement. See Making Prisons Private: An Improper Delegation of a
Government Power, supra note 3.

Appendix 3

Private Prisons

3

A.

The Current Limits of Federal and State Delegation Principles

Delegation principles at the federal and state level safeguard against the transfer of
government powers to public and private bodies without proper constraints. However, reflecting
political reality as much as legal doctrine, courts today are unlikely to invalidate an entire prison
privatization arrangement on delegation grounds.5 The reason is twofold. First, in the modern era,
courts have given political leaders wide discretion in determining how to implement policy decisions,
including reliance on the private sector.6 The tendency is to view most of the functions transferred to
private agencies (as well as subordinate public agencies) as largely or purely administrative in nature.
Only delegated rule making and adjudication functions -- which directly purport to exercise a
government power -- are usually deemed to require special judicial scrutiny.7
Second, the delegation doctrine at the federal level has effectively become moribund over the
last half-century,8 a telling reflection of the rise of the administrative state. Delegations to
administrative agencies have been so liberally construed that private delegations have tended to be
viewed in the same light despite their significantly different factual and ethical attributes. Reflecting
this trend, the Supreme Court has sometimes enunciated the principle of nondelegation, but has almost
without exception sustained the constitutionality of challenged delegations. As discussed below,
rather than invalidating private delegations outright, the Court has used the due process provisions of
the Fifth and Fourteenth Amendments of the Constitution to protect the fundamental interests of those
affected by a delegation. At the state level, courts have breathed more life into state constitutions
concerning delegation questions, but their analysis has generally tended to recapitulate the due process
concerns identified by federal courts.
The weakness of the federal delegation doctrine is starkly indicated by the fact that in the area
of private delegations, as opposed to delegations of power to executive branch agencies, the Supreme
Court has not invalidated a delegation since the New Deal-era case of Carter v. Carter Coal Co.9
Since that time, the Court has only haphazardly addressed the delegation doctrine, and even then, only
in public delegation contexts where property, rather than liberty, interests were at stake.10 Insofar as
these cases collectively have any indirect application to the privatization context, they simply stand
for the proposition that delegations of legislative and policy-making power must be narrowly

5

Possible exceptions may include privatization arrangements where due process is violated or where little or no statutory authority
for the delegation exists. See pp. 5-6 infra.

6

See e.g. Berman v. Parker, 348 U.S. 26, 33-34 (1954) (upholding reliance on private companies to implement federal urban
redevelopment policy).

7

See generally Lawrence, Private Exercise of Governmental Power, 61 Ind. L.J. 647, 650-653.

8

See generally, e.g., K. Davis, Administrative Law Treatise § 3. See also Robbins, The Impact of the Delegation Doctrine on Prison
Privatization, 35 U.C.L.A. L. Rev 911, 917-920, 926 (1988).

9

298 U.S. 238 (1936). The Court struck down a federal statute allowing a majority of miners and producers to agree on wages and
hours that would be binding on minority members of the industry.

10

See, e.g., Industrial Union Dep’t v. American Petroleum Institute (the Benzene Case), 448 U.S. 607 (1980); Immigration and
Naturalization Service v. Chadha, 462 U.S. 919 (1983). The Court did once, in 1958, allude to the fact that it would apply a stricter
standard for cases involving liberty interests, see Kent v. Dulles, 357 U.S. 116, but this principle, has never received further
elaboration.

Appendix 3

Private Prisons

4

construed and have a clear, unambiguous statutory basis.11 In particular, these cases favor standards
ensuring that elected officials make important policy choices, private officials have a sufficiently clear
“intelligible principle” to channel their delegated discretion, and “ascertainable standards” exist by
which reviewing courts can evaluate such discretion.12
Under a post-Carter delegation analysis, a legitimate delegation challenge might be sustained
if a new prison privatization arrangement were undertaken with no statutory authorization, or pursuant
to a statutory or contractual scheme that left broad areas of sensitive decision-making implicating due
process requirements -- for example, disciplinary proceedings or the formulation of release-related
rules -- to private correctional officials’ discretion. This is generally how various states have applied
their own delegation doctrines: courts have struck down statutes granting rulemaking and adjudication
authority to private parties where the state did not retain ultimate authority to accept, reject, or modify
administrative rules and review adjudicative determinations.13
As a matter of practical and political reality, courts may be wary of invalidating on delegation
grounds fully operational privatization contracts, even if they lack what would otherwise be viewed as
proper statutory authorization.14 Moreover, contract provisions agreed to by governments and private
contractors may obviate otherwise defective statutes, depriving potential litigants of an actual harm or
controversy. Still, in the case of new or planned operations, delegation-type challenges based on
deficient legislation might emerge that require statutory remediation. 15
Both for reasons of litigation prevention and sound public policy, few would argue against the
long-term wisdom of adopting some kind of meaningful and comprehensive statutory authorization.
Adopting some minimal statutory scheme strengthens the public legitimacy of privatization and can
ensure that there is some set of minimal standards below which contractual arrangements may not fall.

11

The underlying purpose of delegation principles derives from the separation of powers notions, which protects the polity from
uncontrolled discretionary power. The weakness of the delegation principles’ relevance to private delegations is underscored by the
fact that insofar as the purpose of the separation-of-powers requirement underlying the delegation doctrine is to protect individual
liberty by dispersing power, then private delegations "actually promote that very goal, because power is spread still further [beyond
state bodies]. One need not carry the argument that far, however, to see that the separation-of -powers principle is a weak foundation
for limiting private delegations." Lawrence, Private Exercise of Governmental Power, supra note 7 at 665-66 (1986).

12

See, e.g., Industrial Union Dep’t, supra note 10 at 686. The meaning of "intelligible principle" has varied somewhat from case to
case but generally implies a well articulated rationale as to why a delegation is warranted in the first place. See, e.g., J.W. Hampton,
Jr. & Co. v. United States, 276 U.S. 394, 409 (1928).

13

See Robbins, The Impact of the Delegation Doctrine on Prison Privatization, supra note 8 at 930-950, for a comprehensive
discussion of state law approaches to the delegation doctrine. For example, in one leading case, a court invalidated the Oregon
Public Service Commission’s attempt to adopt, without discussion, safety code revisions (including future revisions) proposed by
private committees of interested parties. Hillman v. Northern Wasco County Peoples’ Utility District, 213 Ore. 264, 323 P.2d 664
(1958), overruled on other grounds, Maulding v. Clackamas County, 278 Ore. 359, 563 P.2d 731 (1977).

14

This would be especially true in the case of a newly-constructed private facility where millions of dollars of public funds had
already been invested in the project.

15

Although even in that case, the speculative or premature nature of some legal challenges may result in dismissal. See, e.g., Ohio
Patrolmen’s Benevolent Ass’n. v. Repke, 1995 Ohio App.; LEXIS 5876 (1995) (nondelegation issue deemed premature, not ripe for
review).

Appendix 3

Private Prisons

5

To date, however, some 20 states have not expressly authorized private prisons,16 and of those, five
have gone ahead with some form of private contracting for imprisonment.17 The federal government,
though possibly without a broad authorization to privatize,18 has nevertheless secured congressional
authorization to pursue two privatization arrangements.19 The tendency among many jurisdictions
moving ahead with correctional privatization has been to rely on contractual safeguards to address the
due process problems potentially raised by an inadequate statutory framework.
In sum, while sound, detailed enabling legislation may help to create a firmer basis for
private correctional accountability, only in certain situations is a statutory deficiency likely to give
rise to a delegation challenge that jeopardizes an entire privatization enterprise.20 Challenges to

16

The 20 states are Alabama, California, Delaware, Hawaii, Indiana, Iowa, Maine, Maryland, Massachusetts, Minnesota, Missouri,
New Jersey, New York, North Dakota, Rhode Island, South Dakota, South Carolina, Vermont, Washington, and Wisconsin. The
state of Illinois actually expressly prohibits private prisons.

17

Kansas, Hawaii, Idaho, Minnesota, and Missouri. Kansas has required its inmates to be placed in private in-state facilities, while the
other states have sent some of their prisoners to out-of-state private prisons.

18

The general legislative basis for the Federal Bureau of Prisons entering into contracts for the incarceration of prisoners in the private
sector has not been entirely clear. The BOP has, since 1983, maintained that it has the authority to make contracts for the secure
confinement of adult prisoners in the private sector. At least one scholarly commentator, as well as the General Accounting Office,
have argued that BOP lacks such authority. See Robbins, The Legal Dimensions of Private Incarceration, supra note 2 at 396-413;
U.S. General Accounting Office, Private Prisons: Cost Savings and BOP’s Statutory Authority Need to be Resolved, GAO Report
No. GAO/GGD-91-21 (Feb. 1991) at 45-50. After several years of intensive debate, the momentum to create a more explicit basis
for overall federal prison privatization has appeared to die down, and BOP has pushed ahead with new privatization plans specific
statutory authority being employed. See fn. 19, infra.
BOP maintains that 18 U.S.C. § 3621(b) provides the Bureau with independent authority to contract with the private sector for
secure adult prison facilities. The section states:
(b) PLACE OF IMPRISONMENT.-- The Bureau of Prisons shall designate the place of the prisoner’s imprisonment. The Bureau
may designate any available penal or correctional facility that meets minimum standards of health and habitability established by the
Bureau, whether maintained by the Federal Government or otherwise. . . .
(emphasis supplied). Because several other provisions in Title 18 specify the kinds of arrangements that BOP may make to obtain
prison facilities, and allow the use of only one category of nonfederal facilities -- those operated by state and local governments and
furnished to the federal government by way of contract -- some observers charge that the general language in § 3621(b) cannot
extend to private facilities. The clear inference, they maintain, is that Congress failed to provide for any arrangements such as
private incarceration that are not expressly authorized. See 2A Sutherland on Statutory Construction § 47.23 (4th ed. 1984)
(designation of specific provisions for executing certain functions implies exclusion of other, more general provision). Thus, 18
U.S.C. § 4002 provides that the Attorney General may contract with state and local governments for "the imprisonment, subsistence,
care, and proper employment" of prisoners for up to three years, and 18 U.S.C. § 4003 authorizes the Attorney General to build new
prison facilities if such state or local governments are unwilling or unable to enter into such contracts or if suitable facilities are not
available at a reasonable cost. There are no provisions addressing places of prisoner confinement in the private sector. By contrast,
other statutes involving the confinement of persons have explicitly authorized the use of private sector facilities for such
confinement. See, e.g., 18 U.S.C. § 4013 (authorizing the U.S. Marshals Service to contract with private facilities for the housing of
federal pretrial detainees); 18 U.S.C. §§ 5040 (permitting the housing of juvenile offenders in private facilities).

19

The BOP’s privately-managed facility in Taft, California was authorized in a 1996 appropriations bill (Conference Report 104-863
to Public Law 104-208, Sept. 28, 1996), while the privatization dimension of the federal takeover of the District of Columbia’s
Lorton, Virginia facility was provided for in the National Capital Revitalization and Self-Government Act of 1997 (Pub. L. No. 10533 §§ 11201(c)(1).

20

Certainly, without a proper legislative mandate defining with some specificity what can and cannot be undertaken in the way of
private sector operation of government correctional facilities, the allocation of public and private responsibilities may become
confused and public trust eroded. This has been manifested most clearly in the case of states that have sought to gain more control
over the housing of out-of-state inmates within the state, see, e.g., the recently-passed Ohio House Bill 293, effective March 17,

Appendix 3

Private Prisons

6

privatization based on a lack of proper statutory authorization should be relatively rare because most
prison privatization arrangements feature some type of statutory authorization sufficient to reflect a
general policy choice in favor of correctional privatization. A more potent regulatory challenge
involves litigation seeking to invalidate a particular kind of private contractor authority or conduct,
based on due process concerns, as discussed below.21 Most private prisons operate according to
statutes and/or contracts that retain some critical oversight, rulemaking and decision making in the
hands of government correctional authorities.

B.

Addressing Possible Due Process Concerns in Correctional
Privatization

The most serious constitutional challenges to contracting out the management of prisons may
arise when private contractors are deemed to have excessive power over prison rulemaking and
decisions affecting inmates’ basic liberties. These situations indicate issues of possible financial bias
and unchanneled discretion in prison management that raise potential due process concerns. While
these concerns can give rise to specific civil rights challenges by inmates and their legal
representatives, such due process objections may also serve as the basis for courts to invalidate certain
arrangements on “delegation” grounds. As with statutory authorization deficiencies, however, it may
only be certain aspects of the overall privatization scheme that are so invalidated.
1.

Basic Due Process Principles Governing Delegation Scenarios

The Due Process Clause has given rise to an influential line of Supreme Court cases
invalidating the delegation of discretionary rulemaking or adjudication powers to financially
interested parties.22 It has also generated support for the idea that private delegations are permissible
where the private entity’s role is subordinated to adequate public involvement, oversight, and/or
approval.23 These standards are applicable to the federal and state governments alike, since the Due
Process Clause has the effect of making constitutional restrictions on private delegations applicable to

1998, discussed infra in section 7.B.
21

As discussed above, due process considerations and the delegation doctrine share a concern about the exercise of arbitrary,
unaccountable government power. Indeed, private delegation cases tend to rely on a due process analysis, due to fuzziness of the
delegation doctrine, its origins in separation-of-powers jurisprudence and its general disuse. Due process analysis has in fact guided
most of the key federal court decisions invalidating private party delegations, including Carter v. Carter Coal Co. But these
decisions have unfortunately blurred the separate and distinct nondelegation and due process issues and standards. See Abramson, A
Fifth Branch of Government: The Private Regulators and Their Constitutionality, 16 Hastings Const. L. Q. 165, 204 (1989).

22

In the rulemaking context, see, e.g., Eubank v. City of Richmond, 226 U.S. 137 (1912) (striking down local ordinance authorizing
majority of landowners to impose set-back requirements on neighbors); Carter v. Carter Coal Co., supra. In the adjudication
context, see, e.g., Sniadach v. Family Finance Corp. of Bay View, 395 U.S. 337 (1969) (invalidating law authorizing private parties
to garnish employee wages before employees can respond to garnishment in court); Gibson v. Berryhill, 411 U.S. 564, 579 (1973)
(striking down state optometry board’s adjudication of conduct by corporate salaried optometrists where board was composed of
individual optometrists with a financial interest in reducing corporate competition).

23

See, e.g., Currin v. Wallace, 306 U.S. 1, 15-16 (1939) (creation of market controls by producers judged constitutionally acceptable
based on government-imposed requirement that 2/3 of producers waive antitrust restrictions); Sunshine Anthracite Coal Co. v.
Adkins, 310 U.S. 381, 399 (1940) (industry officials allowed to propose non-binding minimum prices where officials functioned
"subordinately" to a public commission); Fuentes v. Shevin, 407 U.S. 67, 80-83 (1972)(state-sanctioned prejudgment repossession of
property by private companies permissible only if property owners have opportunity to resist seizure before a judge).

Appendix 3

Private Prisons

7

the fifty states through the Fourteenth Amendment.24 In general, it would appear that the more
administrative the role of private companies -- that is, the more they are responsible for carrying out
predetermined policy choices by government officials -- the more likely it is that significant
privatization schemes will pass constitutional muster.
While explicit direction from the Supreme Court about due process standards in the private
delegation context is lacking, the best guidance by way of analogy may come from federal appeals
courts that have ruled on regulatory and adjudicatory standards for securities markets adopted by
private securities associations pursuant to the federal Maloney Act. The so-called “Todd Standards,”
first articulated in the case of Todd & Co. v. SEC,25 set forth three factors that enabled the regulatory
scheme at issue to accord with due process. First, the private entity’s rules had to be approved by a
governmental authority before becoming effective. Second, all private decisions regarding rule
violations and penalties had to be subject to governmental review. Finally, all such private
adjudications had to be governed by a de novo standard of review, whereby government officials
would draw their own factual and legal conclusions about the violation and penalty in each case.26
The Todd Standards are by no means dispositive of the application of due process to private
prisons. Indeed, the Todd case dealt with property rights, not the liberty interests that are at stake in
the case of private prisons. Yet perhaps for that very reason -- the more serious liberty interests at
issue in the latter case -- the Todd Standards would seem to suggest by analogy certain baseline
requirements for private prisons.
Accordingly, these standards suggest that private prisons should not have uncontrolled power
to make adjudicatory or rulemaking decisions that would disfavor inmates’ individual liberties. In
particular, as discussed in greater detail below, private contractors presumed financial interests should
disqualify them from having the power to make sensitive release-related decisions, including those
involving sentencing, parole, release dates, furloughs, discipline and other matters affecting ‘good
time’ credits, and specific work assignments and work credits. These types of decisions generally
overlap with the release-related rights that were recently deemed by the Supreme Court in Sandin v.
Connor to define the general contours of due process within prisons.27
Meanwhile, the presumed lack of private contractors’ financial disinterestedness should also
disqualify private contractors from making rules that govern any of these subjects. The reasons are
obvious: because private contractors have a potentially strong (it need not actually be strong) interest
in maximizing their inmate populations, they may have an incentive to write release-related rules and
make release-related decisions that affect good time calculations and ultimately disfavor prisoners. In
some cases, this potential bias can be avoided by having a contract specify a contractor fee that is

24

See Liebman, Delegation to Private Parties in American Constitutional Law, 50 Ind. L. J. 650, 652-54 (1975).

25

557 F.2d 1008 (3d Cir. 1973).

26

Id. at 1012.

27

In Sandin, 515 U.S. 472 (1995), the Supreme Court moved to scale back the scope of prison due process protections arising directly
under the U.S. Constitution to those situations which impose "atypical and significant hardship on the inmate in relation to the
ordinary incidents of prison life." Despite this significant, albeit hazy constriction of due process protections within prison confines,
the Sandin ruling leaves in place any protections that would ultimately affect the timing of a prisoner’s release.

Appendix 3

Private Prisons

8

independent of the number of inmates housed at a facility.28 As the key due process cases have
shown, private rulemaking and adjudication must avoid even the appearance of bias.29
While the “Todd Standards” have never been passed on by the Supreme Court, they provide
general guidance to correctional officials attempting to place limits on private prison rulemaking and
decision making, particularly since prisons affect liberty interests rather than the mere property
interests at issue in the Todd case. Translating Todd in a relatively straightforward manner to a
correctional context suggests the following principles: (1) rules authored by private prison contractors
should be reviewed and approved by government authorities before going into effect; and (2) all
private contractor decisions relating in any way to matters affecting release should be subject to
governmental review and approval.30
2.

Specific Ways of Addressing Due Process Concerns in Correctional
Privatization

Even with the substantial expansion of private corrections over the past decade, most states
and the federal government have yet to craft specific, detailed standards that address each of the
principles embodied in the Todd Standards.31 To be sure, contractual provisions addressing due
process concerns can obviate most, if not all potential deficient legislation. In fact, most jurisdictions
have appeared to be content to allow contracting parties to build due process protections into their
privatization agreements rather than adopt specific legislation on the subject. Most commentators
however, believe that a statutory basis for fundamental due process protections is advisable to obviate
due process concerns.32 Legislatively enacted due process standards may highlight a bright line public
policy choice and ensure that constitutional principles are not sacrificed at the margins by contracting
parties.
Few reported decisions exist that purport to apply any kind of due process analysis to general
matters of private prison decision making. In Tennessee, appeals courts have twice upheld a section

28

This is the approach taken in the BOP’s contract for the Taft, California medium-security prison.

29

See In re Murchison, 349 U.S. 133, 136 (1955) ("our system of law has always endeavored to prevent even the probability of
unfairness").

30

Part of the third Todd standard, requiring a de novo review, would not appear to be easily translated into a private prison context.
First, it would be incredibly burdensome to the system. Second, it is not clear what such a review would look like, since the
Supreme Court only requires that "some evidence" support a disciplinary decision. See fn. 42, infra. However a major question
arises: if the government’s review is not truly de novo, must its review be mandatory and automatic (so that a decision is not deemed
to have been rendered until the public correctional authority acts) or can the government’s review simply be satisfied by allowing
an inmate to appeal as of right the private contractor’s decision? Todd suggests that the de novo review requirement is the
equivalent of an independent decision being entered by a public agency; thus, if there is no de novo review, due process would seem
to require that a first-instance decision in all cases be formally approved or rendered by a public official. This is what many states
require in the case of disciplinary hearings. See text at fns 38-40, infra.

31

Neither the federal government nor eleven different states with prison privatization statutes attempt to place any type of statutory
due process limits on private exercise of government-mandated correctional authority. The states are Alaska, Georgia, Kentucky,
Montana, Nevada, New Hampshire, New Mexico, Oklahoma, Oregon, Pennsylvania, and Utah. For a comprehensive survey of
federal and state regulation of private prisons relative to due process concerns, see Ratliff, The Due Process Failure of America’s
Prison Privatization Statutes, 21 Seton Hall Legis. J. 371 (1997).

32

See, e.g., R. Crane, Wanted: A Model Law to Regulate Privatization, Correctional Law Reporter 83 (April/May 1998); Ratliff, The
Due Process Failure of America’s Prison Privatization Statutes, supra note 31 at 398-409.

Appendix 3

Private Prisons

9

of the relevant state code governing disciplinary proceedings by a “private management company” and
stated, albeit without any real analysis, that the statute comports with due process. For example, in
one of the cases, Davis v. Rose,33 the Tennessee Court of Appeals validated that part of the state code
and associated Department of Corrections disciplinary procedures that allow a private contractor
panel to recommend disciplinary action subject to review and final approval by a designated
Department of Corrections employee.34 In the absence of detailed kind of court guidance, government
corrections officials must use their good judgment and general due process case law to determine
where to draw statutory and contractual boundaries around private contractor rulemaking and decision
making. A brief survey of some of this experience with such line-drawing follows.
a.

Rulemaking Generally

Consistent with due process jurisprudence, some states have barred private prison contractors
from unilaterally developing or adopting rules affecting disciplinary proceedings or pertaining to other
liberty interests.35 A handful of states have also sought to prohibit private operators from formulating
rules or procedures for calculating inmate release and parole eligibility dates or developing and
implementing procedures for calculating and awarding sentence (good time) credits.36 Some
observers have suggested that the best way to deal with rulemaking problems is to require explicitly
that private contractors adopt pertinent rules and procedures employed by the contracting agency.37
While notions of fairness might at first blush suggest a presumption that all or virtually all private
prison rules should parallel those for public correctional institutions, and be approved by government
authorities, such guidance might be ill-advised. According to many privatization advocates, having a
boilerplate transfer of government rules would remove room for flexibility and innovation in many
areas of prison operation.
b.

Disciplinary Decisions

Governments have generally interpreted due process to require that limits be placed on
correctional contractors’ ability to render final disciplinary decisions. Such decisions can affect both
good-time credits (and hence, release dates) and in-prison privileges. While only three states have
explicitly prohibited contractors from taking any disciplinary action against an inmate,38 several states
have mandated that any preliminary decisions that are taken by contractors be reviewed and approved

33

1997 WL 83671 (Tenn. App.. 1997).

34

See also, e.g., Jackson v. CCA, 1997 WL 337021 (Tenn. App.. 1997) (same result).

35

See Fla. Stat. Ann. § 957.06; N.C. Gen. Stat. § 148-37; Va. Code Ann. § 53.1-265; Colo. Rev. Stat. Ann. § 17-1-203; Wyo. Stat. §
7-22-112.

36

See, e.g., Va. Code Ann. § 53.1-265 (barring private contractors from [developing and implementing procedures for calculating
inmate release and parole eligibility dates"); Tenn. Code Ann. § 41-24-110 (prohibiting private contractor adoption of rules for
calculating inmate release or parole eligibility dates, as well as calculation and award of sentence credits).

37

See, e.g., The Council of State Governments (CSG) and the Urban Institute, Issues in Contracting for the Private Operation of
Prisons and Jails (hereinafter "CSG Report") 70 (1987).

38

See Ariz. Rev. Stat. Ann. § 41-1609.01; Tenn. Code Ann. § 41-24-110; Va. Code Ann. § 53.1-265. Two states have a different
formulation to the effect that private contractors may take short-term or emergency disciplinary action subject to final review by
public corrections authorities. See Fla. Stat. Ann.§ 957.06; Colo. Rev. Stat. Ann. § 17-1-203.

Appendix 3

Private Prisons

10

by public correctional officers.39 At least a few others have done so contractually.40 Notably, the
BOP’s contract for its Taft, California facility does not have such a requirement.41 Some
commentators have also gone so far as to argue that due process requires a public official to
participate in disciplinary hearings, at least those involving major rule infractions. 42 Still, no court has
ruled that this is required by due process, nor do the Todd Standards (albeit in a property rights
context) seem to suggest this is necessary. There is little guidance as to the type of review to be
undertaken by public correctional representatives. In the meantime, the most cautious approach would
direct a private contractor to make factual findings and a preliminary determination thereon according
to the Supreme Court’s “some evidence” standard,43 but not treat that decision as final until the matter
is ruled on, or approved by, a public official.44
c.

Parole Decisions

Because most -- but not all -- private prison contractors are paid according to the number of
inmates kept in their custody, such contractors may be seen to have a financial interest in parole
decisions. This creates a potential - even if not actual - incentive to supply parole boards with biased
or misleading information.45 Because of this possibility of bias, due process arguably requires that

39

Ariz. Rev. Stat. Ann. § 41-1609-01,-Colo. Rev. Stat. Ann. § 17-1-203; Fla. Stat. Ann. § 957, 06; La. Rev. Stat. Ann. § 15:1177;
Tenn. Code Ann. §§ 301-265; Wyo. Stat. Ann. § 7-22-112.

40

See Contract Between Corrections Corp. of America and the Commonwealth of Virginia Department of Corrections, July 11, 1996
(hereinafter "CCA-Va. Contract") at § 4.21, p. 20.

41

See Taft, California Low and Medium Security Correctional Facilities Contract Between U.S. Bureau of Prisons and Wackenhut
Corporation, § C.1, Part II.J ("Discipline"). The BOP’s contract simply provides for contractor decisionmaking in the first instance,
with an appeal as of right by the inmate through ordinary grievance channels. It is an open question whether due process requires a
first-instance decision to be automatically reviewed and approved by a public official.

42

That was the position taken by the Council of State Governments (CSG) and the Urban Institute in a jointly authored report issued in
1987. See CSG Report, supra note 37 at 10 (1987). Cf. C. Ring, Contracting For the Operation of Private Prisons: Pros and Cons 21
(1987) (suggesting disciplinary panels could be made up of public and private personnel). This is also the position adopted in the
contract between the Virginia DOC and CCA. See CCA-Va. Contract at § 4.21, p. 20. Note that Ira Robbins, author of an early
study for the American Bar Association of prison privatization, took a more radical position that all disciplinary hearings should be
conducted by public corrections officials. Robbins’ view, shared by some, is that simple reliance on public involvement at the
review stage may in practice turn out to be a rubber stamp on decisions that are effectively made at the hearing stage by private
contractors tinged with a financial interest in the outcomes. Vested with considerable discretion, private prison contractors are seen
as even less predisposed than their public counterparts to permit the calling of certain witnesses. See I. Robbins, The Legal
Dimension of Private Incarceration, supra note 2 at 319. According to this view, the dominance of the private contractor in making
preliminary disciplinary hearing decisions is seen as subverting the mandate of Wolff v. McDonnell. See Note, Inmates’ Rights and
the Privatization of Prisons, 86 Columbia L. Rev. 1475, 1187 (1986) (suggesting that private officials should be excluded from
disciplinary committees).

43

In Superintendent, Massachusetts Correctional Institution, Walpole v. Hill, 472 U.S. 445 (1985), the Supreme Court enunciated a
standard that permitted revocation of an inmate’s good time credits upon a showing of "some evidence" in the record - a lower
standard of proof than a "preponderance of proof" standard.

44

In the end the practical realities of efficient administration and deference to be paid to factual determinations by on-site private
contractor personnel make it unlikely that a court would insist on a de novo standard of review to ensure compliance with due
process -- even though that is arguably what the application of the "Todd Standards" might require in a liberty interest context. If,
however, an equivalent degree of independent public authority decisionmaking is required by Todd in the prison setting, then it
would seem necessary for public authorities at least to approve or ratify every first-instance disciplinary decision made by private
contractor personnel.

45

That this is a largely a possible, rather than actual problem is partly indicated by the fact that most inmates are allowed to review
their files for inaccuracies.

Appendix 3

Private Prisons

11

private companies not be given a prominent role in the parole process, including the ability to make
overall parole recommendations. Several states have endeavored legislatively to prevent private
contractors from being able to play this pivotal ‘yes or no’ recommending role in the parole process.46
Wyoming, for example, provides that prison contractors may not “recommend that the parole board
either deny or grant parole, provided the contractor may submit written reports that have been
prepared in the ordinary course of business unless otherwise requested by the parole board.”47 These
state efforts represent a safeguard for a corrections agency to ensure that decisions about release itself
do not in any way appear tainted. A more relaxed standard might be used, however, if, as in the case
of the BOP’s existing private facility, the private contractor’s fee is not tied to the size of the inmate
population.
d.

Other Direct or Indirect Release-Related Decisions

Public authorities may also need to retain responsibility for decisions concerning the
calculation of good time credits and release dates, as well as work assignments, furlough decisions or
security classification determinations that might directly or indirectly bear on the foregoing. 48 All of
these decisions may accelerate or delay an inmate’s release date or eligibility for parole, directly
affecting his fundamental liberty interests. Even if actual financial bias is not present, constitutional
considerations suggest that even the appearance of self interest should be banished. For example, the
possibility that private prison administrators could manipulate prisoners’ work assignments to affect
their release credits necessitates public decision making. To this end, only a handful of states have
enacted statutory restrictions, or insisted on contractual solutions to the same effect.49

46

See Colo. Rev. Stat. Ann. § 17-1-203; Fla. Stat. Ann. § 957.06; Miss. Code Ann. § 47-5-1225; Ohio Rev. Code Ann. § 9.06; W. Va.
Code § 25-5-14; Wyo. Stat. Ann. § 7-22-112.

47

Wyo. Stat. Ann. § 7-22-112.

48

Although it is unclear how infrequently this might happen, a particular security classification, for example, might determine not
only the specific security conditions of incarceration, but indirectly -- though for example, special work assignments that attach to a
particular classification -- the maximum good-time credits a prisoner might receive. In the case of furloughs, a private prison
operator might be swayed to restrict furlough opportunities to cut down on the loss of per diem revenues associated with an inmate
being absent from the prison. Although in most cases, furloughed prisoners remain part of a private contractor’s designated
population.

49

For example, only eleven states appear to prevent private prison operators from granting, denying or revoking inmates’ good-time
credits. See Ariz. Rev. Stat. Ann. § 41-1609.01; Ark. Code Ann. § 12-50-108; Fla. Stat. Ann. § 957.06; La. Rev. Stat. Ann. §
39:1800.5; Miss. Code Ann. § 47-5-1225; Ohio Rev. Code Ann. § 9.06; Tenn. Code Ann. § 41-24-110; Tex. Rev. Civ. Stat. Ann.
Art. 495.004; Va. Code Ann. § 53.1-265; W. Va. Code § 25-5-14; Wyo. Stat. § 7-22-112. Some state DOC’s like Virginia’s,
contractually insist that release computations be done by public officials. See e.g., CCA-Va. Contract at § 4.24, p. 22.
With the exception of Texas and the addition of Colorado (at Colo. Rev. Stat. Ann. § 17-1-203), the above eleven states also bar
private operators from determining the kind of work that inmates may perform or determining the credits that working inmates may
receive thereby. Although they do not specifically address the impact of classification on release, seven state statutes appear to
prohibit private prison operators from making classification determinations. See Ariz. Rev. Stat. Ann. § 41-1609.01; Fla. Stat. Ann.
§ 957.06; Miss. Code Ann. § 47-5-1225; Ohio Rev. Code Ann. § 9.06; Tenn. Code Ann. § 41-24-110; Va. Code Ann. § 53.1-265;
Wyo. Stat. Ann. § 7-22-112.
Finally, it appears that nine states have prohibited contractors from deciding the type of work or work credits that individual inmates
may receive. See Ariz. Rev. Stat. Ann. § 1609.61; Ark. Code Ann. §§ 12-50-108; Colo. Rev. Stat. Ann. § 17-1-203; La. Rev. Stat.
Ann. § 39:1800.5; Miss. Code Ann. §47-5-1225; Ohio Rev. Code Ann. § 9.06; Tenn. Code Ann. § 41-24-110; Va. Code Ann. §§
53.1-265; W. Va. Code § 25-5-14.

Appendix 3

Private Prisons

12

3.

Liability for Private Prison Conditions

Of central importance to governments contemplating correctional privatization is the question
of how privatization affects liability for the operations of, and conditions in, private prisons. Matters
of liability exposure have a critical bearing on accountability and cost: can prisoners’ rights be
vindicated in a private correctional context, and if so, is there a lesser liability exposure for the
government where a private contractor is the principal defendant? For some time, relatively
extravagant claims were made by both advocates and opponents of privatization that the advent of
private prisons would insulate governments from liability exposure.50 Claims were also made that
privatization would substantially shield private contractors from the most potent and costly form of
inmate litigation -- civil rights suits involving alleged constitutional rights violations. 51
These assertions may have confused some government correctional officials and troubled civil
rights and prisoner’s advocacy groups, but the impact of privatization on public and private
correctional authorities appears to be less dramatic, and potentially more beneficial, than initially
thought. As discussed in greater detail below, privatization will not result in any diminution of
litigation exposure for private contractors. They will be treated as “state actors” to whom federal
constitutional standards apply with equal force. At the same time, due to the nature of most inmate
civil rights litigation, governments involved in contracting out correctional management will generally
find a slight reduction in exposure to liability.
This section addresses first, the impact that privatization can have on private contractors’ and
contracting governments’ liability exposure, and second, the impact that a broad indemnification
arrangement and a contractor’s comprehensive general liability insurance policy -- naming the
government as an insured -- can have on keeping the relative litigation costs of privatization low.
While the second topic is ultimately more salient, its impact cannot be fully appreciated without a
review of how privatization affects the calculus of liability between private and public defendants and
how it specifically reduces government liability exposure. There follows a review of the importance of
monitoring by government, and how effective monitoring can ensure that public correctional
authorities avoid acts and omissions of their own that might give rise to liability.

A.

The Impact of Privatization on Liability Exposure

One cannot understand the impact of privatization on governments’ and private contractors’
liability exposure without appreciating the central place that so-called “Section 1983" cases -- federal
civil rights claims alleging constitutional violations -- occupy in correctional litigation. The
mechanics of Section 1983 affect the liability of governments and contractors in different ways.

50

See, e.g., Wash. Post, March 23, 1986, at F6, col. 2 (describing how some politicians believed that government could escape liability
through use of private prisons); M. Walzer, "Hold the Justice," New Republic, April 1985, at 12 (privatization may result in an
"escape from the enforcement of constitutional norms").

51

See, e.g., Sullivan, Privatization of Corrections: A Threat to Prisoners’ Rights, in G. Bowman, et al., eds., Privatizing Correctional
Institutions 139 (1994).

Appendix 3

Private Prisons

13

1.

Inmates’ Overwhelming Reliance on Section 1983 Cases to Redress Grievances

Prisoners litigating suits involving allegations of deficient prison conditions or individual
harm suffered at the hands of prison authorities can bring a variety of legal claims to court, but only
Section 1983 cases offer an effective means of addressing constitutional grievances and obtaining
significant monetary damages. The reasons are clear: under Section 1983 litigation, inmates can seek
to remedy violations of their constitutional rights by private contractors as well as by state or local
government authorities who have contracted out prison management. Analogous actions against
federal correctional officials may be brought under so-called Bivens actions.52 Inmates can also obtain
not only actual, or compensatory damages (reimbursing them for direct harms),53 but also nominal
damages (if a constitutional deprivation causes no actual injury),54 punitive damages (monetary
awards solely for the purpose of setting an example in especially egregious cases),55 and declaratory
and injunctive relief.56 They may also sue state or local officials in their individual capacities.57
Equally important, prisoners prevailing as plaintiffs in Section 1983 actions can win attorneys’ fees,
which provide a significant incentive for inmates bringing lawsuits. 58 Finally, Section 1983 cases are
heard in federal courts, which are perceived as being more fair and impartial than state courts.
By contrast, state court actions alleging negligence or intentional conduct on the part of
private contractors and state officials usually face numerous kinds of limitations on the types of relief

52

"Section 1983" refers to the provision’s codification in Volume 42 of the United States Code, Section 1983, or as it is most
commonly abbreviated, 42 U.S.C. § 1983. Section 1983 actions are designed to offer redress to individuals who have suffered a
deprivation "under color of state law" of "rights, privileges, or immunities secured by the Constitution and laws of the United
States." Although there exists no comparable statute for suits against federal officials, the Supreme Court recognized in Bivens v. Six
Unknown Named Agents of Federal Bureau of Narcotics, 403 U.S. 388 (1971), that similar actions can be brought for violations
allegedly committed under federal auspices.

53

Compensatory damages can include those for emotional or mental distress, although the 1996 Prison Litigation Reform Act limits
such recovery to cases where mental or emotional injuries accompany a physical injury.

54

See Carey v. Piphus, 435 U.S. 247 (1978).

55

See Smith v. Wade, 461 U.S. 30 (1983). Punitive damages can be awarded as long as the defendant acted with "reckless or callous
indifference to the federally protected rights of others." Id. at 56. Most lower courts have also held that a plaintiff who recovers
only nominal damages in a § 1983 suit can still be awarded punitive damages. See, e.g., Brewer v. Chauvin, 938 F.2d 860, 864 (8th
Cir. 1991) (en banc).

56

See City of Los Angeles v. Lyons, 461 U.S. 95 (1983). A declaratory judgment usually entails a finding that a governmental statute or
regulation is unconstitutional. If a court provides the plaintiff with some type of injunctive relief, the court goes even further -directing a defendant to refrain from taking certain unconstitutional actions in the future or requiring the defendant to take certain
steps to avoid future violations of the Constitution. Compensatory and punitive damages, as well as injunctive relief, are also
available to plaintiffs in Bivens actions.

57

Bivens suits can also be maintained against federal officials in their individual capacities.

58

Fees are authorized under 42 U.S.C. § 1988, although no fees are available for plaintiffs proceeding under federal Bivens actions.
To be considered a ‘prevailing party’ under § 1988, a plaintiff need not win the central claim asserted, but can rather succeed on
"any significant issue in litigation which achieves] some of the benefit the parties sought in bringing suit." Texas State Teachers
Association v. Garland Independent School District, 489 U.S. 782, 791, 792 (1989). Although the Prison Litigation Reform Act has
placed a cap on attorneys’ hourly rates in prisoner suits, see 18 U.S.C. § 3006A(d), as well as the total fees recovered as a proportion
of the relief obtained, see 42 U.S.C. § 1997e(d)(1)(B)(i), the availability of any fees at all still makes Section 1983 an attractive
vehicle for prisoner’s rights litigation.

Appendix 3

Private Prisons

14

that can be obtained,59 especially where the common law or state legislatures may have immunized
states and/or state officials from suit for actions arising from official conduct. 60 Moreover,
governments generally cannot be sued for the negligent acts of their independent contractors,61 and
most federal and state tort claims acts bar claims that involve “discretionary functions,” which could
include both the contracting out of the correctional function62 and the placement of an individual
inmate in a privatized facility.63 Only tort suits involving allegations of negligent oversight or
monitoring of a private prison are likely to succeed under a state or federal tort claims act, as the case
may be.64
Three major questions under Section 1983 and Bivens litigation that arose in the mid-1980's in
the wake of the first significant correctional privatizations were (1) could private prison contractors be
sued as defendants acting “under color of state law?” (or federal law, in the case of Bivens suits); (2)
if private prisons were indeed “state actors” amenable to suit under Section 1983 and Bivens, would
private prison contractor officials be able to share the “qualified immunity” defense that public
officials “acting under color of state law” enjoyed? and (3) if private prison contractors were deemed
“state actors” amenable to suit under Section 1983, what kind of liability, if any, would still attach to
public correctional officials who might still have various kinds of continuing responsibilities toward
private facilities? These questions assumed great significance as policy makers sought to ascertain the
true costs and benefits of privatization, and as the public sought to learn whether private prisons
would permit governments to avoid their constitutional responsibilities toward inmates.
2.

Are Private Prison Contractors Amenable to Suit under Section 1983?

The Supreme Court appeared to answer the “state action” question in 1988 in the case of West
v. Atkins.65 In West, the Court concluded that a private physician who worked in a state prison several
days each week could constitute a defendant under Section 1983. The Court dismissed the notion that
the private doctor was not covered by Section 1983 simply because he was a not a state employee and
worked part-time in the prison. What mattered, according to the Court, was the “function within the
state system” that the doctor played within the prison: he had agreed to assume the government’s
responsibility, delegated specifically to him, to furnish medical care required by the Eighth

59

Ordinary tort actions may be severely circumscribed by state tort claims acts, and usually provide only for the award of
compensatory damages. Except in the possible case of medical malpractice cases, such damages are in any event usually
exceedingly modest, given the status of prisoners and the fact that most of their needs are furnished by the prison.

60

See, e.g., Utah Code Ann. § 63-30-10(j) (explicitly excepting prisoners’ suits from waiver of sovereign immunity).

61

See Restatement (2d), Torts, § 409. This principle is statutorily codified in the Federal Tort Claims Act at 28 U.S.C. § 2671.

62

See, e.g., Gotha v. United States, 115 F.3d 176, 179-80 (3d Cir. 1997)(discussing situations where major government decisions
"susceptible to policy analysis" constitute "discretionary functions" immune from suit under the Federal Tort Claims Act, 28 U.S.C.
§ 2680(a). As such, the ‘discretionary function’ exemption would appear to bar any negligence suit based on a theory of "negligent
selection" of a private contractor. Everyday oversight over a private contractor via a prison monitor may conceivably still expose a
corrections agency to negligence liability, but this will undoubtedly arise in the context of a larger Section 1983 suit.

63

See, e.g., Bailor v. Salvation Army, 51 F.3d 678 (7th Cir. 1995)(pre-release placement of inmate in half-way house deemed
discretionary function immune from suit).

64

See discussion on Section 3.B.2, infra.

65

487 U.S. 42 (1988).

Appendix 3

Private Prisons

15

Amendment to the Constitution. His specific function in the unique prison setting was to discharge
the state’s medical care responsibilities which he “possessed by virtue of state law and [which were]
made possible only because [he was] clothed with the authority of state law.”66
There should be no reason to doubt the application of West v. Atkins to private contractors
running entire prison facilities, either on legal or policy grounds. The special, all-encompassing role
that private contractors play in the day-to-day management of privatized facilities -- their “function
within the state system”67 -- together with the understanding that contractors have assumed
responsibility for protecting inmates’ constitutional rights, make it highly unlikely that a private
prison management company could escape Section 1983 liability. This would seem to be the case
even if the reasoning in another Supreme Court state action case were applied.68 At least one federal
circuit court has explicitly endorsed the application of West v. Atkins to private correctional
management.69
It is important to note that, at present, notwithstanding the West v. Atkins holding, it is an
untested proposition in many areas of the country whether Bivens actions may be brought against
federal prison contractors as “state actors.” Only four judicial circuits have unambiguously
recognized private contractors as appropriate Bivens action defendants, while most of the circuits have

66

Id. at 55.

67

Id. This reference to the doctor’s function in the state system may or may not signal reliance on a line of Supreme Court cases
finding state action where the activity at issue has been a "public function" -- defined by the Court to be "traditionally exclusively
reserved to the state." Jackson v. Metropolitan Edison Co., 419 U.S. 345, 352 (1974). Because prisons cannot, strictly speaking, be
said to have been "traditionally exclusively reserved" to the state due to the 19th century legacy of private prisons -- something the
Supreme Court itself acknowledged in the recent case of Richardson v. McKnight, 117 S. Ct. 2100, 2104 (1997) -- there is still some
debate as to whether the "public function" test is applicable to Section 1983 cases in the new correctional privatization environment.

68

One of the most interesting questions regarding private prisons and ‘state action’ is why the Supreme Court, in the recent decision of
Richardson v. McKnight, 117 S. Ct. 2100 (1997) -- a case that involved a private prison run by Corrections Corporation of America
-- instructed the trial court on remand to analyze the state action question under the holding not of West, but of Lugar v. Edmonson
Oil Co., 457 U.S. 922 (1982). The best explanation is that Lugar actually enunciated a test to determine state action, while West
was less clear about its governing principles. The real question, however, is whether by referring to Lugar, the Court was signaling
a reaffirmation of the increasingly restrictive approach to state action that it began to develop in the 1980's. But the result in the
private prison context is in fact likely to be the same regardless of which decision is relied on. The two-part test in Lugar requires
first, that a constitutional deprivation "must be caused by the exercise of some right or privilege created by the state or by a rule of
conduct imposed by the state or by a person for whom the state is responsible"; and second, that "the party charged with the
deprivation must be a person who may fairly be said to be a state actor." The definition of state actor in the second part of the test
includes a person jointly acting with state officials, or who "has obtained significant aid from state officials," or "whose conduct is
otherwise chargeable to the State." Id. at 937 (emphasis supplied). Thus, although the Lugar decision itself revolved around a
finding of joint activity by public and private parties, the last phrase of the two-pronged test would appear to be expansive enough to
attribute liability to almost any kind of private prison company assuming all or virtually all management functions at a facility.

69

See Skelton v. Pri-Cor, Inc., 963 F.2d 100, 102 (6th Cir. 1991) cert. denied, 503 U.S. 989 (1992), quoting Plain v. Flicker, 645 F.
Supp. 898, 907 (D.N.J. 1986)("[I]f a state contracted with a private corporation to run its prisons it would no doubt subject the
private prison employees to § 1983 suits under the public function doctrine"). See also Street v. Corrections Corp. of America, 102
F.3d 810, 814 (6th Cir. 1996) (private operation of prison is a public function). At least one district court has reasoned similarly.
See Payne v. Monroe County, 779 F. Supp. 1330, 1335 (S.D. Fla. 1991)(Wackenhut Corp., operator of county jail, deemed state
actor performing public function).

Appendix 3

Private Prisons

16

not ruled on the issue. One circuit has refused to allow Bivens suits against private parties.70 Lower
courts confronted with private contractor defendants in Bivens actions generally continue to find ways
to avoid ruling on the issue.71 Still, it is hard to see how these remaining judicial circuits will be able
to resist the logic of West in the Section 1983 context if and when an analogous Bivens case appears
before them.72 Interestingly, the current Bureau of Prisons facility in Taft, California falls within the
9th Circuit, which has in fact recognized private contractors as appropriate Bivens defendants.
3.

Do Private Prison Officials Have a ‘Qualified Immunity’ Under Section 1983?

Even with increasing general agreement that private prison contractors and employees are
“state actors” subject to Section 1983 liability, it remained uncertain until recently whether such
employees could raise the defense of “qualified immunity” that is extended to public officials
defending against monetary claims in such litigation.73 Such immunity shields from liability officials
exercising discretionary functions if, in light of the clearly established law existing at the time, they
could reasonably believe that their conduct was lawful. The immunity is therefore helpful in
protecting officials from liability in situations where the applicable law is in flux and they must
operate with resolve and confidence in volatile environments. Proponents of extending qualified
immunity to private prison officials argued that officials acting “under color of state law” in managing
a prison should have same defenses available to public employees. This would allegedly create an
“even playing field” allowing fair competition between public and private entities.74
Last year, in Richardson v. McKnight,75 the Supreme Court rejected such an approach, ruling
that private prison guards cannot raise a qualified immunity defense to a Section 1983 action. The
Supreme Court’s reasoning relied on both history and public policy. The Court found that the “mere
performance of a governmental function” by a private defendant could not, alone, entitle that
defendant to claim qualified immunity.76 The purpose of qualified immunity is to prevent public

70

The four circuits that have found Bivens actions to be appropriate against private defendants are the D.C., 5th, 6th, and 9th Circuits.
See, e.g., Reuber v. U.S., 750 F.2d 1039, 1053-57 (D.C. Cir. 1984); Dobyns v. E-Systems, Inc., 667 F.2d 1219 (5th Cir. 1982);
Yiamouyiannis v. Chemical Abstracts Service, 521 F.2d 1392 (6th Cir. 1975); Schowengerdt v. General Dynamics Corp., 823 F.2d
1328, 1337-38 (9th Cir. 1987). The one circuit that has found to the contrary is the First. Fletcher v. Rhode Island Hosp. Trust
National Bank, 496 F.2d 927, 932 n. 8 (1st Cir.), cert. denied, 419 U.S. 1001 (1974). Other circuits have either not dealt with the
issue or have purposely left it untreated. See DeVargas v. Mason & Hanger-Silas Mason Co., 844 F.2d 714, 720 n.5 (cataloguing
status of private party amenability to suit by circuit).

71

See, e.g., Smith v. United States, 896 F. Supp. 1183 (M.D. Fla. 1995) (court merely assumes private half-way house is proper
defendant under Bivens action against half-way house and U.S. Bureau of Prisons; McChan v. CCA, 1997 WL 104110 (D. Kan.
1997)(same assumption in case involving privately-managed federal detention center); Gleave v. Graham, 954 F. Supp. 599, 604
(W.D.N.Y. 1997)(court does not rule on issue, where private half-way house is the principal defendant).

72

Indeed, the Supreme Court has noted how liability standards in Bivens cases generally track those in Section 1983 litigation. See
Siegert v. Gilley, 500 U.S. 226 (1991).

73

The qualified immunity defense applies only to actions against officials for monetary relief.

74

See, e.g., Thomas, Resolving the Problem of Qualified Immunity for Private Defendants in Section 1983 and Bivens Damage Suits,
53 La. L. Rev. 449, 454 (1992).

75

117 S. Ct. 2100 (1997).

76

The Court noted that "[h]istory does not reveal a ‘firmly rooted’ tradition of immunity applicable to private employed prison
guards," id. at 2104, and further observed that "[c]orrectional functions have never been exclusively public" in the U.S. by virtue of
the history of private jails and prisons in the 18th and 19th centuries. Id.

Appendix 3

Private Prisons

17

officials, who have civil service protection and less strict oversight, from exhibiting “unwarranted
timidity” in the face of liability. The Court noted, however, that private companies and their
employees have other incentives, generated by state monitoring and the marketplace, to spur them to
perform their duties zealously. Moreover, such contractors can obtain relatively inexpensive liability
insurance for their employees that will further minimize any disincentives for good performance. 77
The Court concluded that its decision to deny qualified immunity should apply to any situation where
“a private firm, systematically organized to assume a major lengthy administrative task...with limited
direct supervision by the government, undertakes that task for profit and potentially in competition
with other firms.”78
Despite the narrow decision in McKnight and an agitated dissent by the minority justices, the
ruling is likely to apply to virtually all private contractors managing prisons, even where competitive
market pressures may be relatively weak.79 The bigger question is whether the case’s significance to
current private prison operations is more modest than it initially appeared. To begin with, qualified
immunity by its terms is generally most helpful to public officials when the legal standards governing
their work are unclear and a reasonable official might believe that his conduct was lawful. In the case
of prison litigation, however, constitutional law standards that might have seemed confused or
unsettled in the 1970s and 1980s have now reached a degree of clarity and relative stability that render
the qualified immunity defense far less valuable than in earlier years. Moreover, in monetary terms,
the impact of McKnight on private prison contractors has been negligible, since the additional
insurance necessary to cover individual prison officials’ liability is only marginally higher than that
necessary to cover the corporate liability of the firms themselves.80 Finally, the McKnight decision
left open the possibility that a potentially equally meaningful “good faith” defense to liability -- rather
than an absolute defense to suit -- still might be asserted by private prison officials who genuinely
believed their actions were in accord with prevailing law.81 In the end, the slight increase in private
contractors’ exposure to liability appears not to have had any discernible impact on those firms’
costs.82

77

See id. at 2106-07. The Court argued that if private companies or their employees flag in their duties, they may be replaced in the
competitive procurement process or through the action of market and reputation forces.

78

Id. at 2107. The McKnight decision would also appear to put to rest the notion that a private contractor could avail itself, in federal
actions against states and their high-ranking officials, of the defense of absolute "sovereign immunity" that would also shield it from
Section 1983 liability. Although the Supreme Court did not address this issue explicitly, its reasoning extends to both types of
immunity for private prison contractors and their employees.

79

While the minority opinion, written by Justice Scalia, emphasized that the private prison industry is dominated by a few major
players and often driven by political lobbying and decision making, it does not necessarily follow that the basic profit motive, and
its incentives for private guards to perform effectively, are weak. There are still significant contract and market pressures at work.
See, e.g., Thomas & Logan, The Development, Present Status, and Future Potential of Correctional Privatization in America, in G.
Bowman, et al., eds., Privatizing Correctional Institutions 233 (1993) (noting that profitability and share value are usually tied to
meeting government contractual expectations and enhancing firm’s public reputation).

80

Not only did private prison contractors not assume that their officials were protected by qualified immunity prior to McKnight, but
their insurance policies before and after McKnight typically indemnify private prison officials for all liability relating to their prison
work.

81

117 S. Ct. at 2108.

82

In this regard, it also does not seem persuasive to suppose, as did the McKnight minority, that private contractors will "down-play
discipline" and otherwise exhibit timidity in their work so as to reduce the cost of § 1983 litigation. Anecdotal evidence suggests
that private firms are not deterred from entering or profiting in the market . See generally Note, Qualified Immunity -- Privatized

Appendix 3

Private Prisons

18

4.

Private Contractor and Governmental Liability Exposure in Section 1983 Litigation

With private prison contractors deemed “state actors” amenable to suit under Section 1983,
and with contractor personnel not immune from suit, it becomes possible to analyze the general
contours of how liability would be distributed among contractors and government correctional
authorities in a privatized correctional environment, and whether, and to what degree, privatization
helps or harms governments and prisoners involved in inmate litigation. The most important factor,
from a cost standpoint, is simply the availability to contractors and governments of a contractor’s
comprehensive general liability insurance policy. Still, even from the standpoint of exposure to suit, it
is clear that the interposition of a private contractor handling the day-to-day management of a
correctional facility will tend to reduce significantly the exposure of governments and senior public
correctional authorities, particularly in cases involving isolated or individual incidents. It should not,
meanwhile, affect adversely the ability of prisoners to obtain relief for legitimate constitutional
complaints. To understand these effects, it is useful to examine the respective Section 1983 liabilities
of private contractors and government entities (i.e., local governments, states, federal government).
a.

Private Correctional Firm Liability

As a private entity functioning as a “state actor,” a private correctional contractor would be
amenable to suit under Section 1983 or Bivens, and subject to both monetary damages and injunctive
relief if it were found to have violated one or more inmates’ constitutional rights. Hence, a private
contractor would be subject to the same liability standards established by the Supreme Court for
public actors, including those relating to causation and state of mind.83 Since, however, there is no socalled respondeat superior liability for any employers under Section 1983 -- meaning, no automatic
liability of an employer for the actions of its employees84 -- a firm and its supervisory employees must
be shown to have been directly involved in the alleged violation, have known about the violation or its
likelihood of occurring and been “deliberately indifferent toward the risk, or have generated or
validated a policy or custom that led to the violation.85
As in the case of public employers, in practice this means that while inmate cases may
succeed against private correctional officers, it may be much more difficult to reach supervisory
personnel. Such higher-level personnel may be insulated from liability in cases involving isolated

Government Functions, 111 Harvard L. Rev. 390, 398-400 (1997) (private firms appeared undeterred by lack of immunity in
aggressively operating prisons). Also, since litigation costs are merely possible future costs, firms will usually prefer to reap savings
through immediate staffing and training efficiencies rather than through presumed litigation savings. See id. at 400 n. 75.
83

In the case of an Eighth Amendment claim alleging safety problems stemming from other inmates, for example, a plaintiff would
need to show that the private correctional officials’ actions or omissions had a causal connection to the injuries alleged, and that the
officials displayed "deliberate indifference to a substantial risk of serious harm." Farmer v. Brennan, 511 U.S. 825, 828 (1994).

84

See Monell v. Department of Soc. Serv., 436 U.S. 658, 694 (1978). Although Monell involved a municipal corporation, all federal
circuit courts have agreed that there is no respondeat superior liability for private corporations as well. See Street v. Corrections
Corp. of Am., supra 102 F.3d at, 818.

85

See, e.g., Street v. Corrections Corp. of Am., supra (private prison contractor and warden not proper defendants for alleged Eighth
Amendment violation by prison guard; no showing of custom or policy); Nelson v. Pennsylvania, 1997 WL 793060, at 3 (E.D. Pa.
1997)(allegations against private prison warden and assistant warden dismissed due to lack of participation in or knowledge of and
acquiescence in activity alleged to have caused constitutional deprivation).

Appendix 3

Private Prisons

19

incidents and lower level personnel (e.g., guards or medical staff),86 depending on their particular
level of knowledge or lack of forewarning of specific problems or risks. Even in cases where a failure
to train correctional officers is at issue, the failure to train must be accompanied by officials’
“deliberate indifference” to the risk that deficient training will result in constitutional violations. 87
Mere negligence -- on the part of line employees or their supervisors -- does not give rise to a Section
1983 claim.88
On the basis of comparative recovery of damages, it can be said that privatization may bring
certain benefits to inmates bringing meritorious Section 1983 suits. First, in the case of state and
federal corrections (although not local government corrections), due to the Eleventh Amendment
and/or statutes on sovereign immunity, state and federal correctional employees are not amenable to
suits for monetary damages except in their personal capacity.89 This means that only damages against
their individual assets may be obtained, thereby significantly limiting the amount of money that can be
obtained in such cases.90 By contrast, damages obtained against private correctional employees can be
paid out of the employing firm’s assets, which could be quite large. Moreover, federal, state and local
governments may have statutory limits on damage awards or be exempted from certain types of
damages, 91 while private contractors are not similarly restricted.
b.

Government Correctional Authority Liability

Contracting for prison management services also means lower overall liability exposure for
governments. This lower liability profile results directly from the Section 1983 causation and state of
mind standards discussed immediately above: insofar as public corrections authorities will have
entrusted day-to-day management of prisons to private contractors, they will be less likely to have
advance notice or knowledge of specific harms alleged to have caused injury to individual inmates.
While reliance on a private contractor will obviously not insulate government authorities from being
named in lawsuits92 or being exposed to liability for widespread or obvious problems with prison
conditions, it will greatly lessen the liability of government supervisory officials for most inmate
claims alleging individual harm. These claims represent the most common type of inmate lawsuit,
and assume a significant proportion of a corrections authority’s litigation budget.

86

Thus, for example, in an Eighth Amendment inmate freedom from harm case, if supervisory officials were to be held liable, an
inmate would have to show not only that front-line guards were "deliberately indifferent" to a strong likelihood that the inmate’s
safety was in danger, but that the supervisory officials had similar knowledge of a risk and recklessly and consciously disregarded it.
See, e.g., Payne v. Monroe County, supra at 1334 (claims against private jail management firm dismissed; no showing of deliberate
indifference on part of firm).

87

See City of Canton v. Harris, 489 U.S. 378 (1989).

88

See Daniels v. Williams, 474 U.S. 327 (1986); Davidson v. Cannon, 474 U.S. 344 (1986).

89

Also by virtue of sovereign immunity, states themselves are not amenable to suit. See, e.g., Will v. Michigan Dept. of State Police,
491 U.S. 58 (1989).

90

Although the government cannot be tapped directly for damages, it can, by statute and based on discretion, choose to indemnify
government employees sued in their personal capacities. States, are, however, liable for attorneys’ fees under 42 U.S.C. § 1988 in
cases for injunctive relief where state officials were sued in their official capacity and lost.

91

In addition, the Supreme Court has held that punitive damages are not obtainable against municipal governments. See City of
Newport v. Fact Concerts, Inc., 453 U.S. 247 (1981).

92

As one commentator has noted, "[T]he government cannot hide behind its private operator. An inmate can sue either party." M.
Gold, The Privatization of Prisons, 28 The Urban Lawyer 359, 380 (1996).

Appendix 3

Private Prisons

20

Governments will still be subject to injunctive relief if they permit obviously unconstitutional
conditions to arise in private prisons.93 Municipalities and county governments, unprotected by
sovereign or qualified immunity, will also face monetary damages, including punitive damages in such
situations where a “policy or custom” bearing on the alleged harm is deemed to exist.94 But
municipalities, together with state and federal officials, will face monetary liability in cases of
individual harm only if, based on Section 1983 state of mind requirements, they authorize improper
conduct by a private contractor, or have knowledge of that conduct and consciously fail to take steps
to correct it. In most cases involving these kinds of allegations, courts will dismiss claims against
upper level officials, who are deemed entitled to rely on the judgment and professionalism of private
contractors reacting to the day-to-day challenges of prison management.95

B.

The Role of Indemnification, Insurance and Monitoring in
Reducing Liability and Litigation Costs

This section addresses the importance to public correctional authorities of (a) being covered
by a broad indemnification agreement covering the contractor’s acts and omissions, and (b) being
named as an insured on a private contractor’s comprehensive general liability insurance policy. These
important advantages are discussed in greater detail below. This section also reemphasizes how
Section 1983 will not impose liability on government officials unless they not only know about a
problem but are deliberately indifferent toward it. Prudent monitoring of private facilities by the
government has the clear benefit of alerting public officials to serious and systemic problems in a
facility that, if responsibly dealt with, will not generate damaging litigation -- not to mention public
relations problems -- in the first place. Perhaps equally important from a liability and cost standpoint,
an effective monitor can prevent public authorities from committing their own acts and omissions that
are not generally covered by the contractor’s insurance policy.
1.

Indemnification and Insurance

It goes without saying that any expectations about privatization’s purported litigation cost
advantages would prove completely illusory without government insistence on an effective

93

See, e.g., Nelson v. Prison Health Serv., Inc., 1997 WL 821531 (M.D. Fla.1997) (county could be found liable for deficient medical
care at private jail based on repeated receipt of monitoring reports identifying problems).

94

Sometimes, courts have viewed municipal government liability broadly as non-delegable where contracting decisions are viewed as
constituting a "policy or custom." See, e.g., Ancata v. Prison Health Services, 769 F.2d 700, 705-06 nn. 9 & 11 (11th Cir. 1985)
(county liable together with private prison health provider for inadequate medical care furnished to inmates); Blumel v. Mylander,
954 F. Supp. 1547, 1556 (M.D. Fla. 1997) (county liable with private contractor CCA for Fourth Amendment violation where
county policy entrusted CCA with policy making authority on subject at issue). Custom has been interpreted broadly to include a
"widespread practice, that, although not authorized by written law or express municipal policy, is so permanent and well settled as to
constitute a custom or usage with the force of law." St. Louis v. Praprotnik, 485 U.S. 112, 127 (1988), quoting Adickes v. S.H. Kress
and Co., 398 U.S. 144, 167-68 (1970). If a custom is shown to be so longstanding and widespread that it is considered to have been
"authorized by the policy-making officials because they must have known about it but failed to stop it," liability may attach under §
1983. See Payne v. Monroe County, supra at 1334 (section 1983 prison suit against county and private contractor dismissed where
no showing that inmate safety problem was known or should have been known).

95

Cf., e.g., Calvert v. Hun, 798 F. Supp. 1226, 1232-33 (N.D. W.Va. 1992) (Eighth Amendment claims against state prison officials
dismissed where no showing that such officials actively or tacitly approved provision of allegedly incompetent medical care to
individual inmate by private health provider).

Appendix 3

Private Prisons

21

indemnification and insurance scheme. By demanding full indemnification and compensation of the
government for all acts and omissions of the private contractor that may give rise to liability
surrounding the operation of a private facility, public corrections authorities can ensure that there is
no litigation cost increase attending privatization. Being named as an insured on a private contractor’s
comprehensive general liability policy can provide governments with the security they need in order
to turn over complete operation of a prison and to focus their attention on specific matters of facility
oversight and proper contractor performance.
To ensure that they deliver their full benefit, indemnification clauses typically indemnify,
defend, and hold harmless the government and all of its employees, agents, and officers from any
claims or losses -- including all damages, costs, fines, and/or settlements -- stemming from the
contractor’s acts or omissions in connection with performance of a privatization contract.96 They are
also typically written in such a way as to apply to all subcontractors, agents, or any other persons
supplying services or supplies in connection with the contract. Beyond these basics, a contract may
explicitly provide for indemnification for specific harms, including prison escapes, failure to comply
with federal, state, or local legislation (particularly labor laws), and unsanctioned use of information
or data collected or used in the course of performance of the contract (e.g., research or personal data
implicating privacy, copyright, or trade secret concerns).97 It may also address government
requirements that the government be a signatory to any settlement or explicitly approve any settlement
in excess of insurance limits.
Insurance provisions in a contract constitute a logical extension of the indemnification
clauses. A contracting government agency will usually be named as an insured on any and all
insurance policies written for the private contractor managing one or more prison facilities.98 By
being named explicitly as a direct insured party -- rather than as merely a beneficiary -- a government
can strengthen its indemnification provisions vis-a-vis the contractor and its rights against the insurer.
Among the most important of those rights is the right to advance notice of any nonpayment or
cancellation of the policy by the contractor. Without these direct rights -- contained in the insurance
policy or in the contract with the private contractor99 -- the government’s protection may be illusory.
The contracting agency may also insist that it be given advance notice of, and opportunity to approve,
a new certificate of insurance that replaces an expiring policy.100

96

It is important that the indemnification extend to performance of the contract in all of its dimensions (including by implication all
manner of omissions), rather than simply the operation of the facility in question, which might otherwise narrow the coverage of the
hold harmless benefit.

97

See generally, e.g., Contract Between Commonwealth of Kentucky and U.S. Corrections Corporation for the Operation of Three
Minimum Security Facilities, December 9, 1993 (hereinafter "Kentucky Contract"), at 33.

98

There appears to be unanimous agreement that third-party insurance obtained by the contractor is the safest and most cost-effective
means that governments have of being properly insured. See Robbins, The Legal Dimensions of Private Incarceration, supra at 854,
n. 644 (discussing problems with contractors self-insuring).

99

See, e.g., Kentucky Contract at 34; CCA-Va. Contract at § 7.4, p. 38.

100 See, e.g., State of California Department of Corrections Standard Contract, Attachment B (General Terms and Conditions), ¶ 19 at 6,
furnished as Attachment 1 to RFP for Expansion and Operation of Medium Security Community Correctional Facility, October 13,
1995.

Appendix 3

Private Prisons

22

In developing specific provisions for insurance coverage, governments generally seek to cover
all possible liabilities arising from the operation of the privatized facility, including both personal
injury claims and Section 1983 litigation. Liability areas include those for personal injury,
professional (malpractice) liability (including medical malpractice), contract liability, property
damage, worker’s compensation liability, theft and fraud, and automobile liability (to protect against
damages stemming from transport of inmates). Insurance policies are also generally sought that
explicitly provide for the payment of litigation defense costs.101 Depending on risk calculations
specific to individual facilities, governments will also seek to prescribe insurance amounts that
insurance policies should contain.102 In addition to liability insurance, contracts may also provide that
contractors furnish performance bonds as a means of ensuring proper performance of the contractedfor services, although the usefulness of this tool is limited.103
Government indemnification and coverage under a contractor’s third-party liability policy go
hand-in-hand with the practical reality that public officials have turned over day-to-day prison
management to private firms that must, as a matter of ordinary accountability and proper market
pricing, bear the direct costs of damages they may cause in executing their responsibilities. Attentive
governments will closely monitor both the quality of insurance initially obtained by the contractor, as
well as changes in coverage and premiums that may have an impact on the overall savings derived
from the management contract. While market pressures and profitability concerns should, over time,
be expected to result in lower premiums for contractors that perform responsibly, governments can
still be expected to monitor contractor performance closely to make sure that there is no offsetting
tendency to cut back on staffing, training, and other inputs affecting the bottom line.
It should be noted that government litigation costs at a particular facility may or may not be
lower with prison management in the hands of a private contractor. A great deal depends on the
nature of the facilities involved, the profile of the inmate population, and the risk management
approaches taken by particular public versus private authorities. The latter approaches can themselves
vary greatly according to the overall management plan, quality of staff, and claims handling
procedures employed by those administering a specific prison facility or system. While it may be
contended that private authorities can exercise greater flexibility than their public sector counterparts
in given situations, this assertion cannot be readily substantiated.104 Thus, even though a prudent

101 See, e.g., CAA-Va. Contract at § 7.3, p. 38.
102 For example, Virginia has required a contractor managing a medium security facility to carry comprehensive general liability
insurance in an amount no less than $2 million per occurrence and $5 million in the aggregate. See CCA-Va. Contract, supra at §
7.3, p. 37. Kentucky requires commercial general liability insurance to be carried by a contractor for a minimum security facility in
the amount of $2 million per person and $3 million per occurrence. See Kentucky Contract, supra at 35. Most specialists seem to
think that $5 million per occurrence is a completely adequate amount for a medium security prison with 500-1000 beds. For medical
malpractice insurance, the U.S. Bureau of Prisons prescribed an amount of $1 million per medical speciality per occurrence. See
RFP for Minimum Security Prison, Taft, California, Nov. 27, 1996 (hereinafter "Taft RFP"), at 72.
103 Although performance bonds have been viewed by some as useful where the contracting agency wishes to protect itself from events
such as contractor bankruptcy, default, strikes, or a takeover by or merger into, a new firm, many believe they are unnecessary since
they usually require replacement of a private contractor by a vendor of the insurer’s choosing, and because it is usually easier and
cheaper for public authorities to resume control.
104 Some privatization proponents have asserted that private firms may prove more responsive to inmates and the general public in
remedying problems that result in prison litigation. They content that a private contractor can more expeditiously to correct a
problem or settle a lawsuit when political pressures are absent, fewer approvals are needed, and funding is more readily accessible.
Also, a private firm’s concern with market reputation and profitability is alleged to create incentives for flexibility and holding costs
down. There remains no solid evidence or study to support these contentions, however plausible they may be.

Appendix 3

Private Prisons

23

government correctional agency will insist as a matter of course that a contractor indemnify and hold
it harmless against all acts and omissions of the contractor arising under a management contract -- and
even though such an agency will similarly insist that it be named as an insured on any private
comprehensive general liability insurance policy covering particular prison facilities -- there is no way
to tell for sure whether its litigation expenditures under privatization will be lower. 105
2.

Monitoring to Safeguard Against Liability Resulting from Government Duties

Indemnification and insurance covering the acts and omissions of the private contractor will
not necessarily shield public correctional authorities for failing to exercise that degree of oversight
deemed appropriate to particular contract circumstances. As the Supreme Court acknowledged in the
case of Logue v. United States,106 government correctional authorities may be liable for their own
negligence even though management of a correctional facility has been given over to an independent
contractor.107 To safeguard against the continuing possibility of liability for their own conduct, as
well as to ensure proper contractor performance and compliance with contract terms, most
governments have statutorily108 or contractually provided for public corrections monitors to oversee
private prison operations. Monitors may be individuals who fulfill roles significantly different from
those performed by auditors, inspectors, and performance evaluators whose interventions are less
frequent and less concerned with day-to-day operations. Without such frequent oversight, it is possible
that a government could be found negligent -- according to an evolving standard of reasonableness in
the corrections community -- for negligent or deficient monitoring.109

105

Clear answers to this question are likely to be exceedingly difficult to obtain. To begin with, public entities do not carry third party
liability insurance, so direct cost comparisons are impossible. Also, public correctional authorities and contractors alike would find
it very hard to isolate with any precision all of their relevant litigation costs (staffing costs, particularly in the public sector, are hard
to attribute exclusively to litigation-related functions), and may be reluctant to disclose their damage awards paid out. Ascertaining
what portion of the contractor’s per diem is attributable to litigation defense would probably be quite speculative. Finally, it would
be extremely difficult to control for potentially wide differences in prison populations, facility conditions, and the past history of
litigation at one or more facilities. Still, one could surmise that public authorities have a number of factors that could aid them in
keeping certain litigation defense costs relatively low. These include the judicious use of salaried government attorneys to handle
virtually all prison litigation, possibly less expensive claims handling, and the lack of a profit mark-up. Private contractors, by
contrast, might feature relatively more expensive claims handling and legal assistance (e.g., though contract attorneys billing at
market rates), as well as that portion of their fee allocated to such work.

106 412 U.S. 521 (1973).
107 See id. at 532-33. In Logue, the Court acknowledged that a U.S. Marshal could be found negligent for his own acts of negligence in
failing to take certain precautions in placing a federal inmate in a county jail under an intergovernmental agreement pursuant to 18
U.S.C. § 4002. While indemnification and insurance covering a private contractor will help with litigation stemming from prison
condition problems, it will not obviate government officials liability for negligent selection or oversight of a contractor. As
discussed above, Section 1983 litigation has in recent years confirmed this same fact in the context of private prison contractors
specifically, see § 3.A, supra.
108 See, e.g., Fla. Stat. Ann. § 957.04(1)(g) (requiring full-time monitor with unlimited access to private facility); Ky. Rev. Stat. Ann. §
197.510(29) (requiring frequent monitoring); N.M. Stat Ann. § 33-1-17(D)(7) (requiring monitoring of compliance).
109 See CSG Report, supra note 37, at 533. Lack of monitoring or grossly deficient monitoring could also figure as a major grounds for
voiding a contract on grounds of excessive delegation.

Appendix 3

Private Prisons

24

In some jurisdictions, and pursuant to certain contracts, monitors are full-time Department of
Corrections employees who work at private facilities, 110 while in other cases, monitors may only visit
intermittently.111 In still other cases, a monitor may be an independent contractor who visits a facility
only several days per month.112 RFPs or contracts may require a private prison operator to set aside
space for a government monitor and to assume all costs associated with the monitor’s work other than
his or her salary.113 The monitor’s mandate in many cases may be a broad and active one, in which
ongoing guidance is provided to the private contractor.114 A wide-ranging, ongoing document review - including paperwork on population, classification, discipline, grievances, serious incidents, and
programming115 -- can complement direct observation. In many cases, the level of oversight received
by a private contractor may ultimately exceed that received by publicly-operated prisons.116
A review of alternative approaches to monitoring and the many deficiencies that may attend
many of these approaches,117 lies beyond the scope of this review. For present purposes, it is
important simply to recapitulate the legal implications of using -- or not using -- a monitor. The
problems of government negligence liability resulting from deficient monitoring or a failure to
monitor at all have already been addressed.118 If a monitor is employed, however, public corrections
authorities should make every effort to take the monitor’s role -- and the information he or she relays
to the public agency -- seriously. Otherwise, public correctional officials may find themselves
exposed to “deliberate indifference” allegations in a Section 1983 action for failing to respond
appropriately to information provided by the monitor.

110 See, e.g., Florida, where this is statutorily required.
111 See, e.g., North Carolina Department of Corrections RFP for Two 500-Cell Medium Security Institutions, Dec. 19, 1995 (hereinafter
N. Carolina RFP), at 19.
112 This arrangement exists in the case of the contract between the District of Columbia and the Northeast Corrections Center in
Youngstown, Ohio, operated by Corrections Corp. of America. Although this arrangement was apparently selected in light of
concerns that the D.C. Government could not perform the job properly, it is far from an ideal situation, given that the monitor
resides out of state and only visits the facility a few days per month. This may or may not be viewed as a modest, though important,
contributing factor in the D.C. Government’s failure to respond appropriately to repeated problems of inmate violence at the prison.
113 See, e.g., N. Carolina RFP at 19.
114 For example, the contract monitor for the U.S. Bureau of Prisons’ Taft, California, minimum security facility is empowered to
provide "technical direction" to "all work performed" under the contract, which can include "[d]irections to the contractor which redirect the contract effort, shift work emphasis between areas or tasks, require pursuit of certain lines of inquiry, fill in details or
otherwise serve to accomplish the contractual scope of work." Taft RFP, supra note 102, at 56.
115 See Keating, Public Over Private: Monitoring the Performance of Privately Operated Prisons and Jails, in D. McDonald, ed.,
Private Prisons In the Public Interest 130, 145 (1990). Many private prison contracts require full access to prison facilities and
private contractor books and records, without notice. See, e.g., CCA-Va. Contract at § 4.38, p. 27.
116 See Calabrese, Low Cost, High Quality, Good Fit: Why Not Privatization, in G. Bowman, et al., eds, Privatizing Correctional
Institutions, supra at 184-85.
117 Among the most thorny issues are monitors being co-opted by contractors, see, e.g., D. Shichor, Punishment for Profit 121-22
(1993); confusion and blurring of the chain of command that occurs when prison administrators must answer to government
officials, id. at 120; and governments having insufficient funds to undertake effective, regular monitoring. See, e.g., J. Prager,
Contracting Out: Theory and Policy, 25 N.Y.U. J. Int’l L. & Pol. 73, 1099-1100 n. 71 (1992). Indeed, the high cost of monitoring
may offset many of the financial advantages of contracting out.
118 See Section 3.A.1,supra.

Appendix 3

Private Prisons

25

Note that other problems can develop with independent contractors or monitors: a public
correctional agency can be unnecessarily exposed to additional liability if an independent monitor
does not agree to indemnify the agency for any and all acts and omissions and if he or she fails to
carry adequate professional liability and comprehensive general liability insurance naming the agency
as an insured.119 In sum, monitoring acts as a double-edged sword: it can create as well as forestall
liability depending on how intelligently it is structured and conducted.

119 The problems of liability and conflict of interest are graphically illustrated by Wisconsin’s having several years ago appointed an
independent prison monitor to oversee a private facility in Texas that housed Wisconsin inmates. The monitor’s salary was shared
by Wisconsin and the private contractor, making the monitors allegiances and responsibilities unnecessarily confused.

Appendix 3

Private Prisons

26

4.

Employment and Labor Relations Issues

Some of the most important legal questions surrounding prison privatization relate to
employment and labor relations. Specifically, can prison employees engage in collective bargaining
under the National Labor Relations Act? Can they go on strike? Unionized employees in public
correctional facilities have traditionally worried about governments contracting out jobs to private
contractors who may not be required to recognize or negotiate with their collective bargaining
representatives. For their part, private correctional management companies have sometimes resisted
taking over a once-publicly run facility if in doing so they may be deemed a ‘successor employer’
bound to the terms of an existing labor contract. These and other questions bearing centrally on labor
relations will be examined in this section, along with a discussion of how public and private
correctional authorities can practically address prison employees’ right to strike. First, however, the
section discusses the impact of privatization on existing civil service arrangements, as well as the
significance of private prison companies remaining independent contractors rather than “joint
employers” with public correctional authorities.

A.

Potential Civil Service Impediments

State courts differ about whether civil service regimes -- constitutional or statutory schemes
designed to eliminate political spoils systems and increase governmental efficiency through the
creation of a merit-based employment system -- can prevent or facilitate the privatization of
government services.120 Several state courts have held that civil service laws in fact neither bar nor
impede privatization.121 These courts have concluded that privatization is not contrary to the purpose
of their civil service laws: privatization is deemed a cost savings measure -- not a reintroduction of the
political spoils system.122 However, at least two prominent state court decisions have held that state
civil service laws bar privatization.123 Both courts found that the states lacked statutory or regulatory
authorization to contract with private service providers and held that privatization is inconsistent with
the purpose of the civil service laws (i.e., to provide full civil service protections to everyone
performing a traditional state function).124
Other state courts have not found civil service laws to bar privatization, but have held that
such laws impose significant restrictions on contracting out state-provided services, and impede

120 See Nancy Buonanno Grennan, Note, A Legal Roadmap to Privatizing Government Services in Washington State, 72 Wash. L. Rev.
153, 157-59 (1997).
121 See Moore v. Department of Transp. and Pub. Facilities, 875 P.2d 765, 771 (Alaska 1994); In re Local 195, Int’l Fed’n of Prof’l
and Technical Employees, 443 A.2d 187, 193 (N.J. 1982); Conn. State Employees Assn. v. Board of Trustees of the Univ. of Conn.,
345 A.2d 36, 40 (Conn. 1974); Kiger v. Nixon, 1996 WL 512031, at *7 (Tenn. Ct. 1996); Mich. State Employees Assn. v. Civil Serv.
Comm’n, 367 N.W.2d 850, 852 (Mich. Ct. App. 1985) (per curiam).
122 See id.
123 See Colo. Assn. of Pub. Employees v. Dept. of Highways, 809 P.2d 988, 995 (Colo. 1991) (en banc); Wash. Fed’n of State
Employees v. Spokane Community College, 585 P.2d 474, 476 (Wash. 1978) (en banc).
124 See id.

Appendix 3

Private Prisons

27

privatization.125 For example, courts have held that a state can privatize a traditional government
function if a public entity controls the private contractor, or the privatization is not a subterfuge to
evade civil service laws or an arbitrary or capricious reorganization.126 Ultimately, whether civil
service laws impede or bar privatization seems to depend not on specific constitutional or statutory
language but rather on a court’s policy choice between purported cost savings and civil service
protections for everyone performing a traditional government service. In any case private contractors
will need, in practical terms, to devise attractive inducements to persuade civil servants to move over
to the private sector.127

B.

The Significance of Private Correctional Firms’ Independent
Contractor Status

Once initial impediments to contracting out correctional services have been addressed, it
remains to analyze a private contractor’s responsibilities relative to employment and labor relations
obligations. The contractor’s status as an independent contractor bears significantly on such
questions.
As an independent contractor, a prison management company would retain full authority and
responsibility for establishing, maintaining, and/or modifying the wages, hours, and working
conditions of its employees, including a presumptive right to select and hire its own complement of
employees. A different result would obtain if public correctional authorities retained significant
control over basic employee relations. In that case, the corrections firm and public authorities would
be considered “joint employers” with important ramifications for employment and labor relations
obligations. In brief, as a joint employer, a firm does not have unfettered discretion over a wide range
of personnel issues. Many observers of privatization believe that many of the benefits of privatization
may be lost if government retains significant control over a private firm’s personnel issues.
From the standpoint of employment law, a private corrections firm with independent
contractor status would have to comply with all legal standards that apply to private industry,
including the Fair Labor Standards Act (FLSA),128 the Service Contract Act if applicable (providing
for the payment of prevailing wages by government contractors with service contracts with the federal
government or the District of Columbia),129 and a host of other applicable employment laws, including

125 See State ex rel. Sigall v. Aetna Cleaning Contractors, Inc., 345 N.E.2d 61, 65 (Ohio 1976) (per curiam); University of Nev. v. State
Employees Assn., Inc., 520 P.2d 602, 606-07 (Nev. 1974); Ball v. Board of Trustees of State Colleges, 248 A.2d 650, 654 (Md.
1968); Crowin v. Farrell, 100 N.E.2d 135, 139 (N.Y. 1951).
126 See id.
127 These could include special retirement buyouts and other incentive packages. In some cases, contractors have offered special
annuity plans for senior public officials nearing retirement, while ensuring that other employees can, at a minimum, roll over their
retirement benefits to the maximum extent possible.
128 29 U.S.C. §§ 201-219. The FLSA requires employers to pay their employees a minimum hourly wage and overtime compensation at
the rate of one and one-half times their regular hourly wage for work in excess of 40 hours per work week.
129 41 U.S.C. §§ 351-358. The Act requires contractors to pay their employees wages and benefits that either meet or exceed the levels
that are "prevailing .... in the locality" as determined by the Secretary of Labor, or that are determined by a collective bargaining
agreement. 41 U.S.C. §§ 351(a)(1), (2).

Appendix 3

Private Prisons

28

laws on discrimination and possible affirmative action requirements. The contractor would also be
responsible for meeting any additional personnel or employment requirements that the contracting
agency may have included in an RFP. These could embrace a number of fundamental standards
promulgated by the American Correctional Association. In general, however, these ACA Standards
for Adult Correctional Facilities (hereinafter “ACA standards”) tend to be generic, requiring the
contractor to have policies regarding various personnel issues, but not addressing their content.130 It is
likely that a privatizing public corrections agency would seek to develop tailored contractual
requirements for personnel issues at a particular facility, including specific guidelines to ensure that
the contractor does not achieve cost savings through unreasonable personnel practices.
Genuine independent contractor status also has important implications for a private
corrections management firm’s labor relations obligations. So long as a firm is not a “joint
employer,” it will qualify as an “employer” subject to the National Labor Relations Act (NLRA) and
therefore as an entity with whom employees can potentially collectively bargain and hold the right to
strike.131 Indeed, since the landmark decision of the National Labor Relations Board (NLRB) in
Management Training Corp. and Teamsters, Chauffeurs, Warehousemen, and Helpers of America,
Local 222,132 -- a decision explicitly supportive of flexibility in privatization scenarios -- virtually any
government contractor can so qualify as an “employer,” even if a contracting government agency
retains some control over wages, benefits and other essential personnel matters.133 The determination
is essentially up to the contracting parties. As such, the private contractor’s employees will
presumptively not be considered “public employees” subject to special public labor regulatory
authorities and potentially barred from striking under federal or state law.134 Rather, they will

130 American Correctional Association Standards for Adult Correctional Institutions (3d ed. 1993 & 1998 Supp.) ACA standards
require such basic policies as confidential personnel records, annual performance reviews, appropriate ethics rules, and a personnel
policy manual covering all major aspects of personnel organization, qualifications, hiring, training, evaluation, promotion,
termination, and dispute resolution. See ACA Standards 2-4054-4078. Only wages are addressed in terms of substance:
"Compensation and benefit levels for all institution personnel are comparable to similar occupational groups in the state or region."
ACA Standard 2-4064.
131 The NLRA has no jurisdiction over public entities. 29 U.S.C. § 182.
132 317 N.L.R.B. 1355 (1995)
133 Until this case, the NLRB had long held, pursuant to the case of Res-Care, Inc., 280 NLRB 670 (1986), that a firm would not be
subject to the NLRA if an "exempt entity" -- like a public agency -- had significant control over, and final approval of, wages and
benefits. With Management Training Corp., the NLRB decided that the prior test was unduly restrictive of possible privatization
arrangements, and it would allow private contractors and their employees to determine for themselves whether collective bargaining
could address any number of terms and conditions of employment, including wages. It specifically noted that government approval
of wage packages left considerable room for private employer flexibility, including possible scenarios where a private employer
might absorb certain costs itself without passing them on to the government. This more flexible, open standard has since been
repeatedly upheld. See, e.g., Teledyne Econ. Dev. v. NLRB, 108 F.3d 56 (4th Cir. 1997); Pikeville United Methodist Hospital of
Kentucky, Inc. v. United Steel Workers of America, 109 F.3d 1146 (6th Cir. 1997); Landscape & Property Management, Inc. and
Service Employees Int’l Union, 324 NLRB No. 44 (1997).
134 Unlike private employees, public employees do not ordinarily have a right to strike. Federal employees are completely barred from
striking under 5 U.S.C. § 7311, and are subject to criminal penalties for violation of the statute. See 18 U.S.C. § 1918. Also, federal
employees may only join "labor organizations" that are defined to expressly exclude organizations that participate in strikes against
a government agency. 5 U.S.C. § 7103. State employees, meanwhile, are usually subject to the same kinds of provisions, although
as many as eleven states have now created specific exceptions to the general rule that allow certain classes of public employees to
strike. See J. Fisher, Reinventing a Livelihood: How United States Labor Laws, Labor-Management Cooperation initiatives, and
Privatization Influence Public Sector Labor Markets, 34 Harv. J. Legis. 557 (1997). Still, virtually all states have retained clear

Appendix 3

Private Prisons

29

constitute private employees who are subject to the broad exclusive jurisdiction of the NLRA and able
to exercise the full panoply of rights under the Act. The major implications of NLRA coverage,
including employees’ right to strike, are discussed below.

C.

Effects of Privatizing an Existing Unionized Facility

If a private prison contractor assumes control of an entirely new facility, it will be subject to
the NLRA and will not be permitted to interfere with any concerted activities taken by its employees.
Even if a private firm takes over an existing public facility, however, it is unlikely that the employer
will have to collectively bargain with an existing union or be bound by an existing collective
bargaining agreement.
Under the NLRA, if a private firm is awarded a public contract, substantially maintains the
operations of the predecessor public institution and hires a majority of its correctional officer work
complement from the unionized workforce, the contractor can be considered a “successor employer”
with an obligation to bargain with the current union representative.135 This requirement can exist even
where the original bargaining obligation may have arisen under a state or other public employees
labor regulatory authority, rather than the NLRB.136 A firm’s duty to bargain arises as soon as a
“substantial and representative complement” of its workforce is in place.137
Despite this general presumption, no bargaining obligation should generally arise in any
private company takeover of a public correctional facility. The reason is that in almost all prison
privatization scenarios, the incumbent public employees’ union will not be able to represent prison
guards in a private employment setting. Under the NLRA, “no labor organization shall be certified as
the representative of employees in a bargaining unit of guards if such organization admits to
membership....employees other than guards.”138 In other words, only special guards unions -- such as
the United Plant Guard Workers of America -- can represent prison guards under the NLRA.139 In a
prison privatization context this means that a union with a more diverse membership that previously

prohibitions against strikes by employees whose functions are essential to the public welfare, a category that would include prison
employees. Id. at 572-73. Only Ohio appears to permit "non-safety" public employees to strike. See Ohio Rev. Code Ann. §
4117.01-4117.23.
135 See NLRB v. Burns International Security Services, Inc., 472 U.S. 272 (1972). See also Fall River Dyeing & Finishing Corp. v.
NLRB, 482 U.S. 27 (1987).
136 See Base Services, Inc., 296 NLRB 172, 174-75 (1989) (NLRB not precluded from finding successorship merely because
predecessor employer not covered by the NLRA); Harbert International Services, 299 NLRB 472, 476 (1990). Still, there is some
question about this requirement if, as in the case of most public labor agreements, there is no successorship clause.
137 See Fall River Dyeing & Finishing Corp., supra note 135, at 51-53. The duty to bargain would arise upon the private contractor
taking over operations if it was "perfectly clear," even before the contractor, as successor, hired its full complement of employees,
that it was going to employ a majority of the predecessor’s employees. See Kingswood Services, 302 NLRB 247 (1991).
138 NLRA § 9(b)(3), 29 U.S.C. § 159(b)(3).
139 Corrections officers have been deemed "guards" for purposes of the NLRA. See Crossroads Community Correctional Center, 308
NLRB 558 (1992). The reason for this provision is to prevent conflicts of interest from developing in situations involving, among
other things, employee theft or deceit, where guards could find themselves as witnesses testifying against fellow union members.

Appendix 3

Private Prisons

30

represented public correctional officers -- i.e. a union representing many kinds of employees, such as
the American Federation of State, County, and Municipal Employees (AFSCME) -- cannot be
certified as a bargaining representative under the NLRA when a private contractor assumes
operational control of a correctional facility. An obligation to bargain collectively with the workforce
would only materialize upon an appropriate new guards union being selected by the employees and
certified under the NLRA.
Even in the extremely rare case that guards at a public facility are already represented by a
special guards union, a private contractor will not necessarily be bound by the substantive provisions
of the existing labor contract. This is true even if a prison management firm states that it may make
employment opportunities available to all existing correctional officers at a facility. So long as the
firm expressly reserves the right to establish initial terms and conditions of employment as well as
initial job assignments, it will not be required to adopt the terms of the current labor agreement.140
Indeed, only if the company were to indicate explicitly and affirmatively that it would hire most of the
correctional officer workforce at a facility -- or alternatively, would explicitly consent to be bound by
the existing guards’ collective bargaining agreement -- would an extant labor contract be imposed on
the firm.141 The private contractor will have to determine what kinds of statements and actions it will
take toward the hiring of existing correctional officers, and carefully articulate whether it will be
establishing new terms and conditions of employment.142 Under these circumstances, it will be quite
rare that a new prison contractor taking over existing facilities will have an obligation either to
bargain collectively, or abide by the terms of the existing collective bargaining agreement.143

D.

The Right to Strike

Any employees covered by the NLRA possess the right to strike. Though the right is not
absolute -- there exists no constitutional right to strike for public or private employees144 -- the
extension of NLRA coverage to a wide variety of possible private contractors in the wake of the
Management Training decision means that the right to strike becomes a presumptive right of prison
guards at privatized correctional facilities. In fact, the NLRB has upheld the right of private prison

140 See, e.g., United States Can Co., 304 NLRB 1127 (1992).
141 See NLRB v. Burns Int’l Security Services, supra note 135. Although a successor employer may be found to "constructively adopt"
a predecessor contract through particular actions or conduct, merely retaining a particular provision of the predecessor contract that
was required by law does not constitute "adoption." See, e.g., EG & G Florida, Inc., 279 NLRB 444, 453 (1986) (paying prevailing
wages as required by statute and provided for in predecessor contract does not constitute "adoption" of such contract).
142 Whether these legal doctrines applicable to successorship situations represents good public policy or not is open to debate by labor
and corrections specialists, but its practical implications appear quite clear. Nevertheless many observers may be displeased by the
degree to which successor employers can keep existing unionized employees in a state of turmoil and limbo as new operations and
policies are put into place prior to a substantial and representative complement of employees being deployed. Moreover, more
generally, as with the debate about out-sourcing correctional management in the first place, see section IV.A, supra, many may view
the avoidance of a civil service regime as a sure way to eliminate the training, professionalism, and esprit de corps that may
characterize certain public corrections departments.
143 It will however, need to permit public union access to the facility if there are public, unionized employees operating a prison
industry on-site.
144 See United Federation of Postal Clerks v. Blount, 325 F. Supp. 879 (D.D.C. 1971).

Appendix 3

Private Prisons

31

guards to strike on at least one occasion in the case of U.S. Corrections Corp. and United Plant Guard
Workers of America.145 There, the Board found that guards were entitled to go on strike at a minimum
security institution without perimeter fences or guard towers, and simply with unarmed guards on
duty.
An important question remaining in the wake of the U.S. Corrections Corp. case is whether
the NLRB would decline to exercise jurisdiction -- and by extension, permit the right to strike -- at a
more secure facility. This question remains unanswered, but significantly, the Board has assumed
jurisdiction and permitted strikes at other types of facilities that involved public safety concerns. For
example, in NLRB v. E.C. Atkins & Co., the Board exercised jurisdiction over private guards at a
defense plant who served as civilian auxiliaries to the plant’s military police. In Career Systems
Development Corp. and Correctional Medical Systems, the Board similarly acknowledged guards’
right to strike in situations involving a juvenile correctional facility and security services provided to
the United States Coast Guard and U.S. Navy, respectively.146 While the Board or a court might
conceivably determine in a given case involving a medium or maximum security facility that matters
of public security warranted a decision to decline jurisdiction, this seems unlikely to evolve as a
general principle. If correctional authorities seek to limit the right to strike by guards, they will do
better to turn to two practical mechanisms discussed below -- no strike clauses in labor contracts, and
mandatory notification and strike delays accompanied by binding arbitration.
1.

No-Strike Clauses

Even if private prison employees enjoy the right to strike, a private prison contractor can seek
to have the employees agree, individually, or collectively (in a labor contract) to the incorporation of a
no-strike pledge. Unions often agree to the inclusion of such clauses in labor contracts in return for a
management agreement to seek peaceful settlement of contract disputes through an agreed-upon
grievance procedure extending to third-party arbitration if necessary.147 As one court has noted, “a nostrike obligation, express or implied, has been seen as the quid pro quo for a grievance arbitration
agreement.”148 Consequently, an employer can seek to enjoin any strike in violation of a no-strike
pledge.149
2.

Strike Notification/Mandatory Delays

Either due to the expiration of a labor contract or the refusal of a union to agree to a no-strike
clause, private correctional employers may also seek to negotiate other special provisions to blunt the

145 304 NLRB 934 (1991)
146 See Champlain Security Services, 243 NLRB 755 (1979); Career Systems Development Corp., 301 NLRB 59 (1991).
147 The BOP’s Taft RFP provides that any collective bargaining agreement entered into during the period of the contract "should
provide that grievances and disputes involving the interpretation or application of the agreement will be settled without resorting to
strike, lockout or other interruption of normal operations." Taft RFP at 14.
148 AFSCME v. Executive Dept., 52 Ore. App. 457, 479, 628 P.2d 1228, 1242 (1980), citing Textile Workers Union v. Lincoln Mills,
353 U.S. 448, 455 (1957); Boys Markets, Inc. v. Retail Clerks Local 770, 398 U.S. 235 (1970).
149 See Boys Markets, Inc., supra.

Appendix 3

Private Prisons

32

impact of a potential strike on prison operations.150 First, in the case of contract expiration, the
parties may agree to use their best efforts to reach an early and peaceful settlement to any labor
dispute that might arise in the interim. Whether or not a contract had expired, the parties might also
agree to a notice provision that would potentially allow the private contractor to make arrangements
for an impending dispute, including arrangements for notify in a public corrections authorities of the
dispute and enabling the latter to prepare to assume possible operation of the facility in the event of an
actual strike.151 Some prison contractors have reportedly also stated that they can enter into special
emergency preparedness agreements with government authorities to permit the National Guard to
intervene in a timely manner. While this scenario is hardly a congenial one -- relocating replacement
guards, supervisors, medical staff, etc. from public corrections facilities to unfamiliar (and sometimes
remote) institutions on relatively short notice152 -- it represents the best that corrections officials can
expect in the absence of a no-strike agreement.

150 See Robbins, The Legal Dimensions of Private Incarceration, supra, at 310-11.
151 See id. Robbins suggests a 60-day notice period in cases where a labor agreement is set to terminate, and immediate notification
upon an employer learning of a potential dispute.
152 This could prove even more difficult if privatization expands and there are fewer available public correctional employees to assume
key positions on short notice.

Appendix 3

Private Prisons

33

5.

Inmate Labor Questions

The question of inmate labor, though it constitutes a fairly straightforward area of the law,
reverberates with the controversial legacy of an earlier era in American history where the shameful
treatment of private prison labor ultimately provided a major impetus for the modern public
corrections movement. Today, the federal government and most states have in place a regulatory
scheme that simultaneously guards against inmate exploitation while ensuring that prison-generated
products and services do not gain an unfair competitive foothold in the marketplace. Against this
backdrop, it does not appear that privatized corrections would raise any significant legal issues.

A.

Constitutional Dimensions

The Thirteenth Amendment to the U.S. Constitution excepts convicted prisoners from its bar
against “involuntary servitude.”153 Courts continue to recognize this “convict labor exception,”
upholding laws mandating prison labor, and finding no constitutionally protected right not to work, or
right to reimbursement.154 Because prison labor has been found under the 13th Amendment to be an
acceptable component of punishment, only pretrial detainees are protected from involuntary labor.155
Meanwhile, the courts have not viewed the Eighth Amendment’s prohibition against “cruel and
unusual punishment” as providing an obstacle to prison labor per se, or a basis for redress of inmate
claims of unfair working conditions.156
While it may be possible to imagine policy grounds on which to mount a Thirteenth
Amendment objection to inmate labor at privatized facilities based on potential financial incentives
for abuse,157 it is not clear whether these arguments would find a receptive audience in the courts.
Even in the event that private prisons were given the freedom to profit from inmate labor -- something
not envisioned at present (see below) -- it is unclear whether the Thirteenth Amendment would
directly speak to the problem at issue.158

153 U.S. CONST., Amend. XIII.
154 See, e.g., Berry v. Bunnell, 39 F.3d 1056 (9th Cir. 1994).
155 See, e.g., Channer v. Hall, 112 F.3d 214 (5th Cir. 1997).
156 See, e.g., Ray v. Mabry, 556 F.2d 881 (8th Cir. 1977); Woodall v. Partilla 581 F. Supp. 1066 (N.D. Ill. 1984).
157 See I. Robbins, The Legal Dimensions of Private Incarceration, supra, at 125.
158 The theory in favor of a Thirteenth Amendment objection would likely be that for-profit facilities would have financial incentives to
exploit inmate workers, and particularly to conceive of ways to keep inmate workers incarcerated. Ultimately, the profit orientation
would be seen to overwhelm the penological justification for incarceration of which prison labor is a subordinate part. Still, it is
unlikely that for-profit prison labor would ever be permitted to become a dominant part of the correctional environment and it is
unclear how useful the Thirteenth Amendment would be as a bar to this phenomenon. It should be remembered that it was labor
movement and business community pressure to eliminate cheap prison labor (i.e., the cold logic of economics), not the Thirteenth
Amendment and human rights concerns, that led to the elimination of private corrections facilities during the early part of this
century.

Appendix 3

Private Prisons

34

B.

Federal prisons and inmate labor

For decades, Congress has delegated the management of federal prison labor to a
governmental corporation known as Federal Prison Industries (“FPI”) or “Unicor.” Unicor’s mandate
is to provide employment for all physically fit inmates, so far as practicable. Currently, Unicor
employs roughly 20% of all federal inmates in some hundred or so individual operations (two or more
may be located at a particular correctional facility). It can, however, only operate under a “state use”
principle, whereby products and services are sold solely to federal departments and agencies and not
to the public in competition with private enterprise.159 Moreover, because they are not working
voluntarily, prisoners are not deemed to be “employees” under the Fair Labor Standards Act, and are
not entitled to receive minimum wage.160
Privatization of federal facilities with Unicor operations should not face major difficulties,
assuming jurisdictional lines are properly drawn. In the context of the federal regulatory framework,
federal prisons are prohibited from contracting out prison labor to “any person or corporation.”161
This provision is interpreted to bar both the leasing of prison labor or the relinquishing of control of
Unicor industries to any private entity. A private federal prison would therefore require either a
‘carve-out’ of Unicor facilities -- which would need to continue to be operated by the federal
government -- or a legislative change to the applicable statutory impediment, 18 U.S.C. § 436. While
the need for continuous, coordinated interaction between federal and private management systems
might otherwise seem to create administrative difficulties that could undercut the purported overall
cost benefits of privatization, in fact such administrative integration has not proven a deterrent.162

C.

State Prisons and Inmate Labor

Each state has different laws governing inmate labor. Some distinguish between county and
state prisoners and specify in detail the kinds of work that they may perform.163 Others are more
vague. To prevent inmate exploitation, as well as to mollify private business and labor interests
concerned about uncompetitive impacts, many states prohibit the use of prisoners in the private
sector.164 In virtually all cases, these statutes were intended to reach situations where prisoners are

159
Moreover, even in the context of sales to the federal government, Unicor must operate its shops in such a way that no private industry shall
be forced to bear an undue burden of competition from the prison-made products and services.
160
See, e.g., Vanskike v. Peters, 974 F. 2d 806, 810 (7th Cir. 1992), cert. denied, 507 U.S. 928 (1993); Nicastro v. Clinton, 882 F. Supp. 1128
(D.D.C. 1995), aff’d 84 F.3d 1446 (D.C. Cir. 1996).
161 18 U.S.C. § 436.
162 A "carve out" of this nature is precisely the arrangement adopted at the Bureau of Prisons Taft facility. See Taft RFP at 4-5.
163 See, e.g., Fla. Stat. § 951.01 (1987) and Fla. Admin. Code § 33-8 (1987) (convicted county prisoners can be employed at a facility,
on public works projects, or on projects that benefit a county or municipality); Fla. Stat. §§ 946.002, 946.40 (1987) and Fla. Admin.
Code. § 33-3.017 (1987) (state Department of Corrections ("DOC") prisoners may be used to perform prison labor, as well as
contracted out to other government agencies, political subdivisions, and nonprofit corporations for use in public works projects, with
compensation to be determined by DOC rules).
164 See, e.g., Fla. Stat. § 951.10 (1997).

Appendix 3

Private Prisons

35

used to produce commercial goods for sale (including instances where prison labor is leased out to
private for-profit businesses ), not where they are tasked with prison maintenance or service work.
Whatever the particular construction of state rules, the federal government has significantly
preempted local legislation in this area by generally restricting the flow of prison products and
services in commerce. Thus, the Sumners-Ashurst Act165 makes it a federal offense to transport
prisoner-made goods in interstate commerce -- thereby restricting the use of prison labor to the state or
non-profit sector. More recently, though, the Prison Industry Enhancement (PIE) Act of 1994
explicitly permits prison-made goods from 50 non-federal pilot programs to be sold in interstate
commerce if the prisoners used at such facilities participate voluntarily, receive prevailing or
minimum wages, and are covered by workers’ compensation.166 It is unclear to what degree
increasing prison privatization and potential use of inmate labor will have on federal and state
regulation. Certainly state correctional authorities and private prison contractors can apply for PIE
Program certification, which, if granted, will require inmates to be adequately remunerated in a PIE
Program, based on specific guidelines.
Meanwhile, even if state laws prohibiting the contracting out of prison labor to ‘private
interests’ are deemed not to address the situation where a private corporation takes over the operation
of an entire correctional facility, concerns will persist about potential exploitation of inmates in
ordinary prison maintenance jobs. Will inmate labor increasingly be used by private for-profit
contractors to handle more aspects of prison operation -- substituting for tasks traditionally performed
by government employees in public correctional settings?
The latter scenario could raise significant questions about private companies’ commitment to
prison welfare rather than their own bottom line. These companies’ profit margins are known to be
quite modest. All of them, particularly those whose stock trades in public markets, confront quarterly
pressures to show improving profits. The temptation to save on certain prison maintenance costs
through targeted use of inmate labor could make it hard to distinguish between normal prison labor
undeserving of special protection, and work specifically helping prison contractors meet their profit
targets. Lines between these should not be too difficult to draw,167 but public policy should ensure
that inmate work responsibilities are non-exploitative and are properly accounted for, and reflected in,
lower per diem rates charged to the government by private contractors.

165 18 U.S.C. § 1761.
166 18 U.S.C. § 1761(c). For a detailed examination of the PIE Program, see Note, Factories with Fences: An Analysis of the Prison
Industry Enhancement Certification Program in Historical Perspective, 33 Am. Crim. L. Review 411 (1996). Note that ordinarily,
states exclude state inmates from minimum wage and workers’ compensation coverage. See, e.g., Fla. Stat. § 946.002(5) (1996).
Kofoed v. Industrial Commission of Utah, 872 P. 2d 484 (Utah 1994) (no workers’ compensation coverage).
167 Contractors may be required to "build up from the bottom" their costs for each maintenance position and justify the use of prison
labor therefore. One easy way to draw lines is use the same inmate staffing plans for private prisons that are used in public facilities.
This creates a level playing field and allows for relatively easy cost comparisons. This is the contractual approach taken by the
Commonwealth of Virginia, see, e.g., CCA-Va. Contract at § 4.27, p. 24 ("contractor will be allowed to use Inmate labor for facility
operations and maintenance to the same extent Inmate labor is utilized in department facilities. However, Contractor shall not
receive a direct financial benefit from the labor of Inmates") (emphasis supplied). Still, if this method is not used, line-drawing may
be complicated by differing notions of what constitutes ‘normal’ or traditional areas of inmate maintenance responsibilities within a
‘traditional’ prison, as well as varied opinions about what qualifies as an acceptable threshold of non-exploitative ‘basic’ work
functions for inmates.

Appendix 3

Private Prisons

36

6.

Other Important Legal Issues

A number of other legal issues have important implications for correctional privatization.
These include public access to private prison records, private contractor access to criminal history
records, the impact of bankruptcy laws on a possible private contractor insolvency, private contractor
use of force, and contractors’ environmental responsibilities and liabilities. These topics receive brief
treatment in turn.

A.

Access to Private Prison Records

As government correctional authorities move to privatize one or more prison facilities,
concerns have emerged that the public may lack access to private prison records under state or federal
freedom of information acts. The shift to private prisons is viewed as potentially cutting off a major
source of information necessary for proper public accountability. While public prison officials have
traditionally sought to limit the types of records obtainable through freedom of information act
requests, a more fundamental impediment to records access due to privatization would both frustrate
the purpose of such acts and undermine public confidence in prison systems that undertake
privatization initiatives. While a review of the impact of privatization on state public records
legislation lies beyond the scope of this report due to the number and variety of such statutes, this
section seeks to examine this general issue by focusing, by way of example, on the federal context,
i.e., the degree of accessibility to federal private prison facility records afforded by the federal
Freedom of Information Act (FOIA).168
Based on accumulated case law interpretation, a number of fundamental records access
questions have reached a certain level of clarity over the past decade. First, not surprisingly, records
relating to privatized facilities that the U.S. Bureau of Prisons itself generates and holds in its
possession would be subject to FOIA disclosure. Private prison operators, as mere federal
contractors, however, would not be directly subject to FOIA, since they would not find themselves
subject to the kind of day-to-day supervision by the BOP necessary to characterize them as part of a
federal “agency.”169 As to private prison records generated specifically for the BOP by a prison
contractor, they would be presumptively accessible under FOIA only if they were actually obtained
and held by the BOP at the time of the FOIA request.170
Even if private prison records are accessible by virtue of their retention by BOP, a key FOIA
exemption might limit disclosure of certain types of information. Exemption Four of the FOIA
protects from disclosure privileged or confidential “financial or commercial information.”171
168 5 U.S.C. § 552.
169 See Forsham v. Harris, 445 U.S. 169, 180 n. 11 (1980) (research funded by the Dept. of Health, Education and Welfare and
supervised by the Food and Drug Administration did not constitute "substantial federal supervision" triggering FOIA requirements);
Public Citizen Health Research Group v. Dept. of Health, Education & Welfare, 668 F.2d 537, 554 (D.C. Cir. 1981) (agency
contracts mandating numerous requirements for privatized medical peer review group does not render private entity an "agency" for
FOIA purposes).
170 See Department of Justice v. Tax Analysts, 492 U.S. 136 (1989).
171 5 U.S.C. § 552(b)(4).

Appendix 3

Private Prisons

37

Confidential commercial or financial information has in turn been interpreted to constitute data “of a
kind that would customarily not be released to the public by the person from whom it was
obtained.”172 Even if such information is voluntarily provided to the BOP by private contractors, it
can be withheld by the BOP for no reason other than its confidential nature.173 Meanwhile, if
transmission of information of this nature is compelled by statute or contract, the BOP could still
prevent disclosure if either the agency would find it hard to obtain the information in the future or
disclosure might cause “substantial harm to the competitive position” of the private contractor in
question.174
How these FOIA doctrines -- and potential state law analogues patterned generally on them -would affect the general flow of private prison information to the public remains somewhat uncertain.
Certainly, if there is to be maximum public scrutiny of private prison operations, the BOP and other
public correctional authorities would want to compel contractually the production of, and retain, the
widest variety of private prison records consistent with physical and information management
constraints. This would result in at least certain key types of information being regularly collected and
retained.175 At the same time, considerable portions of the overall information obtained might end up
being withheld if it were interpreted by the agency to constitute confidential financial or commercial
information. A flexible interpretation of the latter might extend to a variety of supposedly
“commercially relevant” information covering staffing and procurement -- i.e., well beyond such
customarily withheld data such as sensitive cost information contained in proposals submitted in
response to RFPs.
Because public oversight might suffer under the foregoing defects even if public officials
mandated the collection of voluminous private prison records, one commentator has suggested the
adoption of federal legislation to expand disclosure.176 A more practical and congenial option,
employed by the BOP, is to subject a private contractor contractually to FOIA requirements.177 While
this kind of solution addresses the critical problem of public agency retention of the records in
question, the fact is that privatization itself may continue to shield from disclosure many kinds of
privately-generated, “confidential” information that might otherwise be available for public scrutiny if
produced by a public corrections department.

172 Critical Mass Energy Project v. Nuclear Regulatory Commission, 975 F.2d 871 (D.C. Cir. 1992) (en banc), cert. denied, 507 U.S.
984 (1993), quoting National Parks and Conservation Association v. Morton, 498 F.2d 765, 766 (D.C. Cir. 1974).
173 The reasoning is that the flow of information of this kind that is voluntarily provided to an agency might dry up if it were released to
the public, or information might end up being less candid and accurate. See Critical Mass Energy Project, 975 F.2d at 879-80.
174 Id. at 878. This justification is frequently cited by public officials in shielding from disclosure financial and other commercial
information submitted by private contractors in response to an RFP.
175 Even a fairly selective information collection mandate would have to steer a difficult course between information overload on the
one hand and significant blind spots in agency and public oversight on the other. See N. Casarez, Furthering the Accountability
Principle in Privatized Federal Corrections: The Need for Access to Private Prison Records, 28 U. MICH. J. L. Reform 249, 293-96
(1995) (noting the enormous burdens that a comprehensive records delivery program would place on the public agency).
176 See id. at 196-301. Casarez would have federal legislation mandate that all prison records created and maintained by private
companies be subject to FOIA. The processing of FOIA requests could continue through the channels already in place at the public
agency, but the costs of processing the additional requests could be allocated under the privatization contract with the private prison
operator.
177 See Taft RFP at 17-18.

Appendix 3

Private Prisons

38

B.

Private Contractor Access to Criminal History Records

To perform its prison management duties properly, private correctional contractors need
access to criminal history records for two purposes: (1) to classify inmates properly and take other
steps for the proper safety and care of prisoners; and (2) to screen potential private correctional
employees. By implication, both of these needs affect the privacy rights and expectations of inmates
and private citizens seeking employment in the field of law enforcement and corrections. As it turns
out, there exist two related, but distinct legal regimes at the federal level that safeguard these privacy
rights. The federal mechanisms established to permit transfer of selected federal criminal history
records information (“CHRI”) to private firms such as private prison contractors represent a
compromise between full disclosure and heavily redacted or edited data that would hamper the
contractors’ ability to perform their contractual obligations. At the state level, the policies are
governed by state law, which reflects varied approaches too numerous to summarize within the scope
of this report. At least some of the mechanisms at the state level, however, roughly parallel those
employed by federal authorities.
The regime governing the sharing of federal CHRI for law enforcement and corrections
purposes is governed by regulations found at 28 C.F.R. Part 20. These address functions related to the
“administration of criminal justice.” Under 28 C.F.R. § 20.31, the Department of Justice empowers
the Federal Bureau of Investigation to operate a National Crime Information Center (“NCIC”), which
is able to link local, state, and Federal criminal justice agencies for the purpose of exchanging NCICrelated criminal history information. Pursuant to 28 C.F.R. § 534, and 20 C.F.R. § 20.21(c)(5), the
NCIC is able to share full criminal records information with these law enforcement bodies. Those
agencies, however, may only summarize or characterize that information if they transmit it to private
entities like prison management companies for law enforcement or correctional purposes (such as
inmate classification).178 While access to such information has not appeared to be a problem for
private contractors in terms of the quality or promptness of the information provided, private
contractors may ultimately benefit from planned revisions to the federal regulations that acknowledge
the growth of privatization and that will make more CHRI available to government contractors
directly.179
A separate federal regulatory regime governs the use of CHRI for uses unrelated to “the
administration of criminal justice,” which includes employment screening. Since the 1971 decision in
Menard v. Mitchell180 and the 1972 passage of legislation in P.L. 92-544,181 CHRI can only be used for
employment screening and licensing purposes pursuant to state and local statutes or regulations that
are certified by the FBI’s Criminal Justice Information Division, Access Integrity Unit, and that are

178 28 C.F.R. §§ 20.21(b)(3), (c)(3). Transfer of information to private entities such as private prison contractors can only occur
"pursuant to a specific agreement with a criminal justice agency to provide services required for the administration of criminal
justice pursuant to that agreement. The agreement shall specifically authorize access to data, limit the use of data to purposes for
which given, insure the security and confidentiality of the data consistent with these regulations, provide sanctions for violation
thereof." 28 C.F.R. § 20.21(b)(3).
179 Legislation is reportedly being developed to accomplish these objectives.
180 328 F. Supp. 718 (D.D.C. 1971) (disclosure of criminal suspect’s arrest record for employment purposes enjoined).
181 P.L. 92-544 is not codified.

Appendix 3

Private Prisons

39

administered by designated state and local agencies. Generally, disclosure parameters are narrowly
drawn for the specific purposes involved. Here too, private prison contractors appear to have had few
problems in obtaining all relevant CHRI necessary for hiring correctional officers. On the other hand,
reports have surfaced of problems with the speed with which such information is transmitted,
prompting some private contractors to “end run” the regulations by asking public correctional
authorities to make CHRI requests on their behalf.
State regulation of CHRI is too diverse to summarize easily. Suffice it to say that while many
of the fundamental principles utilized in the federal scheme are recapitulated at the state levels. To
take just one example, Florida has a general statute providing that upon the furnishing of identifying
information, “persons in the private sector” may obtain criminal history information -- but not direct
records information -- “upon tender of appropriate fees.”182 For employment screening purposes,
Florida law requires all private correctional officers to submit to fingerprint and background checks,
and all counties and private entities running incarceration facilities are entitled to CHRI provided that
they are not the direct recipient of criminal records.183 Although most states will ensure the
availability of CHRI,184 one alternative to having private contractors seek the information -- with its
possible attendant bureaucratic delays -- is to have government corrections authorities retain some or
all screening duties. Under the BOP Taft RFP, for example, the BOP conducts key criminal
background and records checks. This would operate to maintain consistent screening procedures.185

C.

Bankruptcy

Of the many topics mentioned in connection with correctional privatization, few seem to
generate as much concern and controversy as bankruptcy. Yet, as discussed below, these concerns
appear greatly exaggerated. Since privatization became a reality in the mid-1980s, many observers
have worried that the security-oriented, politically sensitive nature of the corrections business would
make it especially difficult for a government to grapple with a private contractor bankruptcy and
ensure a smooth transition to a successor arrangement. They especially worried about the potential
for a serious interruption in the provision of correctional services pending the engagement of a new or
reorganized provider, as well as the loss of contractual control that government authorities might
suffer in the formal bankruptcy process if it were invoked. These concerns have persisted to some
extent, even though governments can, as discussed below, protect themselves with certain contract
termination clauses and even though private contractor bankruptcies have been very rare.186

182 Fla. Stat. § 943.053(3).
183 Fla. Stat. § 951.063.
184 See generally U.S. Dep’t of Justice, Compendium of State Privacy and Security Legislation: 1997 Overview (1998).
185 See Taft RFP at 11-12. Of course, the government would have to bear the liability for any mistakes omissions in the screening
process. Also, the government would have to determine whether to offset any costs under the contract as a result of undertaking this
work.
186 Bankruptcies have been virtually non-existent except in the case of companies that have concentrated on building prisons on a
speculative basis.

Appendix 3

Private Prisons

40

In fact, the bankruptcy of a private prison contractor should not in most instances pose any
insuperable problems for the responsible public entity, especially if the prison is owned by the
government. Of course, a good public correctional authority that has carefully selected a private
service provider based on a thorough investigation of its capabilities and financial status (and that has
properly monitored its contract with that provider) should not have to face a provider bankruptcy in
the first place, at least not without sufficient advance warning to permit contract termination and other
contractual arrangements to be made. Annual financial investigations provide a baseline for
monitoring, while interim signals, such as complaints by certain vendors about missed payments,
provide additional guidance that may militate in favor of prompt contract termination and selection of
another contractor (or resumption of public control). Public corrections authorities may also insist in
their contract with a private contractor that the latter maintain a certain net worth at all times 187 and
have purchased business interruption insurance that names the relevant public entity as an insured.
As described in greater detail below, even if serious financial problems with a private
contractor were to develop quickly, public authorities that own their facilities and that have contracted
properly should be able to terminate their agreement with a troubled provider quickly before filing of
a bankruptcy petition. It should then be able to settle on a successor arrangement -- either a
resumption of public operation or selection of a new contractor -- and have enough funds on hand
(from monies that would otherwise have been used to pay the defaulting provider) to be able to pay
for the new arrangement and offset any losses incurred in the process. Even in the case of a
contractor-owned facility, a sale of the facility to the government at a low price, but sufficient to pay
off creditors, should be feasible due to the alignment of interests of all relevant parties.
How is the foregoing possible? First, public correctional authorities must draft a contract that
allows for rapid termination in the event of incipient or current financial problems. A general
“termination for convenience” clause can serve this purpose. Assuming proper operation of this kind
of termination clause, most contracts provide for a ninety-day phase-out or transition period, and may
specify special procedures and requirements for the transition.188 Quickly disengaging from a contract
allows the public authority adequate time to resume management of a facility or find another private
operator. Today, the private corrections industry is sufficiently active and aggressive, even if
concentrated, that reasonably good alternative providers can be accessed relatively quickly.189
Perhaps the most vital input to prison management -- trained human resources in the form of
incumbent correctional officers -- live in the surrounding community and will be ready and willing to
continue working for a successor provider.190

187 See, e.g., Contract between Corrections Corp of America and Metropolitan Gov’t of Nashville and Davidson County, May 30, 1997
(hereinafter "CCA-Nashville Contract") at § 7.6.
188 See, e.g., M. Gold, The Privatization of Prisons 379 (citing Puerto Rico prison contracts); see also CCA-Nashville Contract at § 8.9.
189 This might of course, be problematic at any given time in an expanding private prison market with tight labor conditions. Many
observers would further contend not only that the market as a whole is quite concentrated, but that the higher quality-end of the
market is exceedingly concentrated, resulting in less than ideal competitive conditions. Others, however, would point to the youth
of the industry as a reason why competition has been so healthy thus far and signs of true entrenchment have yet to appear.
190 Public authorities can also take special measures, such as payment of overtime to officers or the hiring of additional law enforcement
personnel, to help keep operations intact during a transition period.

Appendix 3

Private Prisons

41

The reason that a pre-petition notice of termination based on insolvency or “convenience”
will be so effective is that it removes from the bankruptcy process what would normally be deemed an
“executory contract”191 otherwise subject to the process’s automatic stay and to control by the
bankruptcy trustee.192 Accordingly, a contract providing for automatic termination with or without
notice to the private contractor, and prior to the contractor filing for bankruptcy, deprives the
contractor of protection under the Bankruptcy Code, since the contract terms are enforceable and
preempt the Code.193 By having a contract clause mandating notice of the filing of a petition in
bankruptcy by the private contractor several days or weeks in advance, public correctional officials
can ensure that they can invoke the termination for convenience provisions of the contract beforehand.
Such a clause becomes critical if a contract monitor is unable to detect incipient financial problems
on the part of the private contractor.
Note that, as alluded to above, a 90-day or so transition period also provides a government
facing an impending private contractor bankruptcy with a significant financial resource with which to
engineer a successful transition and cover any expected or unexpected damages. Because private
contractors typically invoice their services at least a month after they are rendered, and then demand
payment up to 60 days or so later, public authorities can withhold payment and use the sums that
would otherwise be due for the purpose of covering current operations, securing a new contractor, or
offsetting damages caused by the bankrupt contractor. The amounts involved are usually very
significant and place public corrections authorities in an advantageous position relative to other
potential creditors. Meanwhile, a public authority may also benefit from a contractor’s business
interruption insurance, if the latter had purchased such insurance and named the public entity as an
insured. However, the withholding of amounts payable to the contractor may well prove to be the
more significant financial cushion.194

191 Generally, a contract is deemed "executory" if performance remains due on both sides throughout the life of the contract. Prison
management contracts would fit this definition. See C. Holley, Privatization of Corrections: Is the State Out On a Limb When the
Company Goes Bankrupt? 41 Vand. L. Rev. 317, 330 (1988).
192 A termination based "solely" on grounds of the debtor’s insolvency, so long as it is invoked prior to the filing of a bankruptcy
petition, has been repeatedly upheld as legitimate, and outside the reach of the bankruptcy process. See, e.g., In re LJP, Inc., 22
Bankr. 556 (Bankr. S.D. Fla. 1982). Moreover, a termination at will clause may also avoid the effect of Section 365(e)(1) of the
Bankruptcy Code, which generally invalidates bankruptcy termination clauses by providing that an executory contract may not be
modified or terminated after the commencement of an action in bankruptcy based on the insolvency or financial condition of the
debtor. See Ruben, Legislative and Judicial Confusion Concerning Executory Contracts in Bankruptcy, 89 Dick. L. Rev. 1029,
1059 (1985). Even if a termination clause were invoked after a petition for bankruptcy was filed by a private contractor, a
Bankruptcy Court might sanction termination of a prison management contract for reasons involving "other factors" besides the
insolvency itself -- reasons that could rest on a variety of pubic policy grounds. Cf., id. at 1043.
193 One commentator has suggested use of all purpose termination at will clause that allows for a pre-petition notice of termination by a
government entity:
[The state] may terminate this agreement upon thirty days written notice to [the private contractor]. [The state] must inform [the
private contractor] of its intention to file a petition in bankruptcy at least five days prior to filing such a petition. Debtor’s filing
without confirming to this requirement shall be deemed a material, pre-petition incurable breach.
Ruben, supra note 192, at 1058-59.
194 Business interruption insurance may not have been purchased with very large policy limits, either as a business decision of the
private contractor or due to the public authorities’ unwillingness to have a certain level of premium charges passed through to them
as a segment of the overall per diem cost of operating the prison facility.

Appendix 3

Private Prisons

42

Although the case of bankruptcy of a contractor-owned facility is much more serious, even
here the system should work to the ultimate advantage of public authorities. To begin with, public
correctional authorities should be able to terminate the private prison contract and keep it outside of
the bankruptcy process, as described above. Insofar as that process ultimately results in a
reorganization or liquidation of the private contractor, a resolution acceptable to the government is
eminently possible. It should be kept in mind that the interests of public correctional officials and
creditors in this situation are not likely to be in conflict. The former seek control of the facility or
seek to have it taken over by another qualified private contractor; the latter simply want to be paid
debts owed to them. It is highly likely that the debtor in possession (assuming a Chapter 11
reorganization) or the trustee in bankruptcy (in the case of a Chapter 7 liquidation), would agree with
the government and the creditors -- on financial as well as public policy grounds -- that the facility
should be sold to a new management company or to government authorities at a price advantageous to
the public and sufficient to pay off all or most outstanding debts to creditors.
Finally, if for any reason (despite the foregoing) a government did not properly contract to
have a termination at will clause -- or failed to invoke it and became subject to the bankruptcy process
-- the government would still potentially retain some important rights. To begin with, a government
may still have been alert enough to have inserted a non-assignment clause in its contract. This will
prevent a trustee in bankruptcy from assigning a prison management contract to other vendors who
may or may not be qualified to assume that responsibility, notwithstanding the pressures from other
creditors to do so.195 Moreover, a government is not stayed by the bankruptcy process from actions to
enforce its police or regulatory power.196 Consequently, any action brought by public correctional
authorities regarding the custody and control of prisoners would fall squarely within the police power
and should not be stayed by the bankruptcy of the private contractor. This would include legal or other
actions to enjoin continued operation of a prison facility or otherwise to safeguard the safety and care
of inmates.197 Only if a private prison contract were actually assigned or rejected by a trustee -something a trustee would be loath to do on public policy grounds -- would government corrections
authorities be truly legally disadvantaged.198

D.

Use of Force

Legal questions about the use of force often revolve around subtle matters of proper statutory
authorization. The major issue for private prisons is whether the use of force generally, or specific
variants thereof -- including the use of deadly force -- are properly mandated by the relevant law of the
jurisdiction. Without proper enabling legislation or contractual provisions authorizing the use of force

195 See, e.g., In re Pioneer Ford Sales, Inc., 729 F.2d 27, 29 (1st Cir. 1984).
196 See 11 U.S.C. §§ 362(b)(4), (5).
197 Note, however, that a stay would suspend all prison inmate litigation or special claims by the government against the firm. See 11
U.S.C. § 362(a).
198 In the case of a trustee rejecting the contract -- where, for example, the contractor faces extraordinary liabilities for civil actions
under the contract -- the government would become an unsecured creditor with a claim for damages for breach of the contract as if
the breach occurred prior to the filing of the petition. See 11 U.S.C. § 365(g). In this event, the government might not be able to
recoup all of its expenses incurred in resuming operation of the private prison facility.

Appendix 3

Private Prisons

43

by designated private prison officials, it is possible that such personnel and the private contractor
could face criminal and civil liability.
The applicable legal standards for the use of force vary tremendously from place to place.
Some states’ laws may adequately treat the use of force generally, but insufficiently address the use of
deadly force in a variety of situations. Other states, for example, recognize a common law right of any
citizen to use deadly force in certain circumstances, including the apprehension of one previously
arrested for a felony.199 For states that have adopted the Model Penal Code, which permits
correctional officers to use deadly force under certain conditions, correctional firms may be so
authorized if their employees are statutorily brought within the definition of “correctional officers.” 200
A still different approach is taken by some states like Louisiana, which, in its correctional
privatization statute, specifically authorizes private firms to use force and further prescribes standards
and procedures for licensing private correctional employees to use weapons.201
Even if a contractual solution is adopted on a case-by-case basis -- delegating certain use of
force and weapons use powers to private correctional management firms202 -- care must be taken to
explore all of the nuances involved in various scenarios, including escapes and actions taken outside
of a correctional facility. ACA Standards for Adult Correctional Institutions address many
dimensions of the use of force in generic ways,203 but specific policies and procedures are necessary to
comply with state statutes and case law that have evolved under the Eighth Amendment. So-called
rent-a-cop statutes delegating certain use of force powers to private security companies may not be
sufficient to cover many of the circumstances encountered by private contractor guards. Also,
correctional officials must be sensitive to interjurisdictional issues about the use of force, particularly
in the case of escapes where use of certain types of force may not be authorized beyond a facility’s
perimeter. All of this requires government correctional officials and their legal staffs to think through
very carefully all of the possible scenarios requiring different kinds of force, and to ensure that there
is proper authorization for their use.

199 This is ostensibly the case, for example, in Michigan. See, e.g., People v. Whitty, 96 Mich. App. 403, 292 N.W. 2d 214 (1980).
200 See Model Penal Code §§ 3.04-3.07, cited in Wooley, Prisons for Profit: Policy Considerations for Government Officials, 90
Dickinson L. Rev. 307, 323 (1985).
201 See La. Rev. Stat. Ann. § 39:1800.4(D)(5), (6). Note, for example, that the state of Ohio similarly requires a private correctional
officer working at a correctional facility housing out-of-state prisoners in Ohio to carry and use firearms only if the officer is
certified as having satisfactorily completed an approved training program designed to qualify persons for positions as special
policemen, security guards or persons otherwise privately employed in a police capacity. Ohio Rev. Code Ann. § 9.07(E).
202 For example, Virginia’s DOC specifies the use of non-deadly and deadly force by private contractors in its contracts with private
prison management firms. See, e.g., CCA-Va. Contract, supra at § 4.22, pp. 20-21.
203 See, e.g., ACA Standard 2-4206, for example, which simply requires a written policy and procedures to restrict the use of force as a
last resort in cases of justifiable self-defense, protection of others, protection of property, and prevention of escapes. Elaboration is
left to prison officials.

Appendix 3

Private Prisons

44

E.

Environmental Law Concerns

Prison privatization appears to raise no significant environmental issues except whether and
how governments and private contractors allocate responsibilities for environmental review of prison
construction and the risks associated with hazardous waste cleanup. Some guidance exists as to
various allocation arrangements.
At the federal level, Section 102(2)(c) of the National Environmental Policy Act (NEPA)204
requires federal agencies to prepare detailed statements of environmental impact for major federal
actions that could significantly affect the quality of the human environment.205 Federal actions
include projects undertaken directly by an agency, agency support of private projects through
contracts, grants, subsidies, loans, or other forms of funding assistance, agency issuance to a private
project proponent of a lease, permit, license, certificate, or other ‘entitlement for use,’206 or agency
sale of government land to a private project proponent.207 Although a project proponent could be
public or private, the federal agency most involved with a project is solely responsible for the
preparation of an environmental impact statement (EIS).208 In the private prison context, if a
construction project involving the Bureau of Prisons and a private prison contractor triggers the EIS
requirement of NEPA, then the BOP would be responsible for EIS preparation.
Responsibility for environmental review is less certain at the state and local levels. Most
states have state environmental policy acts (SEPAs) modeled after NEPA; however, some SEPAs
contain important differences.209 Unlike NEPA, some SEPAs require private project proponents to
prepare an EIS if the EIS requirement of the SEPA is triggered.210 Additionally, some SEPAs require
different EIS preparations depending on whether a project is considered public or private in nature.
For example, the Washington State Environmental Policy Act (SEPA)211 requires state agencies to
consider alternative project sites only if the project is public; if the project is private, then state
agencies need not consider alternative sites. 212 Under NEPA, federal agencies must always consider
alternatives to a proposed project.213 Under the SEPAs that distinguish between public and private

204 See 42 U.S.C. §§ 6901-6992k.
205 See 42 U.S.C. § 4332(2)(c).
206 See McGregor, Environmental Law and Enforcement 10 (1994).
207 See, e.g., Lockhart v. Kenops, 927 F.2d 1028, 1035, (8th Cir. 1991) ("When the government sells or exchanges federal land, it must
consider the environmental impact of the proposed use of the federal land by the private purchaser . . . .").
208 See Seattle Audubon Soc’y v. Lyons, 871 F. Supp. 1291, 1318 (W.D. Wash. 1994) ("Agencies may not delegate the responsibility of
preparing an EIS to private parties . . . .").
209 See McGregor, supra note 206, at 14.
210 See id.
211 See Wash. Rev. Code. §§ 43.21C.010-43.21C.914.
212 See Wash. Rev. Code § 43.21C.030.
213 See 42 U.S.C. § 4332(2)(c)(iii).

Appendix 3

Private Prisons

45

projects, courts would almost certainly consider private prisons to be public projects because they
perform a governmental function.214
Due to the transfer of land and property rights that may occur between governments and
private prison contractors in a particular construction, sale or lease arrangement, hazardous waste
liability presents some concerns. Under the Comprehensive Environmental Response, Compensation,
and Liability Act (CERCLA or Superfund),215 liability for environmental cleanup costs is strict, joint,
several, retroactive and prospective: any former, current, or future party who generated, transported,
treated, stored, or disposed of hazardous waste is liable for all of the associated cleanup costs without
regard to fault. 216 CERCLA does not bar indemnification agreements; however, a liable party can
transfer only its responsibility for cleanup costs -- not its underlying liability to the Government. 217
Experience has shown that contracting parties need to state indemnification agreements expressly to
minimize the risk of non-indemnification.218 This applies to private prison contracts no less than other
business agreements. Meanwhile, state environmental laws vary, but almost every state has statutes
compelling cleanup of hazardous materials.219 In this arena as well, both government agencies and
private prison contractors could be liable for cleanup costs and might rely upon indemnification
agreements to shift those costs to the other party.

214 Cf. Weyerhaeuser v. Pierce County, 873 P.2d 498, 504-07 (1994) (en banc) (holding that a private contractor’s construction and
operation of a landfill is a public project because the handling and disposal of waste is a governmental function).
215 See 42 U.S.C. §§ 9601-9675.
216 See 42 U.S.C. § 9607.
217 See 42 U.S.C. § 9607(e); Purolator Prod. Corp. v. Allied-Signal, Inc., 772 F. Supp. 124, 128-30 (W.D.N.Y. 1991).
218 See Rhonda K. Hill, Contractual Indemnifications for Environmental Liabilities, 21 Colo. Law. 943, 943-44 (1992).
219 See McGregor, supra note 206, at 40.

Appendix 3

Private Prisons

46

7.

Regulating Interjurisdictional Private Prison
Contracting

Prison privatization, as commonly conceived , involves a government corrections agency
contracting out correctional management and/or construction work to a private contractor that
operates within the agency’s political jurisdiction and is subject to statutory and regulatory authorities
promulgated by that jurisdiction. Thus, for example, the U.S. Bureau of Prisons selects a contractor to
build and operate a new minimum security prison on federal land; a state department of corrections
enters into an agreement with a private corrections company to take over the management of an
existing maximum security facility in one of the state’s metropolitan areas; a county government,
pursuant to state law, authorizes its sheriff to contract out the operation of a county jail. In each case,
there exists a close legal and contractual correspondence between the public agency doing the
privatizing and the private entity responsible for providing the correctional services.
Correctional privatization can, however, involve other arrangements between public and
private correctional specialists that span multiple jurisdictions. These arrangements, which involve
the leasing of bed space by public authorities facing prison overcrowding, can implicate the transfer
of prisoners to private contractors in distant localities as well as the political and legal interplay of
several different jurisdictions that have a stake in such a transfer. Definite legal confusion and
ambiguities can result. Two basic factual patterns characterize this phenomenon of interjurisdictional
contracting out. The first encompasses indirect arrangements where public corrections authorities
send inmates to other jurisdictions via intergovernmental agency agreements (IGA’s) -- and the other
jurisdictions in turn rely on private contractors to run the particular facilities where the inmates are
housed. One example of this arrangement is the State of Montana, which for two years sent prisoners
to Dickens County, Texas through an IGA before terminating the agreement for contract violations. 220
Dickens County had in turn contracted with a private contractor, BRG Operations, a division of the
Bobby Ross Group, to run the prison where the Montana prisoners were sent.
The second arrangement involves the leasing of bed space directly from a private contractor,
but in a foreign jurisdiction. Many of these prisons are operated on a speculative basis, without
having been chartered by their home jurisdictions or having been opened on the basis of a contract
providing for a certain minimum level of occupancy. The motivating force behind the opening of
several such private facilities may have been to reap the supposed economic development benefits of
correctional privatization. A prominent example of this phenomenon is the District of Columbia
Department of Corrections’ contract with Corrections Corporation of America to house medium
security prisoners in a new facility -- the Northeast Ohio Correctional Center (NOCC) -- that was built
on a speculative basis in Youngstown, Ohio.221

220 The IGA is denominated an "Inmate Housing Agreement" ("Dickens County IHA") between the Montana Department of Corrections
and Dickens County, and runs for three years, with an option for annual renewal for a period not to exceed seven years. See Dickens
County IHA at 2.
221 As discussed in more detail below, the City of Youngstown entered into a "Development Agreement" with CCA whereby the city
transferred land to CCA and provided for a three-year tax exemption in return for construction and operation of a private prison
within Youngstown city limits that was expected to provide construction and management jobs for the surrounding area.

Appendix 3

Private Prisons

47

Both types of arrangements have generated considerable controversy. Relatives complain
about family members housed in distant localities. Concerned citizens express their displeasure with
the lack of accountability that attends the transfer of prisoners hundreds or thousands of miles away to
places where no one from their home jurisdiction may be present to monitor the governing contract or
represent or acknowledge their interests and grievances. Perhaps most fundamentally, correctional
specialists criticize the fact that the receiving facility may not have legal authorities sufficient to allow
it to cope with a variety of critical situations, including inmate escapes and prison emergencies. The
law may simply lag behind new correctional realities. This is in fact what happened in the widelypublicized case of two Oregon prisoners who a few years ago escaped from a private prison in Texas,
and yet were unable to be charged with a crime since Texas law at the time did not recognize escape
from a private prison as a felony.222 Lack of legal oversight mechanisms also played a prominent role
in the decision of the Ohio legislature to pass a law earlier this year subjecting the NOCC and any
future private prisons in Ohio housing out-of-state inmates to a variety of additional statutory
requirements.
Across the country, it is these inter-jurisdictional private prison arrangements that have
seemed to generate the worst performance problems, spawned the most critical articles on private
prisons and reflected most poorly on the correctional privatization movement as a whole. While
many of these troubles can be traced to poor or hasty arrangements or the difficulties of ensuring
proper contract monitoring from afar, a surprisingly large number of problems stem from inadequate
legal or contractual frameworks, unsuited to the role of regulating private either contractor conduct
outside a particular jurisdiction or the incarceration of out-of-state inmates within that jurisdiction.
This section examines briefly a few of the major legal problems that affect interjurisdictional
contracting out of prison services, as well as the kinds of legislative and contractual provisions that
public corrections authorities have adopted to minimize these problems.

A.

Interjurisdictional Contracting Out and Legal Problems

Few public corrections authorities view the transfer of prisoners to other jurisdictions as
anything but an unfortunate necessity caused by overcrowding and inadequate or inappropriate
facilities at a particular time. Certain transfers to public prisons have occurred for many years
pursuant to different statutory mechanisms, including the Interstate Corrections Compact.223 Still,

222 See, e.g., Associated Press, Private Prisons Shackle Texas with Confusion; State Laws Haven’t Caught Up with New Phenomenon,
Chicago Tribune, Nov. 7, 1996 at 34. Further reflecting the huge legal loopholes often opened up as a result of cross-jurisdictional
prison arrangements, even the State of Oregon was unable to prosecute the inmates for escape, since Oregon law did not cover
escapes outside of the state. See Prison Brake; Texas Needs More Scrutiny of Private Prisons, Houston Chronicle, Sept. 3, 1996 at
18A.
223 Various states have adopted the Interstate Corrections Compact, which allows the transfer of convicts to penal institutions in other
states that are parties to the compact. See, e.g., 730 Ill. Comp. Stat. 5/3-4-4; 61 Pa. Cons. Stat. § 1601. In addition, there are state
statutes authorizing transfers of state prisoners to federal penitentiaries, see, e.g., Mass. Gen. Laws., ch. 127, § 97A; Conn. Gen.
Stat. § 18-91. Such transfers do not, by themselves, trigger due process requirements, see, e.g., Olim v. Wakinekona, 461 U.S. 238
(1983), or implicate the equal protection strictures of the Constitution. See, e.g., Tucker v. Angelone, 954 F. Supp. 134 (E.D. Va.),
aff’d, 116 F.3d 473 (4th Cir. 1997) (Virginia inmate who lost certain benefits upon transfer to Tennessee facility no denied equal
protection of the laws).

Appendix 3

Private Prisons

48

many jurisdictions have been surprisingly willing to go one step further and enter into arrangements
whereby their inmates are housed in private facilities outside the jurisdiction. In the case of
agreements with other jurisdictions, public contracting through an IGA or similar contract mechanism
may allow correctional authorities in the sending jurisdiction to avoid procurement requirements
mandating competitive bidding for private contracts. There may also be the perception that public
authorities in the foreign jurisdiction have developed some kind of legal framework for private prison
operation on which the sending jurisdiction can rely. Finally, the sending jurisdiction may believe
that the interposition of the receiving jurisdiction between it and the private contractor can act as a
buffer relative to various kinds of inmate litigation.
Actual experience with these arrangements may or may not bear out the foregoing
suppositions. Certainly contracting can be more rapidly effected through a sole-source agreement with
another public entity that may only seek to address the most important legal and inmate care topics. It
is understood that certain details can be left to a separate agreement (which may or may not be
explicitly referenced in the intergovernmental agreement) between the receiving jurisdiction and the
management company actually providing the incarceration services. But that other agreement, and
the surrounding legal framework established by the receiving jurisdiction, may prove deficient in a
number of significant ways, such as the provision of liability insurance and the handling of escapes.
Since even the most egregious, systemic misconduct in the out-of-jurisdiction facility may be beyond
the ken of the sending correctional authorities due to the difficulties of long-distance monitoring and
responding promptly to problems identified by the contractor, the public authorities may find
themselves facing more liability exposure than expected.224 Finally, the sending jurisdiction may
continue to have potential liability for its own negligence in sending inmates to such a facility in the
first place or in overseeing the long-distance arrangement.
The experience of Montana correctional officials in the case of the Dickens County, Texas
private prison at least partially illustrates these problems. Montana ran out of prison space in 1995,
and was forced to look for out-of-state solutions when county jails could no longer accommodate the
overflow of inmates. Because out-of-state public facilities were unavailable at that particular time,
Montana explored private facilities and eventually settled on the Dickens County facility. In June,
1996, Montana entered into a three-year Inmate Housing Agreement with Dickens County to house up
to 400 Montana inmates. It was explicitly understood that actual operation of the prison facility
would be in the hands of BRG Operations, part of the Bobby Ross Group, which had a separate
Management and Maintenance Agreement with Dickens County. The Inmate Housing Agreement
was deemed to be "assigned" or "delegated" to BRG Operations through the Management and
Maintenance Agreement.225 Ultimately, about 260 Montana prisoners were incarcerated in Dickens
County.

224 Even if the sending jurisdiction were in the process of taking steps to remedy problems in the foreign jurisdiction, "deliberate
indifference" could still be demonstrated in a Section 1983 action if tangible signs of remediation were delayed or lacking in the
facility at issue.
225 Inmate Housing Agreement at 17-18.

Appendix 3

Private Prisons

49

Montana terminated the contract well short of its 3-year completion after it found problems
with a wide variety of prison conditions, including inadequate food, medical care, and security and
housing arrangements inconsistent with a long-term, medium-security facility.226 Montana’s
designated prison monitor ended up spending much more time and money than expected addressing
these problems. From the standpoint of liability exposure, it was clear to Montana officials that the
situation in Dickens County could generate a number of legal claims but the state would not be in the
best position to extract prompt compliance from the private prison operator, due both to problems of
distance and lack of a direct contractual relationship.
A number of legal and contractual problems common to foreign jurisdiction contracting are
illustrated by the Dickens County example. The biggest problem is a lack of communications and
reporting requirements. According to Montana officials, the lack of these in the contract meant that it
would often take many days before correctional authorities were alerted to the fact that serious
disturbances had occurred. This problem loomed very large given the absence of an on-site monitor.
Another problem was the failure of the contract to prescribe any level of liability insurance to be
provided by the county or the private contractor acting in its stead, something that is standard in instate private contracts. Also, insofar as the Dickens County contract appeared to allow for Texas
county personnel or private contractor personnel to pick up inmates in Montana for transfer to
Texas,227 it is not clear that Montana law permitted either type of officials to use force while
transporting inmates within the state or attempting to apprehend them following an escape.
Even more fundamental and serious legal questions may attend direct contracting by a
corrections authority with a private contractor in a foreign jurisdiction, particularly if the contractor is
operating a prison on a largely speculative basis. Many of the problems are the same as those noted
above, such as proper authority for use of force and apprehension of escaped inmates. But direct
contracting with a contractor in a foreign jurisdiction highlights even more numerous legal problems
that may bedevil the receiving jurisdiction if local public corrections oversight is inadequate. These
legal problems include potential deficiencies in prison operation and governance, facility standardsetting, state corrections authority oversight, and coordination with local law enforcement.
Many of these problems have emerged into public view recently with the much-publicized
troubles surrounding the NOCC, run by CCA in Youngstown, Ohio. The NOCC was essentially built
on a speculative basis through a "Development Agreement" between CCA and the City of
Youngstown in March, 1996. When CCA later began operations in May, 1997 on the strength of a
five-year, $182 million contract with the District of Columbia to house medium security prisoners, the
City of Youngstown was essentially a bystander; no local government or law enforcement agency was
a party to the contract. This only heightened the legal problems and problems of accountability that

226 Telephone Interview of Tom Rolfs, consultant to Abt Associates, with Janice Wunderwald, Contracts Manager, Montana Dept. of
Corrections, March 27 and 30, 1998; April 15, 1998. It was reported that there were also drug problems at the facility, and a small
riot involving the Montana prisoners and inmates from Hawaii.
227 The overall delegation of prison management under the contract left unclear whether private contractor personnel might be tasked
with pick-up of the Montana prisoners.

Appendix 3

Private Prisons

50

are inherent in this type of three-way arrangement (unlike the case of the Dickens County contract
where at least the public entity had some responsibilities for the activities of the private contractor).
These problems received increasing attention in the NOCC’s first year of operation. The
District began sending prisoners to NOCC as part of its gradual shutdown of the Lorton prison
complex, a facility plagued for years by lax discipline and mismanagement. Over the next nine
months, the prison witnessed a significant inmate disturbance that prompted prison guards to use gas
on inmates, allegations of deficient medical care, and a number of violent assaults on prisoners by
other inmates, including two fatal stabbings. A lawsuit was brought in the fall of 1997 alleging use of
excessive force, failure to protect inmates, and substandard medical care. The suit also questioned the
authority of CCA to operate the prison under Ohio law. After the second fatal inmate stabbing, the
judge in the case ordered a preliminary injunction requiring CCA to reexamine the classification of all
inmates who had been involved in violence at the institution; and halting the further delivery of
prisoners from the District to NOCC pending a satisfactory classification system being put in place.228
The lawsuit and Youngstown community fears of further prison disturbances raised a host of
questions about the legal basis for CCA’s operation of the prison, and about the legal framework
necessary for any state or locality to regulate properly the incarceration of out-of-state prisoners. Like
many states, Ohio had a private prison statute that only envisioned the incarceration of state prisoners
by its own corrections authorities. The lawsuit also drew particular attention to the problem of
localities in home rule states taking it upon themselves to enter into private prison contracts -- usually
for reasons wholly centered on economic development -- with little or no oversight by the state
department of corrections.
Based on community pressure, state legislators highlighted a number of legal gaps and
deficiencies overlooked in the Youngstown community’s haste to reap the benefits of its new prison
facility. These included inadequate standards for: acceptance by private prisons of foreign inmates;
coordination of law enforcement activities in the case of riot, rebellion or escape, and reimbursement
of state and local governments for the costs therefor; qualifications in hiring of private prison guards;
state inspection of private prisons for compliance with applicable standards of care; and release of
foreign inmates in their home jurisdictions. These kinds of legal inadequacies have spurred a number
of states to consider or enact special legislation giving state corrections authorities and local law
enforcement authorities greater regulatory authority over private prisons that may choose to accept
foreign inmates. Two examples of such legislation, which recently passed in Ohio and Texas,
respectively, are discussed below.

228 Washington Post, February 23, 1998 at B1.

Appendix 3

Private Prisons

51

B.

Some Regulatory Solutions

Two states -- Ohio and Texas -- have recently passed legislation that attempts to remedy some
of the more egregious legal deficiencies that private prisons have exposed over the past several years
as some of them incarcerated out-of-state inmates. Both enactments seek to give state correctional
authorities greater overall authority to prescribe minimal operational standards for such facilities and
to eliminate gaps and inconsistencies that emerged between the treatment of public and private
prisons, and between private facilities that do and do not accept foreign prisoners.
The Ohio legislation passed in March 1998229 represents a version of a bill introduced in late
September, 1997 in the wake of the filing of the Youngstown lawsuit.230 The bill enacted into law
generally limits the operation of correctional facilities housing out-of-state prisoners 231 and provides
comprehensive criteria for the establishment and operation in Ohio of such prisons. The most salient
provision of the bill created a new statute, Ohio Stat. §9.07, that specifically requires the direct
involvement of a local public entity and the indirect involvement of the state’s Department of
Rehabilitation and Corrections (DRC) in any private prison arrangement with a foreign jurisdiction.232
Moreover, the local public entity is permitted to enter into a contract with an out-of-state jurisdiction
to house out-of-state prisoners in Ohio only if the locality and the foreign jurisdiction jointly submit to
the DRC a statement that certifies the facility’s intended use, projected prisoner population, custody
level, and if the DRC reviews and comments on the plans for the design or renovation of the facility
concerning its intended use.233 A contract for the housing of out-of-state prisoners must also contain a
provision requiring the sending jurisdiction to assume all responsibilities for housing and
transportation of its prisoners should the facility close or cease operations for any reason.234
The new legislation then mandates a considerable number of provisions for inclusion in any
private prison contract entered into by a local public entity with a private contractor for the operation
of a correctional facility. These requirements, which have also been made applicable to CCA at its
NOCC facility,235 obligate the private contractor to do the following:

229 Ohio House Bill 293, effective March 17, 1998.
230 The original bill, Senate Bill 174, was introduced by Senator Robert Hagan of Youngstown, who emphasized the need for private
prisons housing out-of-state inmates to meet certain minimal state-prescribed standards and to permit adequate access to, and
coordination with, private prisons by local law enforcement authorities in the case of illegal activities or prison disturbances.
231 Among other things, the legislation keeps the state out of the business of accepting out-of-state prisoners in private facilities and
instead limits such arrangements to Ohio localities, subject to the considerable oversight requirements of the new statute. See Ohio
Rev. Code Ann. § 9.06; Ohio State Legislative Service Commission, Final Analysis of Am. Sub. H.B. 293 at 3.
232 The new statute also seeks to create greater accountability by requiring a local entity and private contractor contemplating a prison
contract to hold a public hearing in the locality and by specifically requiring the local public entity to authorize the location and
operation of the facility. Ohio Rev. Code Ann. § 9.07(C)(3).
233 Ohio Rev. Code Ann. § 9.07(C)(1).
234 See id. at § 9.07(C)(2).
235 The legislation requires the parties operating under a "preexisting contract" to reach agreement on inclusion of these provisions no
later than 180 days after the act’s effective date -- in the case of CCA and the City of Youngstown, by mid-September, 1998. Ohio
Rev. Code Ann. § 9.07(I). Reportedly, however, CCA was supportive of the legislation, although it did engage in considerable
lobbying over several specific provisions.

Appendix 3

Private Prisons

52




















seek and obtain ACA accreditation within 2 years after accepting its first out-of-state
prisoner;
report to local law enforcement agencies all criminal offenses or delinquent acts
committed in or in connection with the facility, and report to the DRC all escapes or
disturbances at the facility, as well as “unusual incidents,” according to reporting
rules applicable to state correctional facilities;
enter into a written agreement with the DRC setting forth a plan and procedure to
coordinate law enforcement activities with state and local police; and permit law
enforcement officers to enter into the facility to investigate any criminal offense or
delinquent act allegedly committed there;
reimburse the state or any political subdivision for any costs incurred by state or local
law enforcement in responding to an escape or disturbance at the facility;
maintain a staffing plan adequate to ensure inmate supervision and security
maintenance, as well as other operational needs, including programming and
transportation;
indemnify, hold harmless, and reimburse any legal defense costs of the state or the
local public entity and their employees, officers, and agents in connection with claims
arising from operation of the prison facility;
develop a security classification system appropriate to the institution;
reject any out-of-state inmate for whom it has not received an institutional record and
medical records; or whose institutional record shows a pattern of violence involving
use of a deadly weapon;
cooperate with the state’s Correctional Institution Inspection Committee, whose
jurisdiction is extended to private facilities housing out-of-state inmates;
not employ any person until a criminal records check is performed, and not employ
any person whose records check indicates previous malfeasance;
return out-of-state inmates who have completed their term to the sending jurisdiction
and agree with that jurisdiction to return to that jurisdiction any inmate who commits
a crime while housed in Ohio; and
certify its correctional officers to carry firearms consistent with Ohio law
requirements.236

236 Ohio Rev. Code Ann. §§ 9.06, 9.07. Other important obligations placed by the new law on private prison contractors for the
protection of local public entities and the general public include the following: (1) compliance with all federal, Ohio, and local laws,
rules or regulations, including specifically laws on sanitation, food service, safety, and health; (2) sending of all copies of all

Appendix 3

Private Prisons

53

The legislation also makes it clear that any act or omission considered a criminal offense or delinquent
act or criminal offense if committed at a state correctional institution will be deemed as such if it is
committed at a private correctional facility; and that an escape from a “detention facility is
specifically deemed to include escape from a private prison.
The Ohio legislation governing private contractors housing out-of-state inmates is one of the
most ambitious attempts by public authorities to gain greater control over a phenomenon that appeared
recently to overwhelm state and local authorities with legal and operational confusion. The Ohio act
consciously sought to patch up all manner of gaps and inconsistencies in a statutory framework not
originally intended to cover interjurisdictional contracting for private correctional services. Quite
likely, a number of other jurisdictions will pass similar types of acts in the next few years.
Ohio’s legislation itself followed by a few months Texas’ effort to accomplish some of the
same purposes. Texas, which has the greatest concentration of private prisons in the country, had
been plagued for several years by problems of accountability and a legal framework riddled with gaps
and loopholes concerning the operation of private correctional facilities. The Texas legislation,
effective September 1, 1997, similarly provides for interjurisdictional contracting only if a local
public entity is directly involved and submits to the Texas Commission on Jail Standards information
on custody level capacity and availability as well as a law enforcement coordination plan. The
Commission, in turn, is supposed to inspect the facility and certify to the local public entity that the
facility is appropriate for the intended use.237
Some of the Texas provisions represent noteworthy variations on, or additions to the type of
scheme adopted by Ohio. The Texas act requires any private prison contract involving out-of-state
inmates to be sent to the Commission on Jail Standards, and empowers the Commission to adopt rules
regulating the number of federal and out-of-state prisoners housed in private facilities in the state. The
act also requires all employees at the private facility to maintain certification as required by the Texas
Commission on Law Enforcement Officer Standards and Education.238 From a financial standpoint,
the act allows the Commission to require a sending jurisdiction, rather than merely the private
contractor, to reimburse the state for any cost incurred by a state agency in responding to a riot,
escape, or other emergency situation,239 and also permits the Commission to charge a private
contractor a reasonable fee to cover the costs of regulating a privately-run facility.240 Perhaps most

inspection reports to the DRC’s Director and to the local public entity; (3) provision of appropriate perimeter and internal security
necessary to protect the staff and prisoners at the facility, as well as the general public; (4) agreement with the local public entity for
a conversion plan that will be followed if for any reason the facility is closed or ceases operations; and (5) agreement to include in
its contract with a local public entity provision for fines to be imposed if the private contractor fails to perform its contractual duties.
See id.
237 Texas Senate Bill 367, May 7, 1997, § 9 (a)-(e). The Texas act., like the Ohio act, also requires certain minimal facility standards to
be met, and certain information to be supplied to the facility by the sending jurisdiction before foreign inmates can be accepted
there. See id. at § 9(d).
238 See id. at § 9(f).
239 See id. at § 9(i).
240 See id. at § 9(g).

Appendix 3

Private Prisons

54

significant, the act requires the Commission to review the qualifications of any private contractor
prior to that vendor entering into a contract with a local public entity.241
The unmistakable trend of states and localities extracting greater accountability from private
contractors who seek to house out-of-jurisdiction inmates can be said to mark the beginning of a third
major phase of the prison privatization experience. The first phase was characterized by heated
political debates and tentative state experiments. The second phase featured a number of significant
state and federal initiatives to rely on private contractors at selected facilities, and witnessed the rapid
growth of the private prison industry. The third phase represents a thoughtful and nuanced stocktaking of correctional privatization (both intra- and interstate), with an emphasis on using legal and
other tools to ensure local political control of the contracting process and adequate oversight of
prison operations. As experience grows, so does the sophistication of those responsible for contract
negotiations. Increasingly, a number of experienced lawyers, both public servants and independent
consultants, are providing expert advice to correctional authorities on all manner of contracting issues.
Some are providing model contracts and legislation that are gaining wide notoriety.242 Based on this
advice and the publicity accorded many privatization failures -- including the interjurisdictional
incarceration examples discussed above -- it appears less and less likely that states and localities will
enter into private prison contracts without a significant array of legal protections at their disposal.

241 See id. at § 9(h).
242 To take just one prominent example, Richard Crane, a former public corrections lawyer and former general counsel for CCA who
now does independent consulting on contracting for correctional services, has published a widely circulated model law for
regulating private corrections. In it, he prescribes many of the regulatory tools adopted by the Ohio and Texas legislatures, as
discussed above. In addition, he has some important other recommendations, including requirements for competitive bidding for
private prison contracts, basic requirements for private contractor qualifications, delegating contract monitoring functions to local
authorities when the sending jurisdiction is distant and may be unable to provide sufficiently frequent monitoring by its own
personnel, and state review and approval of all private correctional facility construction, so as to curtail the building of speculative
facilities. See R. Crane, Wanted: A Model Law for Regulating Privatization, Correctional Law Reporter 83 (April/May 1998).

Appendix 3

Private Prisons

55

8.

Legal Dimensions of Contracting

A significant number of legal issues with a direct bearing on contracting approaches and
provisions have already received some discussion in this review. Some of them, relating to reduction
of liability exposure -- including matters of indemnification, insurance, and monitoring -- have already
been discussed in Section 3, above. This section addresses somewhat more general legal concerns
associated with contracting, such as careful drafting, the use of objective standards, contract duration,
termination, other enforcement mechanisms, and dispute resolution. While contracting parties can
easily reduce most policy choices to simple legal provisions, a few of these policy matters have a
determinative impact on the efficacy of private prison contracts as a whole. This section will discuss
just a few of these critical contracting matters.

A.

Regulatory Purposes and Contract Specificity

Some of the most important contracting questions relate to the broad purposes to be served by
a particular prison privatization contract and the degree of specificity believed to be necessary to
effect those purposes. Contracts may vary substantially in their approach and detail depending on the
correctional ideas animating public officials and the kind of regulatory framework and capacity that
exists in the public sector. Is there a dense web of regulatory requirements applicable to private
prisons, along with capable governmental monitoring, inspection and/or licensing bodies? If not,
then a contract may have to import from the outside and treat in detail a number of provisions and
mechanisms that would otherwise be expected to govern, at least in part, a private prison
arrangement243 Are standards for private prisons similar to those guiding public prison operation?
And is a level playing field desired to promote fairness, increase competition, and allow for
meaningful cost comparisons?244 If not -- because in part, a host of innovations are required -- a
contract will need to cover more uncharted territory for the parties than might otherwise be the case.
Finally, how new are the parties to contracting? Depending on their experience and track record,
public and private authorities alike may be more comfortable with more -- or less -- detail in a contract
and definition of terms.
Some of the most critical contents of private prison contracts may be imposed at two distinct
levels. First, legislation can require mandatory contract provisions, and provide for a competitive
bidding process that prevents industry concentration and entrenchment as much as possible. Second,
public corrections officials can use RFPs to dictate both what subjects a proposal must address and
which provisions will form the basis of an eventual contract. Of great practical importance are RFPs
that mandate inclusion of certain provisions in a contract, and that insist on contractors providing
evidence of how they would address certain requirements. This allows for considerable adaptation on

243 For example, statutes or regulations may specify requirements for topics such as the procedures or grounds for termination of a
contract. See, e.g., N.M. Stat. Ann. § 33-3-27(F) (specifying bases for termination of a contract); N.M. Stat. Ann. § 33-3-4
(requiring inspections of private prisons by governing bodies of counties and municipalities).
244 For example, public authorities need to be especially careful not to permit private contractors to make money from inmate phone
service, commissary sales and the like. Income from these should be reflected in lower per diem rates or should be paid into the
general fund.

Appendix 3

Private Prisons

56

the part of both contracting parties. It also permits both sides an opportunity to question the
judgments of each other early on in the contracting process. Contractors can politely suggest possible
changes to the statement of work where demands are viewed as unrealistic. Government officials,
meanwhile, can ask for practical illustrations and plans that reveal strengths and weaknesses well
before they are applied to actual prison operations. Better contracts, speedier contract negotiations,245
and a more candid contractual relationship may result.
Based on the degree to which public officials view privatization as specifically benefitting the
public, officials may insert into a contract more or less specific language concerning expected cost
savings. Cost savings requirements may be phrased in terms of a provider delivering correctional
services at significantly less cost or significantly better services at the same or less cost than that
which public corrections authorities would need to pay if it did the job itself.246
The possibility of intentionally or inadvertently bestowing third-party contract beneficiary
status on inmates -- allowing them to sue for contractual deficiencies alongside possible tort and
Section 1983 actions -- should also receive careful attention from public corrections officials.
Prisoners do not automatically possess such rights. A contract must intend to grant them legally
enforceable rights to sue under the contract, which can occur through conscious adoption of
contractual language or by a court implicitly reading such an intention into the contract and its
surrounding circumstances.247 If government corrections officials wish to confer such rights on
inmates as an means of bolstering the policing (and inevitably raising the cost) of the contract, they
can make an explicit statement to that effect in the contract.248 If they do not wish to encourage
additional litigation,249 even for a laudable monitoring purpose, they can deprive prisoners of third-

245 However, inclusive -- even if not necessarily overly detailed -- RFP’s may or may not establish proper expectations and more
efficient negotiations, depending on the specific circumstances.. If RFPs ask for illustrative approaches to a variety of subjects, they
can indeed result in informative negotiations. If, on the other hand, specific approaches are mandated up front, public corrections
officials may artificially circumscribe the negotiations and deprive themselves of otherwise valuable information that could be
presented.
246 The State of Tennessee has enacted a statutory requirement to this effect. See Tenn. Code Ann. § 41-24-105(c)-(f).
247 See Restatement of Contracts (2d) §§ 302(1)(b), 308. There is some precedent for courts reading an implicit intent to bestow thirdparty beneficiary status on prison inmates. In Owens v. Haas, 601 F. 2d 1242 (2d Cir.), cert. denied, 444 U.S. 980 (1979), wherein
a federal prisoner was beaten and injured while housed in a county jail, a court found such an intent by reference to a statutory
policy by the U.S. Bureau of Prisons to safeguard federal prisoners while in the care of county governments pursuant to an
intergovernmental agreement under 18 U.S.C. § 4002 (providing for "suitable quarters for the safekeeping, care, and subsistence" of
federal prisoners). See also Cherry v. Crow, 845 F. Supp. 1520 (M.D. Fla. 1994) (intent to benefit inmate found in contract between
county and prison health service provider). But see, e.g., Oliver v. Vose, 1991 WL 97453 (D. Mass. 1991) (no third-party
beneficiary rights conferred on federal prisoner in custody of Massachusetts Department of Corrections; contract language merely
embodies constitutional right to be free of "cruel and inhumane treatment," rather than specific guarantee of a particular level of
care).
248 Some commentators have specifically viewed such third-party suits as a valuable, though by no means dominant, component of an
overall contract enforcement scheme. See, e.g., Gentry, Note, The Panopticon Revisited: The Problem of Monitoring Private
Prisons, 96 Yale L. J. 353, 367 (1986).
249 While such litigation will not usually involve expensive relief -- actual contract damages will be small, and most actions will be
seeking to enjoin the private contractor to deliver certain services as promised -- it will create additional transaction costs for the
contractor and government agency, since such suits may be independent of tort and Section 1983 actions, and may require a lesser
showing of harm.

Appendix 3

Private Prisons

57

party beneficiary rights by curtailing certain language in the contract 250 or by explicitly renouncing
any intent to allow third party suits.251 The important thing is for public authorities to choose their
words as carefully as possible to avoid inadvertent ratification or endorsement of these rights.

B.

The Importance of Objective Standards

Substantive standards of inmate care and treatment under applicable constitutional provisions
and state law will apply to private prisons just as they do to public correctional institutions, thanks to
the “state action” doctrine. Still, public correctional authorities have reason to be concerned that
private prison managers, operating outside of direct public scrutiny and the political process, and
animated by a powerful cost-consciousness, will attempt to reduce staffing and resources in such as
way as to spark increased litigation over alleged constitutional harms. This concern remains strong
despite the protection afforded by the contractor’s private liability insurance and the contractor’s
market-driven pressures to provide adequate service. Potential harms resulting from defective due
process mechanisms have already been discussed in the broader context of delegation concerns. Other
harms, resulting from the excessive use of force, poor inmate safety, or inadequate medical care, may
cause unnecessary Eighth Amendment violations to occur. As discussed below, contracting
governments can attempt to reduce the frequency of these problems by adopting contractual standards
that provide a superior foundation for inmate care and treatment.
Like monitoring, having correctional standards in place may act as a powerful means to judge
contractor performance and prevent conditions from developing that could prompt litigation. Some
states, like Florida, have an elaborate statutory framework that prescribes some operational
standards.252 Most relevant standards, however, will derive from contract terms specifically made a
condition of contracting by an RFP. While a detailed discussion of standards lies beyond the scope of
this legal review, it is important for present purposes to acknowledge the importance of standards
generally, and particularly the importance of standards that constitute measurable, objective
management norms.253 These can form the basis of meaningful performance evaluations as well as a
means -- if backed up by fines or other enforcement mechanisms -- of ensuring compliance with
fundamental constitutional obligations.
Most available norms address only the minimum operational requirements. Even the ACA
Standards for Adult Correctional Institutions, many of whose requirements are mandated by some
contracting agencies to be included in private prison contracts, usually provide only general criteria

250 See, e.g., CCA-Va. Contract at § 9.7, p. 53. In this case, officials will also have to be alert to the use of preambular language
appearing specifically to benefit inmates and provide them with a certain standard of care -- arguably a laudable purpose otherwise
consistent with one of the major premises of successful privatization.
251 See, e.g., Oklahoma Dept. of Corrections Contract with Capital Correctional Resources, Inc., Dec. 3, 1996 at 18.
252 See, generally Fla. Stat., Chapter 944. Some other states address individual standards, albeit vaguely, by statute. See, e.g., N.M.
Stat. Ann. § 33-3-6 ("independent contractors shall supply prisoners with food of a good and wholesome quality and sufficient in
quantity for the proper maintenance of life").
253 For example, food requirements can be phrased in terms of caloric intake and types of food rather than generalities like "three
balanced meals per day."

Appendix 3

Private Prisons

58

for the treatment of inmates. The standards are general in nature, are intended for institutions of all
types and varieties, and tend to leave prison administrators considerable discretion in developing
specific policies and procedures. Moreover, only 41 of the code’s nearly 500 provisions are deemed
“mandatory.” However, public correctional authorities frequently specify in a contract that a
contractor shall achieve ACA accreditation within 18 months or 2 years of contract signature, and that
it comply with 100% of the mandatory ACA standards.254 Specific application in a particular facility
and contract requires that the standards be refined and adapted to local needs. At the same time
government may chose not to spell out requirements in detail in the belief that innovation and
alternative approaches would be stifled.
1.

Special Attention to Staffing and Training

Adequate, quality staffing and training constitute two of the most critical contract provisions
public corrections authorities may address in private prison agreements. Both represent likely areas
in which private contractors will seek to reduce expenditures, either through the payment of lower
wages and benefits, the hiring of less experienced personnel, the deployment of innovative staffing
patterns, or the introduction of special technology (e.g., specialized surveillance equipment).
To prevent unnecessary and unwanted risk-taking in this area, public authorities have often
addressed staffing matters in detail. Some observers have suggested including provisions on minimum
required personnel per shift, the number of days per week to be worked by each person, and the
positions that, for reasons of security, must be filled at all times.255 Others believe such provisions are
sometimes difficult to prescribe up front in an RFP (particularly in the case of new prison
construction) and may remove opportunities for creative staffing proposals.256 Accordingly, public
officials have crafted RFPs that define “key personnel”257 but then specifically require illustrative
staffing patterns to be included in private contractors’ proposals, so as to help ensure that contractors
can deliver the kind of adequate staffing sought by the contracting agency.258
Training requirements may loom as large as staffing level requirements in ensuring that
private contractors meet their inmate care responsibilities. Some governments have, like the state of

254 See, e.g., CCA-Va. Contract at p. 11.
255 See Gold, supra at 393. Gold cites Puerto Rico private correctional contracts that he has negotiated in which these and other matters
are addressed through a formula to yield the total number of persons required to fully staff each position in a facility. Such contracts
also define which positions are required to be filled, which can remain vacant for some periods, and which are to be filled on an
independent contract basis.
256 Richard Crane, a veteran negotiator of private prison contracts, believes that governments are wise to give weight to contractors’
competing staffing plans during the proposal evaluation process. Telephone interview with Mr. Crane, June 2, 1998.
257 The BOP’s Taft RFP, for example, sets forth 14 positions deemed "essential personnel" with specific qualifications and background
experience specified. Taft RFP at 15-16.
258 Contracting officials can also incorporate into RFP’s and contracts requirements that private contractors meet applicable ACA
standards on staffing. These address the need for ongoing review of staffing requirements, the development of an appropriate
formula of time and persons necessary to yield a proper staffing pattern, and policies and procedures governing all types of staffing
needs. See ACA Standards for Adult Correctional Institutions 4-4072, 4-4073, 4-4074, and 4-4075. See also, e.g., the states of
Tennessee and New Mexico, which specifically provide that prison employees do not need to have the same training as sate
correctional officers. See Tenn. Code. Ann.. 41-24-111(3); N.M. Stat. Ann. 33-3-28(B).

Appendix 3

Private Prisons

59

Florida, mandated that private operators meet the same training requirements that public authorities
must satisfy. While many contractors and some correctional authorities have wrongly believed that
such a requirement is too inflexible,258 having the same training standards in fact, ensures that proper
safety requirements are met and creates a level playing field between public and private facilities,
building public trust in the latter and allowing for helpful cost comparisons.258 While ACA standards
address a host of training issues,259 public authorities can add specialized training requirements for
particular purposes or types of employees. And while some corrections authorities may require that
the private contractor pay for all necessary training, others like the U.S. Bureau of Prisons, may view
certain types of training as so unique, and uniformity of service so important, that they will supply
certain types of training themselves at no cost to the contractor.260

C.

Effective Contract Performance and Enforcement

Effective contract performance and enforcement encompasses far more than the liability
reduction mechanisms discussed earlier in Section 3 above. A well-designed and well-written private
prison contract obviously looks to superior contractor performance on a variety of fronts, including
cost-effectiveness and delivery of services. It also seeks to avoid liability problems and litigation
altogether by encouraging flexible preventive systems to operate effectively.
1.

Contract Term and Renewal

A good contract seeks to prevent contractor entrenchment and complacency with
respect to the agreement as a whole. Rebidding contracts at regular intervals and having
shorter-term contracts forces an incumbent firm to emphasize good performance and
undertake innovative practices that can keep its costs low and position it advantageously for a
re-bid or contract renewal. A shorter contract term will also, particularly at the onset of
privatization, relieve the parties of having to foresee and write into an initial contract all types
of eventualities. If, however, public authorities set a contract term that is too short, contract
incentives will be lessened and contractors will not have sufficient opportunity to demonstrate
innovations or returns on certain human or capital investments. Contracting agencies’
"availability of funds" restrictions may also prohibit contracting that extends beyond a certain
budget period. Most observers seem to believe that three or four years represents an ideal

258 By the same token, however, if public standards themselves are viewed as lax, others concerned about inmates’ rights may believe
that such training parity is inadequate. See, e.g., Robbins, The Legal Dimensions of the Prison Privatization, supra at 276-77.
Robbins suggests that contractors meet the higher of applicable ACA Adult Correctional Institution standards or locally applicable
state, local, or agency-specific training requirements.
259 See ACA Standards 2-4079 through 2-4101.
260 See Taft RFP at 16-17 (highlighting as many as 19 different types of training to be provided to a contractor on a one-time basis upon
request).

Appendix 3

Private Prisons

60

contracting period.261 Consistent with an overall preference for relatively shorter contracting
terms, renewal periods should also be limited in number and duration, although they may
provide for renegotiation of the contract price.262
2.

Bolstering Good Performance

A variety of contractual mechanisms can support good performance on the part of contractors.
These include promulgation of clear objectives, and where possible, measurable standards, active
monitoring of the contract by government corrections authorities, inmates, and the public,263 and
progressive compliance with certain target indicators or compilations of standards. Requiring a
contractor to comply over time with a collection of ACA standards or obtain ACA accreditation264
represents another approach to encouraging good performance that acknowledges the time necessary
for some contractors to demonstrate adequate or superior service delivery in particular institutions. In
addition to reliance on generic standards such as these, public corrections officials may fashion
performance benchmarks of their own, focusing on indicators like numbers of complaints against
staff, numbers of disruptive or violent incidents among inmates or against the staff, number and nature
of disciplinary actions against inmates and staff, rates at which inmates complete certain programs
successfully, etc. These could conceivably serve as the basis for certain kinds of reward or incentive
payments, although the record of employing such bonus payments is relatively poor, and their use
tends to muddy comparisons with the private sector, which receives no such bonuses. Also, to prevent
complacency, most correctional specialists believe that in a market context, renewal of a contract
should serve as a perfectly adequate incentive or award.265
3.

Contract Enforcement through Termination

Contract termination remains the ultimate lever to ensure contract compliance. Most
contracts recognize the following as possible grounds for termination: nonperformance,
noncompliance with monitoring or auditing findings, failure to meet minimum standards and
conditions of incarceration specified in the contract, replacement of key personnel with unsuitable
successors, serious conflicts of interest,266 the filing of a petition in bankruptcy, reorganization or

261 See, e.g., Robbins, Legal Dimensions of Private Incarceration supra at 186 (three years balances competing interests in efficiency
and practicality); D. Shichor, Punishment For Profit 113 (governments often prefer short contracts up to three years in light of
possible changes in political administration). Tennessee and New Mexico statutorily limit the duration of contracts to three years.
262 Shorter renewal periods will act against unnecessary entrenchment while allowing a successful contracting relationship to continue
without unnecessary disruption or additional costs. Adequate notice, however, should precede a contracting agency’s determination
to either exercise a renewal option or re-bid a contract. See generally CSG Report supra at 98-100 (noting that in no case should
automatic renewals exceed five years).
263 Many observers believe significant public access to, and monitoring of, private prisons can support monitoring efforts by public
officials and inmates. See, e.g., Gentry, The Panopticon Revisited, supra at 364-67; Keating, Monitoring Private Prison
Performance, supra at 151-53.
264 See text at fn. 254, supra. The BOP’s Taft RFP calls for a 24 month period to obtain accreditation.
265 Many believe that causation and measurement problems make reward payments of this nature problematic.
266 Some contracts may include detailed conflict of interest provisions. These provision loom more important today, given controversies
that have erupted in recent years about unseemly relationships developing between private contractor officials and public
authorities, including both legislators and correctional officials.

Appendix 3

Private Prisons

61

liquidation, or unavailability of funds. Virtually all government authorities will also insist on
termination without cause, or “for convenience,”267 as well as specific steps and emergency plans to
be deployed in the event the government reassumes control of a facility or turns operation over to
another party. As is obvious, however, termination represents an exceedingly blunt instrument for
remedying contractual problems that fall short of default or grossly deficient performance. “Even the
opening up to rebidding of an existing contract can prove traumatic if the contract involves heavy
capital commitments, considerable staff, and a significant number of prisoners.”268
4.

Contract Enforcement Through Fines and Penalties

Fines or performance penalties provide contracting agencies with a more flexible tool to spur
contractor compliance with a prison contract. Fines can be set in such a way that they tap marketoriented motivations and provide contractors with a continuous incentive to innovate and improve
conditions over time. Although some commentators enthusiastically endorse penalties for significant
acts or omissions such as deaths in custody, escapes or instances of recidivism, or incidence of
illnesses or injuries, 269 others may question whether these are proper measures for penalties, and
whether more objective criteria related to day-to-day performance might be utilized.
Given the pervasiveness, acceptability, and relative neutrality of the ACA Adult Correctional
Institution Standards, some have built a performance penalty regime around them. For example, one
contract negotiator has recounted the use of monetary payments to the government tied to criteria
such as failure to meet a mandatory or essential ACA standard or failure to fill a particular position.270
While correctional specialists can disagree about reliance on specific ACA or staffing criteria to
generate penalties, incorporating some measures like these into a private prison contract can at least
provide an early ‘tripwire’ for informed discussion by contractor and public corrections personnel
about performance concerns. This represents an enormous advance over adversarial termination

267 Both Arizona and Tennessee have formally enacted termination for convenience statutes. Ariz. Rev. Stat. Ann. § 41. 1609.01(C);
Tenn. Code Ann. 41- 24-104(a)(4). Each of these statutes provides for 90 days notice to the contractor.
268 Keating, Monitoring Private Prison Performance supra, at 142.
269 See, e.g., Gentry, The Panopticon Revisited, supra at 362. Gentry also believes that fines can be set in such a way "as to counter the
perverse incentive for the firm to create demand for its own product [which it could do] by fomenting violence among current
inmates in order to scuttle parole chances, arbitrarily reducing good time, or "dehabilitating prisoners so that they are more prone to
recidivism. Such actions will, either directly or indirectly, increase the firm’s fine costs." Many would question whether fines are
necessary to create the proper incentives for contractors in these particular areas, or would question whether contractors have
sufficient control over certain environmental factors to make such penalties as effective and targeted as are generally presumed.
270 See, e.g., Gold, The Privatization of Prisons, supra at 395-97. In the Puerto Rico contracts that he showcases in this article, the
author states that "monetary payments to the government can be thought of either as performance penalties, liquidated damages, or
payment adjustments for the operator’s failure to provide services or staff or to meet standards being paid for by the government."
Id. at 395. In one contract, Gold alludes to monetary penalties of $750 per day per failure for failure to meet a mandatory ACA
standard. He also states that a failure to obtain ACA accreditation may "incur several such penalties per day since all of the thirtyeight [now 41] Mandatory ACA Standards must be met for accreditation." Finally, Gold cites as a typical penalty for an unfilled
position -- e.g., failure to staff a fixed post -- "the amount of the salary not paid, time a multiplier for fringe benefits, plus a penalty
of 10 percent." Id. at 397. The penalty rises to 20 percent for vacancies in excess of 30 days for a nonprofessional and 45 days for a
professional employee.

Appendix 3

Private Prisons

62

discussions by lawyers, which usually occur long after performance problems have manifested
themselves or shown themselves susceptible to effective solutions.
5.

Dispute Resolution

Given the popularity of alternative dispute resolution (ADR), it is to be expected that
the parties to a correctional privatization contract might contemplate incorporation of certain
ADR mechanisms into such an agreement. While mediation effects no major change in the
rights of either party -- so long as there is some time limit on its use and the parties retain their
rights to utilize the judicial process should mediation fail -- arbitration is another story. Under
arbitration, parties typically agree to submit a dispute to neutral third parties (typically
industry experts) who will enter a decision that, based on the specific contractual terms, will
be either binding or non-binding on the parties. While arbitration is generally seen as less
costly and formal than recourse to judicial mechanisms, its use by public correctional
authorities should be weighed carefully.
One drawback is that arbitration provides for a less formal discovery process, so that a
public correctional agency could find it difficult to obtain promptly all of the materials and
evidence it needs to prove its case. In many cases there is no substitute for prompt and
effective injunctive relief -- something that only a court can render. There is also no substitute
for contract termination in many circumstances. Finally, there are also many situations where
public policy dictates use of an open, public forum to resolve a dispute, with a detailed,
complete elaboration of issues on the record that cannot be obtained through arbitration. In
general, public correctional authorities are likely to want to retain as much flexibility as
possible to avoid ADR and the extended discussions and delays it can engender. Only on
matters involving payment disputes (where industry standards might be important), or
relatively inconsequential disagreements (e.g., involving non-mandatory ACA standards) will
a public correctional agency be likely to consider submitting disputes to arbitration.

Appendix 3

Private Prisons

63

Appendix 4
State and Federal Prisoners in Government and
Privately Operated Facilities, December 31, 1997

Appendix 4

Private Prisons