RESEARCH IN

Corrections

Volume 2, Issue 1

February I989

Supported by the Robert J. Kutak Foundation and the National Institute of Corrections

THE COST OF CORRECTIONS: IN SEARCH OF
THE BOTTOM LINE
Douglas C. McDonald, Ph.D.
THE COST OF CORRECTIONAL SERVICES: EXPLORING
A POORLY CHARTED TERRAIN
Alan M. Schuman

Page I

Page 27

CORRECTIONS “PROPORTIONATE VALUE”
Chase Riveland

A

Page 35

series reporting the results of applied research in corrections for administrators and practitioners

EARLIER VOLUMES IN THIS SERIES

Statistical Prediction in corrections, by Todd Clear, Ph.D., Vol. 1, Issue 1, March
1988.
The Effects of Diet on Behavior: Implications for Criminology and Corrections, by
Diana Fishbein, Ph.D., and Susan Pease, Ph.D., Vol. 1, Issue 2, June 1988.
Pretrial Release: Concepts, Issues, and Strategies for Improvement, by Stevens H.
Clarke, Vol. 1, Issue 3, October 1988.

FOREWORD

In the economic climate of the 1980s any discussion of correctional program
choices rapidly focuses on the costs associated with sanctioning offenders in different ways. Both the increased cost of services and the problem of crowding
throughout the corrections system demand that administrators become more
adept at defining and analyzing costs. In this monograph, the fourth in the
Research in Corrections series, Douglas McDonald offers both substantive cost
information on prisons, jails, probation, and parole and a tutorial on the
shortcomings of popular methods of computing corrections costs.
Dr. McDonald provides important insights into why corrections expenditures
grew so rapidly in the decade between 1975 and 1985. He defines a framework
for understanding the real or full costs of correctional services and discusses the
widespread underreporting of costs.
In the practitioner responses, Alan Schuman, Director of the Social Service Division of the Superior Court of the District of Columbia, explores the benefits of
community corrections from a cost perspective and calls for remedies to the generally poor state of cost information on community corrections programs. Chase
Riveland, Secretary of Corrections, Washington State, places the question of an
improved understanding of costs in the perspective of broader public perceptions of the value of correctional programs.
It is our hope that the monograph will be useful to both corrections practitioners
and decisionmakers in the broad public policymaking arena.
Larry Solomon
Acting Director, National Institute of Corrections

EDITOR’S NOTE

No issue has brought criminal justice more to the forefront of public policy than
that of corrections costs. State spending for corrections throughout the nation
has grown by more than 50 percent since 1980-the greatest increase of any
state-funded service-and next year, Americans will spend close to $10 billion
to operate the nation’s prisons.
As the costs of corrections continue to spiral upward, most states are asking:
How much “punishment” can we afford? The answer depends on how much
various types of punishment cost. Most policy discussions of the costs of
imprisonment relative to those of other alternatives have been overly simplistic.
The annual operating cost of prisons (usually quoted to be about $14,000 per
inmate) is generally compared with the annual cost of probation (about $2,000
per probationer). Comparing these two figures fuels the popular notion that
probation is far cheaper than prison.
But as Douglas McDonald makes painfully clear in this comprehensive monograph, reaching the “bottom line” on corrections costs is much more complicated.
The annual operating cost of a prison does not reflect capital costs, fringe benefits and pensions, and other necessary expenditures; adding in these items more
than doubles the annual cost. Other sanctions, such as jail, probation, and
parole, involve cost complications as well.
Dr. McDonald, who has written extensively in criminal justice, reviews current
knowledge on corrections costs, pointing out the critical gaps. His review
should be of particular interest to budget analysts, state legislators, and corrections practitioners who are concerned with the cost implications of corrections
policy.
Chase Riveland and Alan Schuman, in their practitioner reviews, provide
another critical dimension to the cost issue. Both of these corrections officials
have had extensive experience in managing their own operational budgets, and
have been responsible for informing and educating those who fund correctional
services. Here, they review Dr. McDonald’s monograph to examine how his
findings relate to the day-to-day concerns of the corrections practitioner. Mr.
Riveland discusses the ‘proportionate value” of different corrections services and
considers what aspects of corrections seem worth paying for from the public’s
point of view. Mr. Schuman expands the cost discussion to include community
corrections and discusses the impact of litigation on prison operating costs.

The next issue of Research in Corrections will present ‘Intensive Supervision in
Probation and Parole,’ by James Byrne, Arthur Lurigio, and Christopher Baird.
Complimentary copies of this monograph and earlier volumes in the Research in
Corrections series can be obtained by writing to the National Institute of Corrections’ National Information Center, 1790 30th Street, Suite 130, Boulder, CO
80301.
Persons wishing further information on the series are asked to write to Joan
Petersilia, The RAND Corporation, 1700 Main Street, Santa Monica, CA 904062138.
Joan Petersilia

THE COST OF CORRECTIONS: IN SEARCH OF
THE BOTTOM LINE
Douglas

C. McDonald’

INTRODUCTION

Several years ago, in testimony before a state legislative committee that was
considering a change in sentencing law, I referred in passing to the $15,000
per-inmate annual cost of operating the state’s prison system. The commissioner
of the state’s prisons followed me to the microphone and disputed my figures:
“Don’t believe everything researchers tell you. I’ve just calculated that the
annual cost is more like $11,500. That has to be right; just divide my budget by
the average daily population, and that’s what you get.” Meeting him in the corridor afterwards, I asked about pension and fringe benefit costs. “Oh yes, you’re
right,” he said. “I forgot about them. They’re not my costs.”
The idea of “my costs” reflects a much broader problem in public accounting,
one which complicates enormously the task of determining what correctional
services-and other public services as well-actually cost. The task of the public manager is, in part, to administer appropriated funds. Consistent with this
emphasis on budget administration, public accounting methods were designed
primarily to track dollars so that fraudulent use of public funds could be
prevented or detected. Cost analysis has taken a back seat to funds control, and
the way expenditures are counted and reported reflects this. For example, the
unit of analysis is the agency rather than the service performed. Where different services are provided by a single agency, the costs of these services are
often not disentangled. Where different agencies contribute to the same governmental service, public accounts generally fail to consolidate the costs of that service. Rather than separate capital from operating expenses, many reporting systems conflate the two or draw the line dividing them in different ways. Because
determining the actual cost of a service is not a central preoccupation of government, rules about how debt should be capitalized, or depreciation calculated,
have not been adopted. The result is widespread variation in how costs are
computed (when somebody actually makes an attempt to compute them) and a
pervasive underreporting of costs.
In the environment in which correctional managers now work, it is important to
have a clear and precise understanding of the fiscal consequences of different
choices. During the decade between 1975 and 1985, the cost of operating
correctional institutions in the United States rose by nearly 240 percent, and all
levels of government spent large sums-a total of about $10.5 billion-to
1
Several people provided helpful reviews, assistance, and unpublished data. I am especially
grateful to Sue Lindgren of the Bureau of Justice Statistics, John Curry of the Bureau of the Census’
Government Finance Division, and Joan Jacoby of the Jefferson Institute.

2

DOUGLAS C. MCDONALD

expand their correctional systems. 2 Since then, growth has remained strong,
and prospects appear quite bleak for a significant reduction in spending in the
foreseeable future unless dramatic changes are made in sentencing policies. Rising faster than the gross national product, spending for corrections will have to
be fueled either by raising tax revenues or by spending less for other government programs, a painful choice for political leaders and taxpayers alike.
Even if correctional managers were faced with a zero-growth future, just bringing existing facilities up to standards is and will continue to be extremely costly.
As of October 1987, forty-five states, the District of Columbia, Puerto Rico, the
Virgin Islands, and an undetermined number of county and municipal governments were under court orders to remedy conditions in their prisons and jails.3
In mid-1984 (according to the most recent survey available), 46 percent of all
jails holding at least 100 inmates-which, taken as a group, held about 74 per-.
cent of all the nation’s jail inmates at that time-were under court orders to
reduce overcrowding or improve their recreational and/or medical facilities and
services. In most, if not all, instances, the courts are making demands that cost
money to satisfy. In one dramatic case, Judge Frank M. Johnson ordered, in
James v. Wallace, that the Alabama prison system be improved. This order
resulted in a corrections budget in 1976 of $55 million, $35 million of which
was said to be caused directly by the judge’s decision4
This essay presents a discussion of correctional costs, how they are conceived or
misconceived, calculated and miscalculated, and how they behave-that is, why
they vary from place to place and over time, why they have increased so
dramatically in so many jurisdictions, and how they differ among the various
types of correctional services. Because my expertise is primarily in elucidating
the implications of different policy choices, and because I have little training in
professional accounting and none in administering correctional facilities, I have
steered almost clear of a review of accounting rules. More detailed and competent discussions of them can be found elsewhere. I have also chosen to avoid
the topic of cost-inefficiencies, waste, and how to control them. My aim is,
rather, to explore the gross costs of correctional services, including per-capita
expenditures, to indicate the problematic nature of the data commonly used and
published, and to explore how these costs might be estimated more accurately. I
shall present some information that has not been published elsewhere and indicate where I see softness in the published data. My assumption is that the kind
of information discussed may be of interest to decisionmakers in the correctional
policymaking arena, and not just in correctional administration.
I have not been able to avoid a tension in this discussion between presenting
substantive information and discussing methodological matters. This paper is
intended to be part sourcebook and part tutorial. Some readers may find the
result maddening. A straightforward presentation of costs and cost-related
1

See Table 1 on p. 13.
Status Report: State Prisons and the Courts,’ The National Prison Project Journal, ACLU
National Prison Project, 13, Fall 1987, p. 24.
‘Reported in Charles Prigmore and Richard T. Crow, ‘Is the Court Remaking the American
Prison System?’ Federal Probation, 40, June 1976, p. 8.
2

THE COST OF CORRECTIONS. IN SEARCH OF THE BOTTOM LINE

3

issues, with only a few qualifications and quibbles, would perhaps make easier
and more satisfying reading. However, such a discussion would reinforce the
notion I want to challenge, namely, that the data we commonly use for costs in
corrections are misleading. The corrective for this is to take at least a short journey into a more abstract discussion of how costs should be conceived of and
computed. In the following, I have tried to balance the discussion somewhere
between the two poles of abstract methodology and substantive cost information.
This is not intended as a review of the entire field or all the literature on the
subject. Instead, it is a selective discussion of points that I hope have general
significance.

ESTIMATING THE COST OF CORRECTIONAL PROGRAMS
Prisons. According to the Census Bureau’s survey of audited expenditures, state
governments spent between $4.28 and $4.80 billion for their correctional systems during 1985.5 This amounted to an annual average operating expenditure
6
of $9,550 to $10,700 per inmate. During the same year, the federal government spent roughly $501 million to operate its adult prison system, or $12,500
per inmate.7 According to other widely used sources, however, the expenditure
was substantially higher. The American Correctional Association’s annual survey reports that state and federal budgets for operating the nation’s prison systems totaled $7.9 billion in that same year.’ Translated into per-capita terms,
the average annual budgeted amount was $16,400 for each inmate. The Corrections Yearbook computed from its survey of agency budgets that operating costs
averaged $14,591 per inmate that year. The highest estimate was more than
half again as much as the lowest of the three.

The actual cost was probably higher than all three, largely because we can be
assured that some categories of direct costs were not picked up in any of the
surveys. Given the existing data, however, it is difficult to estimate how much
higher the real expenditure was. The Census Bureau’s figures are probably the
best estimates of base operating costs, simply because they were derived from
audited expenditures rather than budgets and because Census Bureau examiners
followed uniform rules to distinguish different types of costs. They also
5
State figures represent aggregate expenditures for institutional operations, reported in Government Finances in 1984-1985, Washington, D.C.: U.S. Department of Justice, 1985, Table 47. The
total of $527 million reported spent by state governments for institutional operations is not separated
into adult and juvenile corrections, but is categorized as ‘other and combined’; hence, a range is
represented, rather than a single figure.
6
Average number of prisoners from Bureau of Justice Statistics, Prisoners in 1985, Washington,
D.C.: U.S. Department of Justice, 1986, Table 2.
7
Annual budget reported in George M. Camp and Camille Graham Camp, The Corrections Yearbook, South Salem, N.Y.: Criminal Justice Institute, 1986, p. 23. Average daily population is
reported in Bureau of Justice Statistics, Prisoners in 1985, Washington, D.C.: U.S. Department of Justice, 1986, Table 2.
8
American Correctional Association, 1986 Directory: Juvenile and Adult Correctional Departments,
Institutions, Agencies, and Paroling Authorities, College Park, Md., 1986, pp. x, xi.

4

DOUGLAS C MCDONALD

followed the same rules in deciding what to include and what to exclude. But
because there is wide variation from state to state in activities and costs that
agency budgets cover, and because the annual budget surveys don’t tell us what
is or is not in them, it is difficult to know how much confidence to place in
these figures.
Jails. Reliable estimates for the costs of the nation’s 3,300 local jails are even

harder to come by. The Census Bureau’s survey of government finances
estimated that county and municipal governments spent a total of $3 billion in
fiscal year 1985 to operate jails and detention centers.9 This amount included
both adult and juvenile institutions, but extrapolating from limited data, we can
estimate that county and local expenditures for adult jails alone totaled roughly
$2.3 billion that year. 1 0 On a per-prisoner basis, this averaged $8,679 per year,
or $23.78 per day.11 Again, some cost elements were systematically excluded
from these figures (principally, pension fund contributions), so the costs are low
by some undetermined percentage.
Few other surveys of jail costs exist. One undertaken by the Criminal Justice
Institute for its Corrections Yearbook polled all jails in the United States holding
more than 250 prisoners in 1985. Four dozen responded, about one-fifth of all
12
These 48 jurisdictions reported spending an
jails meeting the size criterion.
annual average of $13,803 per prisoner, or $37.82 per day. Costs ranged from a
low of $1,080 per year in Alabama’s Jefferson County jail to a high of $40,077
per year in New York City’s adolescent detention center. Again, it is difficult to
determine how much confidence should be placed in the accuracy of these figures, absent a description of the accounting methods used by each county to calculate rates. Moreover, the selected sample consisted of larger jails and may not
necessarily be representative of the entire set of adult jails in the country. The
bottom line is that none of the estimates of average jail costs seems strong
enough to stand on.
Probation and Parole. Very few data are available on nationwide expenditures

for probation and parole. The Bureau of Justice Statistics (BJS) reports that the
total direct expenditure during fiscal year 1985 for both activities at the ‘state
and local, but not federal, levels amounted to $1.5 billion. Of this total, all but
$10.4 million went for agency operations; the remainder represented capital

9
Bureau of Justice Statistics, ‘Justice Expenditure and Employment Final Report, 1985, unpublished document, Table 42.
10
According to the same Bureau of Justice Statistics report, 72 large county governments sampled
by the Bureau of Census reported that 77.5 percent of their expenditures for institutional corrections
went to adult facilities. Applying that proportion to the total combined expenditure for adults and
juveniles in both county and municipal jails yields $2.3 billion for adults alone. (Ibid., Tables 42 and
50.)
11
At midyear 1985, there were approximately 265,000 persons in U.S. jails (Bureau of Justice
Statistics, Jail Inmates, 1985, Washington, D.C.: U.S. Department of Justice, 1987, Table 1.
12
Camp and Camp, The Corrections Yearbook, 1985, pp. 67-68. The sample reported here
represents about half of all jails having more than 250 persons, but Charles Dewitt counts 214 jails
of this size in 1983. See Dewitt, New Construction Methods for Correctional Facilities, Washington,
D.C.: U.S. Department of Justice, 1986, Fig. D.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

5

expenses13 No information is given for probation activities alone, or parole
separately. Again, these data underestimate the actual operating costs because
they, like all data developed by the Bureau of the Census for its survey of
government finances, omit certain categories of fringe benefit and retirement
fund contributions. These and other local sources also tend to overlook certain
categories of direct expenditures made for either probation or parole, because
those expenditures are picked up in other departments’ budgets.
The Criminal Justice Institute, in its Corrections Yearbook (1985), reports that its
national survey of probation agencies found the estimated cost of supervising
one probationer for a year to range from $156 in Connecticut to $1,500 in the
federal system. The average cost for the 44 agencies reporting was $584.14
Because the Yearbook provides no description of the methodology followed to
compute these costs, we cannot assess their accuracy. Probation is a local
responsibility in many states, and it is likely that the state-level agencies that
reported the costs to the authors of the Yearbook passed on estimates of varying
quality obtained from local departments.15
Probation costs vary in relatively direct proportion to the caseload size-that is,
the number of probationers supervised by a single probation officer-and to the
salary levels of officers. Various forms of beefed-up probation supervision programs (e.g., intensive-supervision probation) usually involve smaller caseloads,
which come at a higher per-capita cost. The annual cost of intensive supervision in Georgia during 1982, for example, was reported to be $1,595 per probationer, compared with $275 for regular supervision during the same period.16
To the best of my knowledge, no national survey of parole supervision costs
exists. The cost is often difficult to identify, because in many jurisdictions
parole supervision costs are not separated from the larger agency costs, which
often include those of the parole board and, in many jurisdictions, probation
services as well. To give some idea of the magnitude of the costs, a study of
New York expenditures in 1978 estimated that parole supervision cost about 7
percent of the average per-capita cost of incarceration in the state’s prisons.”
The Need for Special Studies. An extensive list could be assembled here of the

various estimates for different forms of correctional programs in different jurisdictions. But it is more useful, I think, for the reader to understand that the vast
majority of estimates used to represent total costs are of questionable accuracy.
13

Justice Expenditure and Employment Final Report, 1985, Table 43.
Camp and Camp, The Corrections Yearbook, 1985, pp. 74-75.
15
Determining the cost of probation supervision is difficult because of all the general problems
with public accounting systems and their reports, but also for reasons peculiar to the structure of
many probation departments. Probation agencies are multiservice agencies, providing-in addition
to correctional supervision-services to adult courts in the form of presentence investigations, operating intake in the family courts, and other activities. If separate subdivisions do not exist to perform
each of these activities, and if the accounting systems do track these functional areas separately, singling out the cost of probation supervision is quite difficult.
“Billie S. Erwin, Evaluation of Intensive Probation Supervision in Georgia, Atlanta: Georgia Department of Offender Rehabilitation, 1984.
“Douglas McDonald, The Price of Punishment: Public Spending for Corrections in New York,
Boulder, Colo.: Westview Press, 1980, pp. 62-63.
14

6

DOUGLAS C. MCDONALD

Most are derived from agency expenditures-or worse, from budgets-and consequently they ignore several important categories of costs. What is needed is
better agreement and understanding about what should be counted as costs.
Because the cost of a service is so dependent upon how that service is delivered
in each jurisdiction, and because there is such variation among jurisdictions in
how costs are shared among agencies, special studies must be undertaken if
accurate estimates of costs are to be obtained.

WHAT SHOULD COUNT AS A COST?

The basic problem in correctional cost analysis is that many expenditures that
should be counted as costs of providing a particular service are overlooked. The
relationship of several broad classes of costs is illustrated in Figure 1.
Moving from right to left in Figure 1: The total cost of a particular correctional
service should logically include both direct costs, or those expenditures made by
an agency of government to provide the service in question, and indirect costs,
those costs borne by government or nongovernmental parties to support a particular correctional activity. Although these latter costs-the indirect governmental and indirect social costs-are real, their calculation is often difficult;
speculative, and controversial, and will not be discussed further in this paper.
Direct costs consist of both operating costs and capital costs. It is important to
distinguish between the two, because each provides benefits to different clients
and publics. The rule is that operating costs are incurred to provide a service to
the current client population, whereas expenditures for the expansion or replacement of capital stock benefit future generations as well.
Amortizing Capital Expenditures. The true cost of delivering a service in a
particular year should include operating costs incurred during that year and
some portion of the capital expenditures made that year. Determining precisely
what proportion of the capital cost should be incorporated into the current cost
estimate depends how long the capital resource will be used-how many years
the cost should be spread over. The simplest method is to spread the cost of the
facility evenly across all the years the facility provides useful service-the socalled ‘straight-line” method of allocation.18 Most analyses of correctional costs
fail to amortize capital spending and therefore understate the actual cost of
operations.
Financing. The costs of purchasing correctional assets, if financed by borrowing,

should be considered current operating costs, although they rarely are in public
accounts. The portion that should be assigned to the annual (or daily, or
18
For a clear discussion of the concept and the various methods, see Gail S. Funke and Billy L.
Wayson, "What Price Justice? A Handbook for the Analysis of Criminal Justice Costs." unpublished
document written for the National Institute of Justice-supported National Baseline Information Project of the Jefferson Institute and the Institute for Economic and Policy Studies, Washington, D.C..
pp. 64-69.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

Agency expenditure for
a correctional purpose
Expenditures by other
agencies for correctional
purpose

Operating costs

Expenditures by other
governmental accounts
for correctional purposes

Direct
government
costs

Repair and maintenance
Financing costs (interest
portion of debt service)
Cost of constructing/
renovating the physical
plant, including
replacement of equipment

Capital costs

Cost of services dedicated
to correctional purposes
by overhead governmental
organizations
Cost of public assistance of
families of persons jailed
or otherwise prevented from
supporting them

Total costs
Indirect
government
costs

Cost of not being able to
use correctional resources
for other purposes
(“opportunity costs”)
Losses/injuries due to
particular correctional
practice

Indirect
social costs

Impact on environment

Fig. l-Components of correctional costs

8

DOUGLAS C MCDONALD

whatever) operating cost is the amount of interest paid on the debt during the
time period measured.
Expenditures by Other Agencies. The direct costs other agencies incur in serving a correctional agency’s mission must also be counted as direct costs,
although they are typically overlooked because of the exclusive focus on the
correctional agency’s expenditures. For example, teachers in prisons and jails
are sometimes paid not by the correctional agency, but by local school districts.
Doctors and other medical workers might be paid by local or state departments
of public health and mental hygiene. In-hospital care is often charged to the
public hospital rather than to the correctional agency. Utility bills in correctional
facilities are sometimes paid by departments of public works. Departments of
transportation often provide the vehicles used to move prisoners. In some
instances, a local department may have a special program that is financed not
by the local government but by a state or federal government, and these costs
might not show up in the department’s expenditure reports. Failure to count
the costs of services provided by other agencies and other levels of government
may result in signficant undercounting of the total direct cost of a correctional
service.
Disentangling Costs Within an Agency. In instances where a single agency
performs a variety of different services, the costs of those services are often conflated with one another because the control of the agency’s funds is more
important than a precise determination of what each separate service costs. For
example, probation departments typically provide both a correctional servicethe supervision of offenders-and an investigative service to the court in support of the judges’ sentencing function. Likewise, local sheriffs’ departments are
often structured as multipurpose agencies. They operate jails, patrol highways,
and answer calls for service that do not involve law enforcement duties. When
public accounts do not track expenditures for these separate functions (indeed,
that is often difficult because the same people perform many different duties),
teasing out the cost of a discrete correctional service usually requires not only a
reanalysis of raw expenditure reports, but often new data collection as well.
Expenditures from Other Accounts: Fringe Benefits and Pensions. Finally,
some costs are paid not out of any department’s budget, but from general
government accounts. In many jurisdictions, fringe benefits and contributions to
employee retirement funds are not assigned to agency budgets but are
transferred from the general fund into a separate account for all government
employees as a group. Because personnel expenses represent such a large proportion of expenditures for correctional services, failure to count the added costs
of benefits and pension plans results in the loss of a major expense item.

Whereas determining the cost of fringe benefits is quite straightforward, putting
a true dollar figure on pension costs is much more difficult. This is due largely
to the special nature of these costs. Pensions are a form of additional compensation to employees which is deferred either until employees reach retirement
age or until they put in a specified number of years of service. As such, these
obligations constitute a debt to employees. Funds to pay off these debts come

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

9

from government contributions to retirement systems and earnings received on
investments of these contributions.
Government accountants use different methods for measuring the cost of retirement compensation, and inadequate financial reporting practices compound the
confusion.19 Some governments recognize the cost of pension benefits at the
moment when cash payments are paid out to the retired worker. Others recognize it when the liability for a future benefit is created-that is, when the worker
earns an increment of his or her future benefit payment. Determining the actual
cost of a government service is presently obscured by the first method of computing pension costs, because costs in the form of liabilities are not being
counted when they are created. (Lacking a method of matching liabilities to
fund contributions, pension systems so designed also tend to build up large liabilities that are not offset by regular government contributions. The unfunded
liabilities have to be paid, consequently, not by those who used the government
service, but by later generations of taxpayers.) The method preferred by the
standard-setting authorities involves making predictions about the size of those
future costs and contributing some portion of them in the present. Contributions can be made to the retirement fund in amounts that will ensure that an
employee’s retirement benefits will be paid out of the funds that were set aside
by the government for that purpose during his working lifetime.
The virtue of the latter method, from our point of view, is that it provides a way
of measuring a piece of correctional spending for current operations. If the estimates are accurate, one needs only to track down the amount contributed by the
government during any given year to the correctional staff person’s retirement.20
Whether or not these contributions can be taken as accurate indicators of actual
cost is thrown in doubt, however, by studies of current funding practices in
public pension systems. According to some studies, current contributions in
many places do not match the liabilities that employees are incurring for future
benefits. A 1975 study by the U.S. General Accounting Office of state and local
pension systems estimated that the unfunded liabilities of those systems-the
gap between what governments will have to pay and the assets that will be on
hand to cover them-was between $150 and $175 billion, about the same as the
entire outstanding credit market debt of state and local governments at that
time.21 This gloomy picture was reinforced by Frank Arnold’s 1983 survey of all
144 state-administered pension plans. He estimated that assets covered only
about 45 percent of total projected liabilities. The unfunded liability in these

19

For a discussion of the great variability and deficiencies in the reporting practices of public pension systems, see John E. Peterson, Public Pension System Financial Disclosure, Washington, D.C.:
Government Finance Research Center, 1983.
20
Public budgets and published expenditure audits rarely indicate the amounts contributed on
behalf of employees in specified agencies, but those who administer the pension funds can usually
compute the amounts.
21
Comptroller General of the United States, Funding of State and Local Government Pension Plans:
A National Problem, Washington, D.C.: General Accounting Office, August 1979.

10

DOUGLAS C McDonald

state plans alone was estimated to be $170 billion.22. (This exceeded by two and
a half times the total long-term state debt in 1978.23)
Uniformed employee pension systems, at least those for police and firefighters,
tend to be more fully funded than others, but it is not known, in the aggregate,
how much smaller the unfunded liability in these systems is or, correspondingly,
how much larger the actual cost of current corrections services is, properly computed. Studies of unfunded liabilities indicate that the cost could be substantially larger than is commonly perceived. It is safe to assume, consequently, that
pension contributions for correctional workers represent a low estimate of retirement costs.
Correctional pension contributions in New York give some idea of how large
these costs can be. During 1976-1977, contributions by state government to the
retirement fund for state prison employees equaled 23.6 percent of all salary and
wage payments-none of which came from the department’s budgeted funds.24
(This was in addition to expenditures for other fringe benefits, which added
another 12.7 percent.) During the same year, pension contributions for city jail
employees in New York City averaged 35 percent of all salary and wage payments, all of which were paid by a general government account. Some additional fringe benefits were paid out of that general account, the result being that
these extradepartmental costs equaled 46 percent of all salary payments that
year. 2 5 In fiscal year 1987, these costs accumulated at a rate that was only
slightly lower.26 Excluding pension contributions from estimates of correctional
spending in these jurisdictions would have undercounted greatly the actual
direct cost of those services.
The proportion of all pension and fringe costs covered by correctional department budgets, and how much larger the cost of operations would be if these
were fully included in jurisdictions where agency budgets exclude them, has not
been systematically determined. According to the American Correctional
Association’s Vital Statistics in Corrections, 1986, the average fringe and pension
rate of the 86 jurisdictions that responded to a survey was reported to be 24.14
percent of salary. Whether these costs were included in department budgets
was not indicated, nor do we know the method by which these rates were calculated, or whether the withholdings cover 100 percent of the liabilities.

22
State and Local Public Pension Funding: Theory, Evidence, and Implications, Ph.D. dissertation,
Harvard University, 1983. cited in Herman B. Leonard, Unchecked Balances: The Quiet Side of Public
Spending, New York: Basic Books. 1986. pp. 34-35. Based on his review of other evidence, Leonard
suspects that Arnold’s estimates may even be too optimistic.
23
Leonard, Unchecked Balances, p. 35.
24
McDonald, The Price of Punishment, note 12, Chap. 2.
25
Ibid., pp. 112-113.
26
The total pension cost and part of the fringe costs that were not included in the department’s
budget represented 46 percent of the uniformed officers’ salary payments, and 30.5 percent of the
civilians’. (Letter dated January 20, 1988, from Judith Keefer, City of New York Department of
Correction; the equivalent rates in fiscal year 1978 were 42.5 percent and 32 percent, respectively.)

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

11

HOW MUCH HIGHER ARE REAL COSTS LIKELY TO BE?

Several studies give an indication of the difference between true costs and what
correctional authorities perceive as the costs of jailing. Funke and Wayson
(1977) estimated that operating costs of local jails in the state of Washington
were underreported by 28 percent.27 A 1982 study by the Center for Justice
Planning, conducted for the National Institute of Corrections, computed costs
and found an identical underreporting of expenditures for 34 “constitutional”
jails in the figures correctional administrators provided.28 McDonald found that
about 36 percent of estimated total expenditure for New York City’s jail operations was paid not by the Department of Correction, but by other agencies and
other accounts, and that state prison costs were consequently underreported by
about 22 percent. (The degree of underreporting-and the estimated total
cost-would have been even higher had estimates of depreciation and the
interest portion of the debt service been included.)29 Roughly similar findings
were also obtained in other studies30
If audited and reported costs equal about 60 to 75 percent of total direct costs
across the nation (which is not unreasonable to expect), the actual cost of
operating public correctional programs is about 33 to 66 percent higher than is
usually reported.31

HOW MUCH AND WHY HAVE COSTS GROWN SO RAPIDLY?

Even though the periodic surveys of correctional costs provide incomplete information on total costs of operations, they do provide a good picture of how costs
have changed in recent years. The most reliable source is the Bureau of the
Census‘s annual survey of government finances, whose accounting methods
have been consistent from year to year, except for a change adopted in 1971.
Operating costs nationwide, exclusive of capital expenditures, increased 470 percent between 1971 (the first year when operating and capital costs were
reported separately in government reports) and 1985. Growth was fastest at the
federal level, where a 532 percent increase was posted during this period,
although spending in the federal system is only about l/20 of the expenditure
at state and local levels. Growth at the state level was also explosive: 519
27

G. S. Funke and B. L. Wayson, Comparative Costs of State and Local Facilities, Washington, D.C.:
Correctional Economics Center, 1977.
28
The Costs of Constitutional Jails, Washington, D.C.: National Institute of Corrections, 1982,
p. 24.
29
McDonald, The Price of Punishment, p. 112.
30
Coopers and Lybrand, The Cost of Incarceration in New York City, Hackensack, N.J.: National
Council on Crime and Delinquency, 1978; and G. P. Falkin et al., Revising Connecticut's Sentencing
Laws: An Impact Assessment, Alexandria, Va.: Institute for Economic and Policy Studies, 1981.
30
The use of these two percentages may seem contradictory, but if 75 percent of the total +F 1
cost of a program is covered by a department’s budget, the 25 percent not counted is equivalent to
33 percent of that budget. If 60 percent is budgeted, the uncounted 40 percent equals 66 percent of
it.

12

DOUGLAS C. MCDONALD

percent during this 14-year period. This is especially si gnificant because states
shouldered the major burden of correctional spending.32 Local spending for
correctional operations also increased substantially, by 385-percent (see Table 1).
The big boom, however, was in capital spending. Correctional agencies at all
levels of government scrambled to purchase, build, and rehabilitate facilities to
handle their swelling client populations. Between 1971 and 1985, state governments spent a total of $6.05 billion for capital outlays (not including the cost of
servicing the debt on borrowed capital). Local governments during this period
spent approximately $3.8 billion, most of it for prisons and jails. In 1977, 92
percent of all capital outlays in corrections by state and local governments combined went for institutional expansion and improvement. In 1985, institutions
commanded 99 percent of all capital spending.33 By that year, annual capital
expenditures were about 700 percent higher than they were in 1971.
What accounted for these enormous increases? In large part, the increase was
illusory: Costs rose as the dollar’s purchasing power dropped. The value of the
dollar eroded only slightly between 1952 and the mid-1960s, but an inflationary
surge was ignited in the latter years of that decade when federal spending for
domestic and military purposes far exceeded available tax revenues. Inflation
proceeded apace throughout the 1970s and into the early 1980s, so that by
1985, a 1971 dollar was worth only 38 cents. Three-quarters of the increase in
correctional spending between 1971 and 1985 was due, consequently, to inflation and.
the dollar’s corresponding loss of purchasing power. The "real" increase in correctional spending, deflated for inflation, was therefore 122 percent for operations,
rather than 470 percent (the increase not adjusted for inflation). Annual capital
spending was really about 200 percent higher in 1985 than in 1971, not the 700
percent reflected in the dollar differences.
This real doubling of spending for operations between 1971 and 1985 resulted
mainly from the growing offender population and correctional workforce. The
numbers of offenders (adults and children) incarcerated increased 102 percent.34
Probation and parole populations increased significantly, but the size of that
change cannot be computed, because no surveys of probation and parole populations were made between 1965 and 1976. The workforce grew correspondingly. In October 1971, the Bureau of the Census counted 177,600 correctional
32
In fiscal year 1985, state governments paid the lion’s share of all corrections costs-62 percent.
Counties footed the bill for 24.5 percent, and the remainder was split between municipal governments (8.1 percent) and the federal government (5.4 percent) (see Table 1). In all, corrections expenditures represented about 28 percent of all civil and criminal justice spending that year (which
totaled $45.6 billion) and approximately 0.8 percent of direct governmental expenditures for all purposes. (Computed from data in "Justice Expenditure and Employment Final Report, 1985," Tables 1
and 2.)
33
Expenditure and Employment Data for the Criminal Justice System: 1977, Tables 48, 49, and "Justice
Expenditure and Employment Final Report, 1985," Tables 42, 43.
34
State and federal prison populations increased 154 percent and jail populations increased by
about 77 percent, while the numbers of children in custody declined by 14 percent. Bureau of Justice Statistics Bulletins, Prisoners 1925-81 (1982). Prisoners in 1985 (1986). Children in Custody (1986),
Washington, D.C.: U.S. Department of Justice; 1971 jail data estimated. from 1970 and 1972 census
figures, reported in Margaret Cahalan, "Trends in Incarceration in the United States Since 1880,"
Crime & Delinquency, 25, 1979, p. 14; 1971 figure on children in custody and 1985 jail population
from personal communication with BJS personnel.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

13

Table 1
DIRECT EXPENDITURES FOR CORRECTIONAL OPERATIONS
AND CAPITAL OUTLAY: 1971-1985
(In $ thousands)

Actual Costs

Inflation-Adjusted Costs
in Constant 1971 Dollars
Operations

Capita1

4,777
143,283
68,953

-

-

125,524
1,270,238
840,037

7,748
107,538
71,245

121,507
1,229,590
813,156

7,500
104,097
68,965

1973
Federal
State
Local

149,474
1,435,259
943,285

21,380
98,661
92,149

136,171
1,307,521
859,333

19,477
89,881
83,948

1974
Federal
State
Local

170,133
1,688,071
1,091,427

44,396
124,458
121,911

139,679
1,385,906
896,062

36,449
102,180
100,089

1975
Federal
State
Local

196,009
2,015,826
1,283,389

20,769
177,174
150,147

147,399
1,515,901
965,109

15,618
133,235
112,911

1976
Federal
State
Local

242,886
2,276,335
1,485,243

13,466
198,448
169,134

172,692
1,618,474
1,056,008

9,574
141,097
120,254

1977
Federal
State
Local

253,555
2,561,067
1,609,475

45,163
285,953
178,853

169,375
1,710,793
1,075,129

30,169
191,017
119,474

1978
Federal
State
Local

305,314
2,855,318
1,813,931

31,860
321,645
194,642

189,295
1,770,297
1,124,637

19,753
199,420
120,678

1979
Federal
State
Local

324,582
3,173,302
1,958,984

29,379
358,197
195,376

181,582
1,770,703
1,093.113

16,393
199,874
109,020

Level

Operations

1971
Federal
State
Local

106,024
1,179,821
788,215

1972
Federal
State
Local

Capital

14

DOUGLAS C MCDONALD

Table 1 (continued)

Actual Costs
Level

Operations

1980
Federal
State
Local

375,000
3.693,931
2,030,232

1981
Federal
State
Local

Capital

Inflation-Adjusted Costs
in Constant 1971 Dollars
Operations

Capital

12,000
563,578
226,010

184,125
1.813.720
996,844

5,892
276,717
110,971

403,000
4,203,970
2,321,014

10,000
639,887
290,951

179,335
1,870,767
1,032.851

4,450
284,750
129,473

1982
Federal
State
Local

499,000
4,887.192
2,638,840

14,000
672,600
337,315

209,580
2.052.621
1,108,313

5,880
282,492
141,672

1983
Federal
State
Local

555,000
5,610,585
2,990,917

21,000
712,655
512,206

225,885
2,283,508
1,217,303

8,547
290,051
212,131

1984
Federal
State
Local

632,000
6,313,548
3,328,554

49,000
847,262
598,537

246,480
2,462,284
1,298,136

19,110
330,432
233,429

1985
Federal
State
Local

670,000
7,304,000
3,826,OOO

50,000
1,066,000
633,000

251,920
2,746,304
1,438,576

18,800
400,816
238,008

NOTE: The figures for 1985 differ from those published by
the Bureau of Justice Statistics in Justice Expenditure and Employment, 1985. BJS figures for 1985 were obtained from a special
survey by the Bureau of the Census, using the methodology of
the 1971-1979 Expenditure and Employment in the Criminal Justice System series. That series was discontinued in 1979, so
expenditure amounts were obtained from the Bureau of the
Census’ general surveys of government finances. The Bureau
modifies these data each year for the BJS to publish in its post1979 Justice Expenditure and Employment series, but the
methods of data collection and analysis differ from those used in
the earlier series. For the sake of consistency, I have used the
1984 and 1985 Bureau of the Census figures from the general
government finance survey rather than the 1985 figures from the
special survey done for the BJS.
SOURCES: All 1971-1979 data from various annual reports
in the Expenditure and Employment Data for Criminal justice System
series, Washington D.C.: U.S. Department of Justice (1971:
Tables 4, 5; 1972: Tables 4, 39; 1973 through 1975: Tables 4, 39;
1975: Tables 4. 40; 1976: Tables 4A. 45; 1977: Tables 4. 47;
1978: Tables 4A, 47; 1979: Tables 4, 44). 1980-1984 federal data
and 1982-1984 state/local data from unpublished sources provided by US. Department of Commerce, Bureau of the Census.
1980-1981 state and local spending from justice Expenditure and
Employment Extracts: 1980 and 1981, Washington, D.C.: U.S.
Department of Justice, 1985, Tables 2 and 21. 1985 data from
U.S. Bureau of the Census, Government Finances in 1984-1985,
Washington, D.C.: U.S. Government Printing Office, 1986,
Table 8.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

15

workers at all levels of government in the adult and juvenile systems. By 1985,
the number had grown by 120 percent, to 394,700. The cost of that larger labor
force increased at a slightly slower rate. October payroll costs had risen 115
percent in real terms between 1971 and 1985, from $129,119,000 to
$277,285,000 in constant 1971 dollars.35 Higher manpower costs thereby
account for much of the increase in the real cost of operating correctional facilities during that period.
Changes in capital spending are explained by a different dynamic. In the late
1960s and early 1970s, the combined state and federal prison population was
stable, having declined from the 1961 peak of 220,000 prisoners. During the
mid-1980s, in contrast, prisoner populations were swelling at a fast clip, and
money was needed to build or acquire tens of thousands of new cells each year.

THE IMPACT OF LITIGATION ON CAPITAL SPENDING
AND OPERATING COSTS

Capital spending was also spurred by court orders to improve conditions in prisons and jails and to relieve overcrowding. Precisely how much more costly
correctional services have become as a result of judicial intervention is not
known in the aggregate, but we have some indications. Harriman and Straussman examined spending for capital and agency operations in 14 state prison systems that were ordered by the courts to improve conditions prior to 1979. They
found marked increases in capital spending in 12 of the 14 in the years following imposition of the order, above and beyond what would have been required
to accommodate larger inmate populations at the levels of service that prevailed
before the orders. The increase in annual capital expenditures per state prisoner
during the third through fifth years following the court order averaged 164 percent.36 (Wyoming increased its per-prisoner capital expenditure by 744 percent.)
The 14 states as a group had significantly more ambitious plans on the drawing
boards in 1979 than were found in other states that were not laboring under a
court order at that time.
Operating costs also increased, although Harriman and Straussman measured
this increase not in per-capita terms but in the proportion of total state operating
expenditures dedicated to state prisons. During the third through fifth years following the courts’ decisions, state prison budgets represented an average of 10.7
percent of the total states’ budgets for operations, an increase of 7.1 percent, on
average, over budgets adopted during the five years before the decisions came
35
Payrolls, in constant dollars, did not rise as high as the number of workers. Adjusting for inflation, the average per-capita payroll cost in October 1985 was 4 percent less than that in 1971. This
means that the correctional workers’ pay raises did not keep ahead of inflation during this period.
(Computed from data in U.S. Department of Justice, Bureau of Justice Statistics, Justice Expenditure
and Employment in the United States, 1971-79, Washington, D.C.: U.S. Government Printing Office,
1984, Table 7, p. 41; and "Justice Expenditure and Employment Final Report, 1985," Table 4, p. 3.
36
Computed from data in Linda Harriman and Jeffrey D. Straussman, ‘Do Judges Determine
Budget Decisions? Federal Court Decisions in Prison Reform and State Spending for Corrections,’
Public Administration Review, 43, 1983, Table 1.

16

DOUGLAS C. MCDONALD

down. Litigation was more prevalent before 1979 in those states that spent less
on a per-capita basis, and the litigation brought expenditures in these states
closer to the levels found in states not faced with court orders. (Per-capita costs,
degree of-overcrowding, and incarceration rates were strongly correlated, according to a study of prisons in 1977 by Mullen and Smith. States with high incarceration rates-most of them in the South-were more likely to be overcrowded
at the time and had lower per-capita prison costs.)37

THE EXTENT AND CAUSES OF VARIATION IN PRISON COSTS
Even though litigation may have reduced the variation among states in percapita expenditures for imprisonment, wide differences remain. Table 2 shows
the approximate fiscal year 1985 per-prisoner operating costs in state and federal
departments of correction. These costs were derived from information about
agency budgets rather than audited expenditures, and no adjustments were
made to improve the estimates of the full direct costs because of the lack of sufficient information. 38
Although the deficiencies in the data obscure the true range of variation in percapita costs in the U.S. prison systems, it is nonetheless clear from Table 2 that
the differences are substantial. The most expensive system in the country
(Alaska’s) was 4.5 times more costly than the least expensive (Mississippi’s).
The estimated average per capita expenditure, based on the available data, was
$18,217. Several states showed per-capita costs substantially below and above
that mark.
These differences stem primarily from the cost and utilization of staff labor-a
key resource in corrections. Table 3 compares the most expensive systems with
the least expensive during fiscal year 1985, based on the average starting salary
in each state, after the probationary period of employment, and the ratio of
line-level, nonsupervisory staff to inmates in each state.39 The average postprobation starting salary in the most expensive states for which we have data
was $18,222, and the average inmate/staff ratio was 3.5 to 1. In the least costly
systems, the average starting salary was $13,488 and the average inmate/staff
ratio was 6.3 to 1.
A statistical study of interstate variation in per-capita imprisonment costs conducted in the late 1970s found that 85 to 90 percent of this variation could be
accounted for by differences in inmate/staff ratios and average entry-level
37

Joan Mullen and Bradford Smith, American Prisons and Jails, Vol. III, Washington, D.C.: U.S.
Government Printing Office, 1980, p. 117.
38
They also had to be recomputed from tables in The Correctional Yearbook because capital costs
were not excluded from operating expenditures in the table and were consequently included in the
average per-capita cost.
39
Average staff salary would have been the better indicator of labor cost, but these figures are
not readily accessible. A survey published in January 1987, "COs in Demand in Many States,"
Corrections Compendium, pp. 9-11, probably undertaken in late 1986, lists average salary and starting
salary, but the response rate from state departments was too low for the survey to be useful here.

17

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

Table 2
APPROXIMATE PER-CAPITA OPERATING COSTS TO
DEPARTMENTS OF CORRECTIONS: STATE AND
FEDERAL ADULT PRISON SYSTEMS, FY 1985
State

cost ($)

State

cost ($)

Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana

12,213
39,822
20,792
8,709a
21,589
18,891
16,109
21,349
12,402
13,334b
17,915
13,459
20,946
13,978c
24,859
14,524
19,804
10,416
30,430
14,317
25,052
23,842
20,412
8,501
12,170
12,353d

Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Federal BOP

20,718
9,910
23,529
24,174
33,635
19,611
13,622
13,174a
10,873
17,626b
22,608
13,491
19,638
11,056
11,926
14,332
10,097
31,486
27,189
15,920b
24,370
11,926
19,524d
13,255e
14,806

NOTE: Coats computed, unless otherwise specified,
from operating budgets and average daily populations
reported by departments to the Criminal Justice Institute
and published in The Corrections Yearbook, South Salem, NY:
Criminal Justice Institute, 1986, pp. 8-9, 22-23.
a
Computed from operating cost in American Correctional
Association, Directory 1986, p. xi.
b
FY 1985
c
Computed from data provided by the Indiana Department of Corrections.
d
Per-capita cost could not be computed for lack of information on operating costs of adult system; figures here were
submitted by departments and reported in The Corrections
Yearbook.
e
Computed from figures for state penitentiary only;
excludes female and juvenile corrections.

18

DOUGLAS C. MCDONALD

Table 3
PER-CAPITA COSTS, STARTING SALARIES FOR CORRECTIONAL
OFFICERS, AND LINE-LEVEL NONSUPERVISORY STAFF IN MOST
COSTLY AND LEAST COSTLY PRISON SYSTEMS: FY 1985

State

Per-Capita
cost ($)

Average
Starting
Salary ($)

Inmate/Staff
Ratio

Most Costly Prison Systems
Alaska
New Mexico

39,822
33,635

26,520
14,576

4.1
2.7

Maine
Utah
Vermont
Massachusetts

30,430
31,486
27,189
25,052

NA
17,701
13,208
25,206

NA
5.0
3.4
2.6

New Jersey

25,174

19,312

3.2

Iowa
Washington
Michigan

24,857
24,370
23,842

17,118
18,432
19,377

3.0
4.0
NA

New Hampshire

23,529

15,652

3.6

Least Costly Prison Systems
Missouri

12,170

14,292

West Va.

11,966

12,279

4.8

So. Dakota
So. Carolina
Ohio

11,926
11,056
10,873

11,274
12,890
15,038

5.1
6.2
8.2

Louisiana
Texas
Nevada

10,416
10,097
9,910

13,620
18,792
17,201

3.5
5.2
6.0

8,709
8,501

13,316
12,857

5.8
6.5

Arkansas
Mississippi

33.0

SOURCES: See notes to Table 2 for per-capita costs; salary and
staff/inmate ratio information from Criminal Justice Institute, The
Corrections Yearbook 1986, pp. 40-42.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

19

salary. 40 Whether this relationship still holds so powerfully has not been tested,
although it is likely that the findings would be similar.

COST VARIATION WITHIN A SYSTEM: IS THERE
AN OPTIMAL SIZE FOR FACILITIES?

Costs also vary substantially from one facility to another within a single prison
system. These differences are clearly related to many of the same features that
produce differences among state systems. In a study of fiscal year 1978 costs in
New York State’s prison system, which had 32 facilities at the time, McDonald
found that larger prisons were less expensive on a per-prisoner basis, on average, than smaller ones. In addition, the average per-capita cost of operating
maximum-security prisons was lower than the cost of minimum-security camps,
which in turn were less expensive than medium-security facilities. Communitybased work-release facilities were the most expensive to operate that year, on
the whole. These cost differences resulted in large part from variations in the way
each type of facility was staffed. Maximum-security prisons were larger, on average, and had fewer staff persons for each inmate than other facilities. As the
staff/inmate ratio increased, so did cost. 41
Examining cost variation among different types of facilities within a single system raises an interesting question: Is a prison of a particular size more costeffective than others? The unit cost of a wide variety of goods and services produced by private firms decreases as output grows, up to a certain level of production, and then begins to increase. In other words, increasing output permits
a firm to achieve economies of scale up to the point where the cost per unit stabilizes. The finding that larger prisons are cheaper on an average per-capita
basis suggests that economies of scale can also be achieved in the provision of
correctional services. Is there an optimal size for a prison, from the point of
view of operational costs?
Because prisons differ from one another in a variety of important ways in addition to population size, this question cannot be answered unless one holds constant the effects of all dimensions except for size and cost. One early attempt to
do this was Block’s application in 1976 of microeconomic theory to data on California prisons and jails. He found some evidence that unit costs varied according to the size of the prisoner population when several other features of prisons
were held constant. He recognized, however, that his analysis was extremely
tentative and that the findings were only suggestive.42 In their 1984 study,
Schmidt and Witte analyzed facilities in the Federal Bureau of Prisons and
40

Mullen and Smith, American Prisons and jails, p. 369.
McDonald, The Price of Punishment, pp. 17-23. Depreciation and the interest potion of debtservice charges were not included in the computations of costs. Had they been, the operating costs
of maximum-security facilities would have been somewhat higher, and the gap between them and
the cost of other facilities wider.
42
Michael K. Block, Cost, Scale Economies, and Other Economic Concepts, Washington, D.C.:
Correctional Economics Center, American Bar Association, 1976.
41

20

DOUGLAS C. McDONALD

concluded that "prisons are cheapest to run when they are quite large but not
behemoth."43 The statistical tests they applied to cost and program data
revealed that the cost per capita decreased as a function of population size until
the number of prisoners reached about 900. Above 1,300 prisoners, per-capita
costs began to increase. Schmidt and Witte concluded, therefore, that the
optimal size for a prison, from the perspective of cost, is one that houses
between 700 and 1,500 inmates. Moreover, their findings indicated that percapita costs are lower, other things being equal, if inmates are given relatively
ample living space, are housed in single cells, and have fewer sanitary facilities
available. These data support the notion that it is not cost-effective to squeeze
more prisoners into smaller spaces (probably because staffing costs will rise in
these circumstances). The authors indicate that prisons of the commonly preferred size, i.e., holding approximately 500 persons, will be less efficient than
larger-sized ones, but they also recognize that such prisons may be desirable for
reasons other than pure cost-based efficiency.44 These studies raise important
questions that are not fully answered, because the logic of the analyses-using
microeconomic theories of profit-seeking firms to understand public agenciesinvolves taking some large leaps. Rather than accepting the finding that 700- to
1,500-bed facilities are more cost-efficient in all systems and adopting this as a
design principle, corrections policymakers should undertake similar studies of
other systems to clarify the relationship between size and cost.

THE COSTS OF IMPRISONING ADDITIONAL PEOPLE
AND SAVINGS FROM DECARCERATING OTHERS

A related matter that deserves systematic research, and one that is now poorly
understood, is the marginal cost of imprisoning additional persons and the marginal savings of imprisoning fewer. (Also poorly understood are marginal costs
and savings of all types of correctional services, institutional and noninstitutional.) Estimates of these marginal costs are frequently computed to quantify
the implications of changing policies. The most common method of such calculation is to determine the average annual or per diem operating cost of incarceration in the facility or system in question and then multiply this cost by the
number of beds needed or not needed. Even if it were possible to determine the
actual direct operating cost of a facility, this method would yield gross overestimates of costs or savings. Many costs are quite fixed and vary little with
changes in inmate population. For example, adding ten, fifty, or even a hundred inmates to a medium-sized prison generally requires no larger administrative and support staff. The cost of equipment, for example, remains relatively
fixed. But other costs, e.g., those of food, medicines, and personal effects, vary
more directly with changes in population levels. Because so many of the costs
of imprisonment are relatively fixed, the cost of adding additional prisoners is
actually much less than the average cost per prisoner in a facility. The same
43
Peter Schmidt and Ann D. Witte, An Economic Analysis of Crime and Justice, Orlando, Fla.:
Academic Press, Inc., 1984, p. 355.
44
Ibid., pp. 357-363.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

21

holds for the reverse dynamic: The marginal savings incurred by not incarcerating or decarcerating prisoners is much less than the average cost.
To determine the marginal cost of producing one more unit of output-in this
case, incarcerating one more prisoner-one must know how each of the various
cost components in the correctional system varies as the number of inmates
changes. Because certain economies of scale might be achieved for some of
these components, we cannot assume that the marginal cost is exactly the same
at any and all population levels. For example, if a facility is fully staffed but
operating at 80 percent of capacity, the marginal cost of housing more prisoners
will be very low. The marginal cost of adding prisoners to a facility operating at
or above capacity will typically be higher, especially if staffing levels have to be
increased. Moreover, the way costs vary will depend upon a variety of factors
that are peculiar to the correctional system in question-staffing policies are
especially important here because ‘staff salaries, fringe benefits, and retirementfund contributions represent such a large proportion (typically around threequarters) of all operating costs. It is therefore not possible to assert that the
marginal cost of adding one more prisoner (or the marginal saving of subtracting
one) is some specifiable percentage of the average cost in all jurisdictions.
Analysis of costs in the particular system or facility in question is needed.
Given the intense interest- in correctional costs and the consideration given to
different penal strategies, it is remarkable that so little attention has been given
to the analysis of marginal costs.45

THE FUTURE OF CORRECTIONAL COSTS

In 1984, Rich and Barnett produced estimates of future state prison populations,
using a mathematical model that extrapolated crime, incarceration, and demographic patterns found in eight sample states during the early 1980s to U.S.
Bureau of the Census projections of population growth between that time and
the year 2020. They concluded that prison populations would continue to rise
into the early 1990s. The ‘birth dearth” that followed the post-World War II
baby boom would then begin to affect prison admissions, and the number of
persons behind bars would decline slightly for about a decade. Around the turn
of the century, levels were predicted to rise again and continue upward through
2020.46 This paralleled and extended Blumstein, Cohen, and Miller’s projections
for Pennsylvania, which forecast a similar decline starting in around 1990.

45
For a plain-English discussion of marginal costs, see Funke and Wayson, "What Price Justice?"
More technical discussions are given in Block, Cost, Scale Economics, and Other Economic Concepts,
and Schmidt and Witte, Economic Analysis. For an example of estimating the savings incurred by
diverting specified numbers of offenders from jails into nonincarcerative sentencing options, see
Douglas Corry McDonald, Punishment Without Walls: Community Service Sentences in New York City,
New Brunswick, N.J.: Rutgers University Press, 1986, pp. 190-263.
46
Thomas F. Rich and Arnold I. Barnett, "Model-Based U.S. Prison Population Projections," Public
Administration Review, 45, November 1985, pp. 780-789.

22

DOUGLAS C. McDONALD

(Blumstein and his colleagues made no guesses about what would happen in the
twenty-first century.) 47
Reality has had a habit of outstripping forecasts, however, at least with regard
to prison population projections. Rich and Barnett forecast a slightly less than
10 percent increase between 1983 and the point in the early 1990s when the
predicted downturn is to occur. But by 1985, only two years later, state prison
populations had increased by almost 17 percent. They estimated that by 2020,
prison populations would grow between 20 and 25 percent over 1983 levels.
But between 1983 and 1986 alone, those populations grew 30 percent.48 Similarly, Blumstein, Cohen, and Miller’s earlier projection of Pennsylvania prison
populations, based on models that fit 1970-1977 data quite well, badly underestimated subsequent growth. In contrast to the forecast population level of about
9,500 in 1985, there were 14,227 persons in prison by the end of that yearapproximately 50 percent more than estimated.49
The projections fell short partly because the forecasters could not anticipate
changes in sentencing policy and partly because they did not capture adequately
the subtle and possibly changing interactions among age, race, crime, and criminal justice processing. The prison population may taper off in the 1990s, but
there are other forces at work which have not yet been included in model-based
projections, and these forces may continue to propel population levels forward.
For example, the baby boom never really stopped in the black and Hispanic.
communities, and these groups will constitute an increasingly large proportion
of the young male cohort in the coming decades. The number of young white
non-Hispanic men in their twenties began to drop in 1984, but the cohort of
blacks in their twenties will continue to grow until 1989 and will then diminish
only very slowly. In 2005, for example, that number will be only slightly lower
than current levels-about the same as in 1982.50 (Projections of Hispanic
populations were not available at the time of this writing, but Hispanic
birthrates have been higher than those of white non-Hispanics.) Because young
black and Hispanic men have higher arrest and incarceration rates than whites,
and because there is some evidence that those rates are increasing51-perhaps
the result of the worsening socioeconomic conditions of blacks and Hispanicsthe slowdown in prison populations that has been forecast may not come to
pass.
47

Alfred Blumstein, Jacqueline Cohen, and Harold D. Miller. "Demographically Disaggregated
Projections of Prison Populations," Journal of Criminal Justice. 8. 1980. pp. 1-26.
48
BJS, Prisoners in 1985, p. 1.
49
Ibid., p. 2.
50
Computed from data in U.S. Bureau of the Census, Estimates of the Population of the United
States, by Age, Sex. and Race: 1980 to 1985, Current Population Reports, Series P-25, No. 985, Washington, D.C.: U.S. Government Printing Office, 1986, Table 1.
51
Van Alstyne’s study found that the proportion of black males aged 16 to 19 arrested for a
felony in New York State increased 23 percent between 1970 and 1984. Even more troubling was
the increase in felony arrest rates for those who already have prior felony convictions on their
record: about 150 percent for the 16 to 19 age group, and about 70 percent for those between 20
and 29. (Demographically Disaggregated Male Felony Arrest Trends: New York State (1970-1984). NYS
Division of Criminal Justice Services, 1986, pp. 62, 72.) This latter trend is especially explosive in
New York State because nearly all those convicted of a second felony must be sent to prison.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

23

It is not difficult to imagine the US. prison population doubling in the next ten
years. The current rates of growth are pointing in that direction. Without significant changes in sentencing policies, the demand for prison and jail cells will
continue to rise. If the prison population doubles, state and federal governments will have to-construct in short order as many cells as now exist to handle
the demand, to say nothing of replacing outworn or substandard facilities. The
cost of this construction, based on a reported average cost per bed of $51,000,
will be approximately $26 billion in constant 1986 dollars, exclusive of financing
charges, for prisons alone.52
Jail populations have also been rising quickly: Between 1978 and 1985-the
most recent years for which data are available-the number of persons held in
jail has been increasing at an average of 8.8 percent per year.53 Jail populations
will also double in a little more than a decade, if this rate continues. Assuming
again an average construction cost of nearly $49,000 per bed, doubling the size
of America’s jail capacity would cost approximately $12 billion in constant 1986
dollars. The cost of operating both jails and prisons would thus probably be
about double the current cost, in constant 1986 dollars.
These estimates assume no upgrading of existing facilities. Unfortunately, a
large proportion of the existing correctional facilities do not meet accepted standards and are in need of capital improvement. A Bureau of Justice Statistics
survey found that 62 percent of all inmates housed in state prisons were being
held in units smaller than 60 square feet, the American Correctional Association
(ACA) minimum standard. 5 4 If new cells were built for all these inmates at the
average cost of construction that prevailed in 1985 (as reported in The Corrections Yearbook), the total capital expenditure required would be about $11 billion
constant 1985 dollars, exclusive of financing costs and perhaps such additional
costs as those of site acquisition and development. This is, admittedly, a very
rough estimate, but it is probably in the ballpark. Although we cannot calculate
similar estimates for jails because we lack recent data on space utilization, it
would be reasonable to assume that a very large proportion of all jails fall short
of ACA standards and that the costs to remedy conditions would be counted in
billions of dollars as well.

FOOTING THE BILL
Given these conditions, correctional administrators will find themselves in an
increasingly tight bind, and the dilemma facing legislators will become sharper.
State and local revenues are not increasing quickly enough to pay for these
kinds of increases without other services being sacrificed.
52

Average cost per bed computed using data in Corrections Compendium. February 1988, p. 8.
Bureau of Justice Statistics. Jail Inmates. 1985. Washington, D.C.: U.S. Department of Justice;
1978 data from BJS, Jail Inmates, 1984.
54
Population Density in State Prisons, Washington, D.C.: U.S. Department of Justice, 1986, Table
5, p. 4.
53

24

DOUGLAS C. MCDONALD

Between 1971 and 1985, corrections costs rose faster than any other category of
state and local government spending. Whereas all state and local government
expenditures combined rose at an average annual rate of 10.1 percent during
this period; corrections costs grew 14.9 percent, on average, each year-50 percent faster.55 These increases might have been taken in stride had not both
state and local governments faced tightening restrictions on revenues.
Within a few years of the passage of Proposition 13 in California in 1978, 51
new expenditure controls or revenue restrictions were placed on state and local
government spending powers. Personal income taxes were lowered in 35 states,
and sales taxes were reduced in 19. Federal aid to these governments also
began to shrink in 1980. Whereas the ratio of federal aid to state/local revenue
had been rising for at least the prior 25 years, peaking in 1978 at 32 percent (a
level three times higher than that of 25 years earlier), the federal government’s
general revenue-sharing program took several deep cuts, first during the Carter
administration, and then during the Reagan administration.56 By 1986, the general revenue-sharing program was dead, and many local governments were left
for the first time without any direct federal assistance.57
In the early 1980s, many state and local governments were cutting budgets,
incurring general fund deficits, and laying off workers. This was happening at
the same time the demand for local assistance was increasing because so many
people were being thrown out of work. Increasing corrections costs were consequently covered by pulling money out of other types of services and by raising
local taxes. The trend toward reducing taxes was reversed, and a frenzied
scramble for revenue ensued at both state and local levels.
Unfortunately, the ability to raise revenues locally depends upon the underlying
health of the state’s or locality’s tax base. Many regions have not fared well in
the recent economic recovery. Indeed, growth has been sluggish in many areas,
with quite high unemployment rates (and, consequently, high demands for a
variety of state and local services and assistance). Pains will therefore be the
sharpest in this fend-for-yourself federalism in those many states that have high
incarceration rates, poor populations relative to the rest of the country, and, by
extension, weaker tax bases. Most Southern states are so characterized.
Local governments will fare worse than state governments, because their revenue bases are narrower, and local corrections will thereby face harsher conditions than state correctional agencies. According to the National Association of
Counties’ estimates, county-funded justice and public safety programs receive
approximately one-third of all general revenue-sharing dollars.58 With the
decline of federal assistance, continued increases in local correctional costs will
55
J.S. Bureau of the Census, Government Finances in 1984-1985, Series GF85, No. 5, Washington,
D.C.: U.S. Government Printing Office, 1986, Table B, p. XI.
56
Thomas R. Swartz, "A New Urban Crisis in the Making," Challenge, September/October 1987,
pp. 35-37.
57
John Herbers, "The New Federalism: Unplanned, Innovative, and Here to Stay," Governing,
October 1987, p. 30.
58
Cited in Susan D. Crubb, "Huber supports justice programs before House," County News (newspaper of The National Association of Counties), October 5, 1987, p. 8.

THE COST OF CORRECTIONS: IN SEARCH OF THE BOTTOM LINE

25

have to be covered by cuts in other services or parallel increases in tax revenues,
or both. In the poorer states, the limits on raising tax revenues are such that
increased corrections spending will probably come at the expense of other services.
Barring dramatic reversals, corrections administrators will have to attend even
more to the task of cost management, and legislators will have more difficult
choices to make. It would seem that having more accurate information about
costs, better reporting systems, and accounting methods better oriented to cost
analysis in the public sector would be welcome, and perhaps even imperative, if
corrections managers are to make intelligent decisions about how scarce
resources should best be used.

THE COST OF CORRECTIONAL SERVICES: EXPLORING
A POORLY CHARTED TERRAIN
Alan M. Schuman
Director, Social Services Division, Superior Court, District of Columbia

INTRODUCTION
This review is offered from the perspective of a community corrections policymaker. The primary focus will be on the research findings of Douglas
McDonald’s monograph and their relevance for corrections policies. My frame
of reference spans more than 25 years as a practitioner and manager in all
phases of the criminal justice system. My comments will include some general
reactions to the monograph and a discussion of some of the specific findings
and their potential implications.
GENERAL REACTIONS
McDonald has raised many relevant cost analysis issues that should be considered by every policy decisionmaker in the criminal justice system. Two
words that come to mind when reflecting on this monograph are awareness and
comprehensiveness.
The reader becomes aware of the many hidden costs that must be considered
when calculating the total cost of correctional services and the wide diversity of
cost analysis formulas used to prepare budgets in various jurisdictions. The
author has made painfully clear the lack of any uniform formula for calculating
correctional costs on a local, state, or national level. And I will consider a whole
series of different questions when commenting on or reviewing correctional
costs in my community.
McDonald suggests an accurate and comprehensive formula that should be considered as a standardized approach to analyzing correctional costs. We must all
talk a common cost analysis language at federal, state, and local levels. At the
very least, this monograph points out the lack of any systematic cost analysis of
correctional institution building and correctional operations costs. It challenges
us to abandon our old accounting methods and develop and adapt comprehensive and more accurate cost data.
The monograph is comprehensive in that it addresses implicit correctional costs
that result from current policy concerning significant issues. Such issues include
the impact of litigation and its cost for operating correctional institutions,
optimal correctional facility size, the hidden and often uncalculated pension and
fringe benefit costs, and privately financed and managed correctional facilities.
These issues are rarely discussed in cost analysis and cost effectiveness terms.
Readers are made aware of criminal justice cost issues facing us today and in the
future.

27

28

ALAN M. SCHUMAN

The author graphically pictures the enormously spiraling costs involved in continuing to build correctional facilities at the currently projected rate. These costs
do not even include the cost of improving current correctional facilities to the
level of the American Correctional Association (ACA) standards of acceptability.
This cost was estimated to be $11 billion in constant 1985 dollars, exclusive of
financing costs and other costs such as those of site acquisition and development.
I cannot review this monograph without automatically thinking of sentencing
alternatives as a realistic fiscal necessity for future correctional services. Such
alternatives must provide protection and cost savings to communities while
simultaneously reducing the likelihood of repeat offenses.
Although this cost analysis treatise gives little attention to front-end correctional
services such as probation, the findings reported lead me to think of other
options. The fact that probation issues are addressed near the end of the monograph is significant. As the author indicates, very limited cost analysis data are
available. Nevertheless, the major front-end, or nonincarceration, sentenceprobation-is presented after all of the discussion about incarceration, the last
step in the criminal justice system. This is probably symptomatic of society’s
view of probation as a sentence. It increases my own resolve to address the
major role that the many forms of probation must play in the cost analysis
scenario and to remind the criminal justice decisionmakers of one of the most
viable and cost-effective sentencing alternatives.
EFFECTS OF THE LACK OF COST DATA ON COMMUNITY
CORRECTIONS

The lack of specific cost analysis data pertaining to probation is a reflection of
the limited foresight community corrections leaders have shown. This monograph stimulates me to call for initiatives and indicators that will help to clearly
define the cost of community corrections.
McDonald’s monograph provides community corrections with an opportunity to
improve its image. Precise cost analysis developed for community corrections
through national associations such as the American Probation and Parole Association (APPA) and the National Association of Probation Executives (NAPE)
would provide objective data for legislators to enable them to accurately weigh
the alternatives when making budgetary decisions. This would have a very
positive impact on the image of community corrections and would allow us to
make budget presentations that reflect current sound business practices.
A standardized formula for analyzing and presenting the costs of community
corrections would engender a higher trust level from legislators. Today, because
of current accounting practices, corrections policymakers seldom include all of
the hidden costs. This leads to public skepticism or perhaps even a perception
of mismanagement. When all costs are included in reporting the actual cost of
corrections, and these costs are made public, decisionmakers will be in a better
position to consider community services costs and programs, Surely, this will

THE COST OF CORRECTIONAL SERVICES: EXPLORING A POORLY CHARTED TERRAIN

29

lead to more emphasis on early intervention alternatives. Diverting offenders
from corrections at every entry point is the most cost-effective and safest solution for our ‘communities.
McDonald makes brief mention of intensive supervision programs (ISP), specifically, the Georgia program, which has an annual cost of $1,595 per probationer,
as compared with $275 for regular probation. These costs are dramatically
lower than incarceration costs. Based on available data, the average per-inmate
expenditure for incarceration in state facilities was $18,217. This is approximately 11 times the cost of ISP.
Ironically, ISP is, in many ways, the original modem-day design for probation
services. The new ISP concept actually depicts local communities’ original
image of how probation services should operate. ISP provides the type of
comprehensive surveillance services, restitution payments, drug testing and
treatment, employment verification, and networking with other community services that should be expected of all probation agencies that are adequately
funded.
Analysis of the early data indicates that few new crimes are committed by probationers while in ISP programs. The preponderance of violations are technical
and their enforcement is preventive in nature. Thus, the community is protected from additional crimes. Many ISP programs, including those in Georgia,
New Jersey, Philadelphia, and the District of Columbia, actually select participants from institutional populations. If incarcerated populations are acceptable
as probationers to those states and cities, surely the next logical step in an
expansion of this design is to develop programs for all properly classified probationers.
As the community becomes aware of the fact that probation officers see probationers from five to seven days per week and have current information on their
employment and substance-abuse habits, that probation officers randomly spotcheck probationers in their residences at any time of the day or evening, and
that probationers are required to put in many hours of community service as
one of the conditions of probation, a new sense of confidence in community
corrections will begin to evolve.
In her report, Expanding Options for Criminal Sentencing, 1 Joan Petersilia disclosed
projections of prison populations in the larger states, showing an increase of
from 25 percent to 98 percent over the next eight years. The accompanying
spiraling costs of incarceration make the development of effective community
correctional services an attractive and ultimate conclusion.
The challenge really falls on us, the community corrections managers, to insure
that relevant program goals, objectives, and evaluation mechanisms are in place.

‘Published by The RAND Corporation, Report R-3544-EMC, November 1987.

30

ALAN M. SCHUMAN

Another of the findings. in McDonald’s monograph dramatizes the futility of
placing major emphasis on incarceration as the primary solution for treating and
reducing future criminal activity. He reports that, using as the constant the
1971 dollar, the average 1985 correctional institution payroll is 4 percent less
than it was in 1971. If salaries, correctional services, and institutional programs
are not increased and staff/inmate ratios are becoming more unbalanced, how
can we have any expectations for rehabilitation of the incarcerated populations
who will eventually be released to the community? If we can be relatively sure
that institutional resources are not increasing, then we can be equally sure that
we will demonstrate no improvement over our 100-year history of not rehabilitating incarcerated offenders. We can also assume that the same approximate
percentage of released offenders will again be incarcerated at additional costs to
the taxpayers.
Since we are dealing with a finite number of dollars allocated to the criminal
justice system, our direction should be toward more sophisticated front-end services. For every offender we can treat successfully in a community setting short
of incarceration, we will be saving the taxpayers a considerable amount of
money. Packaging our programs properly and using the cost analysis data
presented in this monograph, we can be successful. Given McDonald’s findings
that local governments have a narrower or smaller revenue base from which to
build or update institutions, community sentencing alternatives provide the only
reasonable option.
LITIGATION: JUDICIAL RESPONSIBILITIES
The section of McDonald’s monograph that discusses the impact of litigation
and the cost of operating correctional facilities raises a provocative and fascinating issue. The data reveal that jurisdictions in which correctional systems are
under court order evidence a considerable increase in correctional services and
capital spending costs. We find that the courts-like the legislative and executive branches-are placing their emphasis on institutional correctional facilities
and services, which constitute the last, most. severe, and most costly phase of
the criminal justice system. What we continue to see is court-ordered services
being placed on our least effective and most costly phase of the system. Should
the judiciary, with its power to impose change, address the total system’s needs?
If the judiciary becomes involved, should it play a more proactive role in
reviewing the quality of services of the total system, beginning with programs
oriented toward diversion from the criminal justice system? With probation,
pretrial services, and other community-based programs under the direct supervision of the judiciary in many jurisdictions, is the judiciary not, then, in an
advantageous position to see the cost effectiveness of a balanced allocation of
criminal justice resource dollars?
We cannot ignore the importance of judicial sentencing guidelines and their
potential impact on the length of prison sentences. We cannot ignore the fact
that many offenders are incarcerated because adequate probation services, drug
treatment facilities, or proper mental health services are not available to them
while they remain in the community. Many judges would welcome the

THE COST OF CORRECTIONAL SERVICES: EXPLORING A POORLY CHARTED TERRAIN

31

opportunity for offenders to pay their debt to society in a properly staffed and
funded community setting rather than being incarcerated.
The legislative, executive, and judicial branches of local and state governments
should jointly address the criminal justice system’s problems. The cost-analysis
issues McDonald addresses should play a major part in any joint planning
efforts. Each branch of government has a major stake in the solution of the
criminal justice system’s problems and should, at the very least, have all available data present when making major policy decisions that impact so dramatically on the citizenry.
The monograph reports very limited cost-analysis data for probation and parole.
This is a very significant statement in itself. As an administrator aware of the
dearth of cost data for community corrections, I recognize that the lack of precise data could very well be one of the major reasons that community
corrections-probation in particular- does not have the credibility to demand
more funding. According to my experience, McDonald is incorrect when he
reports that parole officers’ caseloads are close to or exceed ISP levels, bringing
the per-capita costs of parole to the approximate cost level of ISP. Most major
ISP programs, including those in Georgia and New Jersey, have officer/offender
ratios of approximately 1:15. Parole caseloads are significantly higher.
Community-corrections administrators must advocate, in collaboration with
national organizations including the National Institute of Corrections and the
Bureau of Justice Statistics, a major effort to capture and tabulate standardized,
consistent, and accurate cost-analysis data on community corrections. At the
very least, communities should be aware of the cost implications to them in
deciding on criminal justice issues.
ISSUES CONCERNING PRIVATELY FUNDED AND MANAGED
CORRECTIONAL FACILITIES
The question of privately financed and managed correctional facilities is a major
issue facing the public sector criminal justice system. This monograph focuses
on the costs of correctional facilities, again the back end of the criminal justice
system. It would have been particularly interesting to see some data on the
costs of services provided by private vendors who contract these services to
community service agencies.
In some communities, the partnerships established between private and public
sector service providers represent some of the most innovative, cost-effective,
and exciting service models. These partnerships have played a major role in
preventing many offenders from being incarcerated. It is probable that information on the costs of private facility services is not readily available, especially at
a national level. This should be a high-priority cost-analysis project.
The Prison Officers Association reported recently that some of its members had
met with managers of private correctional institutions. Association officials
visited three facilities. The first one, a facility in Northern California, houses
about 80 parole violators; the second, a 350-bed detention facility in Texas, is

32

ALAN M. SCHUMAN

operated on behalf of the U.S. Immigration and Naturalization Service; and the
third, a 360-bed prison in Tennessee, houses both male and female prisoners
serving sentences of up to 7 years. The major conclusions reached were that
these facilities (1) were not properly staffed; (2) lacked privacy for prisoners; (3)
did not have adequate physical plants; (4) paid their staffs very low salaries; and
(5) gave staff training low priority. In one facility, there was a not, with some
signs that inmates were cruelly treated. I present these data to caution that costs
are only one of the elements to be considered in operating a correctional institution that meets constitutional standards. Some correctional administrators are
concerned that states will become dependent on the services of privately run
facilities, after which the private institutions will increase their costs to the same
or higher levels as those of publicly run institutions.
Privately financed institutions can cut costs by paying lower salaries and can
reduce costs even further by providing no or minimal pension and fringe benefit
packages. However, lower salaries could attract less-qualified staff who would
require extensive high-quality training, and this would offset some of the cost
savings. My personal observation has been that privately funded agencies offer
lower salaries and have-very high staff turnover rates. In fact, many of the best
qualified private sector staff eventually apply for public sector probation positions that offer more job security and higher salaries. The high turnover rate
must impact the quality of services that are provided, a factor that should be
considered in any cost-analysis formula.
Another problem facing private companies is the cost of insurance premiums.
Insurance costs have more than doubled as a result of legal actions recently
taken against private operators.
FUTURE CORRECTIONAL COSTS: POLITICAL ISSUES

The issue dealt with in the final section of the monograph-future correctional
costs-is, in many ways, the most important issue and the one least likely to be
predicted with any accuracy. The discussion of the projected overall population
decline following the baby boom is quite relevant. An increase among black
and Hispanic populations is projected. This could very well result in an
increasingly larger proportion of young male offenders in the coming decades.
This would have a decided impact on prison population projections.
To a large extent, the factors that work against future cost projections are found
in the realm of political policy decisions. A good example is the issue of drug
usage and distribution in major U.S. cities. The citizens of the District of
Columbia have reacted strongly to rampant use and distribution of drugs, leading to the creation several years ago of an operation known as Project Clean
Sweep. This project is a special-emphasis police operation that has aggressively
joined the battle against street drugs. In many ways, it has been very successful; it has achieved nearly 30,000 drug arrests since its inception. At the same
time, this has resulted in a tremendous overload to all phases of the criminal
justice system-pretrial, probation, institutionalization, parole, and prosecution.
At the end of 1987, approximately 75 percent of all probationers and

THE COST OF CORRECTIONAL SERVICES. EXPLORING A POORLY CHARTED TERRAIN

33

incarcerated offenders were serving time for drug-related offenses. The project
to rid the community of drug users and distributors has been a major factor in
increasing the numbers of offenders serving sentences in institutional settings.
The resulting overcrowding has led to court-ordered institutional ceilings in
nearly all of the institutions operated by the Department of Corrections.
An additional $50 million has been earmarked to build an 800-bed rehabilitation
facility. But the building of the facility is only the initial cost. The costs to
operate and staff it must then be factored in. What if a policy decision were
made to provide in-patient and out-patient intensive treatment services for drug
offenders? This could have a major impact on correctional costs. As it is, the
Department of Corrections’ operating budget increased from nearly $162 million
in fiscal year 1986 to $187 million in fiscal year 1987.
This example is given to emphasize the impact political decisions can have on
correctional costs and on the size of prison populations.
In the District of Columbia, new legislation requires mandatory prison sentences
for persons convicted of drug distribution. This has resulted in shifting the
problem of drug distribution to the juvenile population. The increase in the
number of juvenile convictions for drug distribution has increased nearly 200
percent in the last few years. What cost impact does this have on the juvenile
justice system? How many additional man-days of costly institutionalization
will result from this legislative change?
The movement toward federal, state, and local judicial sentencing guidelines to
insure equity in sentencing practices among judges can also have a dramatic
impact on correctional costs. How can we include sentencing practices in projections of future correctional costs? Various factors result in offenders falling
into certain grid placements on sentencing guideline charts. What dramatic
changes would occur in correctional costs if grids were modified to include more
use of community corrections rather than incarceration? The financial impact
would be considerable.
I have given an example from each of the three branches of government to
emphasize a key point. Each body of government can and does play a major
role in determining correctional costs.
While reviewing McDonald’s monograph, I could not help but think of the
importance of comprehensive planning and cooperation among the different
branches of government in determining criminal justice policies. I could not
help but think about a process in which goals and objectives are mutually
decided by the three branches of government and then costed out in a careful
and consistent manner.
Douglas McDonald is to be commended for giving us useful ideas for improving
the accuracy of cost analysis and for bringing to our attention the importance of
cost analysis in making criminal justice decisions.

CORRECTIONS “PROPORTIONATE VALUE”
Chase Riveland
Secretary, Washington Department of Corrections

Douglas McDonald’s paper adeptly documents the intricacies and dilemmas facing corrections professionals today in explaining to policymakers, and the public,
the cost of doing business. Ironically, Albert Einstein was able to summarize an
entire theory of relativity in the simple formula E = MC2. In contrast, a formula
for the “cost of corrections” today might well look like this:
COST = # of persons in age-at-risk x fixed variables x courtmandated conditions x bargaining-unit influence x sentencing policies x operant correctional philosophy x releasing practices x judicial trends x fringe benefits x management practices x unemployment rate x prosecutorial practices x law enforcement practices x
perceptions of public opinion x economic conditions - # of
diverted offenders, etc.
No wonder, as McDonald points out, we find it so difficult to assure lawmakers
and the citizenry that the product or service they are buying is of value.
COST VS. PERCEIVED VALUE

According to successful marketing and sales persons, the consumer doesn’t mind
paying any price for something as long as he or she feels the cost and the value
are proportionate. An economy subcompact and a Rolls Royce can both move a
consumer from one location to another in about the same amount of time. Yet
some consumers are eager to pay a much larger ‘cost” for amenities that they
value as proportionate, even if some of those amenities are intangibles, such as
status, image, or recognition. Likewise, many citizens find it easier to believe
that their neighborhoods will be safer, that retribution will more fully be
exacted, and/or that crime rates will go down if more persons are incarcerated.
The perception that they will be safer frequently leads citizens and lawmakers
alike to be willing to pay the ‘cost of corrections,” regardless of the dollar
amount.
One optimistic hypothesis suggested in McDonald’s paper is that a citizenry or
lawmaking body enlightened by data on the ‘real costs” of corrections would
more rationally decide sentencing and correctional policy. Some undoubtedly
would, but many would probably continue to pay any ‘cost’ as long as the perception of the ‘proportionate value” remained the same. I would venture to
guess that those persons buying a Rolls Royce would be little deterred by a
$5,000 increase in the cost of the same car. A painting and frame with a ‘real
cost’ value of only a few dollars may well be purchased for millions by a consumer who attributes an intrinsic proportionate value to the item.
In one example of a legislative body‘s reaction to comparative investment of
public dollars, the lawmakers were faced with the option of investing $250,000

36

CHASE RIVELAND

in resources to conclude a condition-of-confinement suit or almost certainly
investing three or four times that amount in legal fees to pursue further litigation. Despite all advice from legal counsel and corrections professionals, they
chose the litigation route. Their collective interpretation was that the ‘proportionate value” of not being “shoved around by a court” was worth the extra
“cost.” This scenario has repeated itself countless times across the country.
SHOULD THE EFFORT BE MADE?

Having made a case for the occasional irrationality of the consumer, I do not
mean to suggest that more clearly articulating the “costs of corrections” specifically, and the cost of government in general, is less than a worthy goal. The
lack of completeness and definition, and the similarity of “cost figures” within
and across jurisdictions has hampered corrections practitioners in their efforts to
compare facility designs and construction methods, actual costs of various
corrections programs, and even such basic data as staffing ratios. Clearly, any
efforts toward developing clarity in the “cost of corrections” would be useful in
the day-to-day operational decisions corrections practitioners make, as well as in
influencing public policy.
SAME DANGERS

If one looks at earlier efforts, however, one can become weary at the prospect of
the process. Federal attempts to develop consistent reporting from states in programs ranging from Medicare/Medicaid to welfare programs to transportation
systems have historically relied on one of three approaches: voluntary compliance, mandated compliance, or ‘you won’t get your money if you don’t tell it to
us this way.” All of these approaches suffer from the different data, information, budgeting, and accounting systems operating in the states. The second and
third options may partially influence changes toward congruent systems if the
threat is sufficient or enough money is available, but few have been notably
successful.
There are occasions when government, in its attempts to get a better handle on
its costs, has simply driven the costs up higher by introducing more sophisticated accounting systems, which in turn require more highly skilled staff to
operate them, driving the cost of administrative overhead continually higher.
Corrections practitioners who have contact with lawmakers-and who attempt
to influence them-have for a long time used the “cost of corrections” in a
variety of ways. On one occasion, I appeared before two different legislative
committees on the same day. At the first-a sentencing policy committee that I
was encouraging to consider a bill that would limit the number of persons incarcerated and would create new diversionary programs--I prophetically warned
that failure to act would result in a future requirement for several new prisons at
$50,000 per cell in construction costs, followed by several millions of dollars
annually in new operating costs. I also pointed out that investing in ‘low-cost”
diversion programs was a much ‘better deal for the taxpayer” fiscally.

CORRECTIONS ‘PROPORTIONATE VALUE’

37

Later that same day, in an appearance before a committee controlling construction funds, I quite proudly described what fine and efficient people we were for
planning and designing a new institution that could be erected for only $48,000
per cell, much less than the national average for the same type of facility.
To use basically the same “cost” figures to encourage less incarceration in the
morning and to enlist support for funding a new prison in the afternoon borders
on being manipulative. Yet this is not uncommon.’ Many jurisdictions and
practitioners are trapped into trying to leverage all solutions at once in order to
control or contain present and projected crises. In reality, I was simply trying to
describe to two different groups of consumers the “proportionate value* of what
they would be buying (or cost-avoiding). Such appeals are frequently made to
persons who are primarily concerned with the ‘proportionate cost” of something
during a two- or four-year term of office, when many programs are competing
for finite resources.
2
Fifteen-year cycle comparisons, “bow wave” costs, and comparative or ratioed
statistics all may have little impact on a person whose priority is the next election. Specific committees, task forces, or staff groups may be well-informed and
may pursue data-driven decisionmaking, but frequently the larger legislative
(and/or executive) bodies are- driven by less rational “perceptual proportionate
values.”

THE OTHER DANGERS OF FIXATING ON THE COST
OF CORRECTIONS

Many jurisdictions, facing spiraling construction and operating costs, procrastinate or simply don’t make the necessary investments, even after establishing
policy and laws that drive more and more persons into correctional custody.
The resources don’t follow the policy decisions. Probation and parole officers
with unmanageable caseloads, overcrowded prisons and jails with high levels of
idleness and a dearth of programming, and the intervention of the courts are the
most frequent results of such behavior., In some instances, the practitioner who
displays “correctional costs” in a well-intended manner is simply told to ‘make
do,” with the underlying theme being ‘wait until it’s a crisis-then we’ll deal
with it.’
Presenting the “cost of corrections” by itself may set up a scenario in which
existing resources are targeted for elimination as a result of an emotional reaction to the presented data. In one jurisdiction, for example, the operating
budget of the correctional system increased by over 850 percent in a 7-year
period. That fact presented by itself could make the system a prime target for
those looking for “efficiency targets of opportunity.” But that jurisdiction’s
entire state budget actually increased by 780 percent, the fringe benefit and
retirement benefits of all state employees went up 910 percent (in total dollars),
1

Only the argument for the new prison was successful on the occasion described.
“Bow wave” appears to be a Pacific Northwest budgetary colloquialism referring to decisions
made today that will incur greater cost or will require increased funding in the future.
2

38

CHASE RIVELAND

and the costs of three other agencies increased at a rate that exceeded that of
the corrections agency.
One also must remember that the hidden “costs of corrections,” as identified by
McDonald, also exist in other governmental programs. Fringe benefits, retirement funding, financing costs, and services from other agencies are in most jurisdictions equally “hidden” in elementary education, higher education, social
service, and transportation agencies.
Finally, overreliance on “cost of corrections” data has driven many lawmaking
bodies to look for cheap alternatives. Early release programs, privatization, boot
camps, and an array of other temporary fixes have been adopted as quick, cheap
fixes. Although some of these may be legitimate long-term pieces of a total
solution, they usually surface in response to “cost of corrections” problems and
allow jurisdictions to put off “biting the bullet” of making sound policy decisions
that are data-driven, futuristic, and value-filled.
In general, the dangers of developing a more sophisticated system of determining the “cost of corrections” should not be prohibiters but, rather, possible results
of which the practitioner must be aware. Conversely, the practitioner needs
increasingly sophisticated cost-benefit information to manage wisely and to fulfill the public official’s responsibility of providing the best service to the public
as efficiently and effectively as possible.
THE COST OF COMMUNITY CORRECTIONS

Although frequently touted as a major part of “the solution” to major capital
construction and ongoing institutional operating costs, community corrections
programs (including probation/parole, community residential facilities, diversion, and early release programs) have risen and fallen in popularity like the
tide.
Probation and parole, in particular, are referred to as part of the problem one
day and part of the solution the next. Although the largest number of offenders
in the country are in this type of correctional custody, the refrain of “probation
(or parole, or rehabilitation) doesn’t work” has sounded loudly. Most jurisdictions have allowed caseloads to rise so high, and have concurrently allowed
standards to drop so low (while maintaining very high public expectations), that
the predictions of the doomsayers are bound to come true. Only the intensive
supervision programs seem to retain a semblance of what most community
corrections practitioners would feel are adequate resources to truly provide for
offender punishment, restitution, rehabilitation, and community safety. Yet,
even when early release and diversionary programs become part of the “solution” to the overcrowding problem, the resources seldom flow proportionately to
allow them to be fully or well done-or even fairly tested. Few courts have
assisted probation and parole programs to obtain resources; rather, the courts
have generally added an element of fear for the community corrections practitioners by finding them to be civilly liable when one of the many thousands of
offenders goes awry.

CORRECTIONS ‘PROPORTIONATE VALUE

39

Ironically, the most touted programs in the country have the resources only to
see a client once or twice weekly, and the average “supervision” probably consists of seeing the average offender once a month or less. If the attempt to monitor, let alone modify, behavior is to be a serious one, a much greater investment
must be made, although that investment would still be modest compared with
the cost of building prisons. The correctional community, lawmaking bodies,
and budget decisionmakers certainly have not followed a dollar-driven
corrections-cost approach to resolving the corrections problem. The “perceptual
proportionate value” has been not in synch with the displayed dollar costs or
cost savings.
Washington State, as an example, chose in 1981 (effective in July 1984) to eliminate parole, the Parole Board, and what would typically be known as probation. With few exceptions,3 probation was reduced to loose “monitoring,” the
collection of court obligations, and a reasonably clear statement that neither
treatment nor rehabilitation were to be expected, nor would violations result in
significant sanctions. Will it work? The jury, and the statistics, are out.
What is known is that offenders will not be called to task for technical violations, so the recidivism statistics will improve. Caseloads have risen;4 public
expectations haven’t changed; yet authority, assumptions, and philosophy have
been a collage of competing beliefs among a variety of interest groups, “the public” not being one of those actively involved.
WHAT IS UNADDRESSED?

McDonald’s observations about unaccounted-for costs and differences between
reported costs are for the most part quite accurate. He might also add several
others, including the following:
1. State/local differences between jurisdictions. The length of time sentenced
persons spend and the number (or percentage) who serve their sentences
in local facilities versus state facilities vary greatly. The numbers of
offenders supervised by state, county, and municipal “probation officers”
also vary widely-as do the reportable costs.
2. Standards. The standards expected by different jurisdictions vary greatly,
depending on jurisdictional expectations, involvement of the judiciary,
state law, local expectations, etc.
3. Litigutive atmosphere. Although the “cost of corrections” has risen considerably due to a nationwide series of conditions-of-confinement/civil
rights suits, at the state level the degree of tort claim liability varies
dramatically. Some jurisdictions (state and county) spend considerable
amounts of money on attorney general, in-house counsel, indigent
defense, court time, and staff costs for defending against tort claims. Millions of dollars-some accounted for, a great deal invisible in correctional
budgets-are expenditures for litigation, which differ widely by jurisdiction.
3

Some sex offenders and first-time offenders.
Although caseloads are now being lowered through “workload reduction plans,” new resources,
and a shift of standards.
4

40

CHASE RIVELAND

4. Transportation costs. In some states, sheriffs deliver newly sentenced prisoners to the state system; in others, the state system incurs the cost of
pickup from county facilities. The geographic spread of correctional facilities also may significantly increase costs.
5. Revenues. Jurisdictions vary widely in the costs they recover. Supervision
fees, board and room charges, corrections industries sales, agricultural
product sales, surplus land sales, rental of “cell space,” and the like all can
lower the cost of corrections, but this revenue is seldom reported or
deducted from expenditures.
6. Intangible cost savings. Some incapacitated offenders do not commit
crimes while incarcerated or while being closely supervised because of
that custody; some are rehabilitated; and some are deterred. How many?
What is the dollar figure or “cost”? No one knows, but neither do we
know the cost of those families thrown on welfare due to incarceration or
that of people who might have been productive, tax-paying citizens if
they were not incarcerated.
IS THERE SOMETHING THAT MIGHT HELP?

Certainly it is unlikely that a national reporting system will be developed that
will assure standardized reporting of ‘costs of corrections” in the near future.
Yet the premise has merit, and the attempt is worthwhile.
It might well be useful to develop one or more models that would list the majority of costs and savings incurred by an average jurisdiction; other jurisdictions
could then adapt the models by plugging in local ratios.
SUMMARY

Unquestionably, the issues that McDonald raises about the failure to explicitly
state the “cost of corrections” are valid ones. And there is no question that practitioners, the public, legislators, and executive branch officials should all become
knowledgeable about the facts relating to costs in order to manage, choose, legislate, and govern wisely and efficiently.
Yet a hidden assumption seems to be that if we have a better financial data, we
all make-or influence-better decisions. Missing; or avoided, in the discussion
are those things that fit into the “proportionate cost” argument. What really are
our values? What do the people of our nation, states, counties, and municipalities expect regarding public safety, treatment of those arrested, convicted, or
sentenced, and subsequent reintegration of offenders into our’ communities?
How do these values rank with public interests in environmental preservation,
health care, education, transportation, and economic development? Can it all be
clarified by more explicit dollar statements? Does the Rolls Royce or fine arts
consumer hesitate over the cost of something he finds to be of “proportionate
value?”
Our problem in the era of the willing consumer may be not so much to affix the
‘proportionate cost” as to better determine the “proportionate value.”

BIOGRAPHIES

Douglas C. McDonald
Trained as a sociologist at Columbia University, where
he received his Ph.D.. Douglas McDonald specialized
in what used to be called "political economy--the
social/economic/political development of nations. He
later became Interested in legal policy issues and began
doing studies of criminal justice reform. In 1978. he
became director of the Citizens’ Inquiry on Parole and
Criminal Justice, Inc., where he published several
books. He then moved to the Vera Institute of Justice.
a private research and criminal justice action organization in New York City. He is currently a senior social
scientist at Abt Associates. Inc., a Cambridge-based
research organization dedicated to research on a wide
variety of social policy matters. His new book on
private prisons and jails will be published later this
year.

Alan M. Schuman
Since 1972, Alan Schuman has been the Director of the
Social Services Division for the Superior Court, District
of Columbia, responsible for adult, juvenile, and family
social services for Washington, D.C. He was a staff
member of the President’s Crime Commission for the
District of Columbia, 1965-67, and earlier served as
Director of Youth Services and Superintendent of the
Lorton Youth Center, Department of Corrections,
Washington. D.C. He received his MSW in Psychiatric
Social Work from Loyola University. Chicago, and is a
Graduate Fellow of the Institute for Court Management. He has served for ten years as Co-Director of
the National Center for State Courts/Institute for
Court Management and Adult Probation Seminar. Mr.
Schuman serves on numerous local and national
boards relating to the criminal justice system and is a
co-founder and former Vice-President of the National
Association of Probation Executives.

Chase Riveland
Chase Riveland received his MSW from the University
of Wisconsin in 1971. He is presently the Secretary of
the Washington Department of Corrections. Prior to
accepting this position, he served as Executive Director
of the Wisconsin Division of Corrections. He has also
held corrections positions of Superintendent, Regional
Chief, Supervisor. and Probation and Parole Officer.
Mr. Riveland has been associated with several major
building and renovation protects. The Washington
Department of Corrections is an integrated correctional system which includes community corrections, prisons, and industries.