Skip navigation

Aman Indiana University Law Private Prisons and the Democratic Deficit 2010

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.


Research Paper Number 160

March 2010

Alfred C. Aman, Jr.

This paper can be downloaded without charge from the
Social Science Research Network electronic library at:

<CT>Private prisons and the democratic deficit<CT>
<CA>Alfred C Aman, Jr<CA>

<T1>Outsourcing is a common form of privatization in the United States—calling on
the private sector to deliver what were once government-provided services. A key
question, centrally confronted in Part III of this volume, is whether certain functions
should be regarded as ‘‘inherently governmental’’ and thus precluded from
outsourcing. This chapter examines privatization in the US prison context with a view
to mapping out an answer to that question. Is there something ‘inherently
governmental’ in the realm of prisons that should set restrictions on outsourcing?
What might prison privatization teach us about other privatized governmental
functions, such as those having to do with military matters? At the state level, there is
currently a range of legislative initiatives involving specification of conditions under
which prisons, specific prison services, and other governmental services may be
contracted out. What can these privatization policies and approaches at the state level,
particularly those dealing with private prisons, teach us about what activities are
‘inherently governmental’, as well as their procedural demands?
Prisons are a useful case for considering these questions of inherent
governmental functions. On the one hand, they are buildings and workplaces like any
other, servicing inmates as if they were clients—with clean laundry, occupations,
education, and so forth. On the other hand, they are settings in which responsibility
for the care of inmates—nutrition, medical care, basic life conditions—reach the level
of human rights concerns. The former functions might easily be outsourced, but the
latter—where the fundamental integrity of the body and the basic dignity of the

person are at stake—pose more challenging questions. This chapter is primarily
concerned with that grey zone, where the vulnerability of the prison population poses
challenges to the current standards by which inherent governmental functions are
Privatization of state and federal prison systems provides a rich example of the
selective shift away from traditional government-provided services to a system highly
dependent on private providers. 1 In the United States, while there remain several
states that do not currently allow for privatized correctional facilities, at least 33 states
house a percentage of inmates in privately owned and operated prisons. 2 Several
states have also outsourced to private companies for the provision of services within
state-run prisons, such as health care, often with less than desirable, or even
disastrous, results. 3


See Paige M Harrison and Allen J Beck, ‘Prisoners in 2005’, Bureau of Justice Statistics











Ibid. 5–6 (at the end of 2005, ‘33 States and the Federal system reported a total of 107,447
prisoners held in privately operated prison facilities … Private facilities held 6.0% of all State
prisoners and 14.4% of Federal prisoners.’). In comparison, the percentage of juveniles housed
in private facilities is much higher than the percentage of adults in private prisons. As of
October 2002, only 40% of juvenile facilities (housing 69% of juvenile offenders) were
publicly operated. Melissa Sickmund, ‘Juvenile Residential Facility Census, 2002’, Juvenile
Offenders and Victims National Report Series, Office of Juvenile Justice and Delinquency










< >.

See ‘Lawsuit over Michigan Death Settled for $3.25m’, Associated Press, 8 July 2008. See

Since the 1980s, privatization has gained prominence as a tool of governance 4
and, as has been argued elsewhere, is of a piece with a neoliberal conception of
globalization that has introduced new ways for states to carry out their duties. 5 This
neoliberal approach, as it has been implemented in the United States, tends to treat
social services almost exclusively as costs rather than as investments in the state’s
human infrastructure. When applied to state functions, such as prison administration,
it treats privatization through a series of technical questions, which emphasize
technical expertise and substantially minimize public participation. A democracy
deficit, however, occurs when public functions are carried out by private actors to the
extent that requirements of transparency and public participation—the keystones of
administrative democracy—are reduced or set aside. 6
The Federal Activities Inventory Reform Act, which prohibits the privatization
of ‘inherently governmental’ functions, specifies that an ‘inherently governmental’
function is one that is ‘so intimately related to the public interest as to require
performance by Federal Government employees,’ including the execution of laws that

also Michael D LaFaive, ‘Privatization for the Health of It’, Mackinac Center for Public
Policy, Winter 2005, 4, available at <>.

See Ruth Hoogland DeHoog and Lester Salamon, ‘Purchase of Service Contracting in The
Tools of Governance’, in Lester Salamon (ed), The Tools of Government: A Guide to the New
Governance (New York: Oxford University Press, 2002) 319-39.


See Alfred C Aman, Jr, Privacy and Democracy: Resources in Administrative Law in
Government by Design (Harvard University Press, forthcoming).


See Alfred C Aman, Jr, The Democracy Deficit: Taming Globalization Through Law Reform
(New York: NYU Press, 2004).

‘significantly affect the life, liberty, or property of private persons.’ 7 The phrase
‘‘inherently governmental’’ (discussed more fully by Simon Chesterman in chapter
nine in this volume) has been explicated more fully in the context of the Office of
Management and Budget (OMB) Circular A-76, which establishes federal policy and
procedures regarding the performance of commercial activities and whether they may
be performed by commercial sources or if government employees must perform
them. 8 The ‘‘inherently governmental’’ test has not been significantly restrictive
when it comes to privatization at the federal level. 9 Much of the federal government’s
contracting is not subject to the restrictions of the OMB Circular A-76, and there is a
lack of concern about outsourcing involving ‘inherently governmental’ functions. 10


US Congress, Federal Activities Inventory Reform Act of 1998, PL 105-270, codified at 31
USC § 501 (1998), s 5.


‘Performance of Commercial Activities’, OMB Circular No A-76 (Revised), White House











generally Paul Verkuil, Outsourcing Sovereignty: Why Privatization of Government Functions
Threatens Democracy and What We Can Do About It (New York: Cambridge University
Press, 2007).

See generally Dan Guttman, ‘‘inherently governmental’ Functions and the New Millennium’,
in Thomas H Stanton and Benjamin Ginsberg (eds), Making Government Manageable
(Baltimore: Johns Hopkins University Press, 2004) 40 (citing Flagg Bros. v Brooks, 436 US
149 (1978)). In Flagg Bros. v Brooks, the Supreme Court found that in the history of the
United States the only two activities that could be termed ‘exclusive public functions’ are
elections and police activities of company towns.


Paul R Verkuil, ‘Public Law Limitations on Privatization of Government Functions’, North

The OMB has liberally granted outsourcing requests made by federal agencies. State
legislation also provides for a variety of approaches to privatization in general and the
privatization of prisons in particular.
The criteria set forth in these various state approaches to privatization can be
read with a view toward understanding some of the key factors that are currently in
play as criteria for outsourcing decisions. As we will see, these seldom yield bright
lines that would clearly delineate what activity is or is not considered ‘inherently
governmental’. In the end, these decisions are largely political and, as this chapter will
also argue, this is as it should be. As we shall see through a reading of these statues,
the question of inherent governmental function is closely tied to procedural values








Administrative law reforms will thus feature prominently in this discussion.
The primary management tool for outsourcing prisons to the private sector is
the contract. Government agencies remain responsible for the outcomes reached by
these private providers under these contracts, but they are no longer involved in the
day-to-day regulation or management of the enterprise. Under these arrangements,
private actors become the dominant players by virtue of their contract with the state.
And unlike public agencies, these private actors are unlikely to be subject to statutes
such as the Administrative Procedure Act or the Freedom of Information Act at either
the state or federal levels. Privatization tends to imply that oversight and

Carolina Law Review, vol 84 (2006) 441. Verkuil observes that if an agency erroneously
classifies an activity as non-’inherently governmental’, that designation is not subject to
administrative review. The designation may only be challenged by ‘an agency official who is
competing against a private contractor to save her job’.

accountability will remain as they would under the government, but this is not always
explicit nor is it consistently the case. Those who operate private prisons may be
considered state actors for purposes of the Due Process Clause of the US Constitution,
but the application of this body of law to prisons generally does not provide much
protection for prisoners. Moreover, the contracting process itself is often some form
of a least-bid contract procedure, one that tends to focus predominantly on cost. 11
Reducing everything to cost considerations would imply that any private provider
capable of making a profit on the activity can, and perhaps should, provide this
service—the assumption being that profitability will be passed onto taxpayers as
The discussion in this chapter follows the framework proposed by Daphne
Barak-Erez in chapter four in this volume, but its main focus is on the first two prongs
of the inquiry Barak-Erez proposes: the boundaries of privatization and the
administrative process of privatization. The next section of this chapter will discuss
how legislatures have sought to deal with the issues of privatization in general and
prisons in particular. The chapter then examines the reaction of courts to various
privatization issues, especially those involving prison disciplinary proceedings. The
reaction of both legislative and judicial public bodies suggests some criteria for
considering whether and how to outsource prison management functions.
Accordingly, we shall explore the outer boundaries of the use of privatization as a
governance tool as well as the need for processes that can accommodate a broad


See generally Alfred C Aman, Jr, ‘An Administrative Law Perspective on Government Social
Service Contracts: Outsourcing Prison Health Care in New York City’, Indiana Journal of
Global Legal Studies, vol 14 (2007) 301.

spectrum of political views in making these decisions.
The chapter thus argues for a new administrative law that ensures
participation, transparency and accountability when outsourcing decisions are made.
The comprehensive marketizing of government services artificially restricts the realm
of ‘‘inherently governmental’’, which ordinarily involves the test of whether the
service is inherently viable. Factors other than cost must also be involved in assessing
what is and is not ‘‘inherently governmental’’. Are there other aspects of running a
prison that limit outsourcing and what are the implications of these limits in other
privatization contexts? Can the government outsource the adjudication of disciplinary
disputes arising from prisoners’ behaviour? Could the government outsource
executions of prisoners already sentenced to death? Are cost and efficiency factors
alone the only considerations in delegating such tasks?
Drawing on the discussion of prisons, the conclusion is that certain democracy
considerations should be important limitations on privatization. There are some issues
that require a collective response at the time they arise. Issues involving life or death
or the loss of liberty are ‘‘inherently governmental’’ because those decisions speak for
us all in a fundamental way. Considering how private prisons should operate to
mitigate the democracy deficit raises the more general question of when a democratic,
collective response requires a public decision-maker. It is at this level that the
legislation affecting private prisons is more broadly relevant to the privatization of
military services, especially in the detention context. The outcry over Abu Ghraib
focused primarily on interrogation methods, but the situation there also revealed
fundamental problems regarding the integration of military and private contractors


from the standpoint of chain of command, oversight and accountability. 12

<H1>Regulating private prisons: Legislative responses<H1>

<T1>There are various examples of state legislatures attempting to regulate
privatization processes and their effects, especially in the prison context. The
legislatures have treated these issues largely in economic terms, with little regard for
the more fundamental question of what can or cannot be outsourced. The processes
have tended not to allow direct public participation outside of legislators themselves.
Statutes involving the privatization of prisons vary widely from state to state
in the protections they afford to the human rights of prisoners. They also vary in the
extent to which they involve the public at large in a timely way. 13 As a result of costsaving concerns dominating the decision-making process for privatization of prisons,
state privatization legislation rarely, if ever, mentions prisoners’ human rights. Even


Two reports recently released by independent commissions note various concerns and
problems with the structuring of direction and command with respect to Abu Ghraib. See
generally James R Schlesinger, ‘Final Report of the Independent Panel to Review Department










‘Investigation of the Abu Ghraib Detention Facility and 205th Military Intelligence Bridge’,
AR 15-6, available at <>.














where the rights are mentioned in the legislation, there is little or no guidance on how
the rights should be respected and enforced as a matter of process. The fact that
decisions to privatize are often made hastily, usually in the face of budget pressures or
court orders to relieve overcrowding at public facilities, further exacerbates the
This section begins the discussion of legislative reactions to privatization with
a review of the cost assumptions underlying the privatization of prisons. This is
followed by a consideration of which prison services can be privatized in ways that
involve the public generally and retain the government’s accountability for the human
rights of prison inmates. Finally, criteria that should signal the outer limits of
privatization in the context of prisons are proposed.

<H2>Privatization, cost, democracy and inequality<H2>
<T1>Much of the pressure to privatize prisons is driven by cost. Even when prisons
remain state-operated, many have turned to privatizing certain financially draining
services, such as inmate health care. Take prison health services, for instance, the
‘biggest for-profit company providing medical care in jails and prisons’ which
‘[d]espite a tarnished record … has sold its promise of lower costs and better care, and
… amassed 86 contracts in 28 states, and now cares for 237,000 inmates, or about one
in every 10 people behind bars’. 14
In the United States, individual states take varied approaches to privatization,


Paul von Zielbauer, ‘As Health Care in Jails Goes Private, 10 Days Can Be a Death Sentence’,
New York Times, 27 February 2005.

subjecting private service providers to differing degrees of public accountability.
Many states have statutes regarding the privatization of prisons, some allowing and
encouraging them, some banning their use altogether. 15 A number of states also have
general privatization statutes covering the privatization of services more generally.
These general statutes reflect a variety of views towards privatization. On the one end
of the spectrum is Colorado’s privatization statute, which provides that ‘it is … the
policy of this state to encourage the use of private contractors for personal services to
achieve increased efficiency in the delivery of government services.’ 16 In contrast, the
Massachusetts legislature ‘finds and declares that using private contractors to provide
public services formerly provided by public employees does not always promote the
public interest.’ 17 The danger of statutes like Colorado’s is that they state as fact the
debatable claim that private providers are more efficient; the statute assumes that
private companies would more efficiently provide any service. This statute operates as
a presumption that privatization is in the public interest, which may stifle debate on
the actual merits of any individual privatization decision.
There is a continuing debate about whether private prisons offer any real cost
benefit. It is common, however, for some privatization statutes to require a minimum-


See, eg, NY Corrections Law § 72(1) (McKinney 2003) (requiring the department to house
and care for all inmates); Ark Code Ann § 12-50-106(a) (Michie 2002) (authorizing contracts
for construction, financing, and operating facilities).


Colo Rev Stat Ann § 24-50-501 (West 2003).


Mass Gen Laws Ann ch 7, § 52 (West 2003).

percentage cost savings from private prison operators. 18 Some states set a bar to
outsourcing at a certain level of cost savings. While there would be little point in
privatizing if some savings were not anticipated, the risks to prisoners’ rights would
seem to increase directly with the scale of the cost-savings requirement. Contrary to
popular wisdom, many states operate their own prison systems with great frugality
even without outsourcing. Alabama’s public prisons, for example, spend $1.08 to feed
each prisoner each day. It is hard to imagine how private companies could best such a
low level expenditure without compromising the nutritional status of the inmates—a
fundamental human rights concern. In addition, in Alabama prisons, much of the
work required to run the facility—maintenance work such as painting, weeping, lawn
mowing, staffing the kitchen, unloading supplies, and even growing vegetables—is
performed by prisoners, further reducing the costs of operation. 19
Privatization of prison services is not a monolith. In some areas, prisons
involve the basic living conditions of inmates and their human rights status. In others,
they do not—and are much like institutions outside the criminal justice context. Thus,
it is necessary to ask what costs the legislators have specified in their statutes as ones
that would or should be lowered by the private entities and what gains they anticipate
as a result of the freedom given private companies in making those provisions.


See, eg, Ky Rev Stat Ann § 197.510(13) (Banks-Baldwin 1998) (requiring at least 10%
savings); Tenn Code Ann § 41-24-104(C)(2)(B) (2002 Supp) (requiring at least 5% savings);
Ohio Rev Code Ann 9.06(A)(4) (West 2002) (requiring at least 5% savings); Fla Stat Ann §
957.07 (West 2003 Supp.) (requiring at least 7% savings).


Carla Crowder, ‘Alabama Feeds Prisoners on $1.08 a Day’, Birmingham News, 26 November

The Colorado privatization statute explicitly allows private providers to adjust
worker wages and benefits: ‘The general assembly recognizes that such contracting
may result in variances from legislatively mandated pay scales and other employment
practices that apply to the state personnel system.’ 20 In contrast, Washington DC
requires a private provider to offer displaced workers a right of first refusal for jobs
with the private company and further requires that the private company comply with
the government pay-scale for six months. 21 The DC statute restricts a private
provider’s ability to meet cost targets by hiring more efficient workers or changing
incentive structures; any required efficiency gains must, therefore, come from the
reduction of other costs. The Colorado statute, like many others, requires that
‘privatization of government services not result in diminished quality in order to save
money.’ 22
In addition to addressing cost-saving concerns, privatization statutes need to
address the questions of quality benchmarks: how should quality be assessed and from
whose perspective? Unless prison contracts entered into under the statute contain
verifiable measures of quality, such statutory terms are, in all likelihood,
unenforceable in practice. With a few exceptions, states in the United States do a poor
job of specifying what the state desires (besides low cost) to get for their agreement to
employ a private prison provider. Michigan, in its statute allowing the privatization of
juvenile-correction facilities, mandates that private providers of prisons require


Colo Rev Stat Ann § 24-50-501 (West 2003).


DC Code Ann § 2-301.05(d)(2)-(4) (2003).


Colo Rev Stat Ann § 24-50-501 (West 2003).

prisoners without high school degrees to receive a general education certificate
(GED). 23 Colorado requires that private providers guarantee education services (and
other services including dental, medical, psychological, diet, and work programmes)
of at least as high quality as public prisons. 24 The problem with Colorado’s provision
is that it does not provide any specific, concrete standards by which a comparison of
quality can be judged.
Part of the reason for concern about the potential for human rights abuses in
private prisons is that private prisons are unlikely to be able to sustain the conditions
that would support a ‘relatively strong relationship between democratic forms of
government and the protection of human rights’. 25 Even in the case of activities that
would not require government retention of responsibility in other settings (such as the
provision of food services in government buildings) the prison context might call for
such attention. Furthermore, the axiom to the effect that ‘the civil liberties usually
associated with democracies … enable citizens and opposition groups to publicize
abuses’ 26 loses its force in the prison context. In prisons, where a defined population
has lost some of its civil liberties (including, often, the right to vote after release from
prison), prisoners are therefore excluded from the bargaining necessary to prevent


Mich Comp Laws Ann § 791.220g (West 2001).


Colo Rev Stat Ann § 17-1-202(f) (West 2003).


Wesley T Miller, ‘Economic Globalization and Rights: An Empirical Analysis’, in Alison
Brysk (ed), Globalization and Human Rights (Berkeley, CA: University of California Press,
2002) 83.



oppression, and, especially in a privatized context, the general public usually is
excluded from regular information about the treatment of inmates. Even if adequate
democratic checks exist for prison privatization decisions, most states do not subject
providers to oversight sufficient to ensure accountability with respect to those issues.
As discussed above, privatization of prison services by definition involves
shifts in cost allocations and direct or indirect subsidies—for example, on the part of
prisoners whose conditions of confinement will be affected so as to meet the cost
savings threshold. But it should not alter basic democratic processes. For example, the
question of how costs are to be shifted (for example, to workers) should be subject to
the same procedural checks (participation, oversight, accountability) as any
‘inherently governmental’ service would be. This is why the procedures
recommended below take on special importance in contexts such as these. But the
point also has wider relevance—for example, in issues of wages and other cost
savings that directly affect workers or the quality of services. Privatization of prisons
sidesteps important statutory public law protections such as the Freedom of
Information Act. Even when the due process clause of the Fourteenth Amendment is
brought to bear, courts are most reluctant to intervene, except in the most extreme
More broadly, a major criterion for determining what is (or is not) ‘inherently
governmental’ should be guided by procedures in which those directly affected by the
decisions participate—even if indirectly through representation. In the prison context,
this might refer to workers or to inmates; indeed, democracy concerns should not be
suspended for the inmate population as if their exclusion from the management of
their own affairs were a normal condition of their confinement.
Determining what is ‘inherently governmental’ is a political process involving

basic questions of democracy. Administrative procedures are thus key to our concerns
in this chapter. The need for public participation is stressed including, but not limited
to, those directly affected by the service area in question. Without public input, there
is no real check on the powers of the providers. Cost is a check, too, but only
superficially at best, since without other safeguards, costs can easily be shifted to
segments in the population that cannot participate in or defend their own interests.

<H2>Private prisons, transparency and accountability<H2>
<T1>A major limiting factor on privatization is the need for the government to ‘step
back in’ and take over the functions from the private company, for example, if the
private providers involved go bankrupt or if the provider is incompetent or for various
reasons cannot perform the contract agreed upon. Prisons are integral to a state’s
criminal law—so continuity of service and ‘step-in’ ability are needed.
In order for a government to retain legitimate accountability for private
prisons ‘[t]he state must retain and be able actually to exercise “step-in” rights—that
is, to reclaim any privatized portion of its prison system—and to do this it needs to
have ongoing capacity and skill levels of its own. This can only be done if it remains a
direct service provider in relation to at least some part of its prisoner population.’ 27
For some services, such as janitorial services, the ability to ‘step-in’ would not require
the retention of a pool of state workers, even if such services were deemed to be a
high human rights priority. For positions such as prison guards, in contrast, the state
would have to retain a pool of public guards. Otherwise, if human rights problems


Richard Harding, ‘Private Prisons’, Crime & Justice, vol 28 (2002) 282.

developed at a private facility because of inadequate training, the state could not
‘step-in’ with workers to take over security provision, and the state would likely have
also lost the institutional ability even to train workers to perform the duties of a prison
guard. In both 1982 and 1997 the Tennessee legislature considered, and rejected,
proposals to privatize the state’s entire prison system. 28 The Correctional Corporation
of America promised the state that it would be able to save $100 million per year with
a completely privatized system. Not only would a state lose the ability to ‘step-in’ if it
accepted such a proposal, but if the private provider was really able to save such a
large sum of money, it would be difficult to see how a state would be able to quickly
take back a prison system. States have shown a propensity to spend or return in taxcuts any surplus money, and Tennessee, as a mid-sized state, would be hard-pressed
to come up with the money necessary to pay for the re-establishment of a public
prison system in the event the private provider was unsatisfactory. The ability of a
state to ‘step-in’ can be ensured, at least in part, through a ‘contestability’ process
where public providers are able to bid against private providers for contracts. Such a
procedure, however, risks the removal of a governmental check on the dominance of
economic concerns over human rights issues, insofar as the public providers are
subjected to the same dominance of economic issues over non-economic values in the
provision of services.
While there are over one hundred privately run prisons in the United States,
their performance is often not compared with those of public prisons because most
government agencies ‘have been satisfied with monitoring compliance with the terms


Ibid. at 281.

of the contracts’. 29 The fact that most government oversight of private prisons
concerns the monitoring of contract terms necessitates the inclusion of human rights
provisions in privatization contracts. 30 This is especially true given the increasing
judicial indifference in the United States to prisoner suits, 31 though the Supreme
Court has recognized that private prisons and their employees are less immune from
suits than their public sector counterparts. 32
One way to increase the accountability of private prison operators to the state,
and thereby ensure that the state retains ultimate responsibility for prisons, is to limit
the length of the privatization contract. The Supreme Court recognized this form of
accountability in Richardson v McKnight, where a Tennessee statute limited a
contract’s term to three years. The majority stated that the firm’s ‘performance is
disciplined … by pressure from potentially competing firms who can try to take its


Douglas C McDonald, ‘Private Prisons in the United States: An Assessment of Current










In general, accountability provides ‘a management standard that can work even where
fundamental values are in dispute’. See Sarah Armstrong, ‘Bureaucracy, Private Prisons, and
the Future of Penal Reform’, Buffalo Criminal Law Review, vol 7 (2003) 300. In the prison
context, however, accountability is reduced to ‘reaffirming the functionalist role of prisons’ by
making sure the private contractor meets certain contractual benchmarks that may have
nothing to do with the humane treatment of the prisoners. Ibid. 302.


Harding, ‘Private Prisons’, n 27 above, 320.


See Richardson v McNight, 117 S Ct 2100 (1997) (holding that private prison guards do not
enjoy the qualified immunity, in 42 USC § 1983 cases that public prison guards do).

place.’ 33 Many states, however, do not specify a maximum contract length; some
statutes explicitly allow for long-term contracts. Arkansas, for example, states that
contracts with private prisons ‘may be entered into for a period of up to twenty (20)
years.’ 34 On the other end of the spectrum, Ohio provides that a contract ‘shall be for
an initial term of not more than two years, with an option to renew for additional
periods of two years.’ 35 Shorter contracts increase the potential frequency of public
input into the process, which ideally would be encouraged before contract renewal
takes place.
One criterion of a service’s status as an ‘inherently governmental’ function
that emerges from the statutes is the degree of competition involved in the market. If
privatization only means the replacement of a public monopoly with a private
monopoly, the absence of market competition would call for procedural checks
parallel to those applicable to inherent governmental functions. We must consider the
court’s praise for short term contracts in this light. A problem with the Court’s praise
of the short term provided by Tennessee law is that the Court assumed that there
would be a number of firms available if the private provider should fall short in its
performance. Whether there is any real competitiveness in the privatized prison
industry, however, is questionable because of its oligopolistic nature. By the end of
1998, over 76 percent of the private prison capacity in the United States was
controlled by just two companies: Wackenhut Corrections Corporation (WCC) and


Ibid. at 2106.


Ark. Code Ann. § 12-50-106(d) (Michie 2002).


Ohio Rev Code Ann 9.06(A)(1) (West 2002).

Corrections Corporation of America (CCA), 36 and CCA is a former subsidiary of
WCC. 37 When states privatize they must realize that the benefits of any market-like
effects are concentrated in the period before a contract is entered, ‘The distinctive
feature of contracting out is the element of ex ante competition — competition for the
market as opposed to in it.’ 38 The imposition of long-term contracts between states
and prison providers is likely to further concentrate the industry, by providing fewer
opportunities for new companies to enter a market with a very limited number of
potential customers.
In addition to the provision for a short contract period, Tennessee also
provides that any private prison ‘must agree that the state may cancel the contract at
any time after the first year of operation, without penalty to the state, upon giving
ninety (90) days’ written notice.’ 39 This provision encourages the state to oversee the
running of any private prison more closely, because the delegation can easily be
reconsidered. Private groups that are interested in the privatization of prisons also
have an incentive to monitor the private provider more closely, because at any time
after the first year they can lobby the state to rescind the contract if it becomes


James Austin and Garry Coventry, Emerging Issues on Privatized Prisons (Washington, DC:
Bureau of Justice Assistance, 2001) 4.


Stephen Pounds, ‘Florida-Based Prison Operator Wackenhut Corrections Makes Changes to
Expand’, Palm Beach Post, 24 August 2003, A3.


Simon Domberger and Paul Jensen, ‘Contracting Out by the Public Sector: Theory, Evidence,
Prospects’, Oxford Review of Economic Policy, Winter 1997, 67-8.


Tenn. Code Ann. § 41-24-104 (2002 Supp)

apparent that a different provider (either public or private) would be preferable. All
privatization contracts that represent the possibility of significant infringements upon
human rights should contain a provision allowing the state to cancel the contract if it
believes that human rights abuses may be occurring at a facility without fearing that
the private provider might be able to hold the state liable for breach of contract.
Clearly, prison guards have an important enough role to qualify; and it can be argued
that medical providers do, as well. By contrast, janitorial and secretarial services
would fit within the category of services that may be fully privatized.
Another important factor in retaining government responsibility for human
rights abuses is the political accountability of the entity that actually makes the
contract to privatize, and what involvement other actors have in this process. To the
extent that contracts become immutable, often even to later legislatures, and to the
extent that, as mentioned above, states for the most part stick to contractual issues
when policing private prisons, it is important that the participation of the public and
the public’s representatives be maximized as early in the process as possible.
Tennessee’s statute provides a complex contract approval procedure involving several
individuals and entities, but makes no provisions for the input of the general public.
Any contract must be approved by the state building commission, the attorney
general, and the commissioner of correction. Additionally, all proposals are reviewed
by two congressional committees, which can make comments to those responsible for
approving contracts before such approval takes place. All approved and proposed
contracts are sent to the state and local government committees of both the senate and


the house. 40 While this procedure involves various members of the legislative and
executive branches, it does not provide direct opportunities for the public in general to
affect these officials’ decisions. Nevertheless, at least the privatization procedure
involves members of both the executive and legislative branches, and includes
legislators themselves, which are often very accessible to public input. In Idaho, the
decision to enter into a contract with a private prison provider is left solely to the state
board of correction; 41 most of the public would not know whom to contact to affect
privatization contracts, or whom to hold accountable for the decisions of the board.
While several privatization statutes, such as Tennessee’s, provide for some
participation from the legislative branch in the contracting phase, few suggest any
method for direct involvement from the public. Montana, one of the only states to call
specifically for a public hearing, does so in a statute covering all forms of government
privatization. Montana requires an agency to form a privatization plan before any
programme can be privatized. Additionally, the state provides that:
<EXT>‘The privatization plan must be released to the public and any affected
employee organizations and must be submitted to the legislative audit
committee at least 90 days prior to the proposed implementation date. At least
60 days prior to the proposed implementation date, the legislative audit
committee shall conduct a public hearing on the proposed privatization plan at
which public comments and testimony must be received. At least 15 days prior
to the proposed implementation date, the legislative audit committee shall




Idaho Code § 20-241A (Michie 1997).

release to the public a summary of the results of the hearing, including any
recommendations of the committee relating to the proposed privatization
plan’. 42 <EXT>
<T1>Florida requires ‘public sessions of the Correctional Privatization Commission
to be held at which contract variations are [not only] discussed [they are also]
explained.’ 43 Public hearings produce little benefit, however, if the public is not
provided with adequate information with which to make informed suggestions.
Kentucky law requires the production of information necessary for the public to make
informed decisions about the quality and value of privatized services:
<EXT>‘The private provider shall develop and implement a plan for the
dissemination of information about the adult correctional facility to the public,
government agencies, and the media. The plan shall be made available to all
persons. All documents and records, except financial records, maintained by
the private provider shall be deemed public records’. 44 <EXT>
<T1>Kentucky does not rely solely on voluntary disclosure by the provider to amass
information on the functioning of privatized prisons. The legislature has also required
<EXT>‘[t]he department shall annually conduct a performance evaluation of any
adult correctional facility for which a private provider has contracted to operate. The


Mont Stat Ann § 2-8-302 (2002).


Harding, ‘Private Prisons’, n 27 above, 308.


Ky Rev Stat Ann § 197.510(7) (Banks-Baldwin 1998).

department shall make a written report of its findings and submit this report along
with any recommendations to the private provider and the Legislative Research
Commission.’ 45 <EXT>
<T1>The prison privatization provisions create a large amount of data, and attempt to
transmit most of the data to the public. However, they do not provide any mechanism
for the public to participate in the privatization decision, they do not limit the length
of any contract, and they do not contain many concrete requirements for the actual
running of the prisons, all of which would be necessary for meaningful accountability.
Most US prison privatization statutes require certification by the American
Correctional Association (ACA). As Richard Harding notes, however, ‘[i]mportant as
[ACA] standards are in maintaining a level of accountability in US corrections, they
are primarily processual … and formulaic. Practical, on-the-ground compliance or
breach is seldom clear-cut, yet clarity and predictability are crucial for
accountability.’ 46 Requiring certification, therefore, cannot absolve a state from
responsibility for prison services touching on basic human rights such as prisoner
safety. 47 Britain and Australia do not require accreditation; the British system instead
relies upon government prison inspections:
<EXT>‘[t]he primary source of information, in contrast to ACA accreditation


Ibid. § 197.515 (Banks-Baldwin 1998).


Harding, ‘Private Prisons’, n 27 above, 301.


Privatization is not lawless; it is pluralistic and polycentric, involving many sources and kinds
of law as well as other norms. Democratic safeguards should be set throughout the process,
but so far this is not consistently done.

processes [where the ACA spends most of its time during announced visits reviewing
a prison’s written procedures], is direct observation, discussions with prisoners and
staff, participation in some programs, follow-up interrogation of management. All this
is fortified with detailed scrutiny of documentation and records.’ 48 <EXT>
<T1>By using government inspectors, Britain also maintains direct public oversight
of prisoner conditions, a necessary element in preserving prisoner rights.
Additionally, as Angelina Fisher notes in chapter three in this volume, Britain—as
well as Australia, Canada, and New Zealand—grant prisoners access to an
ombudsman to whom prisoners can complain about violations of human rights by
private prison providers; no such system is found in most US systems. 49
Statutes that clearly specify the required elements of any prison privatization
contract are preferable to those that establish few (or no) concrete requirements of the
private entity in such contracts. In contrast to contract provisions, statutory language
is readily accessible to the public; almost all of a state’s statutory law can be located
free-of-charge through the state’s homepage. 50 Additionally, even if the public cannot
comment, legislative sessions are often open for the attendance of interested
individuals (the Colorado General Assembly website allows the public to listen to the


Harding, ‘Private Prisons’, n 27 above, 318.


Ibid at 319.










<> and Arizona @ Your Service, available at

proceedings in the state house and senate, even including committee meetings). 51
Furthermore, incorporating contract terms into a statute allows interested groups to
focus their efforts on the design of every privatization contract in the state. Trying to
influence each contract individually (even where a group is notified before the
finalization of a contract) might often prove to be too taxing on an interested group’s
or individual’s resources. Finally, legislatures are directly accountable to the public
through elections, while entities such as state corrections boards are not.
In short, there are several factors involved in assessing the ‘inherently
governmental’ nature of an activity. Some differentiating factors are the monopolistic
or oligopolistic character of the private market, the level of government or private
subsidy, and the extent to which there is a need for the government to retain the
ability to step in. The more monopolistic the market, and the greater the other factors,
the more likely it is that we are dealing with a public function that has the quality
inherent to government. In such cases, the conventions of legitimacy demand direct
participation or at least a clearly defined democratic process open to all affected
parties or their representatives. The next section explores this aspect more fully.

<H1>The limits of privatization: Scope for a new administrative law?<H1>

<H2>The previous section considered legislative checks on privatization, reading
them for their implicit tests of inherent government functions. We reviewed


See <>

legislative approaches to the cost criteria normally associated with privatization as a
basis for suggesting some specific democracy and human rights checks on
privatization in the prison context. In this section we will first look at the way courts
have responded to these issues. A review of judicial attempts to sustain democracy
considerations in privatized government functions is also productive, yielding a
number of specific considerations for the question of inherent government functions.
We will then turn consideration to the kinds of procedural reforms that are necessary
to ensure both that these various criteria for determining what is and what is not
‘inherently governmental’ are considered and that key stakeholders in the outcomes
have a chance to participate.

<H2>Judicial responses: The outer limits of privatization and prisons<H2>
<T1>One of the inherent human rights problems created by the privatization of
prisons is the blurred line between inmate discipline and the administering of
additional prisoner punishment. Obviously, in order to run a prison effectively, the
private provider must be able to assert some form of authority without clearance from
a public supervisor. There is a point, however, where the discipline exacted by a
private prison crosses over into punishment and constitutional due process concerns
require public oversight.
In the United States, many ‘disciplinary functions for breach of prison rules
are carried out directly by the private operator.’ 52 For example, in most US
jurisdictions inmates can receive time off their prison sentences for ‘good time’


Harding, ‘Private Prisons’, n 27 above, 275

served; in other words a prisoner’s sentence is reduced for behaving well behind bars.
To maintain discipline, these good time credits can also be revoked, which increases
the length of incarceration. This revocation of ‘good time’ amounts to a deprivation of
liberty significant enough to require an action by the state. As a result, prisoners are
entitled to a disciplinary hearing when they lose good time credits.
When disciplinary hearings occur, public prisons often have an incentive to
credit prisoners for ‘good time’ because of prison overcrowding. Private prisons,
however, can have the opposite incentive, as they are usually paid according to their
daily occupancy. In addition, guards at CCA facilities can have a personal financial
interest in maintaining high occupancy, because they are sometimes stockholders of
the company. Such incentives have produced obvious results. ‘The New Mexico
Corrections Department found that inmates at the [state’s] CCA facility lost “good
time” eight times more frequently than prisoners at the state institution.’ Furthermore,
in Tennessee, ‘CCA guards say privately that they are encouraged to send balky
inmates to administrative segregation; by placing prisoners in the “hole”,’ which
increases the prisoner’s sentence by 30 days, earning the company an extra $1,000. 53
In private prison settings, when private parties with a financial interest in the
outcome conduct the hearings, there is an increased danger of an unconstitutional
deprivation of due process rights. 54 As a result, the private contractor should not be


Michael Welch, Punishment in America: Social Control and the Ironies of Imprisonment
(Thousand Oaks, CA: Sage Publications 1999) 291.


See Kenneth L Avio, ‘On Private Prisons: An Economic Analysis of the Model Contract and
Model Statute for Private Incarceration’, New England Journal on Criminal and Civil

able to adjudicate the prisoners’ liberty rights when its profits are directly impacted by
the decisions made during these disciplinary hearings.55 To address this problem,
‘[s]tate courts generally invalidate statutes and administrative regulations that
delegate adjudicative power to private parties when there is no provision for judicial
review of the private adjudications.’ 56
In the federal prison system, for example, Congress has entrusted the power to
supervise the discipline of federal prisoners to the Bureau of Prisons (‘BOP’) under
the Attorney General’s Discretion. 57 The BOP often contracts with private
corporations to provide incarceration for federal inmates. The courts have upheld the
BOP’s delegation of authority to discipline inmates to the private contractors,
including the disallowance of good time credit, because the BOP provides a final
layer of de novo review of any disciplinary actions. Similarly, Britain retains state
control over disciplinary matters, and in private prisons, ‘disciplinary charges laid by
custodial officers are adjudicated by … public sector officials … who work on-site’. 58
In the United States, it is now common for inmates to be shipped to private

Confinement, vol 17 (1992) 282.

See Warren L Ratliff, ‘The Due Process Failure of America’s Prison Privatization Statutes’,
Seton Hall Legislative Journal, vol 21(1997) 402 (noting that ‘prisoner’s liberty interests …
conflict directly with private prison contractors’ profit motives’).


See Ira P Robbins, ‘The Legal Dimensions of Private Incarceration’, 38 American University
Law Review, vol 38 (1989) 569.


See 18 USC § 4042(a).


Harding, ‘Private Prisons’, n 27 above, 276.

prisons in other states, in an effort to find the lowest cost provider. As Harding points
out, this situation ‘stretch[es] the chain of accountability beyond the breaking point’
since the prisoner’s state of origin has no oversight within the private prisons of
another state. 59 These transfers are often to distant states, increasing the
‘accountability deficit’ created by these arrangements, distancing prisoners from their
community representatives and (as Harding also points out) from their families. 60 For
example, two of the largest prisoner-exporting states are the only two not within the
contiguous forty-eight: Hawaii sends prisoners to Minnesota and Alaska sends
prisoners to Arizona. 61 Furthermore, such exporting of inmates may deter inmate
rehabilitation efforts, as ‘[s]everal recidivism studies have found that convicts who
keep in touch with family members through visits and phone privileges are less likely
to violate their parole or commit new offenses.’ 62 States often fail to consider all of
the possible scenarios in which the switch to private prisons affects the status of the
prisoners. For instance, when Oregon sex offenders housed in a Texas prison escaped
in 1998 (Texas officials were not even informed that they were housing these
prisoners), Texas officials ‘could not charge them with escape because in Texas it was
not yet a crime to flee a private corporation’. CCA stated that it ‘had no legal


Ibid. at 280.


Ibid. at 290.


Ibid. at 280


Moore, Solomon, ‘States Export Their Inmates as Prisons Fill’, New York Times, 31 July




obligation to inform city or county officials’ of the transportation of sex offenders to
their jurisdiction. 63 This lack of information on the part of host states arguably
violates a right to information on the part of citizens of the host state, who deserve to
know potential risks they might be exposed to because of private inmates. While such
a lack of information might occur in prisons housing in-state offenders, the risks are
greatly reduced.
While every state has seen a dramatic increase in incarceration levels,
Arizona’s prison population has had increases of almost 1,000 percent in the last 25
years. 64 The state is also facing a $1.3 billion budget shortfall, leading to massive jail
overcrowding. 65 In the face of budget shortfalls, legislators are reluctant to spend
money on prison construction, and this reluctance translates into a de facto ceding of
ever-greater proportions of inmate housing to private facilities. One issue of growing
national concern—but of greater effect in Arizona—is the housing of increasing
numbers of foreign national prisoners. About 10 percent of Arizona’s prisoners are
Mexican nationals, and state officials have estimated that housing all prisoners in
Mexico could possibly save over $8,000 per inmate each year. 66 Arizona’s Senate last
year rejected a bill that ‘would have required the state to seek proposals from private
prison operators to build and operate a prison … within the neighboring Mexican state


Welch, Punishment in America, n 53 above, 289.


Robert Nelson, ‘Big House Inc.’, Phoenix New Times, 3 April 2003.


Robert Nelson, ‘Clink!’, Phoenix New Times, 23 October 2003.


Howard LaFranchi, ‘Arizona Plans to Export Mexican Inmates’, Christian Science Monitor,
17 June 1997, 8.

of Sonora.’ 67 This issue has been discussed in Arizona for several years, however, and
it is likely that it or another state with high numbers of Mexican nationals (such as
New Mexico) will experiment with a cross-border facility in the future.
Even more worrying, from the standpoint of accountability, is the emerging
trend in the United States of ‘bed brokering’ where private companies (such as Inmate
Placement Services) find a prison bed in another state to house prisoners who the
home state cannot place. 68 With private prisons responsible for small numbers of
inmates from numerous states, there is a danger that no individual state will have
enough incentive to properly oversee human rights compliance, especially where the
housing state does not send any of its own prisoners to the facility. Consequently,
states should be required to house, and therefore retain responsibility for, their own
This observation may point to yet another criterion of ‘inherently
governmental’ functions: ethical limitations on lobbying. A state should prohibit
private providers who either house inmates in the state or accept inmates from the
state within facilities found in other states to lobby the state on criminal sentencing
laws. It is true that private corporations can be more cost-sensitive than public entities
because of their accountability to shareholders. There is no evidence that private
providers attempt to either influence sentencing decisions or statute drafting in order
to increase the number of prisoners, thereby increasing their potential ‘clients’.


Paul Davenport, ‘Senate Panel Rejects Proposal to House Prisoners in Mexico’, Associated
Press, 7 March 2003.


Harding, ‘Private Prisons’, n 27 above, 281.

However, as private prisons begin to saturate markets and deepen their relationships
with state legislators, prudent drafting should require that all prison privatization
statutes make clear that contracting prisons are prohibited from lobbying the
legislature on criminal-sentencing matters. Richard Harding has correctly stated the
danger that ‘regulatory mechanisms in relation to private prisons are more susceptible
to capture—that is, a situation where “regulators come to be more concerned to serve
the interests of the industry with which they are in regular contact than the more
remote and abstract public interest” — than in relation to other, more strictly
commercial activities’. 69 It is consequently more important for the legislature to
circumscribe the administering agency’s discretion than in other privatization

<H2>A need for new approaches to administrative law<H2>
<T1>There are seldom any clearly defined lines separating activities that are or are
not ‘inherently governmental’. It is with the actual contract and contracting process
that the most important reforms are most urgently needed. It is here that a new
conception of administrative law is important, one that is not solely state-centric in its
application to decisions with broad policy effects, but relevant to the public/private
decision making that occurs in various privatized settings. Once an agency decides to
contract out its primary functions, the proposed contract should be noticed to the
public on the agency web site as if it were a rule promulgated for public comment.
The public should have a chance to comment on the goals of the contract, its mode of


Harding, ‘Private Prisons’, n 27 above, 306.

enforcement, the monitoring of its implementation (including what shall constitute
monitoring), and all other issues deemed relevant. As with a rule in a regulatory
proceeding, the agency need not adopt all or any of the suggestions made, but it
should provide its own reasons for accepting the ultimate contract.
An important role for administrative procedure is to accommodate most if not
all of these interests with a process that allows them to speak to one another as well as
the ultimate decision-maker. Once a contract is entered into, it is also important that
these discussions occur with some frequency. The nature of the enterprise requires
ongoing monitoring of the contract terms, as well as opportunities to comment on
implementation, and provision for amendments regarding the duties of the private
actor. Procedurally speaking, the privatizing agency should be willing to treat the
proposed contract more like a rule than a contract negotiated between two parties. It
could be put up on the prison authority’s web site, calling for public comments,
suggestions, alternative language and ways to achieve its substantive reform goals
from whoever wishes to comment. In our extended example, this would include
prisoners and their representatives as well. The agency should also provide extensive
information on the track records of firms competing for the contract. Finally,
government regulators should ensure fair competition among the bidders. All of them
should agree that if they are chosen, they will be subject to regular reporting
requirements and a modified Freedom of Information Act allowing interested
members of the public to make relevant inquiries about their operation while the
contract is in place. That contract should be no more than three years in length,
subject to renewal but only after another round of competitive bidding occurs.
The simplicity of notice and comment procedures when it comes to public
service contracts makes transparency reasonably efficient, and transparency need not

impose undue impediments to the bargaining process. A presumption in favour of the
bargains struck in such contracts can be written into the governing statutes. Courts
need not be involved unless there is corruption or an unconstitutional exercise of
discretion. Indeed, the purpose of these citizen-oriented procedures is to ensure that
the many views and voices involved in such public-regarding private arrangements
are heard. It is not just that there is a public dimension involved; it is that there are
genuine public values at stake that necessitate debate and contest. The various
positions are different formulations of democracy—as inherent in the operations of
the market, or external to the market as a larger framework of critique and reform. 70


<T1>Now we are in a position to consider whether privatization—as illuminated by
the prison context—should have limits, and, if so, how they might best be enforced
through administrative law. There certainly appear to be constitutional limits on the
ability of private providers to adjudicate disputes involving prisoners without
involvement of courts; however, as we have seen, even these limits are relatively
easily met, constitutionally speaking, with a judicial review of these private decisions.
This is not an insignificant check on private decision-making power, but it does not


This article does not address global administrative law as such, but issues including the lack of
an independent judge in disciplinary proceedings, the export of prisoners based on cost and
not on social factors, and the use of private interrogators such as those used at Abu Ghraib,
raise human rights concerns that may have an international significance.

go far enough in terms of meeting the democracy and human rights concerns raised in
the prison context. The ability to engage in processes that enable important
stakeholders to have a say in the contracting process is also of great importance. In the
end, these are almost always political decisions that do not usually lend themselves to
a yes or no answer. Still, there are some outer limits to the ability of government to
outsource. Judicial review, for example, would be an insufficient basis for an account
as to why we should resist a public prison’s effort to outsource the execution of an
inmate sentenced to death. 71 Though a hypothetical scenario, it points to a line that
may not be legitimately crossed in reality—confirming the existence of limiting
conditions in relation to privatization in the prison context and beyond, in addition to
those set forth by courts and legislatures.
In the hypothetical scenario of outsourcing executions, even if there were—by
some stretch of the imagination—a competitive market involving legitimate firms
able to provide the so-called service, by definition those firms would have been
created wholly as an extension of the government’s monopoly on legal executions.
Let us leave aside—if we can—the moral reprehensibility of creating an industry
licensed to kill. More relevant for our purposes would be the reservation that it
separates the execution from the collective judgment claimed at sentencing. This
reservation sheds light on the question of limits under more ordinary circumstances—
in that while it may be possible to separate the management of prisons from
government to the extent that they involve institutions, buildings, workplaces, and


I wish to acknowledge and thank Simon Chesterman for raising this important question.


residences like others outside the prison context, they also entail elements that are
unique to the function of prisons and the consequent vulnerability of prisoners in their
physical status and in relation to their fundamental human dignity. If only the state
can take a life legally, then it follows that any function in which the physical body of
the prisoner is directly at stake—in nutrition, medical care, and the basic living
conditions inside the prison—are also ‘inherently governmental’ functions.
In practice, the question of what is and is not ‘inherently governmental’ is a
decision that must be made democratically, through an open political process that
meets the standards of transparency and accountability set in place by legislatures and
courts. For that reason, it is suggested that a new kind of administrative law can, and
should, be created to respond to the new forms of democracy deficit associated with
privatization. Since prisons involve such a wide range of services, the prison context
is useful for thinking through the promise and limits of privatization in relation to
‘inherent governmental’ functions. The pragmatics of privatization have emerged as
an important terrain where a new administrative law might emerge, assuring public
forums for input and debate and a flow of information that can help create meaningful
politics around private actors doing the public’s business. The democracy problem is,
and should, be one of the primary concerns of a new administrative law as we face the
costs and benefits of the privatization trend in contemporary governance.