From the SelectedWorks of Gerald G Gaes

January 2012

The Current Status of Prison Privatization Research
on American Prisons

Contact
Author

Start Your Own
SelectedWorks
Available at: http://works.bepress.com/gerald_gaes/1

Notify Me
of New Work

The Current Status of Prison Privatization Research on American
Prisons

Gerald G. Gaes. Ph. D.
Florida State University

August 2010

January 2010

Page
|1

Introduction
In many ways, any discussion of prison privatization strikes at the heart of the
fundamental goals and purposes of prison and punishment. The discussion elevates such themes
as the role of the private sector in administering punishment, the importance of metrics to
evaluate and compare how well the privately and publicly operated prisons provide services, the
structure and form of oversight and accountability to insure punishment is just and fair, and the
measurement of cost and efficiency. To be sure, many of these issues are crucial even in the
absence of a privatization debate. However, because there are impassioned proponents and
opponents on both sides of the issue, the prison privatization literature has provoked both earnest
debate and fractious polemic. One might expect that the importance of this topic would have
elevated prison privatization research and encouraged the funding of large scale studies. In fact,
there are very few studies comparing privately and publicly operated facilities. Segal and Moore
(2002) identified about 23 U.S. cost comparison studies and fewer quality studies. Many of those
studies were of questionable value. The most recent review, a meta-analysis by Lundahl et. al.
(2009) only identified 12 studies of cost and quality meeting their criteria for sound
methodology.
Even with this limited set of privatization studies, different reviewers have come to
opposing conclusions about what the research literature shows. Of the five reviews I cover, three
conclude there is no difference between the cost and quality of private and public prisons, and
two conclude private prisons are quite a bit cheaper to operate, and produce a similar level of
quality. At the end of this paper, I discuss why the great majority of studies, especially those
assessing quality, have used inadequate methods to contrast the two sectors.
Private Corrections in the United States
The privatization of prison operations is a fairly recent historical phenomenon.
Corrections Corporation of America (CCA) was the first private prison company in the United
States beginning operations in 1983. It has the largest footprint in American prison privatization
with 17,000 employees and 75,000 inmates and detainees. Its web site makes the claim that CCA

January 2010

Page
|2

is the fourth largest correctional system in America, ranking only behind the Federal Bureau of
Prisons, California, and Texas. The GEO Group, originally Wackenhut Corrections Corporation,
is a close second to CCA with 53,000 inmates and detainees under its supervision in the United
States. The closest competitors to CCA and GEO are Cornell Corrections with about 12,800
inmates and detainees and the Management and Training Corporation (MTC) with about 17,000
beds housing prisoners and detainees. CCA, GEO, and Cornell are publicly traded companies
with stock values in the 22 to 25 dollar range. MTC is a later entry into the business and was first
founded as a jobs training organization operating U.S. Department of Labor Job Corps centers as
early as 1966. While there are other firms providing prison management services, these are the
largest and mange a large proportion of American privatized prisons.
In the next section, I review the recent growth in prison privatization. The data are
compiled by the Bureau of Justice Statistics (BJS); however, they do not include aliens who have
been apprehended and detained under the authority of U.S. Immigrations and Customs
Enforcement (ICE). ICE also uses the private providers listed above by directly contracting for
their services or indirectly with agreements with local governments some of which have
contracted for beds with private companies. At the end of fiscal year 2009, there were 5,400
detainees in directly contracted beds. However, there is no systematic way to calculate the
number of detainees held for ICE in local jails which may be operated by private providers.
Prisoners Held in Privately Operated Prisons
There were 128,524 inmates held in private prisons on December 31, 2008. This
represented 8 percent of the 1,610,466 inmates held in all state and federal prisons on that date.
Of the total number of inmates held in privately operated prisons, 33,162 were imprisoned under
federal authority and 95,362 were confined under state authority. The number of prisoners held
in privately operated prisons grew by 42 percent from 2000 to 2008. In 2000, private prisons
held 6.5 percent of all federal and state inmates. However, the private sector’s share of inmates
grew from 6.5 percent in 2000 to 8 percent in 2008. As a proportion of its inmate population,
New Mexico held the most inmates in privately operated prisons, 45.8 percent, followed by
Montana, 36.4 percent, Hawaii, 35.4 percent, Vermont, 34.3 percent, Alaska, 28.9 percent,

January 2010

Page
|3

Mississippi, 24.2 percent, Colorado, 22.7 percent, Oklahoma, 22.1 percent, and Arizona, 21.1
percent. The Federal Bureau of Prisons had the largest number of privately held prisoners,
33,162. This includes 8,644 inmates held in non-secured community halfway houses. States with
large privately held inmate populations included Texas, 20,041, Florida, 9,158, Arizona, 8,369,
Oklahoma, 5,711, Mississippi, 5,497, Colorado, 5,274, Tennessee, 5,155, Georgia, 5,138, and
California, 3,019. From 2000 through 2008, the Federal Bureau of Prisons growth rate in
privately operated beds, 10.5 percent, has been about three times faster than the average state
growth rate of 3.4 percent. These data come from Bureau of Justice Statistics reports and
statistical tables issued in 2009 (Sabol, West, and Cooper, 2009; West and Sabol, 2009). The
table below comes from West and Sabol (2009) and shows the year-to-year growth from
December 31, 2000 through June 30, 2008.
Insert Table 1 about here
The growth in private prison beds is syptomatic of the the growth in the inmate
population over that time span. The private sector’s “market share” has increased from 6.5 to 8
percent; however, the health of the industry has been driven by the tremendous expansion of
prison capacity over the last decade, and almost half of the growth in private prison beds has
been from contracts let by the Federal Bureau of Prisons.

Meta-analyses
There have been a number of reviews of the literature summarizing the studies which
have compared the cost and quality of privately and publicly operated prisons within a
jurisdiction. My colleagues and I reviewed a large part of this literature (Gaes, Camp, and
Saylor, 1998) previously indicating in detail the methodological problems with the majority of
these studies. The studies that compare private and public prisons on cost and quality are listed
in Appendix II of this paper. I include a legend in that Appendix so that the reader can identify
which study is referenced by which review article. Appendix II also contains other articles that
compose the prison privatization literature.

January 2010

Page
|4

In this section, I review two meta-analyses. In the following section, I cover reviews by
other researchers that are often referenced by privatization proponents and opponents. Metaanalysis is a technique that calculates a common metric called an effect size for each study result
so that the analyst can calculate an average effect across all of the studies. Many scholars
consider meta-analysis to be the most systematic technique to review an area of research.
Lundahl et al., (2009) conducted a meta-analysis of 12 studies compiling costs and
quality effect sizes. One of the studies had two different privatization comparisons allowing the
researchers to develop different effect sizes for each comparison. Lundahl et al., selected studies
that met the following criteria: a study had to have an explicit comparison of a private and public
facility; there had to be sufficient information to calculate effect sizes; the study had to be a
primary analysis, not a citation of another study; the prisons had to house adults; the study had to
be retrievable from recognized databases or references contained in the studies that were found;
the studies had to have actual findings and not hypothetical findings or projections of findings.
The results of the meta-analysis indicated that there was a small cost savings associated
with privatization, about 2.2 percent; however, the quality of confinement indicators favored the
public prisons. However, all of the effect sizes were so small that the authors argued that prison
‘‘… privatization provides neither a clear advantage nor disadvantage compared with publicly
managed prisons.(Lundahl et al., p. 392.)’’
Pratt and Maahs (1999) used 24 studies containing 33 evaluations of private and public
adult male prisons to conduct their meta-analysis. To be included in their meta-analysis, a study
had to have enough information to calculate inmate cost per day as well as institutional
characteristics that were used as statistical controls. The controls they used were the number of
inmates in the facility, age of the facility, the prison’s security level, and the year the study was
conducted to control for the discounted value of the cost. Rather than use the cost per day in
dollars, Pratt and Maahs calculated the z-score of the cost. Presumably this was calculated as the
cost per day minus the mean for all 33 values of cost per day, divided by the standard deviation
of these costs per day. This simply transforms the dollar metric into a normal variate. The raw
dollar costs indicated that the public facilities cost $ 41.09 per day and the private facilities
$38.64 per day. When Pratt and Maahs used a regression equation to control for the number of
inmates, security level, and age of the facility, there was no statistical difference between private

January 2010

Page
|5

and public costs. Finally, Pratt and Maahs conducted t-tests of the costs per day between public
and private facilities controlling for the facility security levels. Separate t-tests were run for
minimum, mixed, medium, and high security level institutions contrasting the public and private
prisons. No test reached statistical significance.
Pratt and Maahs concluded from these analyses that management of the facility (public
versus private) has little or no influence on costs. Rather institution size (economies of scale),
security level, and age of the facility are primary determinants.

Two Opposing Reviews of the Literature
There have been other reviews of the literature and in this section I represent two
opposing interpretations of the same set of privatization studies.
Perrone and Pratt (2003) reviewed the literature comparing privately and publicly
operated prisons. They summarized these studies along two dimensions, quality of confinement
and cost. There were nine studies that had evaluated quality of confinement; however, Perrone
and Pratt emphasize that many of the original comparisons did not match the public and private
prisons on such factors as facility size, custody level, and age. Two of the nine studies they
review for quality of operations matched prisons on all three characteristics. Security level was
always matched across the prison comparisons. However, for 5 of the 9 studies, either there was
not enough information in the study to identify prison capacity or one of the sectors (public or
private) had a higher maximum capacity. For six of the nine comparisons, the privately operated
facilities were newer than their public counterparts, or there was no information provided in the
study to determine facility age.
Charles Logan (1990) articulated domains of prison quality that included: safety,
security, order, care, activity, conditions, and management. Using these dimensions, Perrone and
Pratt looked for specific indicators in each study that could be construed as measures
representing these domains. The security domain was represented by escapes; safety by assaults
on staff or inmates; activity by program enrollments and completions; care by a review of health
services; management by employee sick days used, or employee self reports of stress; order by
disciplinary actions and disturbances; and conditions by indicators of a poorly kept prison. Since

January 2010

Page
|6

Perrone and Pratt compiled this quality of confinement dimensions across seven domains and 9
studies, there were 63 possible comparisons. There was no information available for 31 of 63 of
these comparisons. The private facility outperformed the public facility on 17 of the 63
comparisons. The public outperformed the private facilities on 6 of the 63 dimensions. All of the
remaining contrasts were inconclusive. Perrone and Pratt summarize their quality of confinement
review as follows:
Therefore, at this point it is unclear how the private facilities “measure up” in terms of their relative quality
of confinement. To date, the studies are too methodologically diverse (and often too methodologically
weak) to draw any firm conclusions. They typically do not control for confounding factors such as age and
security level, they fail to employ similar methods of data collection, and they do not assess the domains on
equal measures. Such limitations cloud our ability to determine whether private agencies operate their
facilities at a higher quality than the state. (Perrone and Pratt, 2003 p. 311)

Perrone and Pratt reach a similar conclusion when they compare the per diem costs of
publicly and privately operated facilities. Cost comparisons tended to favor the private sector.
However, there were too many differences between characteristics of the private and public
prisons that would affect cost to draw a firm conclusion. Perrone and Pratt refer to the Praat and
Maahs study which used techniques to control for prison size, custody level, and age, finding no
effect of sector. Perrone and Pratt conclude their review with three recommendations to bolster
the empirical analysis of prison privatization. They note that almost all of the studies that have
been conducted compare only a few prisons. They refer to this as the case study method. They
recommend studies of a large set of prisons with sufficient information to characterize the
important dimensions of prison. This would allow analysts to control for variables that could
affect costs and performance while measuring key dimensions of performance. The closest any
study has come to this recommendation is the research by Blumstein and Cohen (2003), and
Blumstein, Cohen, and Seth (2007). Perrone and Pratt want to see more data collected on prison
management practices so that future analyses can inform more precisely the policies and
practices that may determine why one sector outperforms the other. Perrone and Pratt also called
for a national repository of data that could be used to do a thorough analysis. While the Bureau

January 2010

Page
|7

of Justice Statistics collects data with its separate census of prisons and jails, such a data set
would have to be supplemented to do a thorough cost and quality analysis.
Segal and Moore (2002) review essentially the same body of evidence as Perrone and
Pratt. They conclude without qualification that privately operated prisons were from 3.5 to 17
percent less costly. With regard to quality, they are more circumspect and state “…there is clear
and significant evidence that private facilities provide at least the level of service that
government-run facilities do (Segal and Moore, 2002, p. 9.).”
Segal and Moore divide cost studies into three categories: rigorous, peer-reviewed
studies; mostly government studies presumably of lesser quality, and a third, even less credible
group composed of short reports of findings with very weak methods. In the rigorous category,
Segal and Moore have listed most of the studies reviewed by Perrone and Pratt. Perrone and Pratt
list one Arizona study, while Segal and Moore list two; however, neither makes reference to
more recent analysis by Maximus, a professional services firm, which indicated that privatized
beds were 8.5 to 13 percent more expensive than comparable public beds in Arizona (Maximus,
Inc., 2006). Both the Segal and Moore study and Perrone and Pratt study were published prior to
the Maximus analysis. It is curious that Segal and Moore list the OPPAGA study in the peer
reviewed “A” list. OPPAGA (Office of Program Policy Analysis and Government
Accountability) is a government entity in Florida and would not have submitted their analysis for
peer review. However, OPPAGA’s analyses are typically top notch and deserve to be listed
among the more rigorous studies.
Segal and Moore believe the near universality of cheaper private beds is reason enough to
conclude states will reap cost savings. They dismiss the possibility that there are underlying
differences between the publicly and privately operated beds that could have biased the
outcomes. Segal and Moore are less meticulous than Perrone and Pratt in their attempt to
compare study rigor. They argue rather than demonstrate or analyze how studies may or may not
have controlled for differences in the public and private facilities and the composition of their
inmates.
Many of them went to great lengths to compensate for differences between compared facilities and to
develop useful comparison figures. Moreover, there is nothing inherent in the problem of differences

January 2010

Page
|8

between facilities that we would expect to bias results towards lower costs at private facilities. Thus the
extreme one-sidedness of this literature—near universal findings of cost savings from privatization—is on
its own very persuasive. (Segal and Moore, 2002, P. 3)

Missing from the Segal and Moore analysis is the methodological comparison Perrone
and Pratt conducted when they compared studies on the following dimensions: real versus
hypothetical cost models, identical cost calculation methods, calculation of indirect costs,
security level, maximum capacity, programs provided, and age of the facility. A close
examination of Perrone and Pratt’s table 3 shows that private facilities were more likely to be
newer, that indirect costs calculations were often missing from the analyses and that data on
program services was most likely missing, or if it were present the public and private facilities
were not equivalent. As I show later, based on work by Julianne Nelson, these are data that must
be captured and controlled for in the analysis if we are to build a foundation for cost comparisons
and policy recommendations. Unless you can pin down these differences, the policy maker may
actually be buying a more expensive service, or may not have sufficient detail to demand more
savings from the private contractor.
The quality analysis conducted by Segal and Moore suffers from the same lack of rigor. I
am sympathetic to the difficulty in comparing quality within and across state jurisdictions. It
requires a clear set of quality of confinement criteria, common definitions of variables that
measure these indicators, and a model to insure that quality comparisons are made on an equal
playing field. Perrone and Pratt at least tried to pick quality indicators that represented a broad
spectrum of prison conditions.
A paper by Volokh (2002) is also often cited as a review of the literature that finds that
the private sector outperforms the public sector on both cost and quality dimensions. This paper
actually refers to the Segal and Moore review as support for its conclusions. In the section of the
paper that actually reviews studies on cost and quality, Volokh specifically refers to the findings
of only three studies (see Volokh, 2002, section B.2 Evidence from the Studies).
More rigorous reviews of the literature are ambivalent. Direct comparisons of cost and
quality neither favor the public nor the private sector. Using a sound cost methodology, such as
the one I present later in this paper, jurisdictions should be able to compare the relative costs of

January 2010

Page
|9

private and public prisons. Analysis of quality imposes a much higher hurdle. I will discuss this
problem as well in a later section of the paper.

Prison Privatization and Recidivism
Criminal justice scholars who articulate the goals of prison as punishment draw
distinctions among general deterrence, specific deterrence, rehabilitation, just desserts, and
restorative justice. General deterrence theory suggests that punishment should prevent other
citizens from committing a crime. Specific deterrence is the idea that the person who is punished
should be deterred from committing another crime. Rehabilitation advocates argue that
punishment should be combined with training and other protective interventions to reduce the
likelihood the person punished will reoffend. Just desserts advocates argue that the primary
(some advocates imply the only goal) goal of punishment is to promote fair and equitable justice
through proportionate punishment. Finally, those who advance restorative justice as a goal of
criminal justice argue that by having the offender participate in repairing or restoring the victim
of a crime both the offender and the victim will benefit. Clearly the goals of rehabilitation and
specific deterrence suggest that some combination of punishment and program services ought to
reduce the probability of recidivism. Many prison systems adopt this as a goal of their mission
(Gaes et al., 2004).
In a series of studies involving inmates released from public and private prisons in
Florida (Bales, Bedard, Quinn, Ensley, and Holley, 2005; Farabee and Knight 2002; LanzaKaduce, Parker, and Thomas, 1999, 2000), successive analyses, starting with the studies by
Lanza-Kaduce et al., followed by Farabee and Knight, and ending with Bales et al., used larger
samples, better methods, and more objective matching criteria. While the earlier studies found
privatization was associated with lower recidivism rates, the last of this series of studies
conducted by Bales et al., found no difference in the recidivism rates between inmates primarily
housed in privately operated as opposed to publicly operated facilities. The Bales et al analysis
was the most rigorous of all of these studies.

January 2010

Page
| 10

A recent study by Spivak and Sharp (2008) was conducted in Oklahoma. It compared
recidivism rates for inmates housed primarily in the state’s privately operated medium security
prisons versus similar publicly operated medium security prisons. In most prison systems, and
this is true of Florida and Oklahoma, inmates do not spend time exclusively in public as opposed
to private facilities. Because prison transfers are endemic to correctional systems, it is almost
impossible to isolate offenders who have spent their entire incarceration in one type of facility.
Bales et al., and Spivak and Sharp used many different types of definition of exposure to a
privately managed prison experience to try to capture the impact of privatization on recidivism.
Essentially, the authors of these two studies were trying to measure the proportion of time
inmates spent either in a publicly or privately managed prison. For example, in the Spivak and
Sharp paper they used a large number of exposure definitions and conducted separate analysis on
each one of them. Many of these measures were indicator variables. These variables were coded
1 if there was some kind of exposure to medium security, private prison management, and 0
otherwise.
The Spivak and Sharp definitions of privatization exposure included whether an inmate
was released from a privately operated or publicly operated prison; whether an inmate was
released from a private facility, and served at least 6 months in the private facility and less than 6
months in a public facility; whether an inmate served at least 25 percent of their time in a private
prison, or at least 6 months in a private facility and no time served in a public facility; whether
an inmate served at least 50 percent of their time in a private facility and less than 25 percent of
time in a public facility, or at least 12 months in a private facility and less than 6 months in a
public facility; whether an inmate served at least 75 percent of their time in a private facility, or
at least 12 months in a private facility and less than 3 months in a public facility; and whether an
inmate spent any time in a private facility and no time in a public facility. There were also two
definitions of continuous measures of exposure to privatized management. For anyone released
during the period of the study, the proportion of time served in a private medium security prison
was calculated along with the proportion of time spent in a public medium security prison. Some
of these inmates could have 0 percent exposure time to either public or private medium security
prisons. Finally, they calculated the proportion of time served in a privately managed medium

January 2010

Page
| 11

security prison and the proportion of time served in a publicly managed medium security prison
only for those inmates who had spent at least some time in one of these prisons.
Spivak and Sharp presented the results of these eight different definitions of exposure to
privately managed prisons using a host of control variables that have been found to affect
recidivism. In a sense, Spivak and Sharp were trying as best they could to mimic an experimental
design where inmates were randomly assigned to privately and publicly operated prisons. Their
results indicated that recidivism rates were higher for almost every definition of privately
managed exposure. Two of the definitions of privatization exposure had no impact on
recidivism. These were the definitions associated with whether the inmates were released from a
private or public facility. All other definitions showed a higher hazard of recidivism based on a
greater exposure to privately versus publicly managed prisons. Inmates who were exposed to
privately managed prisons in Oklahoma were anywhere from 12.7 percent to 16.7 percent more
likely to recidivate, depending on the indicator variable definitions of exposure. Using the
continuous definitions, for each unit increase in the percentage of time spent in a privately
operated medium security prison, the likelihood of recidivating increased by about 12 to 15
percent depending on which continuous definition of privatization exposure was used.
Spivak and Sharp recognize the major limitation of their study. Absent randomization or
an even stronger quasi-experimental design, recidivism studies to date have relied on regression
procedures to control for differences among inmates assigned to public and private facilities.
Furthermore, as Spivak and Sharp point out, there may be very subtle selection mechanisms that
cause the results. One of the two authors of this paper had been a case manager in Oklahoma and
had observed, indeed had participated in the practice of transferring the most recalcitrant inmates
to the private sector. Unless sufficient controls can capture the difference in behavioral
dispositions of these inmates, the Oklahoma results may represent the possibility that the inmates
that are the most likely to be behavioral problems are transferred to the private sector.
The safest conclusion to make at this point is that there is no winner in the public versus
private management contest using recidivism as the outcome of choice. Until there is a design
that randomly assigns inmates to either a publicly or privately operated prison, or until there is a
strong quasi-experimental design that produces meaningful counterfactual analysis (Morgan and

January 2010

Page
| 12

Winship, 2007; Rhodes, 2009), we should be very cautious in drawing conclusions about the
benefits of one sector over the other.

Relationship Between the State Level of Prison Privatization and State’s Growth Rate in
Per Diem Expenses for Publicly Held Prisoners
As this heading indicates, Blumstein, Cohen, and Seth used data from different sources to
analyze the relationship between whether a state used privately operated prisons and the growth
in the state expenditures for publicly held prisoners. Blumstein et al., argued that competition
between the public and private sector for the same public dollars should introduce cost savings.
They also claim that state correctional agencies will learn to lower their costs by observing how
privately managed prisons within their state use cost cutting measures presumably because they
are profit driven and state agencies are not. They refer to this as an indirect cost reduction
mechanism and distinguish it from a direct cost savings that they assume will occur when a
private firm takes over management of a previously publicly managed facility. They cite Segal
and Moore’s work stating that “In general, empirical evidence supports the proposition that
privatization brings about direct cost savings ( Blumstein et al., 2007, p. 4).” They also cite
Volokh’s work indicating that the private sector’s quality of operations is on par with the public
sector. Even though they exercise selective citation, in a footnote, they do mention that Praat and
Maahs have challenged the strength of the cost evidence, but fail to cite Perrone and Pratt.
Lundahl et. al., was published after Segal, Cohen, and Seth’s paper.
The emphasis of the Blumstein, Cohen and Seth article, however is “…whether or not
the existence of privately-managed prisons in a state provides competitive or other pressures that
help control costs in the public sector (Blumstein, 2007, P. 6).” Despite their bias, Blumstein et
al’s hypothesis clearly deserves a strong empirical test. These researchers creatively cobble
together data on state expenditures for public and private prison operating costs from 1996

January 2010

Page
| 13

through 2004. For most of their analyses, they have about 400 data points representing states
over fiscal years. The dependent variable in their analyses is the annual growth in public prison
operating expenditures over time. By using growth rates in the public sector, they avoid a messy
set of issues that they would have to address were they to use the actual per diem costs for
privately and publicly operated facilities. Any analysis that uses per diem costs would have to
control for facility custody level, age, scale (capacity), over head rate, and a host of other
important variables. Blumstein, Cohen, and Seth analyze the growth rates using random effects
panel models. These models assume that if there are variables that are omitted from the
regression equation that these factors are uncorrelated with the regressors in the model. This is a
very strong assumption, but not atypical of economic analyses.
Blumstein et. al., evaluate the use of private prisons by using an indicator of whether the
state used private prisons in the same year, one year prior to the growth in public expenditures,
or two years prior to the prison public expenditure growth. They reasoned that the public sector
may not realize savings until several years after privatization. As control variables they used the
initial per diem level costs, growth in non-corrections related expenditures, an indicator of
whether the state’s prisons were over capacity, the proportion of a state’s workers that were in
public unions, whether there were court orders on any prison within the state, the region of the
country in which the state was located, and the year of the data point. They found a marginal
effect (p<.087) for the two-year lag indicating that a state that had adopted prison privatization
did not experience a reduction in growth of public expenditures until two years after the initial
endeavor. Since these analysts argue for a competition and learning mechanism, they should also
have included the scale of privatization in their models. These data were readily available and
could be easily constructed using the proportion of prisoners in a state in given years who were
housed in privately operated facilities. This would have been a stronger test of their hypothesis.
Blumstein et. al., also recognize that they may have had a selection bias problem. States
that privatize may be states that are most likely to economize regardless of whether or not they
used private prisons. The two step propensity score analysis Blumstein et. al., performed to
address this issue is not described very well in their paper. They discuss the estimation of the
propensity to participate in privatization using only a few predictor variables, but not how they
use the results of that analysis to re-analyze the growth in expenditures. They report that they

January 2010

Page
| 14

find essentially the same results when controlling for selection bias as they did without those
controls. This could either mean that there was no selection mechanism or that Blumstein failed
to adequately model and control for the selection process. The control for selection artifacts is
the weakest part of their analysis. While there is no statistical rule that assures the analyst has
developed a selection equation that will eliminate selection artifacts, it is not uncommon for
studies to use 20, 40, 50, or more variables that “balance” the treatment and control groups. It is
unlikely that there is a random mechanism which “decides” which states will privatize some of
its correctional operations. Blumstein, Cohen, and Seth reasoned that states most likely to
privatize would be those who were undergoing fiscal pressures and needed to contain costs, were
experiencing prison expansion, and would not get resistance from public labor unions. They used
variables in their selection equation to represent these three conceptual drivers, as well as
variables indicating the region of the country and the budget year of the growth in prison
operating expenses. Scholars have suggested other factors that contribute to the likelihood that
states will choose prison privatization. The private companies may have a lot more success in
lobbying legislatures in certain states, especially where they can establish ties to key leaders in
the legislative bodies. The public sector within the state may have failed in its obligation to
manage its prison operations. Here I am referring more to its operations than its fiscal health.
The only alternative available to the state may be to use a private contractor. Even fiscally
healthy states may have governors who want to control prison costs by having the public sector
compete for the operation of some of the state’s prison beds. The policy issue that hinges on this
potential selection artifact is whether states who were/are motivated to contain costs could do so
without resorting to privatization of some of its beds. If a state can reign in prison costs without
relying on prison privatization, then what we observe in the Blumstein et. al., analysis is simply
states exercising fiscal discipline, a result they may have achieved without prison privatization.
Cost Comparison Methods
The analysis of the cost comparison between public and private prison operations made
by Perrone and Pratt underscores the inherent complexity of this process. When the government
decides to “make or buy” a service, there must be a proper accounting. The goal should be to

January 2010

Page
| 15

provide a level playing field on which the private sector and government are properly compared.
Julianne Nelson (2002) is an economist who has spent her career doing economic analysis of
public policy specializing in prison privatization. She has worked out a set of principles that
should be used by governments when they are evaluating privatization proposals, or assessing
how well a private contractor has performed. These principles not only can assist governments in
conducting a proper cost comparison, they can also be used by reviewers of the cost comparison
literature to rate the quality of studies that have already been conducted. I will briefly review
these principles.

Within Each Policy Scenario Identify Only Those Expenditures That Are Borne by the
Taxpayers

Some analysts have taken pains to examine the cost of each service that prisons provide.
For example, they try to list the medical expenses, money spent on food and clothing, the cost of
transportation, and sundry other services and programs provided by each sector. The point
Nelson is making is that it really doesn’t matter how the contractor spends money as long as
there is an acceptable level of service provided. This turns out to be a contracting issue as well. If
the government sets performance criteria for a service, then the private or public sector can seek
the most efficient way to provide that service, rather than be tied to a set of criteria that may limit
the ability to innovate.

Work to Capture All Expenditures that Do Change – The Concept of Avoidable Costs

When money for a service is paid to a contractor, the cost is avoided by the government.
Unavoidable costs are those that the government must assume for the service provided by the
contractor. If the private sector operates a prison, but the Central Office takes responsibility for
the training of all new hires, then training is an unavoidable cost to the government. To compare
a public sector prison to a private sector prison, the unavoidable cost of training new hires would
have to be subtracted from the government cost structure since it is not an avoided cost assumed
by the contractor. This issue is typically most problematic when analysts try to compare

January 2010

Page
| 16

overhead costs. It is tempting to assume that most overhead costs will be assumed by the
contractor. I used training as an example, because this is an essential core component for which
governments use specialized facilities, paid for out of Central Office funds. These costs are
actually often avoided when a private contractor operates a facility because the private contractor
bears the responsibility and the cost for training its new hires. Thus, typically this cost is
avoided by the government. This problem of overhead costing is so pervasive and difficult, I
show an example later on in this section of the paper where I indicate how this calculation alone
can tip the balance (often erroneously) in favor of one sector over the other.
Another difficulty is that within the government system of finance, avoidable costs are
not always assigned to institutions where the costs are incurred. For example employer
contributions to retirement are direct costs of prison operations and thus avoidable; however,
they are often funded from the Central Office or even a state level budget. Medical costs in many
state prison systems are borne by the public health department and should be considered a direct
cost of a prison operation; however, they are often difficult to pin down. Other such ethereal
costs include unemployment insurance for prison staff and liability insurance.

Insure that Characteristics of Inmate Populations and Facilities Associated with Different Cost
Structures Are Used to Adjust the Level of Avoidable Costs

Perrone and Pratt consider whether cost comparison studies have controlled for
institution capacity (economy of scale adjustment), custody level, and facility age. There are
other inmate characteristics, such as the age distribution and level of health, that are also
associated with avoidable costs. Younger inmates and sicker inmates typically cost more.
Furthermore, some prison designs are much more efficient than others, reducing the level and
cost of staffing the facility. It would be unfair to compare privately and publicly operated prison
if one sector had a design advantage.

Insure the Level of Services and Programs Provided Is Equivalent

January 2010

Page
| 17

Sometimes prisons specialize in inmate programs such as education or vocational
training. Indeed, this was one of the purposes of prison privatization in Florida. It was anticipated
that increases in program services might promote the goals of rehabilitation. Prison programs add
additional costs and must be included in the avoidable cost analysis. Some states do not allow
private contractors to perform specific services such as sentence computation and disciplinary
proceedings. If these are borne by the government, then they become unavoidable costs. To
make a fair comparison they are either added to the cost of the private contractor or subtracted
from the public provider.

Other Possible Adjustments

If institutions that are being compared are in locations with different costs of living, then
adjustment must be made to the calculations. Other complications arise from the appropriate
treatment of property, sales, or income taxes paid by private contractors, as well as profits from
inmate phone calls and commissary accounts. Taxes to the state or local government paid by the
contractor lower the cost of the privately operated facility, since part of the government expense
is returned to the taxpayer. Furthermore, analysts should not forget to include the cost of
contract monitoring as part of the cost of prison privatization.

Overhead Cost Calculations

To show the amount of detail an analyst needs to decompose overhead costs into its
avoidable and unavoidable components, I present a table below taken from Measuring Prison
Performance (Gaes, Camp, Nelson and Saylor, 2004) that was prepared in FY 2000 by the
Oklahoma State Department of Corrections and was available on their web site. For each
Oklahoma Central Office function listed in the left most column, the cost of that operation also
appears. Adjacent to the cost are two columns signifying who assumes the cost. While the
costing terminology is arcane, anytime a “Yes” appears in the column “Allocated to Private
Operations” this means that the state must still pay for that service and it is therefore an
unavoidable cost. For example, all information technology costs of $2,736,643 were borne by the

January 2010

Page
| 18

state in FY 2000. The last row of the table shows the bottom line. Of the entire Central Office
overhead costs of $25, 760,153, about 62 percent, or $16, 098,778 were unavoidable.

Insert Table 2. about here

I use Oklahoma because it is a state with a large percentage of its inmates housed in
privately operated facilities. The larger the scale of privatization, one would expect the larger the
amount of avoidable costs – more costs should be assumed by the contractor. In the Oklahoma
example, 38 percent of overhead costs were avoidable. Unless the effort is made to break down
these overhead costs into avoidable and unavoidable, large errors in the cost comparisons
between private and public sector comparisons can occur. Overhead costs can typically be from
8 to 20 percent of a prison system’s cost. If we assume a system has a 12 percent overhead rate,
it would be unfair to assume that all of those indirect costs would be avoided through prison
privatization. If a state has a privatization effort similar to Oklahoma’s then it will still have to
assume 62 percent of the overhead for its private facilities. Therefore to make an apples-toapples comparison, the comparison would have to involve a 12 percent overhead for publicly
operated facilities. The overhead rate for the privately operated facilities would be 62 percent of
12 percent or 7.4 percent. Alternatively an analyst could assign the 38 percent savings to the
public sector and no overhead costs to the private sector. One typically finds that this issue is
either improperly handled in the literature or it is ignored completely. What is the consequence
of such an error? If an analyst simply assumed that the 12 percent overhead rate were avoided,
then in the Oklahoma case they would underestimate the cost of privatization by 7.4 percent of
the total cost.

Two Cost Comparison Studies in Search of the Same Results

January 2010

Page
| 19

There are a few occasions in the cost literature on prison privatization in which there
have been competing estimates of public and private operational expenditures. This happened in
Florida when there were competing cost estimates by the Florida Department of Corrections, the
private contractor (Corrections Corporation of America) and the Office of Program Policy
Analysis and Government Accountability, the arm of the state legislature which is mandated by
state law to conduct the analysis. All three produced separate and contradictory results regarding
the per capita costs of public and private facilities in the state. The Florida Department of
Corrections showed that the private contractors were more expensive than the public operators.
The private contractor showed that the privately operated facilities were less costly than the
public counterpart. OPPAGGA showed that the costs of the publicly and privately operated
prisons were about the same. As I have already stated, the devil is in the details. C. Elaine
Cummins (2001) has documented a similar tangle of competing interests with different cost
analyses in Texas.
Even when analysts have no incentive to portray one sector as superior to the other, you
can find different conclusions. As part of the Bureau of Prisons (BOP) mandate to conduct a
demonstration project in prison privatization, the agency provided the National Institute of
Justice money to conduct an evaluation. Abt Associates, Inc. won the competition and the BOP
hired Julianne Nelson, who was working for the Center for Naval Analysis at the end of the
study to conduct an independent cost analysis. I refer to the cost results as Abt and CNA
findings. The final reports examine both cost and quality dimensions (Camp and Daggett, 2005;
McDonald and Carlson, 2005; Nelson, 2005). In this section, I focus on the discrepancy in the
cost analysis results. The following table comes from a brief research article I wrote for the NIJ
Journal comparing the ABT and CNA results (Gaes, 2008). In the table below, the privately
operated facility is Taft. The other three facilities were run by the Federal Bureau of Prisons. All
four of these facilities were funded and owned by the government. They were built at about the
same time using the same architectural footprint, and housing inmates of similar custody levels.

Insert table 3 about here

January 2010

Page
| 20

Clearly, the Abt analysis shows a substantial increase in per diem costs among the
publicly operated facilities across all four fiscal years. The CNA analysis presents a much
different picture. The costs are much more similar. In FY 2002, the average cost of the publicly
operated facilities was 2.2 percent higher than the privately operated facility although two of the
public facilities were $.25 per day higher than Taft. The CNA analysis also contained a
hypothetical estimate of the per diem cost of Taft assuming that the BOP had operated that
facility using the same staffing design as it did in the public facilities. In FY 2002, the cost would
have been $37.55, over a dollar less than it cost the private firm to operate Taft. Why are these
costs calculations so different?
The primary reason why the Abt and CNA analyses were different rests on an
interpretation of overhead costs, and on the issue of how to accommodate the different
population sizes of the inmate population each of the facilities had to manage. Taft had a higher
inmate population throughout the study period. Thus, the private provider benefited from
economies of scale that reduced their average costs. The CNA analysis made adjustments to the
expenditures to account for these economies of scale. The Abt report also fixed the overhead rate
at 12 percent, while the CNA report presented an analysis that determined how much of the BOP
overhead could be avoided because the private provider had assumed the functions of the central
and regional office operated by the BOP. This is equivalent to the approach I showed previously
that Oklahoma used to pinpoint the level of overhead that could be avoided through
privatization. Over a five year period, 1998 to 2002, the average amount of avoided overhead
costs was 35 percent of the total overhead costs. By making these adjustments, what appeared to
be a 21 percent savings based on the Abt analysis was reduced, to at most, a 3 percent savings
per year and possibly even a slight loss had not the BOP been mandated by Congress to privatize
the Taft facility. While the Abt study has appeared in some reviews, the CNA study has not been
cited. The public policy question is, “How many decisions to privatize in other jurisdictions rely
on equally, or even more fragile, foundations?”

MacDonaldization of Prison Labor

January 2010

Page
| 21

Labor is by far the largest component of a prison’s operational costs. It is often 60 to 70
percent of the costs of the annual operating budget. If the private sector is to achieve savings and
produce a profit, it is most likely that it will come at the expense of the employee. In many cases,
private firms pay the same wage but have much lower benefits. Previously, we have drawn upon
the organizational sociology literature to describe a theory of job routinization that could explain
how private firms try to reorganize prison tasks to make them simpler. This has the added benefit
of making the prison labor force more fungible (Gaes et al., 2004, pp 111-115), and was based
on work by Ritzer (1993, 1998). In two books he discussed the MacDonaldization of labor that
simplified tasks, reduced training costs, and allowed companies to hire less skilled workers.
Think of MacDonalds restaurants with cash registers that use symbols of the Big Mac and
French fries, hamburgers that are already cut to specifications, pre-defined bins for each type of
sandwich, and sundry other simplifications that make the routine simple and efficient. While it
is difficult to see how many of the more subtle tasks involved in operating a prison can be
reduced to a very fundamental set of routines, the MacDonaldization hypothesis sets a minimal
reference point against which the more skilled practices in prison must be accomplished.
Compare the correctional officer’s routine of checking cells for contraband, monitoring inmate
movement, and insuring doors are locked against the more subtle and perhaps more important
skills of handling inmate grievances, disciplining inmates, as well as helping them through
family and other personal crises. It is easy to see how the former can be re-engineered for low
skilled, lower paid workers. It is more difficult to see how the latter skills can be simplified using
a cookbook approach.
If there is one aspect of prison privatization that needs more research, it is in the area of
privatization and prison employees. What is the impact of lower labor costs on staff turnover or
staff performance? How do private companies develop training for workers with fewer skills?
Do private companies, in fact, hire lower skilled workers? Have private companies re-engineered
the prison employee’s job? Scott Camp and I did some preliminary research on this issue (Camp
and Gaes, 2000). When we conducted this research in 1999, we found that private operators had
much larger staff turnover rates than their public sector counterparts. Some, but not all, of this
finding was explained by the fact that many of the private prisons we surveyed were going
through their startup phase—typically referred to as “prison activation.” With regard to prison

January 2010

Page
| 22

activation, McDonald et. al. (1998) found that when the privately operated federal prison in Taft,
California started their operation, only 10 percent of the workforce had correctional experience.
A typical federal prison activates with 50 percent of their workforce having correctional
experience. Does that practice of activating prisons using staff new to corrections occur in states
that use private prisons? Is it a common practice? Activating an institution having a large
proportion of employees with previous experience was a BOP policy intended to insure that new
prisons would not experience problems with a naïve workforce. Are there startup problems in
privately operated prisons because they use a smaller percentage of staff with prior prison
experience?
Every correctional administrator will say that the key to their success or failure is the
quality of their labor force. I suspect that most of the top echelons of private prison management
still come from retired public sector workers. Yet, little is known about the extent to which the
wardens or other top level management running privately operated firms are retired or former
employees of the state. One could call this a hidden cost of privatization. The years of training
and experience the state had to sink into the development of the workforce later employed by
private companies are an indirect, although small cost of privatization. More importantly, since
the private sector uses many former public sector employees to guide its management and
administration, have these former public employees found new ways to engineer prison
operations? Or is there evidence that private prison quality suffers from lower paid workers?

Prison Privatization, Market Forces, and “Residual Rights of Control”
There are many advocates of prison privatization who argue that market forces will drive
down the costs of prisons by converting public to private beds (Volokh, 2002). However, unless
there is liberalization of the market in which there is open competition, costs will not be reduced
and quality will not necessarily improve (Starr, 1987). There are many examples of converting
public functions to private enterprise resulting in cost overruns, poor performance, and even
complete and abysmal failures (Starr, 1987). Many economists argue that there are much
stronger incentives for private owners to innovate to reduce costs and increase their profits.
Government employees cannot benefit directly from productivity gains. However, Hart, Shleifer,

January 2010

Page
| 23

and Vishny (1997) propose a mechanism that can lead to inferior private performance. The key
to understanding this mechanism is to compare the explicit performance criteria in a contract to
the complexity of the product. If a vendor has been contracted to produce widgets at a certain
level of quality, then the government purchaser can test the quality. When a contract covers a
complex process such as the operation of a prison, there may be many aspects of the service that
may not be explicitly covered as a performance expectation. Therefore, the contractor may
lower costs maintaining “residual rights of control” without the government fully knowing that
some of the services are not being produced.
One of the comparisons that come to mind is school achievement. There are many
jurisdictions which have allowed open competition between public and private schools.
However, I would argue it is quite a bit easier to measure school achievement accounting for
differences in the children’s backgrounds than to measure the productivity levels of prison
services. My colleagues and I have argued that recidivism ought to be the ultimate benchmark
used to measure prison performance (Gaes, Camp, Nelson, and Saylor, 2004); however,
jurisdictions that buy prison services are most concerned about internal performance measures
such as order, health, case management, program services, and safety. To insure that the private
sector cannot exercise residual rights of control, governments contracting for prison services
must include explicit performance criteria for all of the services that are delivered in a prison
setting. In addition to the contract, governments must insure quality services through continuous
monitoring and oversight. The same mechanism that drives innovation in the private sector,
incentives to make a profit, can undermine prison quality if the contractor can exercise residual
rights of control.

Race to the Bottom
I have previously argued (Gaes, 2005) that one of the possible outcomes in jurisdictions
which allow private contractors to exercise residual rights of control is that this puts the
government competitors at an unfair disadvantage. In my 2005 article, I stated the problem this
way.

January 2010

Page
| 24

Public servants negotiate a contract with a vendor and specify at some level what they expect the private
provider to do. If what they require of vendors is somewhat vague, or if vendors can easily manipulate the
proof of the outcomes, then private providers can exercise a great deal of discretion in how they deliver
their services. Vendors that are inefficient or have underbid for a contract can still make a profit by
reducing services. Unless the government has specific performance standards and monitors the contractor’s
performance, service can quickly deteriorate. Public employees in correctional systems who are in
competition with the private providers, in an effort to meet the lower per diem costs vendors are
offering, must cut staff and/or programs to be competitive. This competition produces the perverse effect of
lowering service delivery throughout the system—a downward spiral. To my knowledge, no one has
investigated this possibility. (Gaes, 2005, p. 86-87)

Instead of productivity gains, jurisdictions may be in a race to the bottom. The perverse
nature of this argument is that it is still based on the profit motive, the same motive that
proponents of privatization argue is the compelling feature of prison privatization. For the profit
motive to improve prison services, you need two compelling features of a free marketplace that
may be missing in some jurisdictions – market liberalization and objective performance criteria.
Market liberalization is an open competition that pits private and public providers against
each other. It is the antithesis of an economic environment rife with lobbying and cronyism. As
in England, governments should be allowed to compete against private companies. Contracts for
prison service provision are re-competed with both the government and private sector having an
opportunity to win or lose the next time around.
In a market where consumers can buy competing products, they use objective
performance criteria to choose one product or service over another. Analogously, unless
jurisdictions make an effort to define service levels and hold private and public service providers
to the same standards, they will defeat the goals of privatization. Government purchasers stand
in as the consumers of the product. This kind of indirect consumerism depends on clearly
articulated performance metrics and means of accountability. I cover the performance metrics in
the following section.

Assessing Prison Quality

January 2010

Page
| 25

The goal of Measuring Prison Performance: Government Privatization and
Accountability (Gaes, Camp, Nelson, and Saylor, 2004) was to articulate a method of prison
performance measurement. We tried to show how each prison jurisdiction could start with its
mission and objectives and develop indicia of performance that would allow prison
administrators, external auditors, and oversight committees to monitor the quality of prison
operations as an ongoing activity. In the book, we showed how improvements in computer
systems, and statistical analysis of data could lead to an unbiased, objective assessment of prison
performance over time. These tools could be used to assess public and private prisons either
against each other, or against themselves over time. Almost every prison system uses audit tools
to inspect their prisons. We showed how easy it would be, with slight modifications, to transpose
audits into meaningful prison performance metrics.
I expect as prison systems develop more sophisticated databases, the development of
performance metrics will become easier. This rationalization of the process can be incorporated
into contracts with private providers making explicit the government’s expectations, and
reducing the contractor’s exercise of residual rights of control. Prior reviews of the research
literature, both meta-analyses and what the field calls vote counting methods of summarizing the
literature, have suffered from the lack of methodological sophistication in the primary studies.
This is true of cost and quality analyses. To make a fair comparison, the institutions that are to be
compared must be equated on any dimension that might differ between the two that would affect
performance that is not linked to the privatization effort itself. The steps to conduct these kinds
of analysis are now an integral part of the prison and program evaluation literature. For each
performance measure, the analyst develops a model that includes characteristics of prison and
inmates that might determine differences among the prisons in the target outcome that would
lead to a spurious conclusion about differences among prisons. Using statistical techniques that
have been developed in recent years (see especially Raudenbush and Bryk, 2002), the analyst
compares the expected level of the outcome versus the actual level of the outcome. Then the
analyst can rank order institutions with higher or lower than expected outcomes depending on the
particular performance measure. For example, if an institution has more misconduct than
expected given the statistical model, this prison would be ranked lower than comparable prisons
with lower than expected misconduct outcomes. Privately and publicly operated prisons can then

January 2010

Page
| 26

be compared based on their relative rankings on these performance measures. We show
examples of this type of analysis in Measuring Prison Performance. This allows for a level
playing field when making comparisons among prisons based on a host of variables that could be
different between the two prison sectors – some variables favoring one sector over the other,
other variables having the opposite effect.
All of the examples in the book come from published studies. One of these examples is
based on Camp, Gaes, Klein-Saffran, Daggett, & Saylor (2002) who showed how to use inmate
surveys in assessing the quality of prison operations at Bureau of Prisons (BOP) facilities. The
surveys included measures of inmate security, safety, threats from gangs, sanitation, and food
service quality. On the gang and the safety/security measures, the private facility performed at an
average level in comparison to the BOP prisons. However, the private facility performed poorly
on inmates’ perceptions of sanitation in both the dining hall and housing units. There were also
more serious problems involving food service measures.
In another paper, Camp, Gaes, Langan, & Saylor (2003) showed how to use these
methods to compare public and private prisons on officially recorded indicators of inmate
misconduct. In this study the privately operated facility had higher than expected levels of total
misconduct than the public facilities to which it was compared. It also showed that the private
facility and one of the three comparison public facilities had higher rates of violent misconduct
than the two other public facilities. This analysis also placed these results in the context of all
federal low security prisons. At the time of this study, almost all BOP prison privatization efforts
were focused on low security prisons.
Camp, Gaes, & Saylor (2002) used inmate surveys and found that BOP institutions
performed better on measures of organizational commitment and fire safety in the housing units;
however the private prison had employees with higher levels of commitment to the institution.
Measures of institutional operations indicated that the privately operated and publicly operated
prisons were equivalent.
While these three studies are among the most rigorous ever conducted comparing public
and private prisons, none has appeared in any of the reviews or meta-analyses published to date.
Whether the outcome is per diem cost, level of prison order, or quality of health care, systems
with well designed performance monitoring tools will be able to compare and contrast their

January 2010

Page
| 27

institutions with less equivocation. In the language of meta-analysis, this will produce effect
sizes with greater integrity (internal validity). Comparisons across jurisdictions will still be
difficult because definitions of institutional characteristics and outcomes will vary by
jurisdiction. Nevertheless, the more precision each jurisdiction uses to makes its within-state or
federal comparison, the more precise the cross jurisdiction comparisons will become.
Performance measurement does not insure accountability unless some independent body
monitors the metrics and investigates issues. Richard Harding (1997, 2001) has been one of the
clearest thinkers on this issue and has articulated different mechanisms and oversight structures
that can hold both the public and private sectors accountable. One might expect governments that
run poor quality prisons may not be the best candidates for monitoring and holding accountable
private companies that serve as public proxies. Harding has advocated public-private
competitions, government monitors that are independent of the public agency administering the
prison system, and periodic re-competition so that both sectors must maintain satisfactory
quality. However, there has been little or no empirical research on accountability structures.

Pros and Cons of Prison Privatization
Absent empirical findings, proponents and opponents of prison privatization have carried
on a spirited debate about the impact on cost and quality. McDonald et al. (2003) summarized
many of these arguments into a set of propositions. In addition to their arguments, I add a fair
number from the paper by Volokh (2002). The arguments favoring privatization are:
•

Private firms are not subject to rule-ridden procurement procedures, although they often
participate in cumbersome procurement procedures when they bid on contracts.

•

Without organized labor, private firms manage labor better.

•

Private firms bear risk unlike public management thus increasing cost efficiency.

•

Competition among private firms lowers costs.

•

Incentives for government managed institutions are structured so as to increase rather
than conserve money.

January 2010

•

Page
| 28

Government managers neither benefit nor suffer from their financial decisions reducing
the incentives to save money.

•

Civil service regulations limit managers’ control of government labor.

•

Private prisons are more accountable to governments.

•

Governments may be less deferential to its oversight of private rather than public prisons.

•

Private prisons are responsive to market pressures to win new contracts making them
more accountable.

•

Private companies save money at the operation stage. The main savings come from
reducing labor costs, both through lower wages and through more efficient use of labor.

•

Because they are not bound by civil service rules in managing their personnel, private
prisons use roughly one-third the administrative personnel of government prisons, and
use incentives to reduce sick time and consequent overtime expenditures.

•

Private firms are free from many bureaucratic purchasing rules and can often buy
supplies at lower cost than the government.

•

The traditional hostility of juries toward corporate defendants, private prison guards’
inability to claim qualified immunity in 1983 civil rights suits, and courts’ unwillingness
to defer to the judgment of corporations all increase private prisons legal accountability
relative to public prisons. The presence of this additional judicial check should in turn
increase private prison quality.

•

Monitors of private prisons (even if co-opted to some degree) are likely to be more
independent than monitors of government-run prisons.

•

Private prisons are more accountable because their contract can be terminated. A private
prison may have its contract rescinded. The government cannot take over its own prison
except by firing civil servants, and it cannot have a private firm take it over except by
opening a new bidding process, which is more difficult than finding someone to take
over an existing contract.

•

Private prisons are more accountable because of financial concerns. A private
corporation is punished financially for bad news, and possibly for mismanagement that
may impose costs in the future.

January 2010

•

Page
| 29

Private companies are more concerned with keeping their stock prices high over the
long term by insisting on sound management. Guards and wardens can be encouraged to
act responsibly through stock ownership in the company.

Arguments against privatization are:
•

Government contracts actually increase expenditures because of hidden contract costs.

•

Rule-ridden procurement procedures insure a fair and open competition for goods and
services.

•

Private firms cut corners to make a profit resulting in inexperienced staff or poor quality
goods and services.

•

Corruption can increase costs and lower quality.

•

Private firms abandon their contracts if they cannot make a profit, leaving the
government at risk.

•

Private firms are subject to strikes.

•

The initial bids of firms are “lowball” offers and private firms either raise prices later,
once they have a foothold in the market, or they will fail to deliver quality services
because the original negotiated price was too low.

•

Private firms may lobby for preferential treatment.

•

Private firms may influence substantive criminal justice legislation by supporting toughon-crime candidates and advocating tougher sentencing.

•

Corrupt politicians may extract money from a contractor influencing a contract decision.

•

Fear of stock price drops may make private prisons conceal their problems.

Almost all of these arguments contrast a benevolent to a malevolent motive brought on
by the incentive to make money. Private firms driven by the incentive to make money will lower
costs and increase performance. This is a loose interpretation of Adam Smith’s “invisible hand”;
self-interest compels actors to beneficial behavior. However, critics of this argument also point
out that this economic law only works when there is a liberal market with real competition. If
firms can make a profit by providing fewer services it is also in their self-interest to do so.

January 2010

Page
| 30

Because our system of imprisonment is not a market in the conventional sense where the inmate
qua consumer can decide among different public and private prison service options, governments
must promote quality. To do so they must articulate the goals of imprisonment, formulate these
as clear performance objectives, and put in place performance metrics to hold all prison service
providers accountable. Absent these mechanics of accountability, governments will produce
poorly performing prisons with or without privatization.

Areas of Future Prison Privatization Research
There are at least three key areas of future prison privatization research. The first is a
systematic analysis of how private and public providers hire, train, and retain their correctional
staff. The study of correctional labor should measure whether the two sectors achieve
comparable results. Furthermore, how do the two sectors structure correctional activities that
might promote efficiency while maintaining quality?
There has never been a systematic analysis of accountability structures across prison
systems. Are there models as Harding suggests that would promote better prisoner welfare? I do
know that states implement oversight differently especially in the way in which governments
may or may not use autonomous government functionaries. Some use legislative oversight
committees. Others use executive branch personnel such as ombudsmen or the inspector
general’s office. Is there a role for private firms in oversight? Since this strikes at the heart of
prison oversight, it would seem to me this would be an important avenue of research for any
correctional system.
Lastly, while Gaes, Camp, Nelson, and Saylor (2004) have provided a template for prison
performance measurement, it could be a goal to advance this activity with research in some key
states. This would be partly an evaluation and partly a demonstration project. There are many
states with sophisticated databases that could be used to develop performance metrics. However,
many of these states lack the sophistication to develop statistical models that can be used to
compare prisons on the performance measures. There are no states, to my knowledge that use the
results of their prison program audits as part of a broader prison evaluation framework. Many
audit tools would have to be modified so their results could be incorporated into the prison

January 2010

Page
| 31

monitoring process. For example, almost all states audit and review institutional security
procedures. However, the audit reports often point out material weaknesses with summary
statements about their findings. These could easily be translated into rating forms, and then
incorporated into a model that compares prisons within their jurisdiction. Even in states which
have relatively few prisons, performance measures generated over time would show how prison
performance is improving or deteriorating.

Implications for State Budgets and Policy Choices
In fiscal year 2000, states spent 3.8 percent of their expenditures on corrections. In fiscal
year 2008, it was 3.5 percent. This amounted to $52 billion (National Association of State
Budget Officers, 2009). Corrections expenditures include capital costs ($1.7 billion in FY 2008)
and may include spending on juvenile corrections and probation and parole. Nonetheless, prison
operations are the most significant cost in correctional expenditures. To save costs in FY 2009,
the National Association of State Budget Officers report indicates that states closed prisons,
released prisoners early, and furloughed employees. If policy makers take the Blumstein et al.
results at face value, they may believe they can achieve, on average, a $13 to $15 million
reduction in expenditures on their public prisons by privatizing some facilities. However, at the
time of their report, there were only 13 states with no private prisons in their jurisdictions. Since
Blumstein et al. did not measure how much could be saved by the scale of privatization, their
research provides no policy guidance on the marginal savings (returns) by increasing
privatization in states that already have private prisons.
Because of the ambiguity in the quality of private prison operations, states that choose to
use privatization should also insure they have a systematic and sound mechanism for monitoring
prison performance. However, states can also achieve large savings by lowering the amount of
time inmates serve, and by using alternative sanctions to prison. For example, a back of the
envelope calculation shows that reducing prison time served by 5 percent in a state that has, on
average, lengths of stay of 2.5 years would lower the average daily population in an “average”
state of 28,180 inmates per day by 1,409 inmates. This reduction might be enough to actually
close a prison. Even if there were only a marginal savings of about 30 percent of the 2004 per

January 2010

Page
| 32

diem cited by Blumstein et al of $87.05, this would be equivalent to the same savings that
Blumstein et al claim has been made through privatization in an “average” state. My goal is to
simply point out that prison savings can be achieved through alternate policies to prison
privatization.

Summary
The research literature on prison privatization does not show that one sector is more
efficient than the other. This may be because any savings to the taxpayer that the private sector
achieves through lower wages and benefits to its employees is diminished by the profits it reaps
from the contract.
Although the research on quality is also one where the two sectors seem to have equal
performance, I am less convinced by this set of results because of the lack of quality
performance assessment and the weak methodology that has been used to make comparisons. I
am optimistic that governments can make legitimate cost comparisons. Julianne Nelson and
others have set out a proper framework for conducting those analyses. I am less sanguine in the
short term that governments have the ability to make valid performance comparisons. I have set
out the conditions for this kind of analysis; however, most jurisdictions do not devote the
resources to this endeavor.
The research by Blumstein, Cohen, and Seth (2007) provides evidence that states that
used private prisons lowered the rate of growth in expenditures for public prisons. This research
is not conclusive as to whether the mechanism was private-public competition or whether these
states simply squeezed the public sector; nor does it address what happened to prison
performance in those states.
Because of the pluralistic nature of state governments, I have no doubt that privately
operated prisons will continue to be used. Governments are seeking productivity gains – more
and better services at a cheaper price. While it is not clear how much of a gain in productivity
governments can achieve through prison privatization, it is clear that they must reduce residual
rights of control by enhancing their performance measurement and insuring mechanisms of

January 2010

Page
| 33

accountability. Absent these controls, private prison operators will squeeze profits by lowering
the quantity and quality of their services.

January 2010

Page
| 34

References
Bales, William D., Bedard, Laura E., Quinn, Susan T., Ensley, David T., and Holley, Glen P.
(2005) Recidivism of public and private state prison inmates in Florida, Criminology and Public
Policy, 4(1), 57-82.
Blumstein, James F., Cohen, Mark A., and Seth, Suman (2007) Do government agencies
respond to market pressures? Evidence from private prisons, Vanderbilt University.
Camp, Scott. D., and Daggett, Dawn M. (2005) Evaluation of the Taft Demonstration Project:
Performance of a Private-Sector Prison and the BOP, Federal Bureau of Prisons, U. S.
Department of Justice, Washington, DC.
Camp, S. D. and Gaes, G. G. (2000). Private Prisons in the United States, 1999: An Assessment
of Growth, Performance, Custody Standards, and Training Requirements, Washington, DC:
Federal Bureau of Prisons.
http://www.bop.gov/news/research_projects/published_reports/pub_vs_priv/oreprcg2000.pdf
Camp, S. D., Gaes, G. G., Langan, N. P., & Saylor, W. G. (2003). The Influence of Prisons on
Inmate Misconduct: A Multilevel Investigation, Justice Quarterly, 20(3), 501-533.
Camp, S. D., Gaes, G. G., Klein-Saffran, J., Daggett, D. M., & Saylor, W. G. (2002). Using
Inmate Survey Data in Assessing Prison Performance: A Case Study Comparing Private and
Public Prisons. Criminal Justice Review, 27(1), 26-51.
Camp, S. D., Gaes, G. G., & Saylor, W. G. (2002). Quality of Prison Operations in the Federal
Sector: A Comparison with a Private Prison. Punishment & Society, 4(1), 27-53.
Cummins, C. Elaine (2001). Private Prisons in Texas, 1987-2000. Unpublished Doctoral,
American University, Washington, D.C.
Farabee, David and Kevin Knight (2002) A Comparison of Public and Private Prisons in Florida:
During- and Post prison Performance Indicators. Los Angeles, Calif.: Query Research.
Gaes, Gerald G. (2005) Prison privatization in Florida: Promise, premise, and performance.
Criminology and Public Policy, 4(1), 83-88.
Gaes, Gerald G. (2008) NIJ Journal, No. 259, March.
Gaes, G. G., Camp, S. D., & Saylor, W. G. (1998). Appendix 2: Comparing the Quality of
Publicly and Privately Operated Prisons: A Review. In D. McDonald & E. Fournier & M.
Russell-Einhorn & S. Crawford (Eds.), Private Prisons in the United States: An Assessment of
Current Practice (pp. 1-38). Boston, MA: Abt Associates Inc.

January 2010

Page
| 35

Gaes, Gerald G., Camp, Scott D., Nelson, Julianne B., and Saylor, William G. (2004) Measuring
Prison Performance: Government privatization and Accountability, Alta Mira Press: Walnut
Creek, CA.
Harding, Richard. (1997). Private Prisons and Public Accountability, New Brunswick, NJ:
Transaction Publishers.
Harding, Richard. (2001). Private Prisons. In M. Tonry (Ed.), Crime and Justice: A Review of
Research (Vol. 28, pp. 265-346). Chicago: University of Chicago Press.
Lundahl, Brad W., Kunz, Chelsea, Brownell, Cindi, Harris, Norma, Van Fleet, Russ (2009)
Prison privatization: A meta-analysis of cost and quality of confinement indicators, Research on
Social Work Practice, 19(4), 383-394.
Lanza-Kaduce, Lonn, Karen F. Parker, and Charles W. Thomas (1999) A comparative
recidivism analysis of releases from private and public prisons. Crime and Delinquency, 45:28–
47.
Lanza-Kaduce, Lonn, Karen F. Parker, and Charles W. Thomas (2000) The devil in the details:
The case against the case study of private prisons, criminological research, and conflict of
interest. Crime and Delinquency, 45:28–47
Logan, C. H. (1990). Private prisons: Cons and pros. New York: Oxford University Press.
McDonald, Douglas and Carlson, Kenneth (2005) Contracting for Imprisonment in the Federal
Prison System: Cost and Performance of the Privately Operated Taft Correctional Institution,
National Institute of Justice, Office of Justice programs, U. S. Department of Justice,
Washington, DC.
McDonald, D., Fournier, E., Russell-Einhourn, M., & Crawford, S. (1998). Private prisons in the
United States: An assessment of current practices. Cambridge, MA: Abt Associates, Inc.
McDonald, Douglas, Patten, C., Fournier, E., and Crawford, S. (2003) Governments’
Management of private prisons. Cambridge, MA: Abt Associates, Inc.
Nelson, Julianne (2002) Measuring the cost of public and private facilities, Bethesda: JW
Partners.
Nelson, Julianne (2005) Competition in corrections: Comparing public and private sector
operations, The CNA Corporation, Alexandria, Virginia.

January 2010

Page
| 36

Perrone, Dina and Pratt, Travis C. (2003) Comparing the quality of confinement and costeffectiveness of public versus private prisons: What we know, why we do not know more, and
where to go from here, The Prison Journal, 83(3), 301-322.
Pratt, Travis C., & Maahs, Jeff. (1999). Are private prisons more cost-effective than public
prisons? A meta-analysis of evaluation research studies. Crime & Delinquency, 45, 358-371.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical Linear Models: Applications and Data
Analysis Methods. Thousand Oaks, CA: Sage Publications.
Ritzer, G. (1993). The McDonaldization of Society: An Investigation into the Changing
Character of Contemporary Social Life. Newbury Park, CA: Pine Forge Press.
Ritzer, G. (1998). The McDonaldization Thesis: Explorations and Extensions. Thousand Oaks:
Sage.
Sabol, William J., West, Heather C., and Cooper, Matthew (2009) Prisoners in 2008, Bureau of
Justice Statistics Bulletin, Bureau of Justice Statistics, Office of Justice Programs, United States
Department of Justice.
Segal, G. F., & Moore, A. T. (2002) Weighing the watchmen: Evaluating the costs and benefits
of outsourcing correctional services. Part II: Reviewing the literature on cost and quality
comparisons (Reason Public Policy Institute Study No. 290). Retrieved January 20, 2009, from
http://www.reason.org/ps290.pdf
Spivak, Andrew L., and Sharp, Susan F. (2008) Inmate recidivism as a measure of private prison
performance, Crime and Delinquency, 54(3), 482-508.
Starr, Paul (1987) The limits of privatization, Economic Policy Institute, Washington, DC.
Volokh, A. (2002) A tale of two systems; Cost, quality, and accountability in private prisons,
Harvard Law Review, 115,1868-1890.
West, Heather C. and Sabol, William J. (2009) Prison Inmates at Midyear 2008 – Statistical
Tables, Bureau of Justice Statistics, Office of Justice Programs, United States Department of
Justice.

January 2010

Appendix I. Appendix table 19 from the BJS report Prisoners in 2008.

Page
| 37

January 2010

Page
| 38

January 2010

Page
| 39

Appendix II. References
Legend:
L= Lundahl, Brad W., Kunz, Chelsea, Brownell, Cindi, Harris, Norma, Van Fleet, Russ (2009)
P= Pratt, Travis C., & Maahs, Jeff. (1999) and Perrone and Pratt (2003)
S= Segal, G. F., & Moore, A. T. (2002)
Anderson, G. M. (2000, November 18). Prisons for Profit. America, 183, 12-16.Retrieved
January 21, 2009, from http://www.americamagazine.org/content/article.cfm?article_id=2321
Anderson, P., Davoli, C. R., & Moriarty, L. (1985). Private corrections: Feast or fiasco? The
Prison Journal, 65, 32-41.
L,P,S

Archambeault, W. G., & Deis, D., R. Jr. (1996).Cost effectiveness comparisons of private
versus public prisons in Louisiana: A comprehensive analysis of Allen, Avoyelles, and Winn
Correctional Centers. Baton Rouge: Louisiana State University, School of Social Work, Office
of Correctional Services.
P

Austin, J., & Coventry, G. (1999). Emerging issues on privatized prisons. Washington, DC:
National Council on Crime and Delinquency.
Blakely, C. R., & Bumphus, V. W. (2004). Private and public sector prisons: A comparison of
select characteristics. Federal Probation, 68, 27-31.
Bowditch, C., & Everett, R. S. (1987). Private prisons: Problems with the solution. Justice
Quarterly, 4, 441-453.
L

Bowery, M. (1996, May). Private prisons in NSW: Junee—year two (Research Publication No.
35, May 1996). Sydney, New South Wales, Australia: NSW Department of Corrective Services.
Calvert-Hanson, L. S. (1991). The privatization of corrections movement: A decade of change.
Journal of Contemporary Criminal Justice, 7, 1-20.
Brister, R. C. (1996). Changing of the guard: A case for privatization of Texas prisons. The
Prison Journal, 76, 310-330.
L

Brown, A. (1994). Economic and qualitative aspects of prison privatization in Queensland.
Annandale, New South Wales, Australia: Pluto Press.
Camp, S. D., & Gaes, G. G. (1998, October). Private adult prisons: What we know and how we
can know more, monograph. Washington, DC: Federal Bureau of Prisons, Information, Policy,
and Public Affairs Division.

January 2010

Page
| 40

Chang, T. F. H., & Thompkins, D. E. (2002). Corporations go to prisons: The expansion of
corporate power in the correctional industry. Labor Studies Journal, 22, 45-69.
Cikins, W. I. (1986). Privatization of the American prison system: An idea whose time has
come? Notre Dame Journal of Law and Ethics, 2, 445-464.
Cox, N. R., & Osterhoff, W. E. (1993). The public-private partnership: A challenge and an
opportunity for corrections. In G. Bowman, S. Hakim, & P. Seidenstat (Eds.), Privatizing
correctional institutions (pp. 113-129). New Brunswick, NJ: Transaction Publishers.
Culp, R. F. (2005). The rise and stall of prison privatization: An integration of policy analysis
perspectives. Criminal Justice Policy Review, 16, 412-442.
L,P

Drowata, C., & Stoughton, D. (1995). Comparative evaluation of privately-managed CCA
prison (South Central Correctional Center) and state-managed prototypical prisons (Northeast
Correctional Center, Northwest Correctional Center). Nashville, TN: Select Oversight
Committee on Corrections. Retrieved January 20, 2009, from
http://www.ncjrs.gov/App/Publications/abstract.aspx?ID=175555
Durham, A. M. (1989). The privatization of punishment: Justification, expectations, and
experience. Criminal Justice Policy Review, 3, 48-73.
Geis, G., Mobley, A., & Shichor, D. (1999). Private prisons, criminological research, and
conflict of interest: A case study. Crime & Delinquency, 45, 372-388.
General Accounting Office. (1996). Private and public prisons: Studies comparing operational
costs and/or quality of service. Washington, DC: Author.
L

Greene, J. (1999). Comparing private and public prison services and programs in Minnesota:
Findings from prisoner interviews. Current Issues in Criminal Justice, 11, 202-225.
Harding, R. (1998). Evaluating private prisons: A reply. The Australian Journal of Criminology,
31, 314-320.
Harland, A. (1997). Editorial introduction. The Prison Journal, 77, 3-5.
L,P,S

Hatry, H. P., Brounstein, P. J., & Levinson, R. B. (1989). Comparison of privately and
publicly operated corrections facilities in Kentucky and Massachusetts. Washington, DC: The
Urban Institute.
Henig, J. R. (1985). Privatization and decentralization: Should governments shrink? Public
Policy and Federalism, 12, 26-53.
Lilly, R. J., & Deflem, M. (1996). Profit and reality: An analysis of the corrections commercial

January 2010

Page
| 41

complex. Crime & Delinquency, 42, 3-20.
Lilly, R. J., & Knepper, P. (1993). The corrections-commercial complex. Crime and
Delinquency, 39, 150-166.
Logan, C. H. (1987). The propriety of prisons. Federal Probation, 51, 35-40.
Logan, C. H. (1990). Private prisons: Cons and pros. New York: Oxford University Press.
P,S

Logan, C. H. (1992). Well kept: Comparing quality of confinement in private and public
prisons. Journal of Criminal Law and Criminology, 83, 577-613.
Logan, C. H. (1996). Public vs. private prison management: A case comparison. Criminal Justice
Review, 21, 62-85.
L

Maximus, Inc. (2006). Report on the evaluation of Arizona Department of Corrections’
operating per capital cost report and private prison cost mode. Phoenix, AZ: Author. Retrieved
January 22, 2009, from http://www.afscme.org/docs/maximusnotes.pdf

McConville, S. (1987). Aid from industry? Private corrections and prison crowding. In S.
Gottfredson&S. McConville (Eds.), America’s correctional crisis: Prison populations and
public policy (pp. 221-242). New York: Greenwood.
McDonald, D. C. (1989). The cost of corrections: In search of the bottom line. Research In
Corrections, 2, 1-26.
L

McDonald, D. C., & Carlson, K. (2005). Contracting for imprisonment in the federal prison
system. cost and performance of the privately operated Taft Correctional Institution. Cambridge,
MA: Abt Associates for the National Institute of Justice.
S

Mitchell, G. A. (1996, December). Controlling prison costs in Wisconsin (Report No. 9).
Thiensville: Wisconsin Policy Research Institute, Inc.
Morgan, Stephen L. and Winship, Christopher (2007) Counterfactuals and Causal Inference:
Methods and Principles for Social Research. Analytical Methods for Social Research.
National Association of State Budget Officers (2009) Fiscal Year 2008 State Expenditure
Report. http://www.nasbo.org/Publications/StateExpenditureReport/tabid/79/Default.aspx
P,S

Office of Program Policy Analysis and Government Accountability. (1998, April). Review of
Bay Correctional Facility and Moore Haven Correctional Facility (Report No. 97-68).
Tallahassee, FL: Author.

January 2010

Page
| 42

L,P,S

Office of Program Policy Analysis and Government Accountability (OPPAGA) Private
Prison Review. (2000, March). South Bay correctional facility provides saving and success;
Room for improvement. (Report No. 99-39). Tallahassee, FL: Office of Program Policy Analysis
and Government Accountability.
Ogle, R. S. (1999). Prison privatization: An environmental catch-22. Justice Quarterly, 16, 579600.
Reisig, M. D., & Pratt, T. C. (2000). The ethics of correctional privatization: A critical
examination of the delegation of coercive authority. The Prison Journal, 80, 210-222.
Rhodes, William (2009) Estimating Treatment Effects: A Primer for Evaluators, Cambridge,
MA. Unpublished manuscript available from the author.
Sechrest, D., & Shichor, D. (1993). Corrections goes public (and private) in California. Federal
Probation, 57, 3-8.
Sechrest, D. K., & Shichor, D. (1996). Comparing public and private correctional facilities in
California: An exploratory study. In G. L. Mays & T. Gray (Eds.), Privatization and the
provision
of correctional services: Context and consequences (pp. 133-151). Cincinnati, OH: Anderson.
Segal, G. F., & Moore, A. T. (2002). Weighing the watchmen: Evaluating the costs and benefits
of outsourcing correctional services. Part II: Reviewing the literature on cost and quality
comparisons (Reason Public Policy Institute Study No. 290). Retrieved January 20, 2009, from
http://www.reason.org/ps290.pdf
L,P,S

Sellers, M. P. (1989). Private and public prisons: A comparison of costs, programs and
facilities. International Journal of Offender Therapy and Comparative Criminology, 33, 241256.
Shichor, D. (1999). Privatizing correctional institutions: An organizational perspective. The
Prison Journal, 79, 226-249.
Shichor, D. (1993). The corporate context of private prisons. Crime, Law and Social Change, 20,
113-138.
Shichor, D. (1995). Punishment for profit: Private prisons/public concerns. Thousand Oaks,
CA: Sage.
Texas Sunset Advisory Commission. (1990, December). Contracts for correctional facilities
and services (Staff Report to the Sunset Advisory Commission). Austin, TX: Author.

January 2010

Page
| 43

L,P,S

Thomas, C. ( 1997, August). Comparing the cost and performance of public and private
prisons in Arizona: An overview of the study and its conclusions. Retrieved December, 2006,
from http://www.anarchistblackcross.org/content/essays/articles/priv/ariz.htm
Urban Institute. (1985, August). Comparison of privately and publicly operated corrections
facilities in Kentucky and Massachusetts, project report. Washington, DC: Author.
Vardalis, J. J., & Becker, F.W. (2000). Legislative opinions concerning the private operation of
state prisons: The case of Florida. Criminal Justice Policy Review, 11, 136-148.
Winn, R. G. (1996). Ideology and the calculation of efficiency in public and private correctional
enterprise. In G. L. Mays & T. Gray (Eds.), Privatization and the provision of correctional
services: Context and consequences (pp. 21-30). Cincinnati, OH: Anderson.

January 2010

Page
| 44

Table 1 Growth in the Private Prison Sector

..

Table 11 . Number of pris0ner5 held in private faci lities, December 31 , 2000·2007 , and
J une 30, 2007 and 2008
Number of pri50rlen

Iotil

~.

=
=

Fedeiill

,

90.542
91.953
93.912
95.707
98.628
107.940
113.697

15.524
19.251
20.274
21.865
24.768
27.046
27.726

85.971

moo~

118.239
125.997

30.379
31.310

87.860
94.687

moo~

126.249

32.712

93.537

~,

~,

~
~

=

75.01 8
72,700
73.638
73.842
73.860
OO.~

~,

=- "

Percentdall
~~

.,.,
.,

6.5%

••
""
,..

7.4%

, ~

"""""I chanQe
A"""'IJE! amual cllange.

12/3112O!JO.1m 1f"l!.XJ7

. ~

Percent chanQe.
~,~

""

10.5%

n

3.4%

..,

Note: Irdudes estmates lor 1I!IOis lor 2006. 2007. and 2008 and Nevada lor Oecerrbe< 31. 2007.
See MelIIodotbgy
Not c.aI!Ualed.

January 2010

Page
| 45

Table 2. Avoidable and Unavoidable Central Office Costs, Oklahoma Department of
Corrections, 2000

Amount
Unit

(Dollars)

Allocated to

Allocated to

State

Private

Operations

Operations

Amount
(Dollars)

Administrative Services

356,869

Yes

No

Adult Basic Education

64,000

Yes

No

Building Maintenance

474,896

Yes

No

653,017

Yes

No

1,453,223

Yes

Yes

1,453,223

430,245

Yes

Yes

430,245

Information Technology

2,736,643

Yes

Yes

2,736,643

Construction & Maintenance

1,432,987

Yes

No

13,100

Yes

No

Day Reporting

124,000

Yes

No

Directors Office

554,242

Yes

Yes

Education Administration

342,908

Yes

No

Electric Monitoring

586,000

Yes

No

Finance & Accounting

1,482,278

Yes

Yes

1,482,278

General Administration

3,864,918

Yes

Yes

3,864,918

Center for Correctional Officer
Studies
Central Transportation
Chief of Operations

Curriculum Development

554,242

January 2010

Page
| 46

Internal Affairs

531,442

Yes

Yes

531,442

Legal Services

1,143,344

Yes

Yes

1,143,344

Offender Program Monitor

607,933

Yes

Yes

607,933

Office Technology

560,651

Yes

Yes

560,651

Oklahoma State Industries

3,001,683

Yes

No

Personnel

1,100,458

Yes

No

Population Management

1,542,864

Yes

Yes

Probation & Parole Equipment

500,000

Yes

No

Public Relations

168,631

Yes

Yes

168,631

Research & Evaluation

698,830

Yes

Yes

698,830

22,100

Yes

No

Sentence Administration

323,534

Yes

Yes

Staff Development Center

848,991

Yes

No

73,700

Yes

No

66,666

Yes

No

Safety

1,542,864

323,534

Statewide & Regional Cognitive
Skills
Tulsa Female Offender Care

$16,098,77
TOTAL

$ 25,760,153

8

January 2010

Page
| 47

Table 3. Abt and CNA Analyses of Cost for Publicly and Privately Operated Federal Prisons
Fiscal Year

1999

2000

2001

2002

Abt

CNA

Abt

CNA

Abt

CNA

Abt

CNA

Elkton

39.72

35.24

39.77

34.84

44.75

36.79

46.38

40.71

Forrest City

39.46

35.29

39.84

35.28

41.65

37.36

43.61

38.87

Yazoo City

41.46

36.84

40.05

34.92

43.65

37.29

42.15

38.87

Taft

33.82

34.42

33.25

33.21

36.88

37.04

36.37

38.62