Identifying Beneficiaries
of PIE Inmate Incomes
Who Benefits from Wage Earnings
of Inmates Working in the Prison
Industry Enhancement (PIE) Program?

The George Washington University
Center for Economic Research
For
The National Correctional Industries Association
GWU Contract Task: PTA24384/1/CCNS20186A

Thomas W. Petersik
Research Professor of Economics
Tapan K. Nayak,
Professor of Statistics
M. Katie Foreman
Graduate Assistant, Economics

July 31, 2003

Acknowledgments
This research identifying, describing, and quantifying the beneficiaries of Prison Industry
Enhancement (PIE) incomes was propelled at every stage by extraordinarily skilled and helpful
support from every participant. It would not have been possible without the initiative, effort, and
support of many organizations and persons who, in sequence, supported the research.
We are grateful to Robert Watkins and Richard Sutton, Ph.D., of the U.S. Department of Justice
Bureau of Justice Assistance (BJA), for their encouragement and financial support making this
National Correctional Industries Association (NCIA) research possible.
The research has been greatly assisted by the efforts of the NCIA’s Technical Coordinator for
PIE, Barbara Auerbach, especially in clarifying PIE guidelines, and of the Prison Industry
Enhancement Research Advisory Board (RAB), including Texas Representative Ray Allen,
Derral Cheatwood, Ph.D., Patrick Henry, Ph.D, Timothy Mann, Jeff Merrill, and Cindy Smith,
Ph.D., for supporting the research in concept, for reviewing and improving its design, and for
constructive, insightful reviews ferreting errors and improving the clarity of the report.
George Washington University faculty and staff were extraordinarily helpful in facilitating the
research, with timely, efficient, and—most appreciated—effective administration, making work
with GWU offices a pleasure. We are grateful to the professional staff and always helpful student
assistants in all support offices, including in the Department of Economics; to GWU’s Office of
Human Research for helpful guidance in protecting confidentiality and meeting University
requirements for protection of human subjects; to Helen Spencer, Assistant Vice President for
Research Services; and to Susan Burke and Peggy Sweitzer for their mastery of research budget
procedures. And we are especially grateful to Professor Anthony Yezer, director of the
University’s Center for Economic Research, who both welcomed and actively encouraged this
research on inmate work, and also guided its progress through the University, especially in
locating outstanding support staff.
The success of the research owes a special debt to the PIE and other correctional leadership and
staff of all of the respondent States selected for the research, for their faith in its value, for all of
their willingness—among PIE staff and in records, inmate accounts, and other offices—to
assemble information, transmit or ship it, and then explain or follow up on seeming endless
questions, and for their abiding sense that their efforts would assist everyone in better
understanding PIE. We were continually impressed with every State’s helpfulness, both those
asked ancillary questions, and especially those sampled for detailed research. The quality of the
resultant information is especially indebted to them.

iii

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

We are tremendously grateful to Gwyn Smith-Ingley, Executive Director of the National
Correctional Industries Association, to Carol Martindale-Taylor, and to all NCIA staff for their
confidence in the research and for their unfailing support of the beneficiaries research from its
inception, through its funding, its design and launch, and through each stage to the project’s
completion.
I am especially grateful to Professor Tapan Nayak, Ph.D., of the George Washington University’s
Department of Statistics, for his useful—and instructive—efforts in the sampling design critical
to this research, and to my assistant, M. Katie Foreman, of the University’s Department of
Economics, for her extraordinary skills editing, organizing, tabulating, and documenting the
research data, and especially for her cheer and patience designing, then redesigning presentations
as the research progressed. And I very much appreciate Ms. Foreman’s insights helping us
understand the deeper meanings of the data.
We are all grateful to the staff of Grammarians, whose editorial and formatting skills have greatly
enhanced the final product.
Finally, I am grateful to my wife and family for allowing me time to pursue the research.
Errors in this work are appropriately all mine.
Thomas W. Petersik, Ph.D.

iv

Acknowledgments

This publication was supported by contract task PTA24384/1/CCNS20186A, awarded by the
Bureau of Justice Assistance to the George Washington University, Center for Economic
Research. The Bureau of Justice Assistance is a component of the office of Justice Programs,
which also includes the Bureau of Justice Statistics, the National Institute of Justice, the Office of
Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. Points of view
or opinions in this document are those of the authors and do not represent the official position or
policies of the United States Department of Justice or the National Correctional Industries
Association.

v

Contents

ACKNOWLEDGMENTS..........................................................................................................................III
EXECUTIVE SUMMARY ........................................................................................................................XI
LIMITS OF THE RESEARCH ......................................................................................................................... XI
WHO BENEFITS FROM PIE INCOMES?....................................................................................................... XII
WHAT IF PIE INMATES EARNED AVERAGE AMERICAN SALARIES? .......................................................... XV
Disposition of Remainder After PIE Deductions .................................................................................xx
I. INTRODUCTION.................................................................................................................................... 1
BACKGROUND ............................................................................................................................................ 2
PRISON INDUSTRY ENHANCEMENT (PIE).................................................................................................... 3
II. DESIGN OF THE RESEARCH............................................................................................................ 7
PURPOSE ..................................................................................................................................................... 7
DESIGN ....................................................................................................................................................... 8
CONFIDENTIALITY ...................................................................................................................................... 9
Inmate Confidentiality......................................................................................................................... 10
Firms and States.................................................................................................................................. 10
Protecting Confidences ....................................................................................................................... 10
Confidentiality and the National Correctional Industries Association ............................................... 11
SAMPLING................................................................................................................................................. 11
PROCEDURES ............................................................................................................................................ 13
Peer Reviews ....................................................................................................................................... 13
Coordinating with PIE States.............................................................................................................. 13
The Permission Process ...................................................................................................................... 15
III. INCOME, PIE, AND DEDUCTIONS DEFINED ............................................................................ 17
INCOME .................................................................................................................................................... 18
Inmate Gross Income .......................................................................................................................... 18
Gross Employer Payout ...................................................................................................................... 18
DEFINING PIE DEDUCTIONS ..................................................................................................................... 18
Reasonable Charges for Room and Board.......................................................................................... 19
Taxes ................................................................................................................................................... 20
Victims Compensation......................................................................................................................... 23
Family Support.................................................................................................................................... 24
The Post-Deductions Residual ............................................................................................................ 24
IV. DEDUCTIONS AND BENEFICIARIES ELABORATED ............................................................. 25
TAXES ...................................................................................................................................................... 25
Federal Income Taxes ......................................................................................................................... 26
State Income Taxes.............................................................................................................................. 29
Social Security (FICA) ........................................................................................................................ 30

vii

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Workers’ Compensation ......................................................................................................................33
Unemployment Taxes...........................................................................................................................34
Victims Compensation (Restitution) ....................................................................................................36
Family Support ....................................................................................................................................41
V. RESEARCH RESULTS........................................................................................................................45
PIE WAGES AND BENEFICIARIES ...............................................................................................................46
The PIE Hourly Wage Rate .................................................................................................................46
The PIE Work Week.............................................................................................................................46
The PIE-Inmate Work Year..................................................................................................................47
Turnover in PIE Employment ..............................................................................................................48
Annual PIE Earnings and Their Distribution ......................................................................................48
Mean PIE Wages and Deductions 1998-2001 .....................................................................................48
Analysis................................................................................................................................................51
Social Security .....................................................................................................................................55
Workers Compensation........................................................................................................................55
Unemployment Compensation .............................................................................................................55
Room and Board ..................................................................................................................................55
Taxes....................................................................................................................................................56
Victims Compensation, Restitution ......................................................................................................56
Family Support ....................................................................................................................................57
VI. ESTIMATING BENEFITS IF PIE PARTICIPATION WERE WIDESPREAD .........................59
ASSUMPTIONS ...........................................................................................................................................59
RESULTS ....................................................................................................................................................61
ANALYSIS ..................................................................................................................................................62
Caveats ................................................................................................................................................65
VII. DISPOSITION OF THE PIE RESIDUAL.......................................................................................67
INTRODUCTION ..........................................................................................................................................67
INCOMES AND EXPENDITURES BEFORE PIE...............................................................................................69
Income .................................................................................................................................................70
Expenditures and Dispersals ...............................................................................................................71
Income and Expenditures During PIE .................................................................................................74
Further View of PIE Effects on Deductions.........................................................................................80
Money From Persons/To Persons: ......................................................................................................82
TENTATIVE CONCLUSIONS FROM INMATE ACCOUNT DATA: .....................................................................83
VIII. RECOMMENDATIONS FOR FURTHER RESEARCH.............................................................85
APPENDIX A. DEMOGRAPHICS AND CRIMINAL JUSTICE CHARACTERISTICS ................91
SEX ...........................................................................................................................................................91
AGE ...........................................................................................................................................................92
INMATE RACE ...........................................................................................................................................92
EDUCATION ...............................................................................................................................................94
ANALYSIS ..................................................................................................................................................95
Marital Status/Parenthood ..................................................................................................................95
Offense .................................................................................................................................................95
Sentence Length ...................................................................................................................................96
APPENDIX B. INTRODUCTORY MATERIALS .................................................................................99
INTRODUCTORY PROTOCOL TO SAMPLED STATES .....................................................................................99

viii

Contents

VOLUNTARY PARTICIPATION AGREEMENT ............................................................................................. 101
Department of Corrections, Informed Consent................................................................................. 101
Thomas Petersik, Ph.D. (703) 323-5272; tkpetersik@yahoo.com .................................................... 101
RECORD LAYOUT ................................................................................................................................... 103
Items Requested: ............................................................................................................................... 104
TIMETABLE ............................................................................................................................................. 105

Tables
Table EXS1.

Estimated Gross Employer Payouts, PIE Wages, and Deductions One Unspecified Year
1998-2001—Weighted Sample State PIE Programs Only
(Revised) ......................................................................................................................... xiii
Table EXS2.
Results: Low Case Assumptions, U.S. PIE Income, Benefits, and Beneficiaries, if PIE
Were Generally Applied in U.S. Federal and State Prisons (Assumed Income per PIE
Inmate $10,500 per Year) .............................................................................................. xvii
Table EXS3.
Results: High Case Assumptions U.S. PIE Income, Benefits, and Beneficiaries if PIE
Were Generally Applied in U.S. Federal and State Prisons (Assumed Income per PIE
Inmate $33,250 per Year) ............................................................................................. xviii
Table 1.
U.S. Inmate Populations, by Level of Jurisdiction and Category of Work (2000).................... 3
Table 2.
Cumulative PIE Wages and Deductions, 1979 Through Fourth Quarter 2001 ($Million)........ 5
Table 3.
Cumulative PIE Wages and Deductions, 1979 Through Fourth Quarter 2001 and Third
Quarter 2002 ($Million) .......................................................................................................... 17
Table 4.
Federal Income Tax Rates, Single Filers Showing Annual Gross Income Thresholds ........... 20
Table 5.
State Income Tax Rates, Single Filers Showing Annual Gross Income Thresholds ............... 22
Table 6.
Total U.S. Gross Income Tax Collections Fiscal Year 2000................................................... 26
Table 7.
Total Federal Income Tax Share, 2000—Individual Taxpayers ............................................. 27
Table 8.
Federal Government Outlays, Fiscal Year 2000 ..................................................................... 28
Table 9.
State Spending, Fiscal 2000—Spending Shares, From All Sources* ..................................... 30
Table 10.
Social Security Beneficiaries and Benefits, 1999.................................................................... 31
Table 11.
FICA, Tax Rates 2002, Contributions 2001 (Percent) ............................................................ 32
Table 12.
U.S. Crime Victims Compensation, 2000 ............................................................................... 36
Table 13.
Iowa Crime Victim Compensation Revenues—Fiscal Year 2001 .......................................... 38
Table 14.
U.S. Cost of Violent Crime, 1995, Including Assault, Murder, Rape, and Robbery—Selected
Components ............................................................................................................................ 39
Table 15.
Shares of Medical Costs for Violent Crime, 1995................................................................... 40
Table 16.
Caregivers, Children of State Inmates 1997............................................................................ 42
Table 17.
Percent Distribution, Hourly Wage Rate Weighted Sample, PIE Inmates .............................. 46
Table 18.
PIE Hours Worked Per Week, Weighted Sample, PIE Inmates.............................................. 47
Table 19.
PIE Weeks Worked Per Year, Weighted Sample, PIE Inmates .............................................. 48
Table 20.
PIE Inmate Gross Incomes: Estimated U.S. State-Level PIE Inmate Population, Single Survey
Year 1998–2001* .................................................................................................................... 49
Table 21.
Mean Aggregate Annual U.S. PIE Wages and Deductions, State PIE Programs Only,
Unrevised (Weighted, as Reported by Sampled States) .......................................................... 50
Table 22.
Estimated Gross Employer Payouts, PIE Wages, and Deductions, One Unspecified Year
1998–2001––Weighted Sample State PIE Programs Only (Unrevised) ................................. 51
Table 23.
Estimated Gross Employer Payouts, PIE Wages, and Deductions, One Unspecified Year
1998–2001—Weighted Sample State PIE Programs Only (Revised) ..................................... 52
Table 24.
Assumptions, Two Cases (Low and High): U.S. PIE Income, Benefits, and Beneficiaries if
PIE Were Generally Applied in U.S. Federal and State Prisons ............................................. 60
Table 25.
Results: Low Case Assumptions. U.S. PIE Income, Benefits, and Beneficiaries If PIE Were
Generally Applied in U.S. Federal and State Prisons (Assumed Income per PIE Inmate
$10,500 per Year).................................................................................................................... 61

ix

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 26.
Table 27.
Table 28.
Table 29.
Table 30.
Table A1.
Table A2.
Table A3.
Table A4.
Table A5.
Table A6.

x

Results: High Case Assumptions, U.S. PIE Income, Benefits, and Beneficiaries if PIE Were
Generally Applied in U.S. Federal and State Prisons (Assumed Income per PIE Inmate
$33,250 per Year) ....................................................................................................................62
Estimated Mean Pre-PIE Monthly Incomes, Typical Month ..................................................69
U.S. Household Consumption Averages, 1998–99..................................................................71
Estimated Mean PIE Monthly Incomes and Expenditures, One Typical Month, Survey Year
1998–2001 ...............................................................................................................................76
Pre-PIE and PIE Month Income, Deductions, and Expenditures, Selected Transactions—
Inmates Incurring Transactions Only, 104 PIE Inmate Records .............................................81
Comparison, Distribution by Sex (Percent) .............................................................................92
Comparison, Distribution by Age (Percent).............................................................................92
Comparison, Race Distribution (Percent) ................................................................................93
Comparison, Education* (Percent) ..........................................................................................94
Most Serious Offense, PIE Inmate Sample and U.S. State Prison Population (Percent) ........96
Maximum Sentence, PIE Inmate Sample and U.S. State Prison Population (Percent) ...........97

Executive Summary
The “Prison Industry Enhancement” (PIE) provisions of the Justice System Improvement Act of
1979 (Title 18 U.S.C. 1761(c)), at the end of 2000 engaging about 3,700 State prison inmates in
open-market jobs, either created or recognized financial beneficiary groups not traditionally
associated with inmate work. By paying inmates prevailing wages and then setting wage
deductions, the combination of market wages and accompanying deductions produces significant
monies available for an array of taxes, room and board, crime victim compensation, and family
support, and for meeting other personal, family, and State-required obligations, as well as for
savings and discretionary uses.
The primary purpose of this research is to identify the major financial beneficiaries of PIE inmate
incomes and to measure more exactly the dollar size of those benefits. A secondary purpose is to
estimate the beneficiaries and dollar size of financial benefits if U.S. State inmate populations
were generally employed in open-market jobs similar to U.S. averages in skills, productivity,
annual work hours, and wage levels. The information is expected to be useful (1) to the public
and policy makers in considering inmate work, (2) to corrections and PIE practitioners in
understanding and effecting PIE programs to the greatest benefit, and (3) to actual and potential
PIE wage beneficiaries in order to assist them in recognizing and representing their stakes in
inmate work.

Limits of the Research
This research is limited to financial benefits and immediate beneficiaries of PIE inmate payrolls
and does not represent a comprehensive assessment of PIE; it does not account for other benefits
or costs associated with PIE. It is not designed to be statistically representative of the gender,
race, or other demographic, criminal justice, or employment characteristics of PIE inmates;
moreover, the research is not necessarily representative of non-PIE deductions or characteristics.
The research addresses State-level PIE and does not address benefits from PIE participation by
local jails.
The research is funded by the U.S. Department of Justice, Bureau of Justice Assistance (BJA),
through the National Correctional Industries Association (NCIA), and conducted by the George
Washington University (GWU) Center for Economic Research (the Center), with participation by
selected PIE State departments of corrections.

xi

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

The research relies on PIE payroll and other records provided by an approximately 10-percent
sample of PIE inmates who worked in PIE at any time in a selected year, from a stratified random
sample of PIE participant States, along with GWU staff estimates of Federal and State income tax
liabilities rather than reported deductions. Sampling is based on the State PIE inmate counts
reported to NCIA for December 31, 2001. These reported static one-day counts for any State are
assumed to have the same proportion of inmates of the U.S. total as that State’s unreported share
of the total number of inmates having worked at any time during the sample year 1998–2001.
The research presents State-level PIE inmate incomes plus employers’ payroll tax contributions
for one unspecified calendar year 1998–2001, and represents PIE inmates working in employer
model (private firms) and customer model (operated by departments of corrections) settings. No
personal inmate or other interviews were conducted. All inmate records are treated as
confidential. While not guaranteeing confidentiality, the research does not disclose sampled
States, firms, or the exact survey year. Participation by States and firms was voluntary.
The research estimates include both “inmate gross income,” the gross wages and salaries paid to
inmates, and “employer gross payroll,” including inmate gross income plus employer payroll
contributions to Social Security, workers compensation, and unemployment compensation.

Who Benefits From PIE Incomes?
Using weighted data reported by participant States, and revised by research staff to reflect employer
contributions, tax liabilities, and other adjustments peculiar to the respondent States, Table XS1
presents estimated U.S. PIE incomes and their deductions for one calendar year 1998-2001.
Virtually every American belongs to a group that benefits from PIE inmate work. Non-inmate
beneficiaries of PIE incomes include crime victims, State and Federal household and business
taxpayers, all persons or businesses paying for Social Security and Medicare, and all persons and
programs dependent upon State and Federal income tax funding or the social safety net (Social
Security, Medicare, Workers Compensation, Unemployment compensation), including
elementary, secondary, and college education, welfare, a wide range of State and Federal
programs supporting medical and retirement services, and other goods and services.
Levels of benefits are highly dependent upon levels of income and assumptions affecting tax rates
and distribution policies. Some benefits have at least dual immediate beneficiaries (for example,
both taxpayers and program beneficiaries), voiding simple summing to 100 percent. Nevertheless,
major beneficiary groups are relatively clear for contemporary PIE incomes:
•

xii

Taken together, others, not the inmate, are the primary beneficiaries of contemporary PIE
incomes. An estimated 53 to 57 cents of every dollar earned by PIE inmates goes to noninmate recipients via PIE deductions. A relatively small additional fraction is deducted postPIE.

Executive Summary

Table EXS1. Estimated Gross Employer Payouts,
PIE Wages, and Deductions One Unspecified Year 1998-2001—
Weighted Sample State PIE Programs Only (Revised)
Category
Gross Employer Payout:
Social Security (OASDI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

($Million)
32.04
1.42
0.33
0.11
0.11
0.76
2.74

Percent
100.0
4.4
1.0
0.3
0.4
2.4
8.6
Percent

Percent

Inmate Gross Income:
Room and Board
Taxes, Federal Income (Liability)
Taxes, State Income (Liability)
Social Security (OASDI)
Social Security (HI—Medicare)
Victims Compensation
Family Support*
Other PIE Deductions*
Total Inmate PIE Deductions

29.30
9.60
1.14
0.12
1.42
0.33
2.64
0.23
0.05
15.53

91.4
30.0
3.5
0.4
4.4
1.0
8.2
0.7
0.2
48.5

100.0
32.8
3.9
0.4
4.9
1.1
9.0
0.8
0.2
53.0

PIE Residual**

13.76

43.0

47.0

*Some sampled States include other court-ordered deductions, interpreting the final deduction category as “family and
other court-ordered deductions.”
**Residuals are not net savings and do not equal the net amount accruing to the inmate because of post-PIE
deductions not accounted by PIE.

Totals may not equal sum of components due to independent rounding.

•

The single largest beneficiary of PIE income is the PIE worker, estimated receiving after PIE
deductions on average 43 percent of gross employer payouts and 47 percent of his or her
gross income. Moreover, total PIE deductions therefore, fall well below the 80 percent upper
bound set by PIE.

•

The single largest non-inmate beneficiary group benefiting from PIE incomes includes State
household and business taxpayers and all State programs benefiting from State income tax
payments, accounting for about one third of PIE inmate incomes. About one third of PIE
incomes reduce State taxpayer costs via the PIE room and board deduction. Another 0.4
percent derives from State income taxes paid. Indirect individual lesser contributions
reducing State taxpayer burdens and increasing incomes for others—to the extent State
taxpayers either supplement or occasionally buttress these programs—are PIE deductions
offsetting child welfare support, contributing to State crime victim compensation programs or
reimbursing crime victims (or the State agencies assisting victims), and to unemployment

xiii

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

insurance and workers’ compensation programs. Using State expenditure patterns as a guide,
PIE deductions for the State treasury primarily support a State’s elementary, secondary, and
higher education programs (one third of State spending), and Medicaid (one fifth). To the
extent increased PIE incomes spell increased State tax revenues, then, the primary noninmate beneficiaries of PIE inmate income are State business and household tax payers,
education, and health care.

xiv

•

Social Security taxpayers and recipients—both for Old-Age, Survivors and Disability
Insurance (OASDI) and Medicare (Hospital Insurance, HI)—constitute the second largest
non-inmate beneficiary groups benefiting from State inmate PIE jobholding, accounting for 6
percent of income or 11 percent of total payout, depending upon whether or not employer
contributions are included. By increasing current contributions, PIE inmate earnings-based
deductions strengthen the Social Security system, relieve some burden on current household
and business Social Security tax payers, and at least contribute to reducing pressures for
higher Social Security taxes or reduced income and health care benefits, particularly for the
retired. Recognizing that Social Security taxes are a significant cost to American employers,
take a larger share of lower income paychecks than do income taxes, and that about twothirds of Americans 65 and over get more than half their income from Social Security, PIE
contributions to Social Security address important American concerns and constituencies.

•

Current crime victims (plus occasionally a PIE inmate’s own victim), along with those
households and businesses bearing the costs of crime, together constitute the third largest
non-inmate beneficiary group benefiting from State inmate PIE jobholding, gaining about 9
percent of PIE inmate earnings. Because of Federal matching funds (currently contributing an
additional 40 cents per State dollar gotten), this group adds an amount equal to about 13
percent of PIE earnings. Victim compensation monies contribute to State-operated crime
victim compensation funds, reimbursing today’s crime victims for unreimbursed direct costs
of crime or income lost; in some instances, State crime victim compensation funds may be
used to meet a specific restitution order, compensating the identified specific victims of the
PIE inmate. PIE inmate contributions to State crime victim compensation programs appear to
constitute virtually 100 percent of incarcerated offender contributions to sampled State crime
victim compensation programs, at a time when all U.S. crime victim compensation programs
together compensate crime victims, on average, much less than 1 percent of the $50 billion
annual direct costs of violent crime. Because State taxpayers and employers also bear
significant shares of crime costs—in taxpayer-provided emergency services and in employerprovided increased health insurance premiums, sick leave, and lost employer productivity—
PIE inmate victim compensation deductions potentially affect important beneficiaries in
addition to the crime victim.

•

Federal Income Tax payers and programs dependent on the Federal budget constitute the
fourth largest beneficiary groups, garnering about 4 percent of PIE inmate incomes. In
addition to providing some relief to other Federal taxpayers, PIE inmate deductions for
Federal income taxes can be said to support National defense, income security (Supplemental

Executive Summary

Social Security, food stamps, Temporary Assistance to Needy Families, the earned income
tax credit), and Federally supported health care (Medicaid).
•

Employee Income Support, Ratepayers and Recipients, including workers compensation and
State and Federal unemployment compensation programs, constitute the fifth largest
beneficiaries of PIE inmate incomes, accounting for approximately 3 percent of estimated
gross PIE employer payouts. Because most PIE inmates cannot collect either workers
compensation or unemployment while they are incarcerated, PIE employer contributions to
Federal and State unemployment and State compensation programs directly relieve a State’s
other employers of some tax burdens while increasing funds available for ill, injured, or
unemployed workers in the State.

•

Children, families, and State child welfare programs appears to constitute the sixth largest
group of beneficiaries of PIE inmate incomes, garnering less than 1 percent of PIE inmate
incomes via court-ordered child support payments; voluntary payments made by PIE inmates
from post-PIE residuals are unknown. Child support payments responding to child-support
orders appear to be the lone trigger for PIE family support deductions, and no evidence was
found of PIE-induced voluntary payments where child support orders do not exist, or of
voluntary PIE deductions for other family members than children, such as to spouses or
significant others, to grandparent or other caregivers of inmates’ children, or to other relatives
or householders of the PIE inmate. Nevertheless, with estimates that PIE inmates, like other
inmates, are likely to be parents of minor children about half the time, and with available
evidence suggesting disproportionate likelihood of poverty or economic need in the
households of children being raised in grandparent or female-headed households, children
and taxpayers, the retired and elderly, and stakeholders representing the economic interests of
female householders constitute critical beneficiary constituent groups for the child support
deduction. State taxpayers providing income and direct services support to dependents,
particularly through welfare (Temporary Assistance to Needy Families), constitute a
specifically crucial beneficiary group.

What if PIE Inmates Earned Average American Salaries?
What if PIE inmates earned annual incomes more typical of average American wage earners? A
secondary purpose of the research is to estimate the beneficiaries and dollar size of financial
benefits if U.S. State inmate populations were generally employed in open-market jobs similar to
U.S. averages in skills, productivity, annual work hours, and wage levels.
Estimating potential incomes, beneficiaries, and benefits is highly dependent upon assumptions,
including assumptions for the percentage of U.S. inmates participating in the PIE program, hours
worked, productivity, wage rates, and deduction policies. Assumptions about post-PIE individual
and social behaviors could also play a part, for example, for State and Federal fines and other
court-orders enforcement policies and voluntary purchase preferences for much higher income
inmates.

xv

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

For this exercise, two groups of assumptions are used to set the range of low-to-high “reasonable”
incomes, benefits, and beneficiaries if PIE work were generally applied in the United States.
Obviously other assumptions could be chosen as well.
Critical assumptions for the low and high cases include the following:
•
•

•

•

In both cases, PIE includes both Federal and State prison inmates but not inmates of local
jails.
For the low case, 50 percent of PIE inmates (700,000 persons) work full-time at an annual
minimum wage income, $10,500 a year, and in the high case, 75 percent (1,050,000 persons)
work at the U.S. average annual wage for calendar 2000, $32,000.
Taxes: In both cases, PIE inmates encounter the 28-percent marginal Federal income tax rate
above $7,150. Because some States do not have income taxes, the low case assumes 50
percent of inmates encounter the average State income tax rate of 4.4 percent above $4,700
annual earnings, whereas the high case assumes 67 percent encounter State income taxes.
Other low and high case assumptions are affected by the proportion of inmates assumed in
employer or customer model programs (affecting payroll taxes) and deduction rates for other
programs, assuming higher deduction rates in the high case.

Tables EXS2 and EXS3 summarize the results of the low and high cases, showing National-level
PIE income and deductions if Federal and State inmates participated in PIE under assumed
conditions:
If assumptions and outcomes occurred as depicted in either scenario, benefits could occur as
follows:
First, were the United States to succeed in engaging a large share of the U.S. State and Federal
inmate labor force, their gross labor contributions would likely be much greater than currently
realized. Whereas PIE inmates’ annual gross incomes nationwide today total about $32 million
per year, generalized State and Federal inmate PIE work suggests total U.S. prison inmate labor
income potential ranging from about $7 billion to nearly $37 billion per year.1
Second, were U.S. State and Federal inmates generally successful in PIE at the assumed levels,
(1) the primary beneficiaries of general PIE inmate work would be Federal and State taxpayers,
Social Security contributors and beneficiaries, crime victims, families, and others than the inmate;
and (2) the greater the inmates’ productivity and incomes, the greater the proportional benefits to
others than the inmate. Said otherwise, the greatest beneficiaries of widespread inmate PIE work
would not be inmates but others than inmates, and the most significant stakeholders negatively
affected by current non-work by U.S. prison inmates are these same groups.

1

Including 600,000 local jail inmates suggests total U.S. adult inmate income potential ranging from $10 billion to $48
billion per year.

xvi

Executive Summary

Of course, many benefits have more than one beneficiary—such as wherever both tax payers and
recipients of tax benefits are affected—suggesting a sum of beneficiary shares much greater than
100 percent. And the exact distribution of many benefits is affected by their movement through
the tax and transfer system.
State taxpayers, including both households and business taxpayers, and the programs supported
by State taxpayers, presumably education, Medicaid, and roads and transportation, would likely
be the single greatest non-inmate beneficiary classes if PIE inmate work were generally
successful, probably accounting for roughly one-third of the total benefits of general PIE inmate
work. Under the assumed income levels, PIE inmates could contribute between $3,000 and about
$10,000 per year (10–33 percent) toward their individual costs of incarceration, particularly if
large shares of these amounts either entered the State’s general fund or explicitly reduced
taxpayer costs. Because many inmates are assumed to remain outside PIE, of course, the
aggregate average inmate contribution would fall short of that percentage; moreover, even under
very aggressive assumptions, State taxpayers would continue bearing the heavy majority of
incarceration costs.

Table EXS2. Results: Low Case Assumptions, U.S. PIE Income, Benefits, and
Beneficiaries, if PIE Were Generally Applied in U.S. Federal and State Prisons
(Assumed Income per PIE Inmate $10,500 per Year)
Category
Gross Employer Payout
Social Security (OASDI + HI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

($Billion)

Percent

7.68
0.28

100.0
3.7

0.03

0.3

0.03
0.33

0.3
4.3
Percent

Percent

Inmate Gross Income
Room and Board
Taxes, Federal Income (Liability)
Taxes, State Income (Liability)
Social Security (OASDI)
Social Security (HI—Medicare)
Victims Compensation
Family Support*
Other PIE Deductions*
Total Inmate PIE Deductions

7.35
2.08
0.66
0.09
0.28

95.7
27.1
8.6
1.1
3.7

100.0
28.3
8.9
1.2
3.8

0.78
0.15
0.00
4.03

10.1
1.9
0.0
52.5

10.6
2.0
0.0
54.9

PIE Residual**

3.32

43.2

45.1

**Total exceeds PIE deductions limit of 80 percent.

xvii

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table EXS3. Results: High Case Assumptions, U.S. PIE Income, Benefits, and
Beneficiaries, if PIE Were Generally Applied in U.S. Federal and State Prisons
(Assumed Income per PIE Inmate $33,250 per Year)
Category
Gross Employer Payout
Social Security (OASDI + HI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

Inmate Gross Income
Room and Board
Taxes, Federal Income (Liability)
Taxes, State Income (Liability)
Social Security (OASDI)
Victims Compensation
Family Support*
Other PIE Deductions*
Total Inmate PIE Deductions
PIE Residual

($Billion)

Percent

36.84
2.57

100.0
7.0

0.34

0.9

0.34
3.24

0.9
8.8
Percent

Percent

33.60
10.85
7.31
0.85
2.57
3.70
2.08
0.07
27.42

91.2
29.5
19.8
2.3
7.0
10.0
5.7
0.2
74.4

100.0
32.3
21.7
2.5
7.7
11.0
6.2
0.2
81.6**

6.18

16.8

18.4

**Total exceeds PIE deductions limit of 80 percent.

Nevertheless, the $2–$11 billion savings to State taxpayers from PIE inmate room and board
contributions could be expected to result in either (a) lower overall State tax rates and/or (b)
increased funding meeting State obligations for education, Medicaid, roads and transportation,
and other State-funded programs.
State taxpayers would enjoy other cost reductions or incomes as well. As PIE workers become
taxpayers and average incomes increase, nationwide, State income tax liabilities could increase
from less than $1 million a year today (Table B3) to $660–$850 million a year under general
inmate PIE engagement. Family support offsets—reimbursing the State for the proportion of
child support payments originally funded by State taxpayers through welfare—would likely refill
State coffers with tens of millions of dollars a year to a maximum near $1 billion (if virtually all
family support deductions offset State-supported TANF payments, which is highly unlikely).
Similarly, State taxpayers may enjoy reduced costs in funding State victim compensation
programs, and in supplementing State worker and unemployment compensation programs.
Repayment of court costs, along with contributions to payoff of fines, could also reduce State
(and Federal) taxpayer costs of administering the criminal justice system.

xviii

Executive Summary

The second-largest non-inmate category of beneficiaries would likely be constituted of Social
Security and other social safety net wage-earner and employer payees, and retired, widowed,
disabled, dependent, and other Social Security recipients. Combining employer and worker
contributions, PIE inmates’ total contribution potential to the Social Security—OASDI and HI
(Medicare)—is between $500 million to $5 billion per year, alone absorbing between 7 and 14
percent of PIE employers’ gross payouts.
Federal income tax payers, and all programs dependent upon Federal income tax collections,
constitute either the second or third largest recipient group. Depending upon income assumptions
for the PIE inmate population, PIE inmate Federal tax payment potential is estimated to be as
little as $660 million a year (about 8 percent of PIE earnings) under minimal assumptions and as
much as $7.3 billion a year to the Federal treasury (about 20 percent of PIE earnings) under the
most optimistic assumptions.
Crime victim compensation funds stand to gain significantly if PIE inmate work were generally
successful. Whereas inmates (with the exception of PIE workers) typically contribute nothing or
almost nothing to crime victim compensation programs today, if PIE were widely applied, total
contributions to U.S. crime victims compensations programs could range from about $0.8 billion
to about $3.7 billion a year, an amount far greater than currently paid out nationwide each year
(about $300 million) but still less than 1 percent of the estimated greater than $350 billion per
year cost of violent crime alone.
Family support, including to meet child support orders, could increase substantially, particularly
if a higher proportion of inmates incurred PIE deductions for family support, potentially
exceeding $2 billion under the most optimistic assumptions here, reaching as much as 6 percent
of total payouts.
Except under higher income scenarios, inmates likely remain the single largest beneficiaries of
PIE work, retaining at the conclusion of PIE deductions somewhere between 17 (under highest
income) and 43 percent (under least income) of gross annual payroll earned in PIE. Therefore,
inmates appear to retain very significant proportions of earnings, at least through the PIE phase.
Under the lowest assumed annual income, $10,500 per year, PIE inmates would be expected to
retain 45.1 percent, more than $4,700 per year; at the high end, inmates earning $33,250 per year,
could be expected to retain about $6,650 (the 20 percent minimum required by PIE), in this case a
function of both the progressive tax system and the assumed higher rates of deduction in the high
income case. The inference for social policy would appear to be that the rewards of inmate work
under assumptions of these scenarios, both generally and as inmate incomes increase, accrue at
least as much to non-inmate beneficiaries as to the PIE inmate workers.
The exercise also suggests that, at least theoretically, the 80 percent upper bound for PIE
deductions could be approached if PIE inmates’ saw their gross incomes reach or exceed $30,000
per year and tax and other deduction assumptions occurred. While unlikely, seeing conditions
under which the bound occurs nevertheless informs policymakers about conditions under which

xix

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

individual PIE workers could hit limits, and of the importance of measuring tax liabilities rather
than merely PIE-defined deductions or voluntary (but refunded) payroll deductions.

Caveats
While attractive, estimating potential benefits of general PIE participation is an incomplete
picture of both the challenge and the potential outcomes of PIE work. Among the issues yet to be
considered are the following:
•
•
•
•

correctional changes needed to provide safe, competitive, efficient, and profitable
workplaces;
techniques providing land, buildings, equipment, and services;
the methods by which correctional locations and PIE firms become attractive and
competitive; and
other obvious issues of widespread integration of inmate production into the civilian
economy.

Accommodating public attitudes and acceptance would be critical. Issues affecting both (a)
increased opportunities for U.S. domestic business expansion and additional civilian hiring and
(b) relationships and conditions of competition with civilian labor would likely need to be
addressed.
Nevertheless, estimating potential benefits of widespread inmate work suggests the order of
magnitude of potential good that could result from inmate work; the exercise suggests that
integrating inmates more fully into the American economy may yield benefits and produce
beneficiaries well beyond those traditionally recognized, and spur those beneficiaries to weigh
public policies in their light.

Disposition of Remainder After PIE Deductions
In order to more fully identify beneficiaries and quantify benefits, optional work for the research
extended to examination of PIE inmates’ prison accounts for a month during PIE and another
month before PIE, identifying and quantifying other sources of income, additional mandatory
deductions after PIE, and kinds of expenditures and net savings occurring with remaining monies.
Resource constraints and challenges in obtaining inmates’ individual records limited available
information. Nevertheless, of 968 inmates sampled for the research, PIE-month inmate account
information was obtained for 462 inmates, with at least some records from every sampled State;
pre-PIE month information was obtained for only 106 inmates, and not every sampled State
provided pre-PIE account information. In 104 instances, matched PIE and pre-PIE accounts were
obtained. Data were analyzed and accounted unweighted for this analysis.
Overall, examination of PIE inmate accounts indicates the following:

xx

Executive Summary

First, States do not appear to deduct additional large amounts from PIE workers’ incomes after
PIE deductions; the average State deduction was an additional 3 percent of inmates’ gross
incomes, or $14–$20 a month. The additional deductions were not able to be separately
distinguished, but appear to go primarily to fines, court costs, and other State or correctionsimposed charges, with other fractions going to crime victim restitution and child support. The
generally small scale of the additional deductions suggests that victim restitution and child
support are not being significantly supported beyond PIE deductions.
Second, most PIE inmates do not appear to shoulder significant additional family or child support
burdens beyond PIE deductions, either by court order or voluntarily. Roughly 90 percent of the
462 inmates for whom PIE-month data was available expended nothing during the observed PIE
month for persons with the same last name (a possible indicator of “family”), and 77 percent sent
nothing to persons with different last names. However, some PIE inmates transmitted significant
amounts; the average transmittal to persons with the same last name for inmates sending money
averaged $195 for the month, and $178 to persons with different last names. Overall, monthly net
transmittals (payments to persons plus reduced receipts from persons) to other persons indicate
additional net benefits to other persons on the order of 9 percent ($60) of inmate gross incomes;
3 percent ($20) to persons with the same last names and 6 percent ($40) to persons with different
last names. Transmittals to persons, however, do not necessarily mean support to other persons,
but could as well represent purchases or transfers to meet the inmate’s own savings or interests.
Third, in contrast to pre-PIE inmates, who appear to save almost nothing, PIE inmates likely
accumulate significant savings during PIE participation. On average, they appear to retain in
savings about $97 a month, which is an estimated 14 percent of their gross incomes and almost
one third of their discretionary PIE incomes after PIE and post-PIE deductions. While extensive
examination of inmate records strongly supports existence of inmate savings, nevertheless the
value reached in this research is a residual after estimated expenses and could be significantly
affected by estimation methods.
Fourth, by and large, inmate accounts do not directly indicate any wholesale movement by PIE
inmates into exercising normal financial responsibilities, such as for mortgages, insurance,
transportation, utilities, health care, or other outlays typical of American wage earners. On the
one hand, absence of such payments is hardly surprising, and even if they occurred, they would
likely occur indirectly, via a family member. On the other hand, the lack of either extensive
money outflows to “family” or to recognizable purchase categories in most instances gives the
appearance of PIE inmates’ continued isolation from normal economic participation or
responsibilities.
Finally, there is a striking difference between inmate incomes and outlays, particularly with
respect to meeting responsibilities, creating benefits and beneficiaries, and creating savings.
Whereas inmates while not in PIE are observed earning virtually nothing (about $57 a month
from all sources, including from family) and spending almost all of it for personal items at the
prison store, PIE inmates earn substantial amounts and contribute very significant shares to

xxi

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

taxpayers, social support programs, crime victims, to families, and for themselves. Comparison of
pre-PIE and PIE incomes and outlays leaves little doubt that PIE work yields significantly larger
financial benefits to important National constituencies than idleness or traditional inmate work.

xxii

I. Introduction
The “Prison Industry Enhancement” (PIE) provisions of the Justice System Improvement Act of
1979 (Title 18 U.S.C. 1761(c)) either created or recognized financial beneficiary groups not
traditionally associated with inmate work. As of December 31, 2000, about 3,700 State prison
inmates worked in open-market PIE jobs. The combination of market wages, deductions, and
normal employer contributions produces significant monies available for an array of taxes, room
and board, crime victim compensation, and family support, and for meeting other personal,
family, and State-required obligations, as well as for savings and discretionary uses. The purpose
of the research is to clearly identify the major financial beneficiaries of PIE inmate incomes and
to measure more exactly the dollar size of those benefits. A secondary purpose is to estimate the
beneficiaries and dollar size of financial benefits if U.S. State and Federal inmate populations
were generally employed in open-market jobs similar to U.S. averages in skills, productivity,
annual work hours, and wage levels. The research is funded by the U.S. Department of Justice,
Bureau of Justice Assistance (BJA), through the National Correctional Industries Association
(NCIA), and conducted by the George Washington University (GWU) Center for Economic
Research (the Center), with participation by selected PIE-State departments of corrections.
The research relies on PIE payroll and other records provided by a sample of participant PIE
States. It represents State-level PIE inmate incomes and employers’ payroll tax contributions for
one calendar year between 1998 and 2001, and represents PIE inmates working in employermodel (private firms) and customer-model (operated by departments of corrections) settings. No
personal interviews were conducted. All inmate records are treated as confidential. While not
guaranteeing confidentiality, the research does not disclose sampled States, firms, or the exact
survey year. Participation by States and firms was voluntary.
The research is limited to financial benefits and beneficiaries of PIE inmate payrolls. It is not a
comprehensive assessment of the net benefits of PIE, and does not account for either additional
financial or non-financial benefits of PIE. Nor does the research address PIE’s purported or actual
costs.
The research is not designed to be statistically representative of the gender, race, or other
demographic, criminal justice, or employment characteristics of PIE inmates. The research does
not address PIE benefits emanating from local jails participation in PIE.

1

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Background
In 2000, about 2 million of the Nation’s 281 million persons were incarcerated in local jails and
in State and Federal correctional institutions, about 1.2 million of them in State institutions (Table
1
2
1). The U.S. adult inmate population in 2000, including male and female inmates, was
approximately as follows:
•
•

•
•

1.1 percent of the adult working-age population ages 16–64 (182 million persons in 2000)
and 1.2 percent of the adult working-age population ages 20–64 (166 million).3
1.4 percent of the adult civilian non-institutionalized labor force age 16+ (143 million persons
either employed or actively seeking employment) and 1.5 percent of the labor force age 20+
(133 million persons).4
2.6 percent of the male civilian labor force ages 16+ (74 million persons), and 2.8 percent of
the male civilian labor force ages 20+ (70 million).5
The nation’s 807,000 black male inmate population alone equaled 10.9 percent of the black
male civilian labor force aged 20+ (7.4 million persons).6

Relatively few inmates have work assignments, and working inmate earnings are generally
meager. Despite more than 70 percent of State and Federal inmates reporting having been
employed before incarceration, 55 percent of State inmates have no work assignment (Table 1). 7
The 45 percent of State inmates with work assignments are reported working an average of 32

1

American Fact Finder, U.S. Census Bureau, “Total Population 2000.” Retrieved February 7, 2003 from American
Fact Finder on the World Wide Web: factfinder.census.gov.
2
Including persons incarcerated in adult and youthful offender institutions, excluding military and juvenile facilities.
3
American Fact Finder, “Total Population 2000.”
4
U.S. Department of Labor, Bureau of Labor Statistics (BLS), Series LNS11000000. Retrieved February 7, 2003 from
the Bureau of Labor Statistics on the World Wide Web: http://www.bls.gov.
5
Ibid.
6
Because inmates are not enumerated in the civilian labor force, percentages are calculated excluding inmates from the
labor force totals. Sources: Derived from U.S. Bureau of the Census, Bureau of Labor Statistics, and Bureau of Justice
Statistics data. Sources: Black Male Civilian Labor Force Ages 20+: U.S. Department of Labor, Bureau of Labor
Statistics, “The Employment Situation: December 2000,” Table 2, “Employment Status of the Civilian Population by
Race, Sex, Age, and Hispanic Origin (Seasonally Adjusted),” accessed February 7, 2003 on the Internet:
ftp://ftp.bls.gov/pub/news.release/History/empsit.01052001.news. Black Male Inmate Population: Derived by GWU
research staff from various sources: (1) Total Inmate population, see Table 1 sources. (2) Inmate distribution by race
and sex, estimated from separate Federal, State, and local statistics: (a) Federal data from BOP Quick Facts (see source
note to Table 1); (b) State data from U.S. Dept of Justice, Bureau of Justice Statistics and Federal Bureau of Prisons,
“Survey of Inmates in State and Federal Correctional Facilities 1997,” accessed February 7, 2003 on the World Wide
Web: http://www.icpsr.umich.edu:8080/NCAJD-STUDY/ 02598.xml.; (c) Local data from BJS Key Facts at a Glance
(see source note to Table 1, below).
7
Source: U.S. Department of Justice, Bureau of Justice Statistics, Special Report Incarcerated Parents and Their
Children, Table 13, “Pre-Arrest Employment, Income, and Homelessness of Inmate Parents in State or Federal Prison,
by Gender, 1997,” Study Number 2598, August 2000, NCJ 182335, retrieved February 7, 2003 from the World Wide
Web: www.icpsr.umich.edu/nacjd/home.html.

2

Introduction

hours a week, getting 56 cents an hour for their work.8 In 1997, inmates working in State
correctional industries averaged earning $716 that year.9

Table 1. U.S. Inmate Populations, by Level of Jurisdiction and Category of Work (2000)
Level of
Jurisdiction
Federal
State
Local
Total

Total
Inmates

Total
Working

123,141
1,193,192
621,149
1,937,482

99,623
537,540
163,983
801,146

Institutional
Maintenance
77,785
450,287
159,635
687,707

Agriculture

Traditional
Industries

150
30,494
0
30,644

21,688
53,057
4,313
79,058

Private
Sector
0
3,706
35
3,741

Sources: Inmate Populations: Federal Bureau of Prisons (BOP) populations, BOP, Quick Facts, “Federal Prison
Population Over Time/Drug Offenders (BOP facilities only), 2000,” retrieved on July 22, 2003 from the BOP Web site:
http://www.bop.gov/fact0598.html. State prison and local jail populations: U.S. Department of Justice, Bureau of Justice
Statistics, Key Facts at a Glance, “Correctional Populations,” retrieved July 22, 2003 from the Bureau of Justice Statistics
Web site: http://www.ojp.usdoj.gov/bjs/glance/tables/corr2tab.htm. State prison populations are derived from totals minus
Federal populations independently obtained. Working Populations: Obtained or derived from: (1) for State and for federal
except traditional industries, Criminal Justice Institute, Inc., 2001 Corrections Yearbook, Middletown, CT: Criminal Justice
Institute, Inc. 2001; (2) for Federal Bureau of Prisons traditional industries, UNICOR, UNICOR 2000 Annual Report,
Lexington, KY: UNICOR, 2000; and (3) estimates for local jail work derived using participation rates from U.S. Department
of Justice, Bureau of Justice Statistics, Special Report, Profile of Jail Inmates 1996, Washington, DC: Bureau of Justice
Statistics April 1998, NCJRS, NCJ164620.

In general, Federal and State laws prohibit inmates from open-market jobs, and inmates are
legally excluded from the U.S. labor force. The initial prohibition was the Hawes-Cooper Act of
1929 (Title 49 U.S.C. 60), which mandated that prison-made goods transported from one State to
another are subject to the laws of the importing State, and permitted the importing State to
prohibit the sale of prisoner-made goods. The primary Federal bar to inmate employment is the
Ashurst-Sumners Act of 1935 (Title 18 U.S.C. 1761(c)), making transport of prisoner-made
goods in interState commerce a Federal criminal offense. As amended, 18 U.S.C. 1761(c)
provides, “Whoever knowingly transports in interState commerce … any goods, wares, or
merchandise manufactured, produced, or mined, wholly or in part by convicts or prisoners …
shall be fined under this title or imprisoned not more than two years, or both [herein referred to as
the Ashurst-Sumners Act].” A second critical restriction, the Walsh-Healey Act of 1936, prohibits
the use of inmate labor to fulfill general government contracts exceeding $10,000. State laws
typically repeat and reinforce the Federal prohibitions.

Prison Industry Enhancement (PIE)
PIE stands as the primary exception to inmate work prohibitions for the open market. Under an
exception to Ashurst-Summers triggered by certification by the U.S. Department of Justice,
8

U.S. Department of Justice, Bureau of Justice Statistics, Survey of State Prison Inmates, 1991, Washington, DC: U.S.
Department of Justice, Bureau of Justice Statistics, March 1993, NCJ-136949, 27.
9
American Bar Association, Inmate Labor in America’s Correctional Facilities, Discussion Draft, April 1998, Figure
12.

3

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Bureau of Justice Assistance (BJA), and in the presence of conforming State law, State and local
corrections departments may engage inmates in open-market jobs in the “Prison Industry
Enhancement Certification” program popularly called “PIE.” Under PIE, State and locally PIEcertified programs may be excepted from the Federal prohibitions, and prisoner-made goods
under PIE may be transported in interState commerce.10 PIE opportunities are not available to
Federal inmates or institutions.
State and local departments of corrections with certified PIE programs may either host private
firms operating inside correctional institutions and managing civilians and inmates on site
(employer model), or departments may establish their own corrections-operated production units,
engage and manage inmates themselves under PIE requirements, and provide the inmateproduced goods to private firms for use and sale in interState markets (customer model).11 The
U.S. Department of Justice, Bureau of Justice Assistance administers PIE, with technical,
marketing, assessment, and training assistance provided under grant by the National Correctional
Industries Association (NCIA).
BJA PIE certification requires that programs meet certain conditions, such as voluntary inmate
participation, avoiding unfair competition with private sector business and labor, hourly wage
rates not less than those prevailing for similar work in the locality, and others.12 Certification also
includes States’ authority—but not obligation—to deduct from inmate workers’ gross wages (1)
Federal, State, and local taxes, (2) reasonable charges for room and board, (3) family support, and
(4) victims compensation. Total deductions, if taken, must equal at least 5 and not more than 80
percent of gross wages; after PIE-authorized deductions, the inmate must receive at least 20
percent of gross wages. However, subsequent to PIE, the inmate may encounter additional
obligations, such as for court-ordered fines and victim restitution.
Since PIE employment began in 1979, through the fourth quarter of 2001, all PIE inmate gross
wages (including State and local programs) have totaled more than $215.3 million dollars (Table
2).13
Deductions for room and board have been rising. In the fourth quarter, 2001, 3,700 State prison
inmates earned nearly $8 million in gross PIE wages, averaging about $2,200 each. As a result,
58 percent of fourth quarter 2001 PIE wages were reported going to deductions, 31 percent to
10

For more information prison labor and PIE, see either the National Correctional Industries Association Web site,
http://www.nationalcia.org, PIE Certification Program—Final Guidelines, section II, “Background of PIECP,” or the
Bureau of Justice Assistance national clearinghouse for information on inmate labor, http://www.correction.org/bja.
11
A variation of the customer model is the manpower model, in which the Department retains control of the production
unit and inmate workers, but the private employer manages production unit operations
12
U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Assistance, Prison Industry Enhancement
Certification Program Guideline, section III(a), “Program Guidance—PIECP Purposes,” Federal Register 64 (66)
(April 7, 1999): 17000—17014. The final guidelines are available on the World Wide Web at
http://www.NationalCIA.org, under “PIE Certification Program.”
13
National Correctional Industries Association, “Prison Industry Enhancement Certification Program: Cumulative
Data—1979 through Fourth Quarter (October 1 through December 31) 2001,” available on the World Wide Web:
http://www.nationalcia.org/pie4th01stats.pdf.

4

Introduction

board and room, and the other deductions not greatly different from their longer term cumulative
shares; the lower PIE residual of 42 percent to inmates for the fourth quarter offsets the higher
share to board and room.14

Table 2. Cumulative PIE Wages and Deductions, 1979 Through Fourth Quarter 2001 ($Million)
Category
Gross Wages
Room and Board
Taxes (All)
Victims’ Programs
Family Support
Total PIE Deductions
Residual

Cumulative $
215.3
55.4
29.2
19.8
12.6
117.2
98.1

Percent
100
26
14
9
6
54
46

Source: National Correctional Industries Association, “Prison Industry
Enhancement Certification Program: Cumulative Data—1979 through Fourth
Quarter (October 1 through December 31) 2001.”

14

Ibid.

5

II. Design of the Research
This Identifying Beneficiaries of PIE Inmate Incomes research was conducted by the Center for
Economic Research, Department of Economics, The George Washington University
(Washington, D.C.), under GWU research contract PTA24383/1/CCNS20186A with the National
Correctional Industries Association. Research began in September, 2002, and the final report was
delivered to NCIA in July, 2003.

Purpose
Through open-market work, nearly 4000 State prison inmates generate annual incomes appearing
to average more than three times traditional prisoner earnings, and, as shown by aggregate
statistics, their incomes and deductions provide substantial individual taxes, compensation,
purchasing and savings power, and support.
Nevertheless, neither PIE beneficiaries nor PIE’s benefits are well known, either by departments
of corrections, by the actual or potential beneficiary groups, or by the public and policymakers.
As a result, the PIE program may miss stakeholder and public or policymaker support available
were the benefits more clearly recognized.
Difficulties in understanding PIE benefits arise for a number of reasons. First, PIE deduction
categories are somewhat ambiguous, obscuring identification of beneficiaries and quantification
of benefits. The extent to which “taxes” includes Federal and State income taxes is unclear, and
the extent to which Social Security and Medicare deductions occur is uncertain, especially with
customer model programs exempted from exacting them. “Family Support” remains undefined
(Who are “family”?), and the extent to which “family support” coincides with mandatory or
voluntary child support, resources provided to spouses, partners, child-care givers (like
grandparents and relatives) and other required or voluntary payments to extended families or
significant others is unclear. The term “victim compensation,” while well defined, is likely
broadly misunderstood, and “board and room,” while generally descriptive, may ultimately reveal
little of who profits—taxpayers or others—from the deduction.
Second, PIE firms incur significant employer payroll costs not revealed in national PIE statistics.
Therefore, there are additional unrecognized beneficiaries for employer-shares of payroll taxes
for Social Security, unemployment insurance, and workers’ compensation programs, beneficiaries
and benefits tied to PIE wages but unaccounted in the National data.

7

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Third, net benefits to beneficiaries may differ substantially from the gross benefits identified by
aggregate deductions. PIE reported income tax payroll deductions may not accurately reflect
actual income tax liabilities, since inmates, like other taxpayers, may qualify for income tax
refunds or owe additional taxes based on annual incomes. On the other hand, retained PIE
incomes may relieve inmate families and others of needs to send monies to incarcerated families
members, such that net benefits to families from PIE work exceeds gross deductions for family
support.
Fourth, with nearly half (47 percent) of inmates’ gross incomes left as an unspent residual and
unaccounted among aggregate PIE statistics, the disposition of PIE incomes beyond PIE is of
interest. Anecdotal information suggests significant post-PIE obligations, including court costs,
fines, victim restitution (separate from PIE victim compensation), court-ordered child support, as
well as voluntary payments to children, families, and for other purchases, and for savings.
Finally, whatever PIE inmates’ current earnings, aggregate PIE statistics for a minor share of
State inmates, most likely working less than full-time or for the entire year, and working with
generally limited skills and in entry-level jobs, reveals little of the potential PIE beneficiaries and
benefits were U.S. State prison inmates fully employed year-round in skills and jobs yielding
more typical U.S. annual incomes.
The purpose of this research, then—as afforded by immediately available records—is to identify
more clearly who benefits (what groups) and the magnitude of their benefits from PIE inmate
incomes. To the extent beneficiary groups can be identified and benefits quantified, those
beneficiaries, other stakeholders, the public, corrections departments, and policy makers will be
better equipped to serve those interests, compare them with other priorities, assess PIE, and
evaluate or modify PIE’s contribution to the U.S. economy. Corrections departments designing,
implementing, and garnering public support can better assess decisions in light of beneficiary
consequences. Further, identification of beneficiaries and benefits can assist efforts modifying or
improving PIE to improve the magnitude, distribution, and equity of benefits available from the
PIE program.

Design
Identifying Beneficiaries of PIE Inmate Incomes implements part of the U.S. Department of
Justice, Bureau of Justice Assistance-funded technical assistance and monitoring mission of the
National Correctional Industries Association, and arises from 2001 proposals within the NCIA
Research Advisory Board (RAB) seeking more precise information on PIE’s meeting its mission
of providing benefits: “Through inmate wage deductions, to increase advantages to the public by
providing departments of correction with a means for collecting taxes and partially recovering
inmate room and board costs, by providing crime victims with a greater opportunity to obtain
compensation, as well as by promoting inmate family support.”1
1

8

U.S. Department of Justice, Prison Industry Enhancement Certification Program Guideline.

Design of the Research

The research relies primarily on records and other information provided by a statistical sample of
State PIE programs and PIE employers. No personal interviews were conducted.
The research focuses on identifying the direct beneficiaries and benefits of PIE inmate incomes in
PIE programs operated by State-level departments of corrections representing about 99 percent of
PIE inmates. For this research, “inmate gross income” means the sum of PIE inmates’ PIE wages
for one calendar year as reported to the IRS. A broader measure, “employer gross payout,”
includes inmate gross income plus any employer- or State-paid payroll taxes (Social Security
including Medicare, unemployment compensation, and workers compensation) and any other
benefits directly tied to PIE employment and wage-earning, such as annual and sick leave, health
and retirement benefits, employee stock ownership plans (ESOPs) or other employee investment
plans, or other employer or corrections provided worker benefits tied to PIE employment (such as
holiday parties, good-time credits, or other privileges).2
“Beneficiaries” are those persons or entities who (1) directly receive the monies deducted from
PIE inmates’ wages, (2) by virtue of inmate’s increased self-sufficiency realize immediate
savings from reduced requirements for outlays to the inmate, (3) are relieved of tax burdens, or
(4) can be reasonably considered ultimate users for end-use consumption. Of practical necessity,
the indirect, intermediate, and secondary beneficiaries (such as increased employment in firms
supplying the businesses producing goods and services for inmates’ families) are not explicitly
considered. Moreover, because the research relies on corrections and business records, ultimate
disposition of incomes transferred from inmate payrolls and accounts to other persons and
accounts remains unidentified. Thus the research offers no definitive information on who actually
receives significant monies transmitted by PIE inmates during or at the end of incarceration or
their uses of those monies.

Confidentiality
The Beneficiaries research was conducted under terms approved by the George Washington
University Office of Human Research, Institutional Review Board (IRB) (Non-Medical), IRB
Number U080222ER.3 The research meets Federal requirements (including University of
Michigan Inter-University Consortium for Political and Social Research standards) protecting
research subjects.4 GWU IRB approval included providing assurances and methods protecting
confidentiality of human subjects, clearly notifying respondents of the purposes, voluntary nature,

2

“Good-time credits” are days credited against a criminal sentence.
The Non-Medical IRB provides oversight for all non-clinical or non-medical research in the social sciences, including
research into human behavior. All research projects at GWU must be reviewed and approved by the Non-Medical
Center prior to enrolling any subjects. GWU institutional review guidelines are available on the World Wide Web at
http://www.gwumc.edu/research/human/nonmedical.htm.
4
Inter-University Consortium for Political and Social Research, “Preserving Respondent Privacy/Confidentiality,”
located at http://www.icpsr.umich.edu/access/deposit/confidentiality.html, and “Guide to Social Science Data
Preparation and Archiving,” located at http://www.icpsr.umich.edu/access/dpm.html.
3

9

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

content, and methods of the study, being accessible to and responding to comments and questions
from respondents, and obtaining documented permission from each respondent.

Inmate Confidentiality
Despite inmate information more frequently treated as public record or even available on State
Web sites, the beneficiaries research includes strict and extensive confidentiality protections for
inmate information. Individual inmate information is offered complete confidentiality. Identifiers
(Names, Social Security Numbers, Inmate ID’s) are removed from information on individual
inmate records contained in publicly accessible data bases maintained by the University of
Michigan Inter-University Consortium for Political and Social Research (ICPSR); other
individually unique or revealing information is either generalized or removed from individual
records.
To further ensure inmate confidentiality, individual inmate demographic and criminal justice
information is not linked to individual’s PIE work and beneficiaries information.

Firms and States
Because some information essential to this research may be unique to specific firms or States and
therefore capable of revealing a State or firm to knowledgeable parties, the sampled States and
firms are not guaranteed complete confidentiality. However, protecting firms and States from
disclosure is a critical operational objective of the research, both for the protection of individual
inmates and to protect firms and States as well. As a consequence, firm and State identifiers are
removed and, to the extent possible, unique data characteristics are generalized, otherwise
obscured, or deleted.
To further protect firm and State confidentiality, the number of sampled States and firms in the
research is not disclosed nor is the specific year studied revealed, except that it covers one
calendar year between 1998 and 2001. As a result, data and tables in the text will describe various
years during the 1998–2001 time period (and occasionally other years), in part because available
data may be limited to those years and in part for protection of confidences.
Anecdotes and other data—particularly descriptive of rules, customs, and outcomes—in the
report may be obtained from PIE or other States whose firms and inmates were not sampled for
the research. The selections are provided primarily to more fully describe relevant characteristics
of PIE programs, but also in order to further obscure the sample selection. Data in research
tabulations and their specific discussion, of course, emanate solely from sampled sites.

Protecting Confidences
Both GWU research staff and State respondents collaborated to protect inmate, firm, and State
data. All information containing identifiers was secured. After assembly, GWU working files

10

Design of the Research

have removed names, Social Security numbers, inmate ID’s, and firm and State identifiers.
Further, GWU staff informed and frequently reminded States and firm correspondents to speak of
the research and exchange information only with persons and offices whose involvement was
necessary for conduct of the research, and then only under the correspondent’s agreement to
maintain confidentiality. Research staff typically referred to sampled States by assigned number
rather than by name.

Confidentiality and the National Correctional Industries Association
For information integrity and to protect PIE States and firms from concerns about NCIA
oversight, the National Correctional Industries Association was not informed of the selected
States. Moreover, both participating and non-participating PIE States were asked to not disclose
their GWU research relationships to NCIA or in any way inquire about, discuss, or disclose the
research as part biennial PIE assessments taking place during the some of the same months as this
research.

Sampling
The Beneficiaries research was designed as an approximately 10-percent stratified simple random
sample of State-level PIE inmates. Reported total PIE inmate populations as of December 31,
2001, were used as a proxy for total PIE inmate populations and State shares of total populations
in State PIE programs at any time in the one selected 1998-2001 survey year. State populations
were stratified by (1) PIE model type, employer model States separate from customer or
manpower model States, and (2) by size, that is, the number of inmates in each State within each
model type. Stratification by PIE model reflects differences in deduction requirements for the two
models; stratification by size acts as a proxy for assumed other program differences. No objective
information is available on either the total number of PIE inmate workers during a year or of their
distribution by individual States; hence the reliance on December 31 counts.
No objective information was readily available affording sampling by demographic (sex, race,
age, prior labor force status) or criminal justice (offense, sentence) characteristics. Despite
theoretical advantages of selecting 10 percent of every PIE State’s inmate workers, research
efficiency and resource limitations instead resulted in instead selecting a small number of PIE
participant States whose PIE inmate workforces together represent approximately 10 percent of
the total U.S. PIE inmate worker population. The consequence of selecting all inmates in some
programs rather than some inmates in all programs, lessens the likelihood that actual samples are
fully representative by demographic, criminal justice, and even PIE characteristics.
Because Federal and State income tax rates reflect annual incomes, the defined measurement
period for the research is the calendar year.
Therefore, the actual sample for the research became all PIE inmate workers receiving a W2
during the defined calendar research year (between 1998 and 2001) in each of a randomly

11

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

selected small number of State-level PIE programs whose PIE populations in total approximately
equaled 10 percent of the employer-model and customer/manpower model PIE worker
populations based on Nationally reported December 31, 2001, statistics.
Sampling occurred based on nationally reported PIE State-level statistics for the fourth calendar
quarter, 2001. While necessary, the sampling offered two additional limitations. First, the actual
survey year might not be 2001, but any single year 1998–2001; because PIE employment can
vary notably from State to State and year to year (and PIE firms differ and their employment
complements change as well), actual samples may or not equal 10 percent of actual numbers for
the actual survey year. More significant, national statistics (against which 2001 totals could be
compared for the survey year, if different) were found to represent only the number of persons
employed at the end of the calendar quarter, or more properly, the number of positions PIE
inmate workers occupied at quarter’s end. Therefore, depending upon worker turnover rates—
unknown for both the sampled and all other PIE States—the actual number of PIE workers
receiving a W2 in the survey year, on which sampling is ostensibly based, is unknown in advance
of the actual survey. While it was obvious that the actual number of PIE workers in any State in a
survey year would exceed the year-end total of positions, the extent of the excess and the degree
of difference in turnover rates among PIE States was unknown. Of necessity, for this research all
States are assumed to experience turnover rates matching that of the sum of sampled States. A
critical assumption for the research is that whatever share any one PIE State’s December 31
(quarterly) PIE inmate population contributes to the U.S. December 31 total PIE inmate
population is the same as its share of all PIE inmates who worked at any time during the survey
year. 5
Sampling also distinguished “small” (few PIE inmate workers on December 31, 2001) from
“large” States to ensure that both were potentially included. Fourth quarter, 2001 PIE statistics
indicated 3,680 State-level PIE positions at the end of 2001, 2,189 among employer-model
programs and the remainder among customer/manpower or mixed PIE programs. PIE program
populations in Nevada, which offers both models, were separated according to model type.
Within employer-type and size groups, and numbering the States in each group, sampled States
were randomly chosen using a table of random numbers to yield approximately 400 inmate
positions and an expectation of some larger number representing all inmates who worked at any
time during the sampled year. Actual numbers of PIE workers among the States differed
substantially from expectations in some cases, such that the actual number of sampled inmates
rose to 968 rather than approximately 400.

5

An additional practical difficulty of sampling without knowing turnover rates is the potentially significant added cost
of collecting, editing, coding, formatting, and processing the unknown number of records in excess of the
approximately 400 indicated by fourth-quarter stock statistics.

12

Design of the Research

Procedures
Peer Reviews
The Beneficiaries design was first introduced to the NCIA Research Advisory Board as an
informal proposal in the Spring, 2001, and refined in response to informal comments during
subsequent months. After NCIA approval, the formal research design was formally circulated to
the RAB in September, 2002, and comments addressed. Draft results were circulated for RAB
and GWU review in early Spring, 2003.

Coordinating with PIE States
From the beginning, research staff were concerned about PIE States’ voluntary cooperation.
Outsiders may be seen as threatening. Rewards for assisting research are distant, small, and
uncertain, and costs are immediate, potentially significant, and fairly certain. Already burdened
PIE-offices take on significant additional work supporting research, in communicating with
research staff, obtaining clearances, locating information, submitting data, and in providing
researchers the surrounding facts and information elucidating the data, all within limited budgets
and time. States must protect some inmate information, making access to inmate data potentially
difficult. PIE staff must protect State and corrections’ interests; moreover, PIE and other
correctional programs continuously risk public wrath and potential restriction or elimination of
programs, raising the possibility that cooperation could yield negative outcomes for participant
States. Risks to State PIE programs increased if information provided in the research were subject
to the oversight and compliance activities of NCIA.
Therefore, in order to meet PIE States’ concerns, minimize their burdens, and also to
communicate and work effectively with them, the beneficiaries research included the following
components:
•

•

•

The research was limited to essential information; nonessential items were dropped. Every
requested data element was evaluated for its contribution to the research and its likely burden
on providing States. Research staff especially considered the likelihood that data elements
were confidential or not releasable without individual inmate permissions—and worked to
avoid such data.
The research was limited to normal record systems possessed by the corrections department
or PIE firms. In particular, personal interviews—which would require obtaining individual
permissions and researchers’ confidential meetings with inmates (raising security issues)—
were rejected.
In order to confirm legitimacy of the research, (1) directors of correctional industries and (2)
PIE representatives were introduced to the research by letter from the BJA, which endorsed
the effort and recommended participation if asked, informed potential respondents of
confidentiality and confidentiality measures, and provided recipients contacts at the

13

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

•
•

•
•

•

University. The letter also referred recipients to the NCIA Web site for corroboration of
NCIA support.
The NCIA Web site posted endorsement of the research, repeating content of the notifying
letter.
NCIA recused itself from knowing sampled States. To ensure preventing a participating
State’s voluntarily supplied information being diverted for PIE compliance, a firewall
between NCIA and all the PIE States was constructed relative to this research. GWU research
staff did not notify NCIA of selected States and all States were asked not to contact NCIA
about the research, whether or not they were involved in it. Participating States were asked
not to discuss the research with one another. And both the PIE States and the PIE assessment
teams conducting biennial assessments coincident with the research were asked to not discuss
the research or expose materials or other information connected to the work.
To test and improve the collection cycle, GWU planned testing and completing one State’s
data collection, formatting, and analysis before proceeding to subsequent sampled States.
GWU research staff telephoned each sampled State’s correctional industries director as the
GWU single point of contact, then provided detailed information outlining the research,
identifying the key record systems and data elements and their uses, suggesting an approach
and identifying key participants, providing easy contact information, providing each
participant a State agreement to participate, and asking for State recommendations for
accomplishing the task. E-mail provided the usual method of delivery to the States, with
signed facsimiles constituting the most common method of transmitting signed acceptances.
See Appendix B for copies of documents provided each State and firm.
The GWU sampling design included identifying substitute States in the event of initial
selections’ withdrawal.

Whether through these measures or for other reasons, none of the researchers’ concerns was
realized. Cooperation by all sampled States, without exception, proved stellar, and emphatically
contradicted research fears. Both PIE and other corrections offices proved not only cooperative,
but insightful and helpful in response, particularly in locating, preparing, transmitting, and
explaining requested data. States’ assistance was particularly noteworthy in the extent of effort
obtaining and providing inmate account information and in explaining details of inmate income
and expenditure categories.
•
•
•
•

No sampled State declined to participate; substitute States were not needed.
No required data was missed for any participating State.
Permissions were relatively straightforward in almost all instances.
Some requested data was found to be of public record; all requested records were determined
to be in the legal possession of the State departments of corrections and releasable without
individual inmate approval.

State PIE staff addressed concerns within their departments and obtained needed data and
cooperation without any intercession by GWU researchers.

14

Design of the Research

Some initial bumps did occur, however, resulting in an almost immediate change in procedures.
Unfortunately, the first State’s permission was delayed, and first reactions by some sampled
States were hesitant. Specters of denial of access were raised, and labyrinthine permission cycles
loomed, prompting GWU researchers to relatively quickly scrap plans to complete the first State
before contacting subsequent ones in favor of immediate and coincident collection from every
State on whatever most-rapid schedule might occur.
Further, the GWU proposal of each State identifying and organizing key participants and
permissions in advance—particularly with prison offices beyond PIE and correctional
industries—failed to occur on any occasion. Instead, (so far as could be observed by research
staff) PIE and industries staff tended to organize permissions and plans within the industries
hierarchy and only later addressed outside assistance—such as from inmate records offices—
thereby occasionally triggering temporary uncertainties (some concerns) and brief delays.

The Permission Process
Participation in the Beneficiaries research is voluntary for all participants. Private PIE firms’
participation is voluntary and independent of the host State’s participation, and the research
design calls for independent contact with each group. Contact with private PIE firms was not
anticipated for customer model firms, since all records on inmate work are in the possession of
the departments of corrections. In fact, little contact with employer model firms was required
except to confirm tax contribution rates and inmates’ non-pay benefits.
The GWU human subjects requirements call for written notification of subjects, description of the
research, contact information for meeting subject interests and questions, and written agreement
to participate. Therefore, after introducing the research, before obtaining any survey information,
GWU obtained written agreement from every participant State.

15

III. Income, PIE, and Deductions Defined
With few exceptions, the little money an inmate receives from outside or from traditional
industries has but one major beneficiary, the inmate. Indeed, inmate accounts suggest general
separation of prison inmates from normal adult financial responsibilities, both from supporting
themselves, and from broader personal, family, and social obligations.
On the other hand, PIE jobholding either recognizes or establishes personal, family, and social
beneficiaries, as a function of (1) normal participation in open-market jobs, (2) special obligations
imposed by PIE requirements, (3) enforcement of other child support and criminal justice
obligations, and (4) voluntary expenditure decisions of PIE participants. Moreover, because PIE
firms incur additional tax obligations tied to wages, employer gross payouts exceed the hourly
wages paid employees and inmate workers.
Published national statistics show significant reallocations of inmate incomes. In the third quarter,
2002, for example (Table 3) 58 cents of every dollar of gross PIE inmate wages was diverted to
room and board (30 cents), taxes (13 cents), victims programs (9 cents) and family support (6
cents), with just 42 cents accruing to the PIE inmate workers. Thus, the primary beneficiaries of
PIE inmate incomes appear to be stakeholders other than the inmate.
But who are these stakeholders and what do the PIE deduction categories mean? And what
additional employer-made payments tied to inmate hourly incomes also reach beneficiaries?
Table 3. Cumulative PIE Wages and Deductions, 1979 Through
Fourth Quarter 2001 and Third Quarter 2002 ($Million)
1979 Through 4th Qtr 2001
Category
Gross Wages
Room and Board
Taxes (All)
Victims Programs
Family Support
Total PIE Deductions
Residual

Cumulative $

Percent

215.3
55.4
29.2
19.8
12.6
117.2
98.1

100
26
14
9
6
54
46

3d Quarter 2002
$
8.25
2.50
1.09
0.75
0.46
4.80
3.45

Percent
100
30
13
9
6
58
42

Note: Data include all reported PIE wages and deductions, including from a small number of non-State level PIE programs
occurring in local jurisdictions.
Source: National Correctional Industries Association, “Prison Industry Enhancement Certification Program: Cumulative
Data—1979 through Fourth Quarter (October 1 through December 31) 2001.”

17

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Income
Inmate Gross Income
For this research, “inmate gross income” means the sum of PIE inmate workers’ hourly wages,
including regular work hours, overtime, awards, and bonuses reported as part of annual earnings
to the IRS. “Inmate gross income” does not include any other employer payouts, such as for
health benefits, retirement, stock options, or benefits in kind (non-monetary gifts, awards,
company picnics or parties).

Gross Employer Payout
A broader term is also used in this research, “gross employer payout,” which includes all of
inmate gross income plus specified employer tax and insurance contributions tied to the PIEworkers’ hourly wage rates, namely all of employers’ Social Security, unemployment insurance,
and workers compensation payments.
Nationally reported PIE data describe inmate gross income but do not account for additional
employer contributions. Therefore, where used, the employer gross payouts used in this research
exceed amounts reported in national statistics by the identified and measured amount of
additional employer contributions.

Defining PIE Deductions
Only a general summary of PIE rules is presented here. Persons interested in detailed
understanding of PIE requirements and guidelines should consult the NCIA Web site, “PIE
Overview” and “Final Guideline” at http://www.NationalCIA.org. Moreover, specific rulings
applicable to individual States’ PIE programs may yield specialized deduction practices.
Under 18 U.S.C. 1761 (c)(2), PIE programs are not required to take deductions from inmates’
wages (Page 25, Guidelines), although other provisions of law, such as Internal Revenue Service
(IRS) requirements, may apply. If programs elect to take deductions from inmate gross incomes,
PIE deductions may only be withheld for purposes and not to exceed amounts specified under 18
U.S.C. 1761(c)(2), including for (a) payment of taxes—Federal, State, local, (b) reasonable
charges for room and board, (c) allocations for family support pursuant to State statute, court
order, or agreement by the offender, and (d) contributions to any funds established by law to
compensate victims of crime. Deductions, if taken, may be no more than 80 percent of gross
wages).1 Total PIE deductions may not exceed 80 percent of inmate gross income.2 PIE inmates
must be paid, credited with, or otherwise benefit legally from, the 20 percent gross remainder,

1
2

18

U.S. Department of Justice, Prison Industry Enhancement Certification Program Guideline, 25.
Ibid., 19.

Income, PIE, and Deductions Defined

although the 20 percent gross remainder may be reduced to settle other post-PIE legal obligations,
including for fines and restitution.3

Reasonable Charges for Room and Board
Deductions for room and board constitute by far the largest single PIE deduction, equaling about
half of total PIE deductions. Moreover, given that the cumulative room and board deduction
averages 26 percent of inmate gross income since 1979, but 30 percent for the third quarter, 2001,
the room and board deduction appears to have been substantially increasing over time as a
percentage of gross inmate income (Table 3).
Room and board deductions must be used to lower taxpayer costs of inmate incarceration. “The
legislative history of 18 U.S.C. 1761(c) reflects a Congressional intent to permit the use of the
room and board deduction to lower costs otherwise incurred by the public for inmate
incarceration.”4
In general, although reserving the rights to review, the Bureau of Justice Assistance defers to
individual PIE State determinations (by the State’s Chief Correctional Officer) with respect to
both the size and the applications of the room and board deduction. Where private prisons take
the deduction, BJA requires that the agreement between the department of corrections and the
private institution be documented. BJA recognizes that broader contractual arrangements between
a private entity and the State represent the interests of the taxpayer and leaves to the individual
State the determination that a specific use of the room and board deduction yields an ultimate
consequence reducing taxpayer costs.
“Room and board” covers far more than sleeping accommodations and meals, and instead
includes the costs of most necessities and some additional provisions covering most items for the
security, justice, rehabilitation, and daily lives of State prison inmates. The costs of inmate
confinement are borne principally by State taxpayers, although some incidental costs are also
borne by outside parties, usually the spouses, parents, other family or friends of inmates, and by
not-for-profit and other groups providing services to inmates. For State prisons, “room and
board” can be considered encompassing almost all security, criminal justice, and living costs,
covering both the capital and the operating costs of (1) structures, equipment, grounds, roads, and
their maintenance, (2) security and enforcement, (3) records, court, and other legal services and
support, (4) utilities, (5) food service, (6) shoes and clothing, (7) health care, (8) education,
libraries, training, rehabilitation, and counseling, (9) religious, recreation, and sports support, and
(10) external relations, including with family, press, and other visitors.

3
4

Ibid., 26.
Ibid., 26.

19

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Taxes
The requirement to pay income taxes arises from Federal and State requirements, not PIE, but
PIE inmates’ income taxes are considered PIE deductions.5 Taxes include Federal, State, and
local income taxes and also Social Security contributions paid by PIE inmate workers (but not
those paid by employers). Income and Social Security taxes garner 13–14 percent of PIE inmates’
gross incomes, and account for one-fifth to one-fourth of taken PIE deductions.

Federal Income Taxes
Federal income tax rates are based on annual gross incomes, and tax rates are progressive. The
combination of exemptions, deductions and progressive rates means that initial levels of annual
income go untaxed (are exempt or excused from taxation), while annual incomes above
thresholds are taxed at increasing rates. Married inmates should file “married, filing separately.”
Although undoubtedly some few PIE inmates file as married, and enjoy business, interest,
retirement, and other incomes, and may also itemize deductions separate from PIE work,
nevertheless, the overwhelmingly weight of anecdotal evidence suggests that most inmates have
few, if any, other tax-affecting circumstances, and that PIE inmates file as single taxpayers. This
research assumes PIE-inmates to be “single” tax filers whose only taxable annual income is PIE
earnings.6 The research also assumes that PIE inmates are normal taxpayers and therefore know
about, calculate, file for, and receive refunds for Federal tax overpayments incurred as payroll
deductions.
This research includes PIE earnings during one calendar year 1998–2001. During that period the
combination of the Federal standard deduction, a personal exemption, and progressive tax rates
yielded the following income thresholds, at or below which single taxpayers were liable for no
Federal taxes, above which they incur a 28 percent marginal to a maximum of $33,263
(arithmetic mean), and finally a 31 percent marginal rate for any reasonably conceivable
additional annual income.
Table 4. Federal Income Tax Rates, Single Filers
Showing Annual Gross Income Thresholds
Calendar Year

0 tax Rate ($) to -

28% tax Rate ($) to -

6,950
7,050
7,200
7,450
7,163

32,300
32,800
33,450
34,500
33,263

1998
1999
2000
2001
Mean
Source: Internal Revenue Service.

5

Ibid., 16.
PIE inmate earnings from institutional support or traditional prison industries are legally considered gratuities and are
not taxed.

6

20

Income, PIE, and Deductions Defined

Calculations for this research, therefore, assume PIE earners pay no Federal income taxes on the
first $7,150 PIE earnings in the survey year, 28 percent for relevant gross incomes $7,151 to
$33,250 per calendar year, and 31 percent above $33,250. For 2001, an intermediate 15 percent
rate was introduced up to $13,450; however, because this intermediate rate was not characteristic
of the overall period, it is not used here.

Earned Income Tax Credit (EITC)
Were they not inmates, inmate tax filers with relatively low incomes could be eligible for the
Federal “Earned Income Tax Credit” (EITC, sometimes referenced as “EIC”), a credit for lowincome working individuals and families. The EITC reduces the amount of Federal tax owed and
may result in a refund, sometimes even if no taxes were paid. The EITC is designed to yield a
7.66 percent refund for incomes up to $4,750 (for a $364 refund) and then decline to nothing
above $10,710.7
Because inmates are Federal taxpayers, usually earn less than $10,710 a year, and often have
children and low-income spouses, the EITC appears opportune for increasing inmate and inmate
family incomes. On the other hand, inmates already receive board and room and other in-kind
benefits. In any event, PIE inmates, despite being Federal taxpayers, are explicitly excluded from
the EITC: “Earnings while an inmate. Amounts received for work performed while an inmate in a
penal institution are not earned income when figuring the earned income credit. This includes
amounts received through a work release program or while in a halfway house.”8
Accounting for progressive Federal income tax rates suggests that most PIE inmates’ Federal tax
liabilities may be closer to zero than suggested by reported data. If all 3,700 or so PIE inmates
reported for the third quarter, 2002, worked all four calendar quarters at exactly the same average
incomes, generating about $33 million dollars for all of 2002, each PIE inmate’s Federally taxed
income would total no more than $1400 and generate, about $400 Federal income tax liability per
inmate, less than the $500 or so suggested by the reported data after accounting for estimated
Social Security deductions. Given that many inmates work much less than a full year, it is
conceivable that actual Federal tax liabilities for most inmates are quite small or zero, and that
PIE deductions for Federal taxes often subsequently result in refunds.

State Income Taxes
State income tax structures are difficult to generalize. Income tax rates for forty-three States and
the District of Columbia differ individually with respect to exemptions, deductions, and tax rates;
seven States have no income taxes. This research located no standard publications yielding
“average” State income tax thresholds and rates. Therefore, section VI of this research,”Benefits
if PIE Participation Were Widespread,” broadly assumes an overall PIE worker’s State income
tax structure as shown in Table 5.

7

U.S. Department of the Treasury, Internal Revenue Service, Publication 596, Earned Income Credit (EIC), catalog
number 15173A (Revision 2002).
8
U.S. Department of the Treasury, Internal Revenue Service, Publication 596, 10.

21

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 5. State Income Tax Rates, Single Filers
Showing Annual Gross Income Thresholds
0 percent to
4.4 percent above

$4,700
$4,700

Source: Derived from State of Wisconsin, Fiscal
Bureau, “Individual Income Tax Provisions in the States
#4,” Informational Paper, Madison, Wisconsin, January
2001.

State-level researchers are encouraged to apply their own State income tax regulations to
estimates of PIE worker income tax liabilities.

Local Income Taxes
This research is not aware of local income taxes being levied on PIE inmate incomes, and no
research on local income tax rates was done or calculations included.

Social Security
Like civilian employees, PIE inmate workers for private firms in employer model programs also
pay Social Security Taxes, including both the 6.2 percent OASDI (Old-Age, Survivors, Disability
Insurance) and the 1.45 percent HI (Health Insurance, Medicare) premium, together totaling 7.65
percent of gross income, to a ceiling of $84,500 (with no income limit on the HI portion).
PIE inmates in customer model programs are not obliged to pay Social Security taxes because
they are considered working for State-owned entities (“Services performed in an institution by an
inmate in the employ of a State, a political subdivision, or a wholly-owned instrumentality are
excepted from social Security employment by 26 U.S.C. 3121(b)(7)”). Both the Social Security
Administration and the Internal Revenue Service concur that inmates working for Department of
Corrections-owned production units (Customer and Manpower model employers) are exempted
from Social Security employment by 26 U.S.C. 3121(b)(7). Section 3121(u)(2)(B)(ii)(II) also
exempts such work from the Medicare contribution.9 Nevertheless, despite the lack of obligation,
PIE inmates in customer model programs may be contributing to both Social Security and
Medicare, and the contributions, when made, are considered PIE contributions. Employer
contributions to Social Security, if made, are not enumerated among nationally reported PIE
statistics.

Unemployment Compensation
Section 26 U.S.C. 3306(c)(21) excepts from including as employment “service performed by a
person committed to a penal institution.”10 As a result, PIE employers are not obliged to
contribute for unemployment compensation under either the employer or the customer/manpower
models, although some do. Because statistics do not account for employer contributions, PIE
9

U.S. Department of Justice, Prison Industry Enhancement Certification Program Guideline, 14.
Ibid., 15.

10

22

Income, PIE, and Deductions Defined

employers’ unemployment contributions, if made, are not enumerated among nationally reported
PIE statistics.

Workers’ Compensation
Employers, both in employer and customer model PIE programs, pay workers’ compensation,
either as members of authorized State workers’ compensation programs or, in States denying
inmates participation in their workers’ compensation programs, comparable provision of such
benefits as provided for public sector and private sector employee.11

Victims Compensation
Overall, deductions for victims compensation consume about 9 percent of PIE inmate’s gross
wages (Table 3).
However, the victims compensation deduction is not designed to meet individual PIE inmate’s
victim restitution obligations. Deductions for victims compensation contribute to Federally
sanctioned State victim compensation funds compensating current crime victims; in some
instances, the PIE victims compensation fund also provides restitution to the PIE inmate’s
specific victim, if the monies are deposited in a State fund established by law to facilitate
offenders’ restitution obligations.12
In general, three important distinctions are made in law and practice distinguishing victims
assistance, victims compensation, and restitution: “Victims Assistance” refers to monies spent by
States, not for offsetting victims’ losses, but for providing direct services to victims and for other
victim-related services, such as for crisis intervention, counseling, emergency shelters, emergency
transportation, and advocacy. PIE deductions may not contribute to victims assistance.
“Restitution” refers to direct payment by an offender to his or her own victims. Restitution
occasionally occurs via a State’s crime victim compensation fund (channeled through the fund)
and can qualify as a PIE deduction. However, court-ordered restitution appears to occur primarily
subsequent to and separate from PIE deductions. There is no defined lower or upper bound on
restitution taken separately from the victims compensation deduction.
“Victims Compensation” refers to a direct reimbursement from a State-operated crime victim
compensation program to or on behalf of specific crime victims to offset victims’ uncompensated
direct costs of the crime, for medical, burial, mental health, and/or lost wages or support. Pain and
suffering are not compensated. Specific costs must be documented. There is normally no direct
connection between an offender’s contribution to a crime victim compensation fund and his/her
specific victim. PIE deductions contribute to State victim compensation funds.

11
12

Ibid., 25.
Ibid., 16.

23

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

The State crime victim compensation funds to which State PIE inmates contribute is a part of the
Federal-State product of the “1984 Victims of Crime Act,” (VOCA, P.L. 98-473) that established
the Crime Victims Fund in the U.S. Treasury. State crime victim compensation programs get
annual financial support from the Federal crime victim compensation fund financed by fines,
penalty assessments, and bond forfeitures from convicted Federal offenders (not from taxpayers).
VOCA’s existence and its distribution of assistance and compensation funds to the States sets the
framework for all 50 States’ (and the District of Columbia’s) victim assistance and compensation
laws and programs. Beginning in fiscal year 2002, with the passage of the October, 2001, “United
and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct
Terrorism Act (USA Patriot Act), the victims compensation fund can also accept gifts, bequests,
or donations from private entities.

Family Support
Family support remains generally undefined, although the Guidelines specify the family support
deduction for “support of family pursuant to State statute, court order, or agreement by the
offender.13 On average, family support garners 6 percent of PIE inmates’ gross wages, the
smallest of PIE deductions.

The Post-Deductions Residual
PIE Guidelines are explicit, stating that, “deductions, in aggregate, cannot exceed 80 percent of
gross wages … PIECP workers must be paid, credited with, or otherwise benefit legally from, the
20 percent gross remainder.”14 However, this guidance only limits PIE-based deductions. In fact,
the guidelines clearly permit additional post-PIE deductions, and the State “…may direct the 20
percent gross remainder to a PIECP inmate worker’s expense accounts, savings accounts, or
toward the settling of the worker’s legal obligations, including the payment of fines and
restitution.”15

13

Ibid., 25.
Ibid., 26.
15
Ibid., 26.
14

24

IV. Deductions and Beneficiaries Elaborated
This section provides additional information defining the content of PIE beneficiary classes,
identifies more clearly who (what groups) constitute the beneficiaries, provides relevant
quantities—monetary and otherwise—affecting them, identifies critical issues affecting the
beneficiary class, and positions the PIE deductions in the context of the economic characteristics
of the beneficiary group.
Of necessity, the research identifies the most immediate beneficiaries of PIE incomes, with little
or no emphasis on identifying secondary or “ultimate” recipients.

Taxes
Given the ubiquity of taxation and the far-reaching influence of Federal, State, and local
government expenditures, the simplest and most accurate generalization is that every person and
every business that pays taxes and every person and every business that receives a benefit from
taxes to some degree benefits from increasing the number of taxpayers. Individuals, households,
and businesses pay taxes. Taxes reduce consumer purchasing power and raise business costs, in
each case reducing business sales and reducing consumer well-being. On the other hand,
householders and businesses share in the benefits of National security, law, and infrastructure,
and in varying degrees share in the benefits of publicly provided goods and services; moreover,
the huge purchasing and employment powers of government create jobs and demands for goods
and services in most sectors.
Overall, taxation can be viewed as a transfer among groups, shifting production and purchasing
power from private individuals and businesses to governments, and then from governments to
selected classes of individuals and businesses. Furthermore, by and large, the progressive income
tax system in the United States, coupled with social safety nets, is thought to yield a general
effect of shifting benefits from higher income to lower income groups, and from the wealthy
toward those with least wealth.
Given that inmates are, by and large, disproportionately (compared with the normal population)
less wealthy, of lower incomes, and more dependent upon the taxpaying population, the addition
of inmates as taxpayers could be viewed as in the direction of offsetting the redistribution of
incomes and wealth, with inmates lessening their dependence upon other taxpayers and
contributing more to their own—and their families’ and communities’—well being, health care,
and retirement, and increasing the funding and stability of the overall tax, income security, health
care, and retirement systems.

25

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Moreover, because adding inmates as taxpayers yields a coincident dual benefit for every dollar
collected—both to other taxpayers (shared burden, proportionately less tax burden per taxpayer)
and at the same time to its recipients (a larger total to be distributed), the percent distribution of
benefits taken from PIE incomes could substantially exceed 100 percent and makes describing the
percentage distribution of every PIE dollar less than transparent.

Federal Income Taxes
Seen from both revenue and expenditure perspectives, the beneficiaries of inmates as taxpayers
include both taxpaying American households and businesses, and, as recipients, all populations
dependent for retirement, income, and health support, as well as for the National defense, interest
on the debt, and a range of other physical, natural resource, and other programs. Retirement,
income, and health support are, of course, heavily supported by social insurance and other
contributions, and supplemented by outlays from Federal income taxpayers.

Who Pays Federal Income Taxes?
Of total Federal tax revenues of about $2 trillion in 2000, Federal income taxes grossed about
$1.4 trillion (Table 6); another $0.6 trillion was collected from various social insurance taxes and
contributions, such as Social Security, railroad and Federal employees retirement, and the like.
The Federal income tax burden is borne both by households and by businesses, and
disproportionately by higher income entities.
Unfortunately, because households and many businesses (proprietorships and partnerships) use
the same tax reporting vehicles, the proportion of Federal income taxes paid by households and
businesses is not known.1 For fiscal year 2000 (October, 1999—September, 2000), nevertheless,
Internal Revenue Statistics clearly reveal that both households and corporate businesses each bear
significant Federal income tax burdens.

Table 6. Total U.S. Gross Income Tax Collections Fiscal Year 2000
Group
Individual*
Corporate
Total

($Billion)
1,137
236
1,373

(Percent)
83
17
100

*Note: Individual returns include proprietorship and partnership businesses.
Source: Internal Revenue Service.

1

William Ahern, The Tax Foundation, personal communication, February 26, 2003. Internal Revenue Service data,
however, may be mildly indicative, given that about 2 million of 130 million individual income tax returns are filed by
partnerships (without noting the dollar value of taxes paid).

26

Deductions and Beneficiaries Elaborated

Therefore, at least 17 percent and presumably a significant additional share of the remaining 83
percent of U.S. income taxes are paid by American businesses.
Individual tax burdens are not small. For 2000, the Internal Revenue Service estimates the
average tax rate for the top 50 percent of income earners to have been 16.9 percent of adjusted
gross income.2 Little surprise that taxpayers typically welcome either reduced tax rates or support
from additional taxpayers.
Overall statistics, moreover, indicate relatively clearly that higher income taxpayers bear a highly
disproportionate share of the Federal income tax burdens (Table 7).
Table 7. Total Federal Income Tax Share, 2000—
Individual Taxpayers
Taxpayer Group
Top 1%
Top 10%
Top 50%
Bottom 50%

Percent of Federal
Taxes Paid
37
67
93
7

Source: Internal Revenue Service.

Who Benefits From Federal Income Taxes?
Federal government expenditures touch virtually every sector of the economy, with the primary
beneficiary classes being defense, income security, and health. Overall, Federal government
expenditures fall into the following categories (Table 8).
Social Security and Medicare Part A are funded separately and treated separately here (below).
Part B Medicare is optional insurance available to pay benefits to elderly and disabled persons for
physician and other medical outpatient services, durable medical equipment, home health care,
occupational therapy, and other services; Federal taxpayers pay about three-fourths of the cost of
Medicare part B, for which contemporary outlays—taxpayer plus enrollee fees combined—are
approximately $100 billion a year.3
The largest single Federal beneficiary category is, of course, the National defense, directly
absorbing 16 percent of Federal fiscal 2000 outlays, plus another $47 billion (3 percent) for
veterans benefits and, in addition, accounting for a sizeable fraction of interest on the National
debt, itself the third largest outlay category ($220 billion, 12 percent) of Federal outlays.

2

Tax Foundation, “Historical Data on Federal Individual Income Tax Returns, 1980-2000,” Table “Average Tax Rate,”
retrieved February 26, 2003 on the World Wide Web: http://www.taxfoundation.org/ prtopincometable.html.
3
Office of Management and Budget (OMB), Budget of the United States Government, Fiscal Year 2001, 252.

27

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 8. Federal Government Outlays, Fiscal Year 2000
Function

($Billion)

Defense
Human Resources
Social Security
Income Security
Medicare
Health
Education/Training
Veterans
Physical Resources
Net Interest
Justice
Other
Total

291
1,125
407
251
203
154
63
47
86
220
26
85
1,790

(Percent)
16
63
23
14
11
9
4
3
5
12
1
5
100

Note: Sum exceeds total because of independent rounding.
Source: OMB, Budget of the United States Government, FY2001.

Income Security constitutes the second largest individual category of Federal outlay for fiscal
2000 (14 percent), and includes payments for Food Stamps, Supplemental Security Income (SSI),
Temporary Assistance for Needy Families (TANF), and the Earned Income Tax Credit, plus
Low-Income Housing Credits, all supported by Federal taxpayers.
And Federal taxpayer-supported health care—other than Social Security and Medicare—
constitutes the third largest beneficiary of Federal taxes (9 percent), most of which goes to
Medicaid, the Federal-State taxpayer supported program providing health care to low-income
Americans. Medicaid provides health services to one quarter of the Nation’s children, and is the
Nation’s single largest purchaser of maternity, nursing home, and other long-term care services,
covering almost two-thirds of nursing home residents. Medicaid is reported to serve at least half
of all adults living with AIDS. Other Federal services for health care include health insurance for
children, Indian Health Services, substance abuse and mental health services, services for the
mentally ill, and other services.4

Contemporary Challenges in Federal Income Taxation
Although receiving incomes from government expenditures tends to be welcomed, paying taxes
reduces household, consumer, and business income and wealth, and diminishes purchasing and
investment power. Paying fewer rather than more taxes is almost universally preferred by all
taxpayer groups. Increasing tax burdens (independent of spending characteristics), is more
generally viewed as slowing economic growth, particularly as individual taxation rates increase.
Tax burdens raise input and production costs, increase prices, leave firms less competitive
(particularly in world markets), and results in paying lower wages and stockholder dividends as
well. Personal income taxes reduce disposable incomes, purchasing power, and consumer well
4

28

Ibid., 245–246.

Deductions and Beneficiaries Elaborated

being. As any year’s headlines attest, taxpayers tend to welcome tax relief, and the oft-repeated
line, “Don’t tax you, don’t tax me, tax that man behind the tree,” captures the taxpayers’
collective fond hope.
Medicaid is of particular interest at both the Federal and State level, as health care costs rise and
numbers of persons covered increases, all the more when economic downturns increase caseloads
while decreasing tax revenues.

State Income Taxes
Who Pays State Income Taxes?
At the State level, taxes tend to be borne first by consumers through sales taxes, and then by
individual, business, and corporate taxpayers through State income tax systems. Of total State
government expenditures approaching $1 trillion dollars in 2000, roughly one-fourth, $227
billion, was obtained from State individual and corporate income taxes.5 Similar to shares at the
Federal level, most income taxes (86 percent in 2000) are obtained from individual household and
business taxpayers, and the remainder is obtained from corporate entities. Nationwide, State
income taxes in 2000 cost every American about $700.6 In addition, State governments receive
significant portions of operating funds from the Federal government.

Who Benefits From State Income Taxes?
Beneficiaries of State expenditures tend to be all persons receiving or paying for education or
who are employed in or provide support to elementary, secondary, and higher education, then
Medicaid delivery services or its recipients, and finally transportation (Table 9).

Challenges in State Taxation
In contrast to the Federal government, State budgets legally cannot incur deficits. Economic
slowdowns mean reduced revenues, reduced State expenditures, and State revenue crises. As a
result, budget crises expose the primary beneficiaries of increasing or decreasing State tax
revenues. According to the National Conference of State Legislatures, “State legislators face a
common problem around the country. Spending needs are outpacing projected budget levels,
particularly in the area of Medicaid and health costs.… Whether the decisions are to raise
revenues

5

The National Association of State Budget Officers (NASBO) cites $945 billion dollars expended by States in fiscal
2000, of which $682 billion was obtained by funding in the States (NASBO, 2000 State Expenditure Report,
Washington, DC: NASBO, 2001). For the same period, the U.S. Census Bureau, “State Government Tax Collections:
2000,” reports total individual and corporate income taxes at the State level totaling $227 billion (U.S. Bureau of the
Census, “State Government Tax Collections: 2000,” retrieved July 22, 2003, from the IRS Web site:
http://www.census.gov/govs/www/Statetax.html. In fiscal 2000, according to NASBO, about $246 billion was provided
to the States by the Federal government.
6
U.S. Bureau of the Census, “State Government Tax Collections: 2000.”

29

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 9. State Spending, Fiscal 2000—Spending Shares, From All Sources*
Expenditure Group
Elementary & Secondary
Education
Higher Education
Public Assistance
Medicaid
Corrections
Transportation
Other
Total

Estimated $Billion

Percent Share

213

23

103
23
184
36
83
303
945

11
2
20
4
9
32
100

*Funding includes State and Federal sources. State general funds are concentrated in
education (47 percent), whereas Federal funding is disproportionately evident in Medicaid
support (40 percent).
Source: NASBO, 2000 State Expenditure Report.

or cut programs, the lives of our constituents will be profoundly affected.”7 Thus, at the State
level, beneficiaries tend to be concentrated among education providers and recipients, with
special sensitivity today to providers and recipients of Medicaid. Medicaid spending is of
particular interest, with rapidly rising costs facing strained State budgets; increasing State
Medicaid costs have been cited as the “biggest single reason that may States’ budgets are out of
balance.”8

Social Security (FICA)
What is Social Security?
Social Security (or FICA, the “Federal Insurance Contributions Act”) is primarily an old-age
insurance program, ensuring that retired American workers have a floor of income. Social
Security also provides income to the disabled and to the families of prematurely deceased
workers (Table 10). Furthermore, Social Security now includes “Medicare—Part A” medical
coverage.
FICA represents payments to trust funds for 3 social insurance programs:
•
•

7

OASI, “Old-Age and Survivors Insurance,” pays retirement and survivor benefits;
DI, “Disability Insurance,” pays disability benefits; and

National Conference of State Legislatures, New National Survey Reports State Budgets Fall $17.5 Billion Short,
November 22, 2002, retrieved from the National Conference of State Legislatures Web site: http://www.ncsl.org.
8
“Governors Finalizing Proposal to Revamp Medicaid,” The Washington Post, June 3, 2003, page A10.

30

Deductions and Beneficiaries Elaborated

Table 10. Social Security Beneficiaries and Benefits, 1999
Beneficiary Group

Number (Million)

(Expenditure $Billion)

Old Age
Survivors
Disability
Total:
Medicare (Part A)

31.0
7.0
6.5
44.6
33.9**

334.4*
*
1.4
385.8
128.8

*Sum, Old Age and Survivors benefits. **Number enrolled.
Source: Social Security Administration, Annual Statistical Supplement 2000.

•

HI “Hospital Insurance” benefits constitute Part A Medicare, and are generally received by
persons over age 65 getting OA or who are disabled. Part A pays for inpatient hospital care,
skilled nursing and other services.9

“Social Security” is the combination of OASI and DI, or OASDI; HI is the part A component of
Medicare; their combination constitutes contemporary FICA.

Who Benefits From Social Security (FICA)?
Social Security protects more than 150 million American workers, and is by far the largest
income support program for American retirees. About two-thirds of Americans 65 and over get
more than half their income from Social Security, and estimates suggest that Social Security
prevents poverty for nearly half of all Americans 65 and older.10
In 1999, the average monthly benefit for retired workers totaled $804, and for survivors and
disabled, $750–$775.11

Who Pays for Social Security?
Employees and employers pay equally for Social Security. In 2001, more than $668 billion
dollars in total FICA taxes were collected. Year 2002 tax rates for each program and total
collections are shown in Table 11.
Funds collected from Social Security contributions are deposited into four separate trust funds, (a)
For OASI, into the Federal Old-Age and Survivors’ Insurance Trust Fund; (b) for DI, into the
Federal Disability Insurance Trust Fund; and (c) for HI, the Federal Hospital Insurance Trust
Fund. A fourth trust fund, the “Federal Supplementary Medical Insurance Trust Fund,” receives
monies collected from premiums for medical insurance coverage. Deposits into the trust funds are
used to pay for the retirement, survivors, disability, hospital, and medical insurance programs,
plus administrative costs. Excess funds are invested in interest-bearing Federal securities.

10

Harris, Diane, “Social Security, The Good News,” My Generation 8 (May-June 2002), 33–38.
Social Security Administration, Office of Policy, Annual Statistical Supplement 2000. Retrieved July 26, 2003 from
the World Wide Web: http://www.ssa.gov/policy/docs/Statecomps/supplement/ 2000/supp00.pdf.
11

31

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 11. FICA, Tax Rates 2002, Contributions 2001 (Percent)
Contributor

OASI

DI

HI

Total

Employee
Employer
Total

5.30
5.30
10.60

0.90
0.90
1.80

1.45
1.45
2.90

7.65
7.65
15.30

DI
$75
11%

HI
$152
23%

Total
$668
100%

Total Contributions 2001 (Billions):
OASI
$442
66%

Source: Social Security Administration, “Summary of the 2002 Annual Social Security Report,” accessed
July 22, 2003 on the World Wide Web: http://www.ssa.gov/OACT/TRSUM/trsummary.html.

What Are The Most Important Social Security Issues Today?
While current retiree and other obligations are able to be met, the most important issue facing all
parts of Social Security today is the threat that sufficient trust funds will not be available in future
decades to meet demands by qualifying insureds; in short, the primary problem facing Social
Security is that current and future income is projected to fall short of future payout needs. Future
retiree coverage is at issue, with reduced services, increased taxation, or both, as options.
According to the Summary of the 2002 Annual Report of the Social Security and Medicare
Boards of Trustees,
•

The combined OASI and DI trust fund continues to fail to meet 75-year tests for actuarial
balance. Bringing Social Security into actuarial balance could be achieved either by a 13
percent decrease in benefits or a 15 percent increase in payroll tax income, or a combination
of the two. When the baby-boom generation begins to retire in about 2010, financial pressure
on the Social Security trust funds will rise rapidly and tax income will fall short of outlays
beginning in 2017.12

•

Medicare faces financial difficulties that come sooner—and in many ways appear to be more
severe—than those facing Social Security. Health care costs per enrollee are projected to rise
faster than wages. Whereas trust funds are expected to remain solvent for at least 75 years, in
fact, the HI trust fund is currently projected to remain solvent only until 2030; projected tax
income is projected to fall short of outlays beginning in 2016. Expressed otherwise
(according to the Trustees’ report), the 75-year actuarial deficit is projected to be 2 percent of
taxable payroll. Bringing HI into actual balance over the next 75 years could be achieved by
raising the HI payroll tax rate from 1.45 percent to 2.46 percent on both employers and
employees, or by reducing outlays (benefits) 38 percent from current levels—or a

12

Social Security Administration, 2002, The 2002 Annual Report of the Board of Trustees of the Federal Old-Age and
Survivors Insurance and Disability Insurance Trust Funds, Washington, D.C.: Social Security Administration, March
26. Retrieved from the World Wide Web April 19, 2003: http://www.ssa.gov/OACT/TR/TR02/tr02.pdf.

32

Deductions and Beneficiaries Elaborated

combination of the two.13 The HI cost rate increases not only because of the projected
increase in the number of beneficiaries per worker, but because of increases in the use and
cost of health care per person. Whereas in 2000 there were an estimated 3.4 workers per
beneficiary, by 2030 the ratio is expected to fall to just over 2.0, and to about 1.8 workers per
beneficiary by 2070.14
Although the parlance of Social Security and Medicare is of trust funds fueled by enrolled firms
and persons, they are in fact considered liabilities of the U.S. Treasury.15 Shortfalls in income
must be made up with (a) increased taxation, (b) increased borrowing, (c) decreased benefits, or
some combination of all three.
OASI costs increase steeply between 2010 and 2030 because the number of people receiving
benefits will increase rapidly as the baby-boom generation retires, while the number of workers
paying payroll taxes grows more slowly because of currently low fertility rates.16

Workers’ Compensation
“Workers’ Compensation” occurs at the State level and encompasses the worker compensation
statutes and programs of all fifty States and the District of Columbia. State programs differ, and
they are not Federally regulated. A workers’ compensation program may be a State fund, a
private insurance carrier, or a firm’s self-insurance certified by the State. Federal, State, and local
government employees are covered by a various government workers’ compensation programs.

Who Pays Workers Compensation?
Employers pay for workers compensation. In calendar year 2000, employers are estimated to
have invested $56 billion dollars for workers’ compensation premiums, spending, on average
$442 per covered worker that year. On average, employer costs are estimated to have equaled
1.25 percent of covered wages in 2000.17 Premiums paid by employers are based on the industries
and occupational classes of workers and also on individual firms’ experiences with claims in the
past.

Who Benefits From Workers Compensation?
Workers’ compensation programs compensate injured wage and salary workers and their
dependents and survivors.
The National Academy of Social Insurance (NASI) estimates that 88 percent of the entire U.S.
workforce was covered by workers’ compensation in 2000, with the self-employed and firms with

13

Ibid., 2.
Ibid., 19.
15
Ibid., 13.
16
Ibid., 8.
17
National Academy of Social Insurance (NASI), “Workers’ Compensation: Benefits, Coverage, and Costs, 2000, New
Estimates,” June 2002, Washington, DC: NASI.
14

33

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

5 or fewer employees the major groups excluded.18 NASI estimates that 126.6 million employees,
94 percent of all wage and salary workers, were covered by workers’ compensation in 2000.19 In
2000, $45.9 billion dollars in workers’ compensation benefits were paid.20
Benefits to injured workers and their survivors vary, depending upon the rules of the individual
States and the characteristics and duration of the injury to the worker. Workers’ compensation
pays for the immediate medical treatment of work-related injuries, for hospital and medical care,
and for rehabilitation. Permanent disabilities result in disability payments whose magnitude
depends upon the impairment, the employees’ pay rate, and a State payment schedule. According
to the National Council on Compensation Insurance, worker compensation claims averaged about
$10,000 in 1997, with about half going to wages compensation and the remainder going to
medical costs.21
PIE inmate workers usually do not receive compensation benefits while they are incarcerated. As
a result, workers’ compensation premiums obtained from PIE employers can be generally viewed
as a transfer payment funding the compensation claims of injured civilian employees.

Unemployment Taxes
Unemployment taxes and benefits are the product of joint Federal and State unemployment
insurance programs. Each State administers a separate unemployment insurance program meeting
Federal Standards.22

Who Pays Unemployment?
Employers fund unemployment insurance with a combination of Federal and State taxes.
Employers’ insurance rates vary based on (1) the amount of wages paid, (2) the amount already
contributed to the fund, and (3) the amount paid from the fund to support discharged employees
from that firm. For 1998, the average employer unemployment tax rate on taxable wages was
1.92 percent, but ranged from a low of 0.32 percent in Kansas to a high of 3.85 percent in
Pennsylvania.23 Differences in State rates reflect laws, wages, unemployment rates, and economic
conditions in the States. Taxes paid are usually based on wage earners’ incomes up to a
maximum. The higher a firm’s payout rate (the more persons dismissed and the higher their
wages), the higher the employer’s unemployment tax rate. State taxes are credited against the
18

NASI, “Workers Compensation, Data Fact Sheet, Workers’ Compensation Coverage by State,” October 2002, No. 1,
retrieved July 26, 2003 from the World Wide Web: http://www.nasi.org/publications2763/
publications_show.htm?doc_id=124291&name=Workers%27%20Compensation.
19
NASI, “Workers’ Compensation: Benefits, Coverage, and Costs, 2000 New Estimates,” June 2002.
20
NASI, “Workers’ Compensation: Benefits, Coverage, and Costs, 2000 New Estimates,” June 2002.
21
National Council on Compensation Insurance (NCCI), “Average Claim Costs,” retrieved July 26, 2003 from the
NCCI Web site: http://www.ncci.com.
22
Cornell School of Law, Legal Information Institute, “Unemployment Compensation Law: An Overview,” retrieved
July 22, 2003 from the Cornell Law School Web site:
http://www.law.cornell.edu/topics/unemployment_compensation.html.
23
U.S. Department of Labor, Employment and Training Administration, ET Handbook No. 394, Data for Calendar
Year 1998: Categories, retrieved July 26, 2003 from the U.S. Department of Labor Web site:
http://www.ows.doleta.gov/unemploy/content/hdbk394_99/lnk98.html#taxempwag.

34

Deductions and Beneficiaries Elaborated

Federal tax. The Federal Unemployment Tax (FUTA) rate is currently 6.2 percent. However,
payment of the State tax (SUTA)—regardless of the firm’s actual State tax rate—results in relief
of 5.4 percentage points of the FUTA rate and therefore reduces the effective FUTA rate to 0.8
percent. The FUTA tax rate has two components, a permanent tax rate of 0.6 percent plus a surtax
rate of 0.2 percent.24 Funds collected under the Federal tax are used solely for administration of
the program; compensation to unemployed persons is paid by State collections. Proceeds of the
unemployment taxes are deposited in the Federally administered Unemployment Trust Fund (the
Fund), in which each State has a separate account. During economic recessions, the Federal
government can also transfer monies to specific States from the “Extended Unemployment
Account” also funded by State programs.

Who Benefits from Unemployment Taxes?
Unemployment insurance provides workers who lose their jobs through no fault of their own
income for a period of time or until they find a new job. Compensation is intended to provide the
worker time to find new employment equivalent to that lost. Compensation is also justified as
maintaining aggregate consumer demand during times of economic downturn. Thus both
employees and their families along with the business communities and others dependent upon
their spending are viewed as the normal beneficiaries of unemployment insurance. Eligibility for
unemployment insurance, benefits, and length of time for benefits is determined by individual
State laws. An unemployed person must have worked for wages or time worked during a previous
year, must be determined by the State to be unemployed through no personal fault, and the person
must likely wait a short period (a few weeks) before actually receiving compensation. Benefits
can be paid to a maximum of 26 weeks in most States, but may be extended during periods of
high unemployment.
PIE inmates who lose their jobs generally do not collect unemployment compensation. Therefore,
unemployment insurance premiums collected from PIE employers constitute transfer payments
solely to the unemployment insurance claims of civilian workers.
For the United States, unemployment premiums and payments in 1998 averaged as follows:
•
•
•
•

average monthly covered employment: 100.2 million persons;
average weekly wage in covered employment: $610 (taxable: $196);
unemployment premiums in 1998: $19.8 billion (plus $3 billion interest earned on trust
funds);
benefits paid, 1998: $18.4 billion;

24

Almanac of Policy Issues, “Unemployment Compensation,” excerpted from the 2000 House Ways and Means Green
Book, retrieved July 22, 2003 from the Almanac of Policy Issues Web site:
http://www.policyalmanac.org/social_welfare/archive/unemployment_compensation.shtml.

35

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

•
•
•

reserves 12/31/98: $48.0 billion;
average weekly insured unemployed in 1998: 2.2 million; and
average weekly benefit amount, 1998: $200.29 (33 percent of wages).25

Victims Compensation (Restitution)
Who Benefits From Victims Compensation?
The direct beneficiary of victims compensation is the current crime victim or the victim’s
survivors, who receive monetary compensation for documented medical, burial, and lost wages
uncompensated by other sources (Table 12). Victims compensation does not compensate for
property losses or for pain and suffering. Awards are limited, with maximum payments varying
by State, but in the range of $25,000. Actual awards typically fall well below maximums, in 1997
averaging about $1,400. About 114,000 claimants were compensated in that year.26
In calendar year 2000, U.S. crime victims received more than $295 million in crime victim
compensation payments from States, with most awards compensating for medical and mental
health costs not otherwise compensated.

Table 12. U.S. Crime Victims Compensation, 2000
Category
Medical
Mental Health
Lost Wages/Support
Funeral & Burial
Crime Scene Cleanup
Forensics & Other
Total Compensation

$Million Paid
141.3
48.7
46.1
39.3
3.5
16.5
295.4

Source: U.S. Department of Justice, Office of Justice Programs,
Office for Victims of Crime, “Crime and Victimization in America
Statistical Overview,” retrieved from the DOJ Web site
http://www.ojp.usdoj.gov/ovc/ncvrw/2002/ncvrw2002_rg_3.html.

25

U.S. Department of Labor, Employment & Training Administration (ETA), ET Handbook No. 394, retrieved July 22,
2003 from the ETA Web site: http://workforcesecurity.doleta.gov/unemploy/ content/hdbk394_99/home.asp#top.
26
Urban Institute, Justice Policy Center, The National Evaluation of State Victims of Crime Act Compensation and
Assistance Programs, Findings and Recommendations From a National Survey of State Administrators, Research
Report, Washington, DC: Urban Institute, March, 2001, 6.

36

Deductions and Beneficiaries Elaborated

Who Pays Victims Compensation?
State victims of crime compensation funds are financed by monies from (a) a State’s current
offenders, and (b) the Federal victims of crime compensation fund. Both the Federal and State
crime victims compensation funds are financed by fines, penalty assessments, and bond
forfeitures. 27
Compensation literature is remarkably silent on sources of victim compensation funds. Anecdotal
information suggests the vast majority coming from civil and criminal fines, possibly
predominantly from white-collar criminal (not civil) fines.28 Some States also provide victims
compensation funding from the State treasury. Except for PIE inmates’ contributions totaling
about $3 million a year, nothing suggests inmates as notable contributors to State victims
compensation programs. In fiscal year 2001, the Iowa Crime Victims Assistance Division
reported $6.7 million in total program revenue, but only a small share from either restitution or
inmate payments. Offender compensation, presumably primarily from probationers, accounts for
about 6 percent of total revenues, and inmate wages, a bit less than 2 percent (Table 13).29 For
Iowa in 2001, it appears that virtually all payments by inmates to victims compensation was
provided by PIE victim compensation deductions; of $116,618 reported published in the
Division’s annual report, $116,518 (99.9 percent) was independently reported to NCIA as PIE
deductions for the Iowa victims compensation program. Inquiries regarding inmate contributions
to victims compensation in the States sampled for this research indicate that all or virtually all
monies collected by those State victim compensation programs was obtained from PIE inmates,
with other State and local inmates contributing nothing or almost nothing to those States victim
compensation programs.30
In fiscal year 2002, the Minnesota Crime Reparations Board paid almost $3 million to crime
victims, of which 10 percent, nearly $300,000, was obtained from offender payments, most of it
from probationers. Another $208,000, 6 percent, was obtained from inmate wage deductions for
victim compensation; comparison with reported PIE contributions for calendar 2000 and 2001
suggests that Minnesota’s PIE inmates contribute at least 80 percent of that State’s overall inmate
contributions.31 Nevertheless, in both States, inmate contributions provide a very small share of
overall victim compensation.

27

Since passage of the USA Patriot Act in October 2001, the Federal fund can also accept gifts, bequeaths, and
donations from private entities.
28
This research located no data—and rare mention—specifying sources of funds among a number of reports by the
Department of Justice, the National Association of Crime Victim Compensation Boards, or others. Business Week,
however, suggests that portions of large corporate fines may find their way to State victim compensation programs
(Business Week, “Crime Pays—the Treasury,” December 23, 2002, 8.
29
Iowa Department of Justice, Crime Victim Assistance Division, Fiscal Year 2001 Annual Report, Des Moines: Iowa
DOJ, 15.
30
Personal communications, research staff with staff of sampled States’ crime victim compensation program offices,
May 2003.
31
Personal communication, research staff with Marie Bibus, Minnesota Crime Victims Reparations Board, December
12, 2002.

37

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 13. Iowa Crime Victim Compensation
Revenues—Fiscal Year 2001
Source
Criminal Fines
Driving Intoxicated (Civil)
Federal Grant
Offender Restitution
Subrogation*
Inmate Wages
Total Revenues

$Million
3.4
1.7
0.9
0.4
0.2
0.1
6.7

*Subrogation is reimbursement of the compensation fund.
Source: Iowa Dept. of Justice, Fiscal Year 2001 Annual Report.

In general, the Federal victims of crime program distributes to a State an amount equal to 40
percent of that State’s expenditure for victims of crime compensation two fiscal years earlier; for
example, if a State spent $1 million for crime victim compensation in FY1998, it would receive
$400,000 from the Federal VOCA (Victims of Crime Act) fund for distribution in FY2000. Thus,
every $1.00 collected from PIE inmates for a State’s crime victim compensation fund effectively
adds $1.40 to monies distributed to the State’s current crime victims.

Who Benefits From Restitution?
“Restitution” refers to the direct payment by an inmate or other offender to his or her victim.
Unlike compensation, restitution requires the identification of an associated convicted offender.
Restitution can occur as a PIE deduction if it occurs through a crime victim compensation fund;
in fact, restitution may also occur as another court-ordered deduction as part of, or subsequent to,
PIE. Therefore, in contrast to victims compensation, the beneficiary of restitution is the offender
or inmate’s specific crime victim or the survivors of the victim.

Who Pays Restitution?
Offenders, including inmates, pay restitution. Restitution is the product of a court order setting
the amount of restitution and setting a restitution payment schedule. However, in most
jurisdictions the court orders restitution only when the defendant possesses some ability to pay.32
When the offender has been ordered to pay restitution in conjunction with a prison sentence, the
State department of corrections serves as the administrator of the program.33
This research found no data quantifying U.S. direct restitution payments.

32

National Center for Victims of Crime (NCVC), Get Help Series, “Restitution Legislation,” retrieved July 22, 2003
from the NCVC Web site: http://www.ncvc.org/gethelp/restitutionlegislation/.
33
Ibid.

38

Deductions and Beneficiaries Elaborated

Issues in Crime Victim Compensation and Restitution
Despite data shortcomings, it is overwhelmingly clear that crime victims are not generally
compensated by their offenders, either via crime victim compensation or by restitution.
Furthermore, it is highly unlikely that inmates are major contributors to compensation, and
instead it appears that payments are primarily obtained from fines, and from probationers and
parolees. Finally, there is some indication that compensation/restitution requirements are not fully
enforced, suggesting the need and opportunity for increased enforcement within corrections.
First, the annual cost of crime is far in excess of crime compensation payments. One estimate of
the annual cost of violent crime is provided by the Children’s Safety Network (CSN) Economics
and Insurance Resource Center (Table 14).34

Table 14. U.S. Cost of Violent Crime, 1995, Including
Assault, Murder, Rape, and Robbery—Selected
Components
Cost Category
Medical Spending
Mental Health
Victim Work Loss
Property Damage
Quality of Life
Criminal Justice Costs/Other
Total Cost of Violent Crime

($1997 Billion)
6.75
4.91
39.72
1.45
274.93
18.29
346.05

Source: CSN Economics & Insurance Resource Center, “Cost of
Violent Crime and of Alcohol-Involved and Drug-Involved Violent Crime
in the USA.”

In addition, the authors add about $19 billion for annual drug crime costs, and other estimates
include nearly $8 billion losses to auto theft, more than $5 billion to larceny/theft, and almost $3
billion in burglary losses in 2000.35 Altogether, $50 billion, then, would probably work as a
general estimate of the annual direct costs of violent crime ($100 billion for major crime),
excluding pain and suffering and quality of life.
And who primarily bears the cost of crimes? According to the National Research Council, private
insurers—particularly including those paid by victims’ employers—plus crime victims and
government agencies, bear most of the costs of violent crime victims’ injuries. Costs borne by
others than the victim are not compensated by victim compensation programs. Thus private
insurers, employers, local, State, and Federal governments, and others bear the direct medical
34
Children’s Safety Network (CSN), CSN Economics & Insurance Resource Center, “Cost of Violent Crime and of
Alcohol-Involved and Drug-Involved Violent Crime in the USA,” 1995, retrieved July 22, 2003 from the CSN
Economics & Insurance Resource Center Web site: http://www.csneirc.org/csneirc/ pubs/violent/us-viol.htm.
35
U.S. Department of Justice, Office of Justice Programs, “Crime and Victimization in America, Statistical Overview,”
(2002 National Crime Victims Rights Week), 2, retrieved from the Office of Justice Programs Web site:
http://www.ojp.usdoj.gov.

39

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

costs of violent crime. According to the Council estimates, the proportions of medical costs of
victims’ injuries are borne as follows (Table 15):
Table 15. Shares of Medical Costs for Violent Crime, 1995
Bearer
Private Insurers*
Government
Victims/Families
Charities

Percentage Borne
44
32
19
5

*Primarily employer-provided health insurance.
Source: National Research Council, Understanding and Preventing
Violence.

Victims and their families bear the costs of pain and suffering, and victims and their families are
estimated to bear nearly 80 percent of the ultimate costs of wage losses from murders and much
of the wage loss for other violent crimes. Employers are estimated to carry large proportions of
violent crime victimization costs, through sick leave, lost production, and health insurance
premiums.36 Thus the costs of crime are primarily borne by persons other than the offender.
By even the most generous of interpretations—and glossing over critical conceptual and
measurement issues (like inmates as a proportion of all offenders and the differing time periods of
offenses and restitution)—it is clear that compensation and restitution are a small fraction of the
annual costs of crime and that the costs are primarily borne by others than offenders.
Further—and most relevant to this research—inmates’ annual contributions offsetting the costs of
crime are miniscule. Even if all crime victims compensation ($295 million a year) were paid by
inmates and if restitution approximated compensation—and if pain and suffering were ignored—
inmates would annually compensate victims far less than 1 percent of the annual costs of crime.
Given that available information suggests that inmates pay only a small fraction of either
compensation or restitution totals (PIE victims compensation totals about $3 million a year) it is
clear that both restitution and compensation, relative to victim costs, are significantly
underfunded, and that inmates of any stripe are generally absent from compensating for crime
costs. By even the most generous of assumptions, inmate repayment in any year equals no more
than 1/100 of one percent of even the annual direct cost of crime, excluding pain and suffering
and quality of life losses.
There is some evidence that criminal justice and correctional policies, however justified,
particularly in eras of restricted resources, contribute to offender and inmate nonpayment of
restitution. After a 2002 review of restitution in Vermont, the State auditor concluded, “What

36

National Research Council, National Academy Press, Understanding and Preventing Violence, (volume 4)
Consequences and Control, Washington, DC: National Academy Press 1994, 145, 146.

40

Deductions and Beneficiaries Elaborated

we’re saying is strategically and in every way…it’s not a top priority for anybody…”37 A 1999
U.S. Department of Justice, Office for Victims of Crime report, Victim Services in Corrections,
report, pointedly notes: “While all 50 States, the District of Columbia and the Federal government
have statutory provisions for victim restitution, it is one of the most under-enforced of all victims’
rights.” In addition to many other recommendations, the report also highlights the following
recommendation, “Correctional agencies must acknowledge, through policies and practice, that
restitution is a basic right that holds offenders financially accountable for their criminal actions,
and provides victims with some monetary compensation to cover their losses resulting from
crime.”38 Arguments for under-enforcement include (a) lack of court orders because cost far
exceeds expected revenues, even when statutes appear to require them. (b) breakdowns in
communication among agencies, (c) lack of resources to enforce, along with other, more pressing
priorities, and (d) a general sense of futility in expecting a return from inmates; as the Justice
Department report notes, “the often misguided belief that ‘you can’t squeeze blood from a
turnip.…”39
In sum, available evidence overwhelmingly indicates that inmates, with the signal exception of
PIE participants, generally contribute little or nothing to compensating crime victims or offsetting
the annual costs of U.S. crime.

Family Support
Although not defined in any detail and conceptually capable of encompassing almost anyone
considered “family,” including spouses or significant others, children, parents, siblings, cousins,
aunts, or even other unrelated household members, “family support” here is interpreted as
financial support for an inmate’s children, whether court-ordered or voluntarily provided.
Presumably, then, the beneficiaries of family support are the children of inmates and the
caregivers of those children. Because some children of inmates may get public support from
taxpayers, beneficiaries of PIE family support deductions may also include taxpayers being
partially reimbursed for welfare payments under TANF (Temporary Assistance for Needy
Families).

How Many Inmates Have Children?
According to the U.S. Department of Justice, Bureau of Justice Statistics, in 1999, 55.4 percent of
State prison inmates reported being parents of, on average, 1.84 minor children.40 As a
consequence, it appears that there are more unsupported minor children of State prison inmates in
the United States (about 1.32 million) than prisoners (about 1.30 million) in 1999. About 1 of
every 50 minor children in the United States is reported to have had a parent incarcerated in a
37

David Mace, Vermont Press Bureau, “State Panel Looks at Crime Restitution” January 17, 2002, retrieved July 22,
2003 from the Times Argus Web site: http://timesargus.nybor.com/Legislature/Story/40828.html
38
U.S. Department of Justice, Office for Victims of Crime, Promising Practices and Strategies for Victim Services in
Corrections, Washington, DC: U.S. Department of Justice, July 1999, 21, NCJ-166605.
39
Ibid., 21.
40
U.S. Department of Justice, Bureau of Justice Statistics, “Incarcerated Parents and Their Children,” Washington, DC:
U.S. Department of Justice, August 2000, 1, NCJ-182335.

41

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

State or Federal prison in 1999, representing 0.8 percent of all white children, 2.6 percent of
Hispanic children, and 7 percent of black children.41 If incarceration and parenting rates are
similar for jail inmates, then about 2 million minor children (about 3 percent of them) in the
United States have at least one incarcerated parent, and corresponding percentages would then
reach about 1 percent of white, more than 3 percent of Hispanic, and more than 10 percent of
black children having at least one incarcerated parent in a U.S. jail or prison.

Who Takes Care of Inmates’ Children?
In the vast majority of cases, the caregiver of a State inmate’s child is the child’s mother; if the
inmate is the mother, the caregiver is typically the child’s grandmother (Table 16).

Table 16. Caregivers, Children of State Inmates 1997
Caregiver
Other Parent
Grandparent of Child
Other Relatives
Foster Home/Agency
Friends/Others

Percent Caregivers
Total*
85.0
16.3
6.4
2.4
5.3

Male Inmates
89.6
13.3
4.9
1.8
4.9

Female Inmates
28.0
52.9
25.7
9.6
10.4

*Sum of percentages exceed 100 because some prisoners had children in different homes.
Note: 55 percent of inmate parents (56 percent of males, 36 percent of females) reported not living with
their children before admission.
Source: Bureau of Justice Statistics, “Incarcerated Parents and Their Children.”

Statistics are unclear about a number of important characteristics of child care givers, including
whether the mothers live as single parent householders (as heads of households) or whether they
live with others, either significant others or with their own parent or parents, important
differentiating characteristics with respect to the child’s probability of living in poverty. Further,
it is not clear whether or not the children live in poverty or what percentage receive taxpayer
support. No doubt, among the 2 million children of U.S. jail and prison inmates, many are not in
significant financial need, and it would be incorrect to assert without evidence that all or even
most children of inmates are in poverty.
Nevertheless, available data strongly suggest that the children of inmates are much more likely
than the general population to live in lower income or even impoverished households:
•

41

42

The (annualized) median income of State inmates in the month prior to arrest—likely less
than $12,000 in 1997—would, if the sole source of income for a family of four, equal less

Ibid., 2.

Deductions and Beneficiaries Elaborated

•

•

•

•

•

than 75 percent of the defined poverty level, $16,600 in 199842 and less than half 1.5 times
the defined poverty level, commonly considered a better measure of minimal annual income
requirements. By the broader standard, 77 percent of male inmates—and 87 percent of female
inmates—appeared to have earned less than a poverty level annual income prior to
incarceration.
One-third of all female headed households in the United States were in poverty in 1998, and
46 percent of children in female-headed households were in poverty in that year.43 With the
vast majority of male inmates’ children cared for by mothers (fathers absent), probability of
poverty for inmates’ children appears disproportionately high.
1999 data for Maryland show 6,720 children, 12 percent of the State’s children on welfare
(TANF) having at least one parent incarcerated.44 There is reason to believe the actual
percentage of Maryland children on TANF having an incarcerated parent may be
significantly higher, however, because only parents incarcerated in Maryland (and not the
District of Columbia, Pennsylvania, Virginia, or elsewhere) are enumerated and the total
population includes parents whose whereabouts are unknown. At an average monthly TANF
expenditure per recipient of $317.30, Federal and State taxpayers likely expended $26 million
dollars in 1999 assisting the children of Maryland State prison inmates.45
If 12 percent of all 5.3 million U.S. TANF children in 1999 were the unsupported children of
jail and prison inmates, at the National average monthly payment of $214.06 per month, US
federal and State taxpayers would have expended $1.64 billion in that year for welfare
payments to the unsupported minor children of U.S. incarcerants.46
Moreover, if the Maryland data are roughly correct, the TANF implication would suggest that
slightly more than 50 percent (51.7 percent) of the minor children of U.S. jail and prison
inmates are in poverty. However, though it would appear safe to conclude that a high
proportion of inmates’ children are poor, one should be mindful of the tenuousness of the
procedure by which exact numbers were developed and remain appropriately skeptical of any
precision in the measure provided.
Anecdotal information for the District of Columbia indicates that about 50 percent of the
District’s children on welfare have at least one parent who either is or was incarcerated.47

Who Gets Child Support?
In most instances, child support payments go to the household and the children. However,
inmates’ child support payments may not go to the children or families of welfare recipients, but
instead are returned to Federal and State taxpayers. Temporary Assistance for Needy Families
(TANF), commonly known as welfare, is the monthly cash assistance program for poor families
42

World Almanac Books, U.S. Bureau of the Census data, in World Almanac 2002, 387, New York: World Almanac
Books, 2002.
43
Ibid., 388.
44
Personal communication with John Cannon, Maryland Division of Human Resources, March 2000.
45
Ibid.
46
World Almanac Books, Administration for Children and Families, Office of Planning, Research, and Evaluation,
U.S. Department of Health and Human Services, in World Almanac 2002, 387, New York: World Almanac Books,
2002.
47
Personal communication with Kate Jesberg, Administrator, Income Maintenance Administration, District of
Columbia Department of Human Services, June 24, 2003.

43

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

with children under age 18. TANF is funded both by State and Federal taxpayers, with State
taxpayers paying from 25 percent to 50 percent, depending upon the State.48 On average, Federal
taxpayers pay about 57 percent of TANF payments.49 When an inmate’s child support payment is
paid on behalf of a TANF (welfare) child, the payment goes instead to reimburse Federal and
State taxpayers, generally in the Federal and State proportions initially paid; in about half the
States, the first $50 or so of child support may be passed directly through to the family without
reducing the TANF amount received (an amount in total taken from the State share of the welfare
payment).50

48

“Federal Financial Participation in State Assistance Expenditures for October 1, 2000 Through September 30, 2001”
in the Federal Register, 65(February 23, 2000, Number 36), 8979–8980.
49
Personal communication, author with Robert Stewart, Office of Health Policy, U.S. Department of Health and
Human Services, April 24, 2003.
50
Roberts, Paula, and Michelle Jordan, “State Policies Regarding Pass-Through and Disregard Of Current Month’s
Child Support Collected for Families Receiving TANF-Funded Cash Assistance,” Center for Law and Social Policy
(CLASP) (Washington, D.C., April 2002), retrieved April 28, 2003 from the CLASP Web site:
http://www.clasp.org/Pubs/DMS/Documents/1013104441.19/Pass_thru0402.pdf.

44

V. Research Results
This chapter quantifies PIE benefits for the United States during a single unspecified calendar
year between 1998 and 2001; county and local PIE payrolls and benefits are not included. It
quantifies actual PIE wages and beneficiaries in three ways:
1. as reported to the National Correctional Industries Association, mean values for 1998–2001;
2. as derived using the sample data as reported, weighted to a U.S. total; and
3. as derived by GWU staff, using the sample data, weighted to a U.S. total, but after
incorporating (a) employer payroll taxes, (b) adjusting for estimated actual Federal and State
income tax liabilities, and (c) where necessary, based on information provided by sampled
States, adjusting PIE deductions.
Throughout, total benefits are measured in two ways, first as “inmate gross income,” and second
as “employer gross payout.” “Inmate gross income” represents the total wages and salaries paid
by the employer to the worker, and either is or estimates the values reported by NCIA as “gross
wages.” “Employer gross payout” includes inmate gross income. When addressing PIE statistics,
inmate “gross income” includes workers’ hourly wages, including regular work hours, overtime
hours, awards, and bonuses reported as part of annual earnings to the IRS.
Sampled States were also asked if any benefits other than wages and salaries were offered, and
were specifically asked about annual leave, sick leave, individual or family health insurance,
retirement programs (other than Social Security), and employee stock ownership, along with
inquiry about unique PIE benefits provided by departments of corrections, such as privileges,
1
housing benefits, good times credits, or other benefits.
No sampled customer model firm offered any of the specified benefits or any other notable
benefit tied solely to PIE work. One employer-model firm reported providing paid holidays and
annual leave; no other notable benefits other than wages and salaries were identified for
employer-model firms.

1
PIE inmates may receive good time or other correctional benefits not addressed by this research (benefits not unique
to PIE but received by all inmates as part of correctional work, training, or other programs).

45

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Therefore, for the vast majority of sampled cases, a PIE inmate’s total financial compensation
appears fairly well accounted by payroll gross wages and salaries, and the payroll cost to the firm
by gross employer payouts.

PIE Wages and Beneficiaries
The PIE Hourly Wage Rate
Recognizing that sampling did not account for hourly wage rates, the sampled States are not
known to be representative of wage rates for the overall PIE inmate population. Moreover, in a
significant subset of sampled cases, hourly wage rates were estimated based on general guidance
provided by the host State (Table 17).

Table 17. Percent Distribution, Hourly Wage Rate
Weighted Sample, PIE Inmates
Hourly Wage Rate ($)
5.15
5.16–5.50
5.51–6.00
6.01–6.50
6.51–7.00
7.01 +

Percent PIE Inmates
71.7
9.5
9.7
5.8
1.5
1.7

Median: $5.15
Mean: $5.35
Maximum: $8.50
Totals may not equal sum of components due to independent rounding

The PIE Work Week
Data from the U.S. Department of Labor, Bureau of Labor Statistics, indicate that two-thirds of
America’s workers work full-time 50–52 weeks a year; another 20 percent work part-time 50–52
weeks per year, where a full-time worker is considered to have worked 35 or more hours per
2
week, including paid vacation hours.
Recognizing that sampling did not account for weekly hours worked, the sampled States are not
known to be representative of work weeks for the overall PIE inmate population. Moreover, in a
significant subset of sampled cases, weekly hours worked were estimated based on general

2
U.S. Department of Labor, Bureau of Labor Statistics, “Work Experience of the Population,” retrieved July 22, 2003
from the Bureau of Labor Statistics Web site: http://www.bls.gov/news.release/work.t01.htm.

46

Research Results

guidance provided by the host State. Nevertheless, the combination of information on wage rates
and annual incomes, plus guidance on work hours per week, along with apparent only minor
variation within PIE programs, suggests relatively accurate conclusions at least for the sampled
States.
On average, sampled PIE inmates worked 28.6 hours per week during the survey year, part-time
work (Table 18). Factors likely contributing to less than full-time weeks include (a) unique or
peculiar non-representative characteristics of the specific State PIE programs selected for the
research, (b) inmates participating in joint traditional industry and PIE programs in which some
work hours contribute to traditional prison industries and others to PIE customer model
employment, and (c) restrictions on work hours resulting from correctional routines and
requirements—meal schedules, required counts, workday security processing, and so forth.

Table 18. PIE Hours Worked Per Week, Weighted
Sample, PIE Inmates
Percentile
Minimum Reported Hours:
25th Percentile:
Median (50th Percentile)
75th Percentile
Maximum

Hours
1.0
23.8
32.5
36.0
48.8

The PIE-Inmate Work Year
The beneficiaries research estimates the proportion of the survey year worked by all PIE inmate
3
workers receiving a W-2 for income earned during that year.
Recognizing that sampling did not account for weeks worked, the sampled States are not known
to be representative of work weeks for the overall PIE inmate population. Moreover, in a
significant subset of sampled cases, work weeks were estimated based on general guidance
provided by the host State. Table 19 provides estimates for PIE weeks worked.

3
The research does not estimate the total or average duration of PIE inmate’s participation in PIE work (multiyear or
across years) but only the totals and averages that occurred within a specific calendar year.

47

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 19. PIE Weeks Worked Per Year, Weighted
Sample, PIE Inmates
Weeks Worked

Percent

0–4
5–8
9–16
17–24
25–32
33–40
41–51
52

18.3
9.7
15.6
9.5
9.4
13.4
7.9
16.2

Minimum Weeks Worked : 1
25th Percentile: 7
Median Weeks Worked: 22
75th Percentile: 40
Maximum Weeks Worked: 52

Turnover in PIE Employment
The reported data also suggest significant within-year turnover in both employer model and
customer (including manager) model PIE employment. Whereas December 31, year-end data
reported to the NCIA suggest about 4000 inmates involved in State PIE-level work in the years
around twentieth century’s end, this analysis suggests that approximately 8500 State prison
inmates working in PIE in a given year.

Annual PIE Earnings and Their Distribution
Reported sample data (Table 20) indicate median annual PIE earnings more than triple the $700
or so earned on average in traditional prison industries but well below civilian averages,
reflecting the combined effects of partial year part-time employment and the Federal minimum
wage. Averages, of course, are affected by very high turnover; nearly one third of the sampled
inmate labor force earned more than $5000 in the survey year.

Mean PIE Wages and Deductions 1998-2001
As of December 31, 2001, 3,719 State prison inmates were reported working in U.S.PIE
programs; during that year, State-level PIE generated $33.1 million in inmate gross income,
suggesting a mean annual income per position of about $8,900. Of course, actual incomes slip
below that amount the greater the average annual turnover rate of the PIE inmate workforce.

48

Research Results

Table 20. PIE Inmate Gross Incomes: Estimated
U.S. State-Level PIE Inmate Population, Single
Survey Year 1998-2001
Annual Gross Income ($)
1–500
501–1,000
1,001–1,500
1,501–2,000
2,001–2,500
2,501–3,000
3,001–3,500
3,501–4,000
4,001–5,000
5,001–10,000
10,001+
Minimum Gross Income
25th Percentile
Median Gross Income:
Mean Gross Income:
75th Percentile:
Maximum Gross Income:

Percent
19.2
11.4
8.4
6.9
7.0
3.2
3.5
2.6
6.5
26.4
5.0
$5
$785
$2,345
$3,755
$6,373
$16,187

Under the PIE program, deductions, up to 80 percent of inmate gross income, may occur for room
and board, taxes, victims compensation, and for family support. During 1998-2001, on average,
deductions for the four categories were reported to NCIA as having absorbed about 60 percent of
PIE inmates’ gross incomes (Table 21), meaning that significantly more than half of every dollar
earned by a State-level PIE inmate rewarded someone other than the inmate; moreover, the
unmeasured but recognized practice of additional post-PIE deductions ensured that the primary
financial beneficiaries of PIE incomes are a collective of public and private beneficiaries other
than the inmate.
Using a sample of employer and customer/manager model programs, GWU staff collected and
reassessed State-level PIE programs for one unspecified calendar year during the same 1998-2001
period and, based on replies from those States, developed estimates for the same categories
(Table 21). Based on sample data, the estimated U.S. totals generally conform with reported
totals, but nevertheless vary slightly from that reported to NCIA initially. Total estimated
incomes derived from sample data yield results somewhat lower than directly reported; estimated
room and board deductions are slightly higher, taxes slightly higher, victims’ programs somewhat
less, and family support significantly less than the aggregate reported by all States. With the
important exception of family support, which from the sample is estimated much less than
actually reported, overall conclusions from the weighted sample data are not greatly different than
would be obtained from the reported values.

49

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 21. Mean Aggregate Annual U.S. PIE Wages and Deductions,
State PIE Programs Only, Unrevised (Weighted, as Reported by Sampled States)
Mean 1998–2001
Reported to NCIA
Category
Inmate Gross Income
Room and Board
Taxes (All)
Victims Programs
Family Support
Other PIE Deductions
Total PIE Deductions
PIE Residual**

($Million)
29.8
8.4
4.4
3.1
1.9
NA
17.9
12.0

Percent
100.0
28.3
14.8
10.6
6.2
NA
59.8
40.2

One Year 1998–2001
Estimated from Sample
($Million)
29.3
9.6
3.9
2.6
0.5
*
16.7
12.7

Percent
100.0
32.8
13.2
9.0
1.7
0.1
56.8
43.2

*Less than $50,000.
**Residuals do not equal either savings or the inmates’ ultimate share because post-PIE deductions occur.
Note: Sum of components may not equal totals because of independent rounding.
Source 1998–2001: Derived from Quarterly and Cumulative Reports to the National Correctional Industries Association
(NCIA), 1997-2001.

The research has examined PIE inmate incomes in more detail, clarifying deduction categories,
adjusting deduction amounts based on additional information, and accounting for additional
employer contributions not included in National reports. Table 22 begins the expansion of
original reports by adding estimated employer payroll taxes not reported to NCIA, but without
any other revisions; Table 22 suggests that, on average, PIE employers expend 8.8 percent more
than reported to NCIA each year, all to social insurance programs. Because all employer
contributions accrue to recipients other than the inmate, the inmate’s percentage share of
employer gross payouts is naturally somewhat smaller when employer contributions are
considered.
The final representation of PIE inmate incomes (Table 23) reflects the full range of GWU
estimated revisions, namely:
•

•

•

50

Estimated Employer contributions to Social Security, workers’ compensation, and
unemployment compensation, reflecting both employer-model and customer model practices
(weighted, as reported by sampled respondents);
calculated (rather than reported) Federal and State income tax liabilities assuming that each
PIE inmate is a single filer with no other income sources, who successfully files for and
receives, Federal and State income tax refunds. At the Federal level, PIE inmates are
assumed to pay no taxes until inmate gross income exceeds $7,150 but incur a 28% Federal
tax rate for all income above that amount; and
significantly revised deductions for family support, reflecting the weighted effects of
revisions to data reported to NCIA to account for (a) return of family support deduction
monies to PIE inmates and (b) actual inclusion with family support deductions of monies for

Research Results

other court-ordered purposes, such that, for some respondents, the family support deduction
actually represents “family and other court ordered” deductions.

Analysis
Sample data reflecting selected States and then weighted to the universe yield results somewhat
different than independently known from population data already available; as a result both
weighted U.S. totals and individual categories of deductions in this research differ in significant
regards. Nevertheless, overall, the GWU analysis of PIE inmate earnings indicates the following:

Table 22. Estimated Gross Employer Payouts, PIE Wages,
and Deductions, One Unspecified Year 1998–2001––
Weighted Sample State PIE Programs Only (Unrevised)
Category
Gross Employer Payout
Social Security (OASDI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

($Million)
32.04
1.42
0.33
0.11
0.11
0.76
2.74

Percent
100.0
4.4
1.0
0.3
0.4
12.4
8.6
Percent

Percent

Inmate Gross Income
Room and Board
Taxes, Federal Income (Deducted)
Taxes, State Income (Deducted)
Social Security (OASDI)
Social Security (HI—Medicare)
Victims Compensation
Family Support*
Other PIE Deductions*
Total Inmate PIE Deductions

29.30
9.60
1.79
0.33
1.42
0.33
2.63
0.50
0.04
16.64

91.4
30.0
5.6
1.0
4.4
1.0
8.2
1.6
0.1
52.0

100.0
32.8
6.1
1.1
4.9
1.1
9.0
1.7
0.1
56.8

PIE Residual**

12.65

39.5

43.2

*Some sampled States include other court-ordered deductions, interpreting the final deduction category
as “family and other court-ordered deductions.”
**Residuals are not net savings and do not equal the net amount accruing to the inmate because of
post-PIE deductions not accounted by PIE.
Totals may not equal sum of components due to independent rounding.

51

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 23. Estimated Gross Employer Payouts,
PIE Wages, and Deductions, One Unspecified Year 1998–2001—
Weighted Sample State PIE Programs Only (Revised)
Category
Gross Employer Payout
Social Security (OASDI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

($Million)
32.04
1.42
0.33
0.11
0.11
0.76
2.74

Percent
100.0
4.4
1.0
0.3
0.4
2.4
8.6
Percent

Percent

Inmate Gross Income
Room and Board
Taxes, Federal Income (Liability)
Taxes, State Income (Liability)
Social Security (OASDI)
Social Security (HI—Medicare)
Victims Compensation
Family Support*
Other PIE Deductions*
Total Inmate PIE Deductions

29.30
9.60
1.14
0.12
1.42
0.33
2.64
0.23
0.05
15.53

91.4
30.0
3.5
0.4
4.4
1.0
8.2
0.7
0.2
48.5

100.0
32.8
3.9
0.4
4.9
1.1
9.0
0.8
0.2
53.0

PIE Residual**

13.76

43.0

47.0

*Some sampled States include other court-ordered deductions, interpreting the final deduction
category as “family and other court-ordered deductions.”
**Residuals are not net savings and do not equal the net amount accruing to the inmate because of
post-PIE deductions not accounted by PIE.
Totals may not equal sum of components due to independent rounding.

1. Because of employer contributions, PIE gross employer payouts are substantially larger than
inmate gross income, an estimated 9.4 percent. These contributions, all to State and Federal
social insurance programs, contribute to reducing tax burdens on others and yield income
support benefits paid directly from PIE employers to workers, retirees, and other dependents
in the civilian population.
2. Because average PIE inmate earnings are below income tax thresholds, Federal and State
income tax liabilities are likely somewhat smaller, on average, than income taxes initially
deducted from PIE payroll to NCIA. Whereas Federal income tax deductions from payroll
indicate $1.8 million dollars paid by PIE inmates in income taxes, estimated Federal income
tax liabilities indicate only $1.1 million owed, based on a single taxpayer who does not
itemize or have other income or deductions; therefore, sample data suggest that 36 percent of
Federal taxes deducted are subject to refund to the PIE earner. Similarly, estimated liabilities

52

Research Results

suggest that nearly two-thirds of State income taxes deducted are subject to refund. As noted
in recent income tax debates, however, most PIE inmate workers significantly contribute
Social Security taxes, both for old-age, survivors, and disability insurance (OASDI) and also
for Medicare (HI) and thereby reduce tax burdens on the civilian population and contribute
substantial additional benefits to workers, retirees, and other dependents in the civilian
population.
3. Family support deductions appear substantially smaller than indicated in National data (table
21) for at least five significant reasons: First, and most important, sample selection suggests
very significant under representation of family support deductions among the States selected
for this research, for which family support deductions average less than 15 percent of the
National average reported to NCIA. Second, analysis shows that family support is, to some
extent, a collection of deductions more fully described as “family and other court-ordered
deductions.” Eighteen percent of “family support” deductions appear to have served other
court-ordered purposes such as fines, victim restitution, and court costs. Third, some family
support deductions may be escrowed, unused for lack of a child support order, and ultimately
returned to the inmate at release. Fourth, in absence of a court requirement, PIE programs
may perceive a lack authority to require the deduction. Finally, neither courts nor child
support interests enforce support obligations.
4. Nevertheless, by any measure, the collective of beneficiaries other than PIE inmate workers
are the primary beneficiaries of PIE inmate work; parties other than the PIE inmates receive
more than half of the financial benefits of either PIE inmate incomes or gross employer
payouts.
5. Although quantifying beneficiaries and benefits is difficult, being subject to significant
interpretation in assignment, and in the case of transfers and taxes such as income taxes and
Social Security, yielding coincident multiparty benefits (by relief) to taxpayers as well as to
recipients receiving the transfers, nevertheless some clear beneficiary classes emerge:
• State taxpayers are the single largest non-inmate beneficiaries of State PIE inmate
employment, estimated on average to receive roughly one third of PIE inmate gross
employer payouts, almost all of it via inmate board and room payments. Many PIE
inmates also pay State income taxes. Where court orders paying fines, covering court
costs, or paying other court-ordered expenses are enforced through PIE deductions (as
part of the “family support and other court ordered deductions), State taxpayers also
appear to gain a small additional benefit.
• Social Security, both OASDI and HI (Medicare) taxpayers and recipients constitute the
second largest non-inmate beneficiary class of PIE inmate employment, on average
estimated to garner about 11 percent of PIE inmate gross employer payouts, more than
half paid by inmates and the rest by employers. Thus both retirees (and the disabled, etc)
and all parties paying into Social Security are beneficiaries of PIE inmate employment.
• Current Crime Victims, and to a much smaller extent, the prior individual victims of PIE
inmate workers, constitute the third largest category of non-inmate beneficiaries of gross
PIE payouts, receiving 4 percent of employer gross payouts. State victim compensation
programs compensate current crime victims; court-ordered restitution compensates the
identified victims of PIE inmates’ crimes. Although the share is not known, a small

53

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

•

•

•

•

fraction of victim compensation deductions is diverted to crime victim restitution as part
of State compensation programs and, in a few instances, deductions reported as “family
support” in fact include court ordered deductions for individual victim restitution.
Federal Income Tax payers and programs dependent on the Federal budget constitute the
fourth largest beneficiary groups, garnering about 6 percent of PIE inmate incomes. In
addition to providing some relief to other Federal taxpayers, PIE inmate deductions for
Federal income taxes can be said to support National defense, income security
(Supplemental Social Security, food stamps, Temporary Assistance to Needy Families,
the earned income tax credit), and Federally supported health care (Medicaid,)
Unemployment and Workers Compensation Payees and Claimants constitute the fifth
largest category of non-inmate PIE beneficiaries, with beneficiaries including both other
private firms paying into unemployment and workers compensation funds having
additional funding provided by PIE firms, and by civilian workers and their families,
idled, ill, or injured, receiving benefits from unemployment or workers compensation
funds, garnering 3 percent of employer gross payouts, all from employer contributions.
To the extent Federal and State taxpayers are relieved of subsidizing these funds, both
groups can also be viewed as beneficiaries of PIE programs.
Families having child-support orders are the smallest and last category of non-inmate
beneficiaries of deductions from PIE inmate employer gross payouts. Even under the
best of circumstances (reported national average rates shown in table 21), child support
constitutes just 6.2 percent of PIE incomes. At worst, child support accounts for as little
as 0.7 percent of gross employer payout. Of course, PIE deductions do not reflect other
voluntary or enforced withdrawals or deductions for child or family support that occur
post PIE.
PIE Residual: National statistics suggest that about 40 percent of each dollar of PIE
inmate gross income devolves to the PIE inmate worker. However, the likelihood that
most PIE inmate workers get refunded most or all of their Federal and State income
taxes—an outcome anecdotally corroborated during this research—and that inmates may
in some instances be refunded some portion of family support deductions, lead to a
revised estimate of the gross PIE inmate residual equaling 43 percent of employer gross
payout and 47 percent of inmate gross incomes. How much of the PIE residual share that
is ultimately retained under the PIE inmate’s control is undetermined at this point,
because additional deductions after PIE are known to occur. Evidence of additional
deductions after PIE, and the beneficiaries of its distribution, will be discussed separately.

Discussion
Working more carefully with sample State data and then comparing detailed information with
reports to NCIA and also NCIA reported data of other PIE States, along with other information,
indicates significant variation in PIE deduction practices among PIE States, such that
generalizations from any small sample, especially when based on a few characteristics, can vary
markedly as a function of the sample drawn. As a result, while the estimated PIE totals appear to
be relatively close to actual National totals and the rank order of individual shares appear to be

54

Research Results

fairly clear, some caution in assigning specific percentages, especially to the smaller deduction
categories, like family support, is warranted.

Social Security
Where Social Security is deducted, both OASDI (6.2 percent) and HI (1.45 percent) are deducted
at the standard rates, both for employees and for PIE inmate workers. However, although
employer model programs must deduct Social Security, in practice customer model programs
occasionally voluntarily deduct Social Security from PIE workers and also contribute the
employer shares as well.

Workers Compensation
The very limited information obtained in this research suggests that the 2.4 percent worker
compensation payment rates paid by participating PIE firms equal or exceed National averages
(estimated 1.28 percent of employee wages), despite PIE workers’ general ineligibility for
workers compensation while incarcerated.4

Unemployment Compensation
No sampled customer model programs, but some (not all) employer model programs contribute to
State and Federal unemployment insurance programs, contributing 0.7 percent of employer gross
payout, less than half the National average rate of 1.98 percent. PIE inmate workers, of course,
are not eligible for unemployment compensation while incarcerated, making the entire
contribution a transfer to the civilian unemployed.

Room and Board
PIE deductions for board and room are significant, averaging about $1,250 per year per PIE
inmate worker, on average contributing an amount equaling roughly 5 percent of the estimated
$25,000 annual cost of incarceration.5 But who receives the PIE room and board deduction? In a
polling separate from the research sample, 15 PIE participating States indicated more specifically
the recipient of the PIE room and board deduction:
•
•
•

sixty percent reported board and room reverting to industries, including PIE;
twenty-seven percent reported board and room reverting to State general funds; and
thirteen percent reported board and room reverting to Departments of Corrections.

4

While exact rates were calculated for some sampled States, in other instances where the contribution rate was not
known, the National average rate was used.
5
Figures adjusted for inflation plus assumed 10 percent real cost increase since 1996 based upon historical trend.
Source: U.S. Department of Justice, Bureau of Justice Statistics, State Prison Expenditures, 1996, Washington, DC:
U.S. Department of Justice, August 1999, NCJ172211.

55

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Board and room deductions reverting to industries appear to be used for the full range of PIE and
industry costs borne by departments of corrections, including for buildings and equipment, for
staff payroll, and for administrative costs.
Taxpayers are the unambiguous beneficiaries of room and board deductions reverting to the State
general fund, presumably also benefiting recipients of State priorities, particularly education,
health care (Medicaid) and transportation. To the extent one assumes that monies reverting to
industries and departments of corrections reduce or compensate for taxpayer-borne correctional
programming, taxpayers are again the beneficiaries of the board and room deduction.

Taxes
This research did not rigorously examine procedures by which Federal and State income tax
withholding occurs in State PIE programs, although by and large the impression gained was one
of inmate discretion in declaring a filing status, just as is normal for civilian employees. However,
PIE-reported deductions for taxes, including income taxes (and Social Security) appear to be
substantially in excess of income tax liabilities. To the extent total PIE deductions and other
policies (like deductions for other purposes) are affected by the size of individual PIE deductions
or post-tax PIE earnings—such as the 80 percent upper bound for the sum of all PIE deductions—
further attention to actual liabilities rather than initial PIE deductions may be useful. It is possible,
of course, that some PIE inmates do not file as single filers or have other income or deductions;
anecdotal evidence, however, suggests that significant variations in tax liabilities are unlikely.

Victims Compensation, Restitution
Clearly, most deductions for victims compensation go to State crime victimization funds and then
to current crime victims. However, some State victim compensation funds also transmit
restitution payments to an offender’s own victim, when restitution orders are in place.
Research staff briefly examined the role of PIE deductions for State crime victim compensation
programs. Taken together, PIE contributions appear to have provided in the neighborhood of 1
percent of all crime victim compensation funds distributed by the host States during the time
period studied, excluding the Federal contribution.6 However, for those same States, despite their
being only about one half of one percent of all prison inmates in those States, 7 PIE inmate
contributions appear to have constituted more than 99 percent of the contributions to victim
compensation made by their States’ inmates in that year. Said otherwise, in the United States, it
would appear that inmates overall contribute little or nothing to crime victim compensation.
However, where inmates are employed and a mechanism of crime victim compensation exists,
6

Source: Personal conversations and information exchange, research staff with sampled States’ State crime victim
compensation program offices.
7
Sources for this calculation: U.S. Department of Justice, Bureau of Justice Statistics, Bulletin, “Prisoners in 2000,”
Washington, DC: U.S. Department of Justice August 2001, NCJ 188207; National Correctional Industries Association,
December 31, 2001 PIE employment, retrieved May 13, 2002 from the NCIA Web site:
http://www.nationalcia.org/qtr0401stats.pdf.

56

Research Results

inmates do materially contribute to crime victim compensation. Moreover, State policies on
inmate employment would appear to have a very direct relationship to State crime victim
compensation funding.
Research analysis showed some, though few, explicit restitution payments occurring as PIE
deductions via court orders categorized in reports to NCIA as “family support and other.”
Anecdotal information obtained in this research suggests that restitution deductions do not
generally occur—except via victims compensation programs in some States—as part of PIE
deductions. Restitution claims are generally exercised post-PIE.

Family Support
No evidence of PIE deductions for “family support” other than child support was found, such as
for parents, grandparents, spouses, or other significant relations other than children, although nonsampled States may feature such categories for PIE deductions.
Five factors appear to propel the small proportion of sampled PIE inmates incurring PIE
deductions for child support, despite the very high proportions of State prison inmates having
minor children. First, and most important, sample selection suggests very significant under
representation of family support deductions among the States selected for this research, for which
family support deductions average less than 15 percent of the National average reported to NCIA.
Second, analysis shows that family support is, to some extent, a collection of deductions more
fully described as “family and other court-ordered deductions.” Eighteen percent of “family
support” deductions appear to have served other court-ordered purposes such as fines, victim
restitution, and court costs. Third, some family support deductions may be escrowed, unused for
lack of a child support order, and ultimately returned to the inmate at release. Fourth, in absence
of a court requirement, PIE programs may perceive a lack authority to require the deduction.
Finally, neither courts nor child support interests enforce support obligations.

57

VI. Estimating Benefits if PIE Participation
Were Widespread
This section estimates PIE incomes, beneficiaries, and benefits if U.S. Federal and State inmates
were generally engaged in PIE work at average U.S. employment and income levels, in jobs at
skill levels and productivity matching some definition “typical” of U.S. labor.

Assumptions
Estimating potential incomes, beneficiaries, and benefits is highly dependent upon assumptions,
including assumptions for the percentage of U.S. inmates participating in the PIE program, hours
worked, productivity, wage rates, and deduction policies. Assumptions about post-PIE individual and
social behaviors could also play a part, for example, for State and Federal policies regarding fines and
other court-orders, and for voluntary purchase preferences for much higher income inmates.
For this exercise, two groups of assumptions are used (Table 24) to set the range of low-to-high
“reasonable” incomes, benefits, and beneficiaries if PIE work were generally applied in the
United States. Obviously other assumptions could be chosen as well.
Prison population assumptions are rounded close approximations of actual State and Federal
prison populations, chosen for a grand total, with jail inmates, equaling 2 million.
For the “high” case, while no explicit assumptions are made about the inmate or program
characteristics, the prison population is assumed to have whatever it takes to succeed at the
specified level, meaning that they are not used for institutional maintenance, are non violent, are
not segregated, and that they possess the physical, mental, motivational, educational, and
technical skills to fully meet the job requirements. Anecdotal evidence heard by research staff
indicates that at least 50 percent of the inmate populations are both willing and able to hold
regular jobs. Furthermore, the higher “upper bound” assumptions implicitly permit such
behaviors as (a) multiple jobholding (b) overtime, and (c) fewer inmates in institutional
maintenance or inmates having dual duties, in both maintenance assignments and a job.
The “low” annual income approximates full-time minimum wage work, while the upper bound
represents the rounded total for 52 weeks’ full time work at the median weekly earnings rate of
male wage and salary workers ages 25 and older in 2000.1
1

Source: U.S. Department of Labor, Bureau of Labor Statistics, in The World Almanac and Book of Facts 2002, New
York: World Almanac Books, 146.

59

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 24. Assumptions, Two Cases (Low and High): U.S. PIE Income, Benefits,
and Beneficiaries if PIE Were Generally Applied in U.S. Federal and State Prisons
Category
U.S. State Prison Population
U.S. Federal Prison Population
Total U.S. Prison Population*
Proportion in Full-Time PIE Work
Annual Income
Taxation
Federal

State

Share of Inmates Paying State Income Taxes
PIE Assumptions:
Social Security Contributions
Employee
Employer
Unemployment Compensation Contributions
Workers’ Compensation Contributions
Board and Room
Victim Compensation
Family Support
Court Orders

Low
1,260,000
140,000
1,400,000
50%
$10, 500
Rate Ceiling
0% $7,150
28%
$33,250
Rate Ceiling
0% $4,700
4% > $4,700
50%

50%
50%
0.7%
0.7%
28.3%
10.6%
2.0%
0.0%

High
1,260,000
140,000
1,400,000
75%
$32,000
Rate Ceiling
Same
Same
Rate Ceiling
Same
Same
67%

100%
100%
1.0 %
1.0%
32.3%
11.0%
6.2%
0.2%

*Although local jail populations are generally excluded from the estimates, on occasions where jails are
included, the total U.S. jail population is assumed to be 600,000 persons.

Compared with a U.S. average unemployment compensation contributions rate for employers of
about 1.28 percent of gross wages, and workers compensation rate of 1.25 percent, the assumed
“low” unemployment compensation contributions rate assumes both low premium rates and
relatively low participation rates, as if many PIE programs were customer model PIE systems not
contributing; “high” implicitly assumes both higher average contribution and higher participation
rates (including more employer model firms).
The board and room deduction rates represent a “low” of 28.3 percent, the rate reported to NCIA
for calendar 2001 (Table 21), and a high of 32.3 percent, the estimated U.S. weighted value for
the research survey year (Table 23).
The victim compensation “low” rate represents the rate reported to NCIA for 2001; “high”
represents an arbitrary value greater than the NCIA rate.

60

Estimating Benefits if PIE Participation Were Widespread

The “family support” rate is exclusive to family support, and not for other court-ordered costs.
“Family support” assumes either child support orders or voluntary payments to support children
or other “ family,” such as parents, grandparents, care giving relatives, spouses, significant others,
or others approved for payment as “family” by the department of corrections or PIE program.
“Low” represents an arbitrary minimum rate exclusively for family support; “High” repeats the
2001 NCIA reported average rate, but assumed exclusively for family.
The “Court Orders” “low” assumes that court ordered payments, such as for court costs and fines,
do not occur as PIE deductions, and for “high” that such payments occur as PIE deductions at
some rate in the range of that observed in the research.

Results

Table 25. Results: Low Case Assumptions. U.S. PIE Income, Benefits, and
Beneficiaries If PIE Were Generally Applied in U.S. Federal and State Prisons
(Assumed Income per PIE Inmate $10,500 per Year)
Category
Gross Employer Payout
Social Security (OASDI + HI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

($Billion)

Percent

7.68
0.28

100.0
3.7

0.03

0.3

0.03
0.33

0.3
4.3
Percent

Percent

Inmate Gross Income
Room and Board
Taxes, Federal Income (Liability)
Taxes, State Income (Liability)
Social Security (OASDI)
Social Security (HI—Medicare)
Victims Compensation
Family Support
Other PIE Deductions
Total Inmate PIE Deductions

7.35
2.08
0.66
0.09
0.28

95.7
27.1
8.6
1.1
3.7

100.0
28.3
8.9
1.2
3.8

0.78
0.15
0.00
4.03

10.1
1.9
0.0
52.5

10.6
2.0
0.0
54.9

PIE Residual

3.32

43.2

45.1

61

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table 26. Results: High Case Assumptions, U.S. PIE Income, Benefits,
and Beneficiaries if PIE Were Generally Applied in U.S. Federal and
State Prisons (Assumed Income per PIE Inmate $33,250 per Year)
Category
Gross Employer Payout
Social Security (OASDI + HI)
Social Security (HI—Medicare)
Federal Unemployment Insurance
State Unemployment Insurance
Workers Compensation
Employer Contributions

($Billion)

Percent

36.84
2.57

100.0
7.0

0.34

0.9

0.34
3.24

0.9
8.8
Percent

Inmate Gross Income
Room and Board
Taxes, Federal Income (Liability)
Taxes, State Income (Liability)
Social Security (OASDI)
Victims Compensation
Family Support
Other PIE Deductions
Total Inmate PIE Deductions
PIE Residual

Percent

33.60
10.85
7.31
0.85
2.57
3.70
2.08
0.07
27.42

91.2
29.5
19.8
2.3
7.0
10.0
5.7
0.2
74.4

100.0
32.3
21.7
2.5
7.7
11.0
6.2
0.2
81.6*

6.18

16.8

18.4

*Total exceeds PIE deductions limit of 80 percent

Analysis
Contemporary PIE engages less than 1 percent of the U.S. prison inmate population, typically less
than 40 hours per week, much less than 52 weeks a year, and in positions paying at or near
Federal minimum wage. Not surprisingly, then, considerable unused productivity appears to
remain in the State and Federal inmate labor forces, and their PIE work holds prospect of sizeable
potential benefits available from their fuller participation and increased productivity, making
pursuit of increased inmate contributions enticing. At the same time, any set of assumptions about
dramatically improved output from the U.S. inmate labor force awaits actual demonstration in the
competitive marketplace and is likely to entail considerable effort and adjustment in the doing.
First, were the United States to succeed in engaging a large share of the U.S. State and Federal
inmate labor force, their gross labor contributions would likely be much greater than currently
realized. Whereas PIE inmates’ annual gross incomes nationwide today total about $30 million

62

Estimating Benefits if PIE Participation Were Widespread

per year, generalized State and Federal inmate PIE work suggests total U.S. prison inmate labor
income potential ranging from about $7 billion to nearly $37 billion per year (Tables 25 and 26).2
Second, were U.S. State and Federal inmates generally successful in PIE at the assumed levels,
(1) the primary beneficiaries of general PIE inmate work would be Federal and State taxpayers,
Social Security contributors and Social Security beneficiaries, crime victims, families, and others
than the inmate, and (2) the greater the inmates’ productivity and incomes, the greater the
proportional benefits to others. Said otherwise, the greatest beneficiaries of widespread inmate
PIE work would not be inmates but others than inmates, and the most significant stakeholders
negatively affected by current non work by U.S. prison inmates are these same groups.
Of course, many benefits have more than one beneficiary—such as wherever both tax payers and
recipients of tax benefits are affected—suggesting a sum of beneficiaries shares much greater
than 100 percent. And the exact distribution of many benefits is clouded by their movement
through the tax and transfer system.
State taxpayers, including both households and business taxpayers—and the programs supported
by State taxpayers, presumably education, Medicaid, and roads and transportation—would likely
be the single greatest non-inmate beneficiary classes if PIE inmate work were generally
successful, probably accounting for roughly one-third of the total benefits of general PIE inmate
work. Under the assumed income levels, PIE inmates could contribute between $3000 and about
$10,000 per year (12-40 percent) toward their individual costs of incarceration, particularly if
large shares of these amounts either entered the State’s general fund or explicitly reduced
taxpayer costs. Because many inmates are assumed to remain outside PIE, of course, the
aggregate average inmate contribution would fall short of that percentage; moreover, even under
very aggressive assumptions, State taxpayers would continue bearing the heavy majority of
incarceration costs. Nevertheless, the $2–$11 billion savings to State taxpayers from PIE inmate
room and board contributions could be expected to result in either (a) lower overall State tax rates
and/or (b) increased funding meeting State obligations for education, Medicaid, roads and
transportation, and other State-funded programs.
State taxpayers would enjoy other incomes or cost reductions as well. As PIE workers become
taxpayers and average incomes increase, nationwide, State income tax liabilities could increase
from less than $1 million a year today (Table 23) to $660–$850 million a year under general
inmate PIE engagement. Family support offsets—reimbursing the State for the proportion of
child support payments originally funded by State taxpayers through welfare—would likely refill
State coffers to the tune of some tens of millions of dollars a year to a maximum near $1 billion
(if virtually all family support deductions offset State-supported TANF payments, probably
highly unlikely). Similarly, State taxpayers may enjoy reduced costs in funding State victim
compensation programs, and in supplementing State worker and unemployment compensation

2

Including 600,000 local jail inmates suggests total U.S. adult inmate income potential ranging from $10 billion to $48
billion per year.

63

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

programs. Repayment of court costs, along with contributions to payoff of fines, could also
reduce State (and Federal) taxpayer costs of administering the criminal justice system.
Social Security and other social safety net wage earner and employer payees along with retired,
widowed, disabled, dependent, and other Social Security recipients would likely constitute the
second-largest non-inmate categories of beneficiaries. Combining employer and worker
contributions, PIE inmates total contribution potential to the Social Security—OASDI and HI
(Medicare)—suggests $500 million—$5 billion per year, alone absorbing between 7 and 14
percent of PIE employers’ gross payouts.
Federal income tax payers—and all programs dependent upon Federal income tax collections—
constitute either the second or third largest recipient group. Depending upon income assumptions
for the PIE inmate population, PIE inmate Federal tax payment potential is estimated able to
contribute as little as $660 million a year (about 8 percent of PIE earnings) under low case
assumptions and as much as $7.3 billion a year to the Federal treasury (about 20 percent of PIE
earnings) under the most optimistic assumptions.
Crime victim compensation funds stand to gain significantly if PIE inmate work were generally
successful. Whereas inmates (with the exception of PIE workers) typically contribute nothing or
almost nothing to crime victim compensation programs today, if PIE were widely applied, total
contributions to U.S. crime victims compensations programs could range from about $0.8 billion
to about $3.7 billion a year, an amount far greater than currently paid out nationwide each year
(about $300 million) but still less than 1 percent of the estimated greater than $350 billion per
year cost of violent crime alone (Table 14).
Family support, including to meet child support orders, could increase substantially, particularly
if a higher proportion of inmates incurred PIE deductions for family support, potentially
exceeding $2 billion under the most optimistic assumptions here, reaching as much as 6 percent
of total payouts.
Except under higher income scenarios, inmates likely remain the single largest beneficiaries of
PIE work, retaining at the conclusion of PIE deductions somewhere between 17 (under highest
income) and 43 percent (under least income) of gross annual payroll earned in PIE. Therefore,
inmates appear to retain very significant proportions of earnings, at least through the PIE phase.
Under the lowest assumed annual income, $10,500 per year, PIE inmates would be expected to
retain 45.1 percent, more than $4,700 per year; at the high end, inmates earning $33,250 per year,
could be expected to retain about $6,650 (the 20 percent minimum required by PIE), in this case a
function of both the progressive tax system and the assumed higher rates of deduction in the high
income case. The inference for social policy would appear to be that the rewards of inmate work
under assumptions of these scenarios, both generally and as inmate incomes increase, accrue at
least as much to non-inmate beneficiaries as to the PIE inmate workers.

64

Estimating Benefits if PIE Participation Were Widespread

The exercise also suggests that, at least theoretically, the 80 percent upper bound for PIE
deductions could be approached if PIE inmates’ saw their gross incomes reach or exceed $30,000
per year and tax and other deduction assumptions occurred. While unlikely, seeing conditions
under which the bound occurs nevertheless informs policymakers about conditions under which
individual PIE workers could hit limits, and of the importance of measuring tax liabilities rather
than merely PIE defined deductions or voluntary (but refunded) payroll deductions.

Caveats
While attractive, estimating potential benefits of general PIE participation is an incomplete
picture of both the challenge and the potential outcomes of PIE work. Among issues to be
considered, correctional changes needed to provide safe, competitive, efficient, and profitable
workplaces are not addressed; techniques providing land, buildings, equipment, and services are
not addressed; the methods by which correctional locations and PIE firms become attractive and
competitive are not addressed, as are obvious issues of widespread integration of production for
the civilian economy. Public attitudes and acceptance would be critical. Issues affecting both (a)
increased opportunities for U.S. domestic business expansion and additional civilian hiring and
(b) relationships and conditions of competition with civilian labor would likely need to be
addressed.
Nevertheless, estimating potential benefits of widespread inmate work suggests the order of
magnitude of potential good that could result from inmate work; the exercise suggests that
considering the challenges of integrating inmates more fully into the American economy may
yield benefits and produce beneficiaries well beyond those traditionally recognized, and spur
those beneficiaries to weigh public policies in their light.

65

VII. Disposition of the PIE Residual

Introduction
The main body of this report answers the question, “Who gets PIE incomes?” with estimates of
PIE inmate income and its distribution through Federally authorized PIE deductions, concluding
that 53–57 percent of PIE incomes go to beneficiaries other than PIE inmates. But the disposition
of the remaining 43–47 percent of PIE incomes is not described; in absence of additional
information, the reader is cautioned against interpreting this unassigned PIE residual as
“belonging” to the inmate. However, by relying on PIE inmates’ prison bank account
information, this section extends the analysis and offers some information on the disposition of
the remainder unassigned after PIE-authorized deductions. It provides introductory estimates of
post-PIE deductions and of the residual amounts ultimately under the inmate’s control. The
section also offers additional information on PIE inmates’ expenditures and on other “non-PIE”
sources of income.
All information in this section is obtained from monthly prison account data for a subset of PIE
inmates sampled for this research. Each sampled State was asked to provide two months’ inmate
account information, first for one “typical” PIE month and another for a “typical” pre-PIE month.
“Typical” was defined to be a month in which incomes and outflows would likely be unaffected
by unusual events, such as by Christmas or by impending release. September was suggested as a
preferred choice for inmates working in that month.
Information for this section was provided by sampled States as optional assistance to the PIE
research; not all requested information was obtained. Obtaining account information tended to be
challenging, given PIE office unfamiliarity with inmate bank accounts, their dependence on
corrections offices beyond their own familiarity or control, and the burden of responding to
research inquiries about unfamiliar information. To minimize reporting burden, respondents
tended to select a specific calendar month and report all PIE inmate accounts for that month
(usually September of the survey year), whether or not the inmate worked in PIE during that
month. Some States were unsuccessful in reporting pre-PIE account information at all.
Respondents providing pre-PIE month information also chose a single month in the year prior to
the PIE sample month, a technique that yielded few acceptable records, given that many inmates
were already PIE workers in the prior year and others were not yet incarcerated in the selected
facility during that month.

67

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

As a result, satisfactory PIE-month information was provided for less than half the total sampled
inmate population—462 from a total sample of 968 persons—and only 104 pre-PIE records were
ultimately found useful for examining pre-PIE accounts.
Interpreting information introduced additional uncertainties. Typically, respondent States
provided paper copies of inmate accounts, in each State containing transaction categories and
codes not immediately meaningful. Some transactions reflected accruals to sub accounts and, in
fact, were neither deposits nor withdrawals to the inmates’ overall holdings; some transactions
repeated PIE payroll and PIE deductions already accounted in the PIE payroll information. And,
of course, interpretation of the kind of transaction and party with whom the transaction occurred
was often subject to research judgment. Although PIE staff confirmed general interpretations (for
example, whether a commonly recorded name represents a person or a business, and what kind of
business?), resource constraints prevented detailed interpretations.
Estimating overall PIE income disposition also involved converting the annual PIE income data
to monthly averages and then combining the separately obtained PIE data with the inmate account
information.
For this section distributions of the data are not weighted to represent the overall PIE population,
and simply represent the distributions and averages for the 968 sampled records for overall PIE
income and PIE deductions and 462 acceptable records or specified subsets for post-PIE
information. Reporting and account acceptability rates differ dramatically among sampled States.
For many categories, overall means for all reporting inmates are greatly reduced because few
individual inmate records feature the specified income or expenditure category. Therefore,
whereas tables display overall means for all inmates, supporting text highlights averages for those
few inmates whose accounts actually feature income or expenditures in the specified categories.
Obviously, caution is appropriate in uses and interpretation of these data on inmate accounts.
Much of the information virtually begs additional examination.
Nevertheless, some characteristics of inmate accounts seem to be so common across all sampled
States—such as expenditures at the prison store—as to confidently indicate important general
features of PIE income and disposition patterns. Other information provides clues for
understanding or additional research. And research staff have tried to be careful in divining
conclusions from such limited information.
In an effort to indicate money exchanges within families, including PIE inmates’ contributions to
family support, research staff separately distinguished money flows to and from PIE inmates and
persons with the same last names, a helpful but incomplete measure to an unknown degree of
such family exchanges given that “family” members—such as significant others, relatives, or
children—can have different last names. Masking may be significant for inmates to the extent
spouses have remarried and they and children or the inmates’ parents have other names.

68

Disposition of the PIE Residual

Inmates’ total monthly incomes are the sum of (1) PIE incomes, (2) additional income from
traditional prison industries, and (3) monies from outside sources, including family. On some
occasions, refunds from unknown sources occur, as well as transfers from financial institutions,
the ultimate sources of which are unknown.
Answers to three additional questions were sought, (1) How did incomes change when PIE work
occurred, (2) How did monies from persons change with PIE work, and (3) to what extent are
source “persons” other family members, however defined?

Incomes and Expenditures Before PIE
To broach these and other questions, PIE inmate incomes before PIE were also estimated.
However, pre-PIE month and PIE month records were provided for only 104 of 462 PIE inmates,
not quite 11 percent of the 968 PIE inmates sampled overall. Table 27 summarizes average prePIE monthly earnings and their disposition.

Table 27. Estimated Mean Pre-PIE Monthly Income,
Typical Month
Dollars

Percent

Income
Traditional Industries
From Persons, Same Last Name
From Persons, Different Last Name
Other/Unknown
Total Mean Monthly Income

37.54
6.88
12.29
0.45
57.16

65.7
12.0
21.5
0.8
100.0

Disposition
Mandatory Deductions*
Prison Store
To Persons, Same Last Name
To Persons, Different Last Name
To Financial Institutions
Medical
Legal/Photocopies
Postage
Press
Other/Unknown
Total Accounted Expenditures

1.72
50.21
0.77
1.04
0
0.66
0.13
0.65
0.39
2.83
58.40

3.0
86.0
1.3
1.8
0
1.1
0.2
1.1
0.7
4.8
100.0

Change, Inmate Savings

–1.24

*Mandatory deductions before PIE include fines, court costs, child support, or other
charges imposed by court order or corrections regulations.

69

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Overall, pre-PIE accounts reveal an extraordinarily simple financial cycle of State prison inmates,
one of virtually total exclusion from the U.S. market economy. Before PIE, inmates obtain a few
dollars from prison work plus a bit more from outside, and then spend almost all of it at the
commissary (primarily for food), with almost nothing spent on anything or anyone else and
without notable savings. Pre-PIE inmates do not pay taxes, do not contribute for board and room,
do not compensate crime victims, and do not provide notable family support. They appear to not
invest in capital goods or purchase the goods or services typical of American wage earners. In
fact, the limited data obtained suggests that responsibilities are reversed for inmates, with
prisoners somewhat dependent on family and others for support.

Income
The very limited data available suggest that before working in PIE, most PIE inmates obtained
some money from work in traditional prison industries or institutional maintenance and also
received some money from family and others. Some did not work.
Mean pre-PIE monthly income from all sources for PIE inmates was $57.16 for the 104 inmates
for whom information was provided. On average, about two-thirds of the income, $37.54, was
obtained from in-prison work in traditional prison industries or institutional maintenance.
However, only 83 (80 percent) of the 104 inmates’ accounts evidenced prison industries earnings
at all, and 21 displayed no earnings whatsoever during the pre-PIE month. The month’s income
for those actually reporting prison industry earnings averaged $47, an amount when annualized
somewhat less than the $716 per year earlier cited as an average annual inmate income from
1
prison industries. .
In addition, some pre-PIE inmates—though not many—received money from outside. On
average, the 104 pre-PIE inmate accounts showed on $6.88 received from persons with the same
last name and $12.29 from persons with other last names, and nothing from outside financial
institutions. However, only 17 (16 percent) of the pre-PIE inmates received any money from
persons with the same last name, and their receipts averaged $42 in the observed month; 25 prePIE inmates received money from other persons, obtaining, on average $51 during the observed
month. Overall, then, most pre-PIE inmates appear to receive no money from outside, though
2
some receive small amounts. Moreover, whether the monies received represent transfers from
the family to the inmate, withdrawals from the inmates’ own savings, income or dividends, or
other sources, is not known, although presumably some significant portion represents unearned
transfers from family and others to the inmate.

1

U.S. Department of Justice, Bureau of Justice Statistics, Survey of State Prison Inmates, 1991, Washington, DC: U.S.
Department of Justice, March 1993, NCJ-136949,, 27.
2
The extent to which families and others expend monies to support inmates through gifts, visits, collect telephone calls,
bearing legal costs, and so forth, is unknown.

70

Disposition of the PIE Residual

Expenditures and Dispersals
More than anything else, inmates’ expenditures highlight the absence of prison inmates from the
demand side of the American economy. Inmate-borne expenses appear to be generally bereft of
housing, utility, transportation, health care, education, and other major charges typifying
contemporary U.S. household expenditure patterns (Table 28), though inmates clearly spend
money on personal incidental food and clothing purchases, and some expenditures are observed
for entertainment, educational, medical and other items. The sapping of consumer demand by
inmate unemployment—particularly by groups typically in family and household formation
years—and consequent aggregate loss of jobs might form at least one major component in the
debate on social benefits of inmate jobholding.

Table 28. U.S. Household Consumption Averages, 1998–99
Dollars
Income Before Taxes
Outlays
Taxes
Housing
Shelter
Utilities
Operations & Supplies
Furnishings
Food/Alcohol
Clothing
Transportation
Vehicle Purchases
Gasoline & Motor Oil
Other/Public Transport
Insurance and Retirement
Health Care
Personal Care
Entertainment
Education/Reading
Tobacco Products
Cash Contributions
Miscellaneous

42, 770

6,519
11,854
6,856
2,394
1,097
1,507
5,256
1,704
6,815
3,136
1,036
2,643
3,408
1,921
399
1,849
761
254
1,160
870

Percent
100.0

15.2
27.7
16.0
5.6
2.6
3.5
12.3
4.0
15.9
7.3
2.4
6.2
8.0
4.5
0.9
4.3
1.7
0.6
2.7
2.0

Source: U.S. Department of Labor, Bureau of Labor Statistics, “Consumer Spending
Patterns in Minneapolis-St. Paul, 1998–99,” Document 9278, Table 1, ““Consumer
Unit Characteristics and Percent Distribution of Expenditures, U.S. Average and
Selected Metropolitan Areas, Consumer Expenditure Survey, 1998–99,” Washington,
DC: Bureau of Labor Statistics, retrieved July 26, 2003, from the BLS Web site:
http://www.bls.gov/ro5/cexmpls.pdf.

71

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Second, observation of inmates’ pre-PIE expenditures highlights that inmates not in PIE appear to
contribute little or nothing to other deserving beneficiaries, with almost nothing in repayment to
taxpayers (or assistance to the beneficiaries of taxes), to crime victims, or for child support. In
fact, departments of corrections do impose deductions on inmates’ earnings, and are understood
to deduct on occasion not only from inmates’ incomes but also from monies sent them by family
and others.

Commissary
The single largest inmate expenditure, both before and during PIE, is for purchases at the prison
store, alone accounting for nearly 86 percent of pre-PIE inmate gross incomes. In fact,
observation of inmate accounts suggests a relatively basic income-expenditure cycle for pre-PIE
inmates: get money, spend it at store. Inmate accounts more closely resemble allowances and
allowance expenditures than worker income.
What items are most commonly purchased in inmate stores (commissaries)? An informal survey
of PIE participant States indicates overwhelmingly that the biggest sellers (most dollars spent,
most frequently purchased) items in inmate stores are (1) food and drink, and (2) where
permitted, tobacco products; other much less frequently cited items commonly sold in
commissaries suggest that (3) personal hygiene products may rank third, followed by (4) writing
3
supplies, including stamps.
Telephone charges do not, by and large, appear to be paid by inmates directly via inmate
accounts. Presumably, telephone charges occur primarily via collect calls. The extent to which
outflows from inmate accounts compensate for telephone costs is unknown.
Who profits from—are the beneficiaries of—commissary sales? Obviously, purchasing inmates
are the primary beneficiaries and can be assumed to capture the vast majority of the benefits of
commissary items purchased. And certainly producers of items consumed by inmates gain benefit
from increased inmate incomes, apparently especially producers of snack foods and drinks and
tobacco products; though relatively small, increased commissary sales in prison stores serve as a
reminder that increased incomes translate into increased consumer demand, and is reflected in
increased sales opportunity for private firms. Commissary sales here also include television
rentals or purchases, a common feature among PIE inmate expenditures.
In order to further identify commissary beneficiaries, the informal survey of sampled and other
PIE States asked the disposition of commissary profits. In a small minority of cases responding,
inmate commissaries were reported operated by private firms that retained profits. And in one
instance commissary profits were reported returned to corrections’ general operating fund,
presumably reducing taxpayer costs of corrections. However, in the vast majority of responding
cases, profits from inmate stores contributed to (variously named) inmate recreation funds, and
are reported used for such items as televisions, weight lifting equipment, sports equipment, and
other recreation supplies; infrequently named uses included for books (including law books),
3

72

Fifteen States responded to the informal survey.

Disposition of the PIE Residual

picnics and parties, laundry equipment, and other goods and services supporting the inmate
population.
Directly or indirectly, then, beneficiaries of PIE inmate commissary purchases would appear to be
first the PIE inmates themselves in the purchase of consumables, second the firms selling the
additional items consumed by inmates, and finally some small fraction (profits) extending
benefits to the private firms operating some commissaries and to the general inmate population
via the recreation fund. The extent to which the recreation uses to which commissary profits are
put serves to the benefit of State taxpayers is unknown, but is presumably positive, increasing
prison safety and reducing taxpayer costs.

Mandatory Deductions
Account records show that inmates are occasionally held responsible for costs even when not
working in PIE. However, because so little is earned, these data suggest that very little money is
likely obtained from State prison inmates who are not working in PIE. Pre-PIE mandatory
deductions, on average, appear to total much less than $2 per month (3 percent of expenditures)
for all purposes, including for fines, child support, victim compensation, and other corrections’
required payments. However, more than 80 percent of the 104 inmates for whom pre-PIE
information was obtained appeared to incur no mandatory deductions whatsoever, whereas 20
pre-PIE inmates on average incurred mandatory deductions averaging $9 that month. Of course,
no taxes, Social Security, workers’ compensation or unemployment compensation occurs from
traditional prison work, for which payment is considered a gratuity.

Money to Persons:
On average, pre-PIE inmates transmit almost nothing to outside persons, family or otherwise, less
than $2 a month altogether, about 3 percent of expenditures. However, the observed data indicate
that most inmates transmit nothing at all to family or other persons when not working in PIE; 101
of 104 pre-PIE accounts show nothing whatsoever sent to persons with the same last name and 96
sending nothing to other persons. For those few inmates transmitting money to persons, the
average sent to persons with the same last name was nearly $27 and to other persons, nearly $14.
The observed data, then, suggest that before PIE very few inmates transmit money home or to
others, but a few may contribute a significant portion of their incomes for that purpose.

Money to Financial Institutions/Savings
Interesting but not surprising, no pre-PIE inmate account showed inmates either receiving or
transmitting money to financial institutions. Further, on average, pre-PIE inmates appear to be
saving nothing or virtually nothing in inmate accounts; in fact, for the pre-PIE month observed,
inmates overall appear to have slightly disinvested in savings (–$1.24 on average). The absence
of notable savings by prison inmates not working in PIE, either in prison savings or in separate
accounts, suggests reason for concern for inmate resources at release.

73

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Medical
Inmates may be charged for sick-call visits and for medications, and 29 of the 104 inmates’
accounts show charges averaging $2.38 during the observed month.

Legal/Copies
Inmates may be charged for copies of legal documents and for photocopying, though few pre-PIE
inmates appear to purchase them and apparently do so sparingly; the 6 accounts showing such
expenditures averaged only $2.30 expended for materials in the observed month.

Postage
Only 25 inmate accounts showed postage expenses, which often appeared to occur for mailing
parcels as well as for purchase of stamps; postage expenditures for those incurring such expenses
averaged $2.69 in the observed month’s accounts. Either inmates rarely communicate, they obtain
stamps via the commissary, or they obtain them by other means not observed in this research.

Press
“Press” includes purchases of books (including religious materials) and periodicals. Identifying
such purchases was particularly difficult, since inmate accounts identify to whom payment was
made rather than the produce being purchased. Therefore, it is highly likely that some press
purchases were not detected by research staff. Nevertheless, available evidence suggests that very
few inmates purchase reading materials; among the observed accounts, only 3 inmates’ records
clearly showed reading materials expenditures, averaging $13.63 that month.

Other/Unknown Purchases, Expenditures
The relatively large proportion of pre-PIE inmate expenditures in the “other/unknown” category
represents mix of other identified and unidentified inmate expenditures—such as to religious and
other organizations, mail order, including shoes, athletic clothing, music, educational and
recreational books and periodicals, and gifts, among other approved items. Inmates also pay for
services—such as attorney services—and donate monies to churches, charities, and other
organizations.
mail order —and payments for which purposes could not be discerned. No doubt, some
unexplained expenditures recorded in this category belong in higher identified categories; hence,
when added to any other specific category, the “other/unknown” could be interpreted as helping
establish the maximum that may have been expended for specific purposes. Nevertheless, only 8
inmates of the 104 (8 percent) incurred such expenditures, though they were significant for these
persons, averaging $36.75 each.

Income and Expenditures During PIE
Methodology
Monthly PIE-period incomes and expenditures are a combination of estimates based on different
inmate samples and methods. In order to estimate total income and expenditures for inmates

74

Disposition of the PIE Residual

during their PIE experience, research staff first converted annual PIE payroll income and
deductions information to estimated monthly PIE income and deductions and then combined the
estimated PIE income and deductions with estimated non-PIE income and expenditures obtained
from a month’s inmate accounts. To convert annual PIE income to monthly estimates, staff first
either identified or estimated each PIE inmate’s duration of employment within the PIE year and
then derived the appropriate monthly earnings and PIE deductions. Other income and
expenditures are the summed and averaged values of information obtained from PIE inmates’
PIE-month accounts. PIE-based income and deductions represent unweighted values for all 968
PIE inmates sampled for the research, whereas non-PIE earnings and expenditures reflect
unweighted information for 462 inmate records provided by respondent States.

Income
Inmates’ monthly gross income increases dramatically upon PIE participation (Table 29). PIE
inmates’ estimated monthly gross incomes from all sources increase from an average of about
$57 before PIE to $693 during a PIE month. PIE inmates’ net income after PIE and post-PIE
deductions ($312.65) also increase very significantly, though less dramatically, again highlighting
the shared benefits of PIE participation.
Not surprisingly, income from traditional industries and maintenance declines upon an inmate’s
participation in PIE, with mean prison industries income for all 462 inmates averaging only $2.63
in the observed month; only 46 inmates (10 percent) of PIE inmates were reported also working
in traditional industries, with their incomes averaging $26 during the PIE month rather than the
average of $47 earned in traditional work before PIE.
PIE inmates’ income from outside sources also appears to decline during PIE work, suggesting
some minor possible decline in external dependency. Overall other mean income from persons is
estimated to drop from about $19 before PIE to about $16 during PIE. Oddly, however, the small
number of inmates actually receiving money from outside persons on average appear to receive
significantly more than before joining PIE; average income from persons for those inmates
getting money increases from an average of $40–$50 per month before PIE to $70–$75 during
PIE; moreover, whereas before PIE inmates on average received no money from other sources or
financial institutions, during PIE inmates on average received about $5.60 from such sources;
those few inmates (12 of 462) getting money from financial institutions and other sources
received, on average, $214. Although the reasons for these increases are not known, a plausible
(but unproven) explanation may be that some PIE inmates circulate funds with family and
financial institutions, depositing and withdrawing funds as needs require.

Expenditures Overall
A dramatic illustration of differences between traditional prison industries and PIE work with
respect to beneficiaries is found in the absence of benefits to parties other than the inmate before
PIE compared with the tremendous increase in benefits for taxpayers, crime victims, and others
when inmates are productive in PIE and contribute to meeting their financial responsibilities.
Whereas during their month in traditional industries, inmates contributed almost nothing, on

75

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

average $1.72, to taxes, crime victims, and others, during their PIE month total PIE and post-PIE
mandatory deductions are estimated to have averaged $380.63.

Table 29. Estimated Mean PIE Monthly Incomes and Expenditures, One
Typical Month, Survey Year 1998–2001*
Dollars

Percent

PIE Inmate Gross Income
Traditional Industries
From Persons, Same Last Name
From Persons, Different Last Name
Other/Unknown
Total Mean Monthly Income

669.25
2.63
6.02
9.82
5.56
693.28

96.5
0.4
0.9
1.4
0.8
100.0

Disposition
PIE Estimated Federal Tax Liability
PIE Estimated State Tax Liability
PIE Social Security
PIE Room and Board
PIE Victims Compensation
PIE Family Support
PIE Other Court Ordered
Non-PIE Mandatory Deductions***
Prison Store
To Persons, Same Last Name
To Persons, Different Last Name
To Financial Institutions
Medical
Legal/Photocopies
Postage
Press
Other
Total Mandatory Deductions (PIE, Post-PIE)
Total Inmate Discretionary Expenditures
Change, Inmate Savings $

9.26
1.05
19.94
275.73
56.09
2.28
0.57
15.72
103.57
19.87
40.95
20.52
1.19
1.32
0.74
2.96
24.72
380.63
215.83
+96.82

1.3
0.2
2.9
39.8
8.1
0.3
**
2.3
14.9
2.9
5.9
3.0
0.2
0.2
0.1
0.4
3.6
54.9
31.1
14.0

Income

*”Typical” is usually September; data avoid December and months of inmate release.
** Less than 0.1 percent.
*** Mandatory post PIE deductions include fines, court costs, child support, or other
charges imposed by court order considered PIE by respondent States.

Post-PIE Deductions
Review of inmate accounts for sampled States overwhelmingly indicates that inmates’ PIE
earnings are not greatly reduced by post-PIE additional mandatory deductions. While admittedly
a small sample, the research found no evidence of significant post-PIE takings from PIE inmate

76

Disposition of the PIE Residual

incomes. Of the 462 inmate records, only 80 (17 percent) incurred any post-PIE deductions
whatsoever, and 380 (more than 80 percent) incurred no post-PIE deductions. On average, then,
the PIE inmates studied incurred less than $16 in observed post-PIE deductions during the
observed month. However, for those persons who incurred them, the deductions were significant,
averaging nearly $91 apiece during the examined month.
In many cases accounts did not clearly identify the specific purposes of the deductions, but postPIE deductions appeared to generally occur for court-ordered court costs, fines, crime victim
restitution, child support, reimbursement for charges (such as for security escorts to attend
funerals), and so forth, leaving taxpayers, crime victims, and families somewhat additionally
better off under PIE than before PIE, and slightly better off than indicated by PIE deductions
alone. Thus the net benefit to taxpayers, crime victims, and families from PIE work again appears
to be slightly larger than revealed solely by PIE deductions.
The research also suggests that court-ordered deductions for reasons other than authorized under
PIE are not generally notable claimants for residual earnings of PIE inmates. Research staff found
almost no evidence that individual crime victim restitution orders or child support enforcement
orders are enforced beyond the view of PIE; instead, available evidence suggests that such orders
either do not exist or not enforced.
The research also inquired informally among sampled and other PIE States about PIE policies
furthering voluntary PIE inmate family support. Among States responding, a strong majority
identified court orders as the exclusive or primary trigger of family support payments, and a few
noted either inmates’ voluntary requests or the mere fact of having legal dependents constituting a
prompt for family support payments by PIE inmates. While States clearly note PIE checking for
and then enforcing child support orders, no responding State explicitly cited PIE procedures
encouraging inmates’ voluntary child support in the absence of court orders. Moreover, no State
referred to family support extending to any persons except with respect to child support,
suggesting that support for parents, grandparents, or other family members may not be attended in
PIE procedures for the family support deduction. Nevertheless, no rigorous review was
conducted, and clear descriptions of State family support and victim compensation deduction
policies await more complete research. Information provided here on State PIE policies for these
deductions is tentative at best and should be treated with caution.

Transfers to Financial Institutions
Researchers informally inquired about department of corrections control of—and possible
insights into—the recipients of inmate money transfers to financial institutions. Limited evidence
in reply suggests that most States do not prohibit inmates’ transferring monies to other persons or
financial institutions, although many institute disbursement controls, such as requiring a
minimum level of savings in an inmate account to ensure having money at release or requiring
approval of the person or institution to whom monies are sent.

77

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

PIE inmates’ accounts show few inmates transferring monies to outside financial institutions
(banks, savings and loans), though those few making transfers on average transmit significant
amounts. Only 34 inmates’ accounts posted transfers to financial institutions, but their transfers
averaged nearly $280 in the observed month. Reasons for the transfers, recipients of the transfers,
and applications of the transferred monies are unknown. The extent to which such transfers
constitute savings, indirect or direct support to children, family, or others, or represent payoffs of
debts (loans), telephone charges, or other inmate costs, or other uses, remains to be discerned.

To Persons With the Same Last Name
About 10 percent of observed inmate accounts (47) showed money transfers to persons with the
same last name during the examined PIE month; for such persons the average transfer was almost
$200, a significant potential contribution to family finances to the extent such monies became
available to meet family costs (or reduce family costs for telephone or other inmate-related
expenses). Nevertheless, the uses to which the monies were put remains unknown, and the overall
data suggest either that the vast majority of PIE inmates (90 percent) do not send monies to
family or that “family” recipients receive such monies under other names or via transfers to
financial institutions. Averaged over all 462 inmates, the average transmittal to persons with the
same last name during the PIE month totaled $19.87.

To Persons With Different Last Names
Nearly one quarter (23 percent) of observed inmates transferred monies to persons with different
last names during the PIE month, on average transmitting nearly $180 to them during the month.
However, most inmates (77 percent) did not send monies to other persons, the relationships
between the inmates and the persons are unknown, and the purposes for the expenditures is not
revealed. Again, transfers to other persons could also represent expenses for goods or services,
such as to defense attorneys or medical professionals, to pay debts, or for other impersonal
expenses as well as to other unrelated or related persons. Taken as an average for all 462 inmates
in the PIE month, the mean transmittal to persons with different last names was $40.95.

Medical and Related Transactions
Inmate accounts, taken independently, or comparing expenditures during PIE compared with
beforehand, also suggest other expenditures and benefits of inmates’ PIE income. Upon PIE
participation, inmates appear to spend somewhat more on sick call visits and medications, on
postage, and on photocopying (often, apparently, for photocopies of legal documents), suggesting
that inmates benefit and taxpayers may enjoy some additional relief as well.
During PIE, inmates’ medical expenditures increased, with nearly one fourth (24 percent)
expending on average $5 during the examined PIE month; whether the increased expenditures
represent increased access to prison medical services or merely increased charges for the same
level of care is not known. Some PIE inmates also appear to invest more in photocopying and
purchase of legal documents, with 36 PIE inmates (8 percent) spending an average of $17 during
the reported PIE month.

78

Disposition of the PIE Residual

Some, but relatively few, PIE inmates also incurred expenditures for books, periodicals, and other
publications. However, discerning such purchases among account entries involved considerable
judgment by GWU research staff and staff best guess is that many purchases were left
unidentified among PIE inmate expenditures. Nevertheless, observed data indicate that at least 38
PIE inmates (8 percent) spent on average $36 during the observed month for such items.
Somewhat surprisingly, relatively few PIE inmate transactions showed postage purchases during
the observed month, with an overall mean for all 462 inmates of much less than a dollar spent;
only 18 percent (83 inmates) recorded postage expenditures at all. Whether such purchases are
hidden among commissary sales or are somehow accounted or obtained elsewhere is not
determined.

Other
Inmates purchase goods via mail order, including shoes, athletic clothing, music, and gifts, among
other approved items; inmates also pay for services—such as attorney services—and donate
monies to churches, charities, and other organizations. While inmate accounts explicitly shows
instances of each occurring, so many instances cannot be quantified as to void any overall
estimates of specific categories. Some expenditures may represent inmates meeting obligations,
although most appear voluntary. Nevertheless, evidence from both PIE and pre-PIE inmate
accounts indicates that PIE inmates both dispense monies and increase those expenditures after
becoming PIE participants. Therefore additional beneficiaries of PIE work include organizations
and firms to which PIE inmates dispense monies.

Other Deductions
“Other” deductions garner a sizeable share of PIE inmates’ known expenditures, no doubt in part
reflecting purchases actually in other categories but left undetected and in part other expenditures
for purposes not categorized elsewhere or infrequent. Forty percent the PIE inmates (181 persons)
posted such expenditures during the observed PIE month, averaging $63.09 during the observed
month. Although staff were unable to identify most expenditures, observed purchases regularly
included clothing, shoes (especially athletic clothing and athletic shoes), art and music supplies,
religious purchases, religious and charitable donations, and distance learning materials.

Savings
Methodology plays a part in estimates of inmate savings during PIE participation. Because (1)
estimates of PIE inmates’ total monthly incomes and total expenditures are a composite of
information sets and (2) “savings” are still a calculated residual—estimated total income minus
estimated expenditures—the estimate of total savings presented here reflects any calculation
errors in the differing estimates used to prepare the account summaries. Nevertheless, extensive
review of individual PIE inmate accounts suggests that the overall conclusions with respect to
savings are correct.
Unfortunately this research does not reveal the extent to which inmates’ money transfers to
financial institutions and other persons represent savings rather than purchases; surely some non-

79

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

trivial portion of inmate financial transfers likely becomes savings. Separate from transfers to
external institutions and persons, however, the research nevertheless indicates that PIE inmates
save a substantial portion of their monthly incomes, with “monthly incomes” here including from
all sources, including and beyond PIE work.
Whereas before PIE, the limited data available suggest that inmates before PIE saved essentially
4
nothing, PIE inmates appear to save substantial amounts in their prison accounts. On average
PIE inmates retained after required and voluntary expenditures $97 during the PIE month, 14.0
percent of their gross income from all sources and 31 percent of their retained discretionary
earnings (estimated at $313) after PIE and other post-PIE mandatory deductions.

Further View of PIE Effects on Deductions
Examination of pre-PIE and during-PIE inmate accounts suggests that, by and large, all
transactions of focus—additional deductions post-PIE, money flows from and to family and
others—are relatively small. Few PIE inmates either received or transmitted significant amounts
or incurred significant deductions under either circumstance.
However, examination of pre-PIE and PIE deductions and transactions strongly suggest
additional external financial benefits beyond those immediately described in PIE.
First, PIE inmates occur additional post-PIE deductions in the range of $16 a month, roughly
equaling 2 percent of inmate gross PIE incomes; moreover, these deductions appear to average
about $14 more per month for external beneficiaries than total average deductions before PIE.
The deductions appear to primarily serve taxpayers by contributing to court costs, fines, and other
legal charges, and to secondarily also contribute to victim restitution and family support.
Second, although with the strong proviso that inmate transmittals may to a significant extent also
serve their own consumption or savings needs and not just contribute to family or dependent well
being, nevertheless, the reduction in average amounts received from persons with the same last
names combined with the average increase in amounts sent to persons with the same last name
suggests that “persons with the same last name” enjoy a net increase per month on average of
about $20 (19.96), equal to about 3 percent of inmate gross income; persons with different last
names enjoy a net increase per month on average about $42 ($42.38), roughly equal to 6 percent
of inmate gross income. Combining net changes in money available to all external persons
suggests an average increase of about $62 money from PIE inmates to persons in the examined
month, roughly equaling 9 percent of inmate gross PIE income.
To gain additional information on the net effects of PIE participation on (a) mandatory post PIE
deductions and (b) inmates dependency on external funds or contribution to family support,
research staff isolated 104 inmate records for which both pre-PIE month and PIE month account
4
Corrections departments may require inmates to save minimal amounts (like $50) to ensure having some money at
release.

80

Disposition of the PIE Residual

information was provided and then compared post-PIE deductions, income sources, and
expenditures for these inmates.
The 104 records form an extraordinarily small sample from 968 total inmate records studied,
themselves a small sample from the thousands of inmates working in PIE during the survey year.
Nevertheless, the examined records may yield some insights into general beneficiary
consequences.

Table 30. Pre-PIE and PIE Month Income, Deductions, and Expenditures, Selected
Transactions—Inmates Incurring Transactions Only, 104 PIE Inmate Records

Money From, Same Last Name
Money From, Different Last Name
Mandatory Post-PIE Deductions
Money to, Same Last Name
Money to, Different Last Name

Pre-PIE
Mean $

Number
Inmates

PIE
Mean $

Number
Inmates

42.12
51.12

17
25

31.50
49.44

8
9

8.97

20

63.27

34

26.67
13.58

3
8

94.83
125.37

12
21

Overall, as illustrated in Table 30, relatively few of the 104 inmates were involved in any of the
five categories of transaction described, with as few as 3 affected (Money to persons with the
same last name) and 34 at the maximum (Mandatory post-PIE deductions). Therefore, in general,
the 104 records indicate that neither mandatory deductions nor external persons are noticeably
affected by inmates becoming participants in PIE.
On the other, the evidence provided by comparing pre-PIE and PIE months for the 104 inmates
indicates that some external beneficiaries may be affected, and may enjoy either additional money
or reduced expenses as a result of the offenders’ participation in PIE, benefits not accounted in
national PIE statistics or apparent solely from PIE data alone.

Mandatory Post-PIE Deductions
Mandatory post-PIE deductions include the full range of court-ordered and sometimes
corrections’ ordered takings from inmate earnings for fines, court costs, victim restitution, child
support, and compensation for specific costs (such as for security reimbursement for inmates
attending funerals) not deducted as approved deductions in PIE. In the aggregate the beneficiaries
of these deductions can be considered primarily taxpayers, but payments sometimes also benefit
crime victims and inmates’ children, among others.

Review of the 104 records show both the number of inmates contributing to post-PIE deductions
and the average size of such deductions increasing significantly with the introduction of PIE

81

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

participation; the mean deduction PIE-period deduction leapt from about $9 prior to PIE to more
than $63 during the measured month of PIE participation, a net $54 increase for the beneficiaries
of inmates incurring the charges. For all 104 inmates, the mean deduction increased from just
$1.72 to $20.69 as a consequence of PIE, suggesting a very small but positive increase (here
$18.97) in dollar benefits from PIE to beneficiaries of post-PIE deductions, equal to 3 percent of
inmates’ gross PIE incomes. The estimate is similar to that gotten when comparing all available
records, $14.

Money From Persons/To Persons:
Observation of money flows to inmates from persons and from inmates to persons also suggests
that most potential external beneficiaries are left unaffected by inmates participation in PIE, but
that some outside beneficiaries enjoy modest monetary benefits from inmates’ PIE work.
Money from outside persons is relatively modest even for the 25 or fewer inmates obtaining
money from outside before PIE; introduction of PIE appears to permit about half to two-thirds of
those outside sources to cease support altogether during the inmate’s PIE participation, while the
remainder continue support at a somewhat reduced level. Applied to all 104 inmates, external
support from persons with the same last name drops about $4.40 a month, and from persons with
different last names drops about $8.00 a month.
At the same time, for some external beneficiaries, inmates’ participation in PIE appears to spark
large increases in money transfers, some of which probably provides external support, and some
unknown proportion of which may effectively occur for the benefit of the inmate and not the
external correspondent. Nevertheless, whereas no more than 10 percent of the 104 inmates sent
any money during their pre-PIE experience, and in amounts averaging $14 to $27 in the examined
month, both many more inmates transmitted money while in PIE and the amounts transmitted
increased tremendously, including contributions which, while not large in meeting overall
household costs, could contribute meaningfully to overcoming household poverty—particularly if
such sizeable provisions were commonplace in inmate work. During PIE, as many as 21 inmates
of the 104 transmitted money to persons with different last names (12 made payments to persons
with the same last names), and forwarded as much as $125 in the observed PIE month. Overall,
for all 104 observed inmates, mean transmittals to persons with the same name increased from
just 77 cents before PIE to an average of $10.94 during PIE, and to persons with different last
names from a mean of just $1.04 before PIE to $25.32 during the observed PIE month. An
unknown part of this overall average increase in out payments to persons of $34.45 can be
expected to serve family, though some undoubtedly goes to serve the inmate.
Combining reduced inflows from persons (–$12.40) and increased outflows (+$34.45) suggests a
net turnaround possibly as great as $46.85 a month if all moneys affected supporting family, not a
large amount of money by most measures, but significant and helpful in the context of inmate
earnings and possible assistance to fragile families. The estimate, though different, is in the same
order of magnitude as the $62 obtained when examining all pre-PIE and PIE records. At the same

82

Disposition of the PIE Residual

time, even this amount is likely significantly reduced by inmate transactions aimed at their own
needs rather than to meeting external interests’ interests.

Tentative Conclusions from Inmate Account Data:
Despite the paucity of data and cautions about the data’s representation of inmate incomes and
expenditures separate from PIE, some tentative overall conclusions are suggested:
1. Post-PIE mandatory deductions from inmate incomes appear to be relatively small, and
observed data provide no indications that significant takings from inmates’ PIE earnings
occur post-PIE deductions. Taken as a share of either total incomes or of PIE incomes alone,
mandatory post-PIE takings appear to be not much more than 3 percent of total PIE incomes,
on average about $14–$20 a month. To the extent other charges, such as for sick call, legal
copies, or other transfers disguise other required expenditures, mandatory post-PIE
deductions could be slightly higher, but not significantly more. The limited information
provided by the sampled States, therefore, appears to strongly contradict any concerns that
PIE inmates’ incomes are significantly reduced by post-PIE deductions. At the same time,
however, the post-PIE deduction information suggests that parties—like crime victims and
inmates’ children and families—who did not obtain significant support via PIE deductions
find no particularly greater support post-PIE, and appear to be generally overlooked at both
stages.
2. Although indicating that PIE participants on average slightly increase voluntary payments to
family members upon PIE participation, nevertheless the PIE and post-PIE deduction data
overall suggest that most PIE inmates, even upon receipt of significant incomes, do not
shoulder significant child or family care financial burdens, either via mandatory deductions or
with voluntary payments. Roughly 90 percent of the 462 PIE inmates for whom PIE-month
data was available expended nothing during the examined month to persons with the same
last name; 77 percent sent nothing to persons with different last names. However, some PIE
inmates appear to expend significant amounts. The average transmittal to persons with the
same last name for the 10 percent of PIE-month inmates making such transmittals averaged
$195, and for the 23 percent sending money to persons with different last names, the average
transmittal was $178. Overall, however, net transmittals to other persons indicate additional
benefits on the order of 9 percent of inmate gross incomes, 3 percent to persons with the same
last names (the narrowest possibility for family support) and 6 percent to persons with
different last names).
3. Transmittals to financial institutions, which may or may not have been accessible by other
family members, increased from nothing before PIE to an average of $41 during PIE, and for
the mere 7 percent of inmates making such transmittals, the average was $279.
4. Somewhat surprisingly, the available data suggest that PIE inmates continue receiving
significant incomes from outside persons during PIE participation.

83

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

5. PIE inmates likely accumulate significant savings during PIE participation, certainly in their
inmate accounts and possibly in external bank and other financial institution accounts. On
average PIE inmates retained after required and voluntary expenditures $97 during the PIE
month, 14.0 percent of their gross income from all sources and 31 percent of their retained
discretionary earnings (estimated at $313) after PIE and other post-PIE mandatory
deductions.
6. By and large, PIE inmate accounts do not directly indicate any wholesale movement into
normal financial responsibilities, such as to mortgages, transportation, utilities, insurances,
health care, or other expenses characteristic of contemporary American households. On the
one hand, recognition of nonparticipation is not surprising, given that it would be much more
likely that such expenses would be handled by the civilian householder even if the inmate
were aware or transmitted money for payment of such expenses; and it is altogether possible
that research staff could not detect such payments among the minority of inmate payments for
“other” and financial transmittals. Nevertheless, the continued absence of PIE inmates from
normal economic behaviors even in the presence of growing incomes may suggest a need for
additional education and guidance for PIE inmates in the future.
7. Finally, the overall difference between inmate incomes and consequences for important
beneficiaries before and during PIE is little short of stunning. First, the almost complete
absence of inmates from economic participation or commerce before PIE should give pause
to anyone whose emphasis is U.S. economic growth—including the rate of growth of
aggregate consumer demand—or reduction in poverty and inequality, given that prime
working age adults not in PIE get and spend little more than snack and personal hygiene
money, and contribute virtually nothing to the economy, compensate almost none of their
victims’ costs, and not only do not support their families but instead receive support from
them reducing net household resources. PIE-participating inmate, in stark contrast, earn
meaningful incomes, contribute substantially to the tax base and reducing taxpayer costs,
contribute to critical social insurance programs, compensate their victims, build savings, and
increase consumer demand and consumption. While improvements in delivery of benefits to
crime victims and families may appear warranted, and increasing incomes may suggest needs
for additional guidance in responsibility and use of increased earnings, there can be little
doubt that PIE employment yields significantly larger financial benefits to important National
constituencies than idleness or traditional inmate work.

84

VIII. Recommendations for Further Research
This research indicates substantial national benefits from prison inmate work in PIE.
Nevertheless, this analysis is also an early effort, and additional beneficiaries research is needed
in virtually every component, including topics, facts and data, assumptions and relationships, and
the structure and execution of the analyses. It is highly likely that additional beneficiaries
research will enlarge, revise, and sometimes refute conclusions of this work, and that the
additional research will ultimately provide superior guidance to citizens, policy makers, inmates,
other stakeholders, and practitioners in inmate employment. For the sake of the Nation’s
economic and social health, fundamental broader areas in PIE and in inmate work in general also
merit additional attention, beginning with further examination of the specific topics in this
benefits research and broadening to more general PIE issues and evaluations.
First, key topics broached in this research should be reexamined and their representation tested,
corroborated, or improved, not only with respect to their conclusions for benefits and beneficiary
groups, but also for assumptions and conclusions with respect to other PIE inmate demographic
and criminal justice characteristics. Any number of important areas merit attention, including –
1. All demographic and criminal justice descriptors of PIE inmate workers, including sex, age,
race and ethnic origin, education, and if possible, work or skill history and parenthood.
Important fundamental descriptors of PIE work are introduced in this research, including an
hourly wage rate near the Federal minimum, large proportions of part time work, and average
weeks worked much less than a full calendar year. Not only is confirmation of these basic
conclusions critical in assessing PIE, but also explanations and possible recommendations for
changed policies or practices could be crucial to PIE’s future fortunes. Knowledge about
racial, educational, and any prior work experience characteristics of PIE inmate workers
differing from the normal inmate population could be significant in evaluations of PIE or in
comparative evaluations of other correctional programs. Is job turnover as frequent as
suggested by the research data, and what effects does higher than average turnover have on
industries appropriate for State prisons or on PIE firm competitiveness? Each characteristic,
especially in comparison with populations of other correctional programs, each State or the
Nation’s general prison populations, or with relevant U.S. subpopulations, could reveal
important explanatory information about prisons, prison work programs, and PIE in
particular. Furthermore, review of characteristics for this research suggests that specific
States’ PIE inmate subpopulations may differ significantly with respect to both demographic
and criminal justice characteristics, making National conclusions from small samples much
less certain and subject to statistical variation. Without additional testing and corroboration,
conclusions drawn from characteristics generalized from this specific sample may, in fact, be

85

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

unwarranted. Moreover, where relevant, additional work needs to be done understanding not
only what characteristics typify PIE workers, but why they exist and what differences or
effects result.
2. Critical assumptions need to be empirically established, both under current PIE
circumstances—such as for inmates’ tax filing and refunds claiming practices—and
especially where potential outcomes in alternative PIE scenarios are examined—such as for
proportions or distributions of inmate populations earning average civilian incomes, and with
respect to “optimal” or other deduction policies. In addition to assumptions yielding welcome
outcomes, challenging assumptions of plausible alternatives also deserve exploration,
including for limited skills, limited public acceptability, and other alternatives.
3. PIE deduction rules and policies may deserve review with respect to requirements and their
implementation across all participating States. Evidence in this research suggests that the
family support deduction may be misunderstood or applied differently than suggested in
current guidelines; similarly, this research suggests that PIE employers’ practices for workers
compensation may be effected differently than would be expected from review of PIE
guidelines.
It would be helpful if State PIE programs in general gathered, organized, and then shared
additional descriptive statistics on PIE programs, program participants, PIE deduction policies,
program purposes and results, and program participants, especially including basic demographic
descriptions of all PIE applicants and inmate workers.
Useful individual analyses and data gathering could occur at the individual State level as well as
for all States under nationwide programs. Nevertheless, some expansion in consistently defined
nationwide reporting merits consideration.
Furthermore, additional research is likely warranted on broader PIE beneficiary issues untouched
or only lightly touched by this work, including on policies, procedures, and objectives of PIE
deductions. This research naturally calls for examining broader economic and social benefits in
the context of recipient groups and issues, both for the motives, goals, principles, rules,
procedures, and outcomes effected by PIE, as well as to outcomes in the context of the recipient
groups.
1. Both crime victim compensation (to victims of other offenders) and victim restitution (to the
PIE inmate’s own victims) deserve additional attention, including the role and scale of PIE
inmate contributions to State victim compensation funds, interactions with State victim
compensation programs, and public recognition of PIE inmate victim compensation
contributions and potential. Reexamining PIE’s relationship to crime victim restitution may
be in order, including legal bases for such deductions, principles for establishing victim
restitution, setting restitution with and without court restitution orders, and flexible restitution
procedures (reducing requirements if PIE work is lost).

86

Recommendations for Further Research

2. Employer payroll contributions for Social Security, unemployment compensation, and
workers compensation may deserve additional consideration for customer (prison owned) PIE
firms, given that such employer-paid deductions might more closely match requirements for
civilian firms, would buttress the finances of the programs, and could build public support.
3. It is difficult to imagine that the PIE family support deduction is meeting envisioned purposes
when 50 percent of inmates have children, yet few inmates pay it, and those who incur it
contribute so little. The family support deduction almost certainly deserves reexamination,
including in concept, in instructions to States, and in extent and scale of use, both for the sake
of the client group and in light of potential public support. Whom should the family support
deduction assist—only natural offspring who lived with the incarcerated parent (What about
offspring now in other households, or mothers remarried?)—or should “family” include
parents, grandparents, other relatives, or unrelated but interdependent others, and under what
circumstances? Of course, the extent to which inmates have anyone reasonably to be
considered “family” deserves review, and realistic review may reveal either the absence of
family or effective estrangement of family in some instances, raising additional issues to be
accommodated in assessing deductions for family support. What is the appropriate role of
PIE or of corrections, in identifying and encouraging or enforcing family support deductions?
Should there be, and what would be, appropriate conditions on recipients and uses of the
deduction, and is any oversight needed? To what extent does enhanced financial support
necessitate other family reconciliation, cohesion, or parenting services? How should or does
the PIE family support deduction relate to child support orders and welfare (TANF) support?
The catch-22 in PIE child support could be reexamined, by which PIE support is deducted
only with a child support order but child support orders are avoided because, once set, impose
impossible financial burdens on inmates left unpaid before and after PIE.
4. Post-PIE deductions deserve additional attention, including the extent to which PIE inmates
have unmet legal obligations for fines, court costs, and repayments for special expenditures
(such as repayment for guard services when attending family funerals).
5. Because PIE inmates can acquire significant savings, efficient and productive use of the PIE
inmate’s residual income deserves additional examination, both to ensure appropriate
accounting and oversight (and to deter misuse), and also to help meet legitimate
accountholder (inmate) objectives, such as for voluntary family support, accumulating
savings, insurances, mortgage payments, or other productive long-term purposes. The
potential role of investment and money planning should be considered for PIE support to
inmate workers.
6. If it is to expand, particularly without taxpayer subsidy, PIE simply must become profitable
for private firms. Therefore, it is likely crucial for future PIE success and the expansion of
stakeholder benefits that the actual and potential benefits of PIE business locations be clearly
understood, publicized, and aggressively exploited. Research including corrections

87

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

departments, PIE programs, and both successful and unsuccessful firms (including interested
firms abandoning efforts) seems particularly apropos at this time. Consequences for host
communities could be examined, including on the local tax base, on civilian employment
both in the PIE firm, for multiplier (added employment) or displacement effects in the
community, if any, and for any ancillary benefits for the community. Research should
consider assessments of the kinds of industries, firms, and jobs most amenable to efficiency
and PIE success, particularly in an unsubsidized competitive marketplace, and possibly with
respect to import substitution (retrieving jobs from overseas or preventing job export).
7. Examination of corrections policies and practices affecting PIE firm success, including (a)
security-efficiency requirements and tradeoffs, (b) contractual terms and procedures affecting
prison-firm relationships, (c) correctional practices facilitating or deterring firms’
participation, (d) public perceptions affecting participating firms and support for PIE, and (e)
explicit and implicit taxpayer support for PIE firms for land, buildings, equipment, utilities,
support services, staffing, and security, and implications for taxpayers and business
competition.
8. Broader labor issues could be examined, including civilian and PIE inmate hiring practices,
wage rates, prevailing wage comparability, pay and benefits (including health and
retirement), education, training, promotion, displacement, civilian and inmate security, health
and safety, labor-management relations, and inmate and civilian organization and
participation rights and issues.
9. PIE should be much more extensively examined with respect to correctional and justice
effects, such as on prison order, prison security, inmate rehabilitation, restorative justice
(including healing for crime victims), and family cohesion. Does the presence of PIE
opportunities increase prison security and reduce danger for inmates and staff, and if so, how,
and how are they evidenced?
10. Support for additional objective research on recidivism effects of PIE participation is clearly
warranted, both for base effects (Does PIE reduce recidivism?) and for distinguishing key
correlates (What characteristics of PIE correlate most highly?). Given that the greatest social
benefits of PIE could be reduced crime, and that earlier research and PIE analysis suggests
reduced recidivism from PIE employment, more extensive and definitive work is clearly
warranted.
11. Either as part of recidivism research or separately, it would be useful to know the extent to
which, if any and why, PIE inmates obtain superior employment after release than do other
inmates.
Finally, PIE deserves additional examination for its implications in the broadest context of
American society and core social values. Should working inmates be considered separately from
others with respect to employment law, including the Fair Labor Standards Act (wages and hours)

88

Recommendations for Further Research

National Labor Relations Act (rights to union participation), and what national benefits or losses
occur given alternative answers? What answers maximize economic growth, employment, output,
and wages, and what answers maximize social cohesion? What subgroups gain and who loses
under alternative answers? Which choices are superior on equity (fairness) grounds? And what
are the implications for offender employment beyond PIE, beyond prisons and jails, for offender
employment policies for pretrial, for probation and parole, and for post incarceration employment
discrimination? Does the PIE experience suggest anything for national and subgroup gains or
losses for employment policies for these offender and ex-offender groups.
At its heart, this research barely begins the work seeking rational, defensible, answers to the
question, “What policies with respect to offender employment participation yield the best results
for America overall and for whom within the society, and how do we know it?” Any research
likely to contribute useful parts to the answer to that question deserves further consideration.

89

Appendix A. Demographics and
Criminal Justice Characteristics
Neither demographic nor criminal justice (offense, sentence) characteristics of the State PIE
inmate-worker population were the focus of this research, and the percent distributions of the
demographic and criminal justice characteristics of the sampled State PIE inmates are not known
to be statistically representative of all U.S. PIE inmates.
Nevertheless, basic demographic and criminal justice information—sex, date of birth, race,
education, principal offense, and maximum sentence are provided.
By and large, however, the sample data suggest a State-level PIE inmate worker population that is
older than the overall U.S. inmate population. The data also suggest that State-level PIE inmate
workers are somewhat more likely than the normal State prison inmate population to have been
sentenced for a longer time and to have been sentenced for a more serious—violent—offense.
Whether these sample results generally reflect PIE participants and, where they are found to
occur, why PIE distributions differ, may be appropriate for further research.

Sex
The sample data corroborate participation of both women as well as men in PIE. However,
because female PIE participants are likely concentrated in a few sites, and because sampling for
this research did not explicitly account for gender, the proportion of women in the sample should
not be interpreted as statistically representative of their overall share of PIE employment, which
may be significantly less or more than shown here. It is safe to conclude, however, that women
are very likely a fairly small share of the overall PIE inmate labor force just as in the overall State
prison inmate population.

91

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table A1. Comparison, Distribution by Sex (Percent)

Male
Female

PIE Inmates

State Inmate
Population 1997

U.S. Population
18+, 2000

91.9
8.1

93.7
6.3

48.3
51.7

Sources: Bureau of Justice Statistics, Survey of Inmates in State and Federal Correctional Facilities
1997, retrieved July 22, 2003 from the National Archive of Criminal Justice Data (NACJD) Web site:
http://www.icpsr.umich.edu/NACJD/SISFCFL; U.S. Bureau of the Census, Census 2000, Summary
File 1, Table DP-1 Profile of General Demographic Characteristics: 2000, retrieved July 22, 2003 from
the U.S. Bureau of the Census Web site: http://factfinder.census.gov.

Age
The weighted sample data suggest that State-level PIE inmate workers are, on average,
significantly older, on average 5 years older, than the general State inmate population.

Table A2. Comparison, Distribution by Age (Percent)
Age
15–24
25–34
35–44
45–54
55+
Total
Median
Mean
Minimum
Maximum

PIE Inmates

State Inmate
Population 1997

U.S. Population
2000

9.2
28.9
39.0
18.5
4.5
100.0

19.8
38.1
29.4
9.8
2.9
100.0

17.7
18.0
20.4
17.0
26.8
100.0

38
38
18
77

32
33
15
89

Note: Sums may not equal totals because of independent rounding.
Sources: Bureau of Justice Statistics, Survey of Inmates in State and Federal Correctional
Facilities 1997; U.S. Bureau of the Census, Census 2000, Summary File 1, Table DP-1
Profile of General Demographic Characteristics: 2000.

Inmate Race
Although all sampled States identified at least two categories, “black” and “white,” sample data
did not generally or consistently identify other contemporary racial classifications, “Asian or
Pacific Islander,” or “Native American/Alaska Native.” Therefore, PIE inmates in “other” race
categories are almost certainly significantly undercounted, and data shown here should be
considered minimum proportions for these groups. Some members of “other” racial groups are

92

Appendix A

likely reported among “white,” with the consequence that the number and share shown here as
“white” is highly likely a slight overrepresentation of that group. Where able to be distinguished
in reported data, “Mexican” and its variations are included here among “White,” following U.S.
Bureau of the Census practice. “Other/Unknown” includes PIE inmate participants identified as
Asian or any other category not able to be categorized as black or white, plus those few for whom
race is not specified.
Information for ethnic (Hispanic) origin was inconsistently reported and cannot be reliably
presented.

Table A3. Comparison, Race Distribution (Percent)
Race
White
Black
Native American
Other/Unknown
Total

PIE Inmates
75.9
17.0
6.0
1.2
100.0

State Inmate
Population 1997
46.6
47.9
3.7
1.7
100.0

U.S. Population
2000*
77.0
12.6
0.9
9.5
100.0

*Based on 97.6 percent of the U.S. population reporting one race only.
Notes: Sums may not equal totals because of independent rounding. “Other/Unknown” includes
Asian and unknown.
Sources: Bureau of Justice Statistics, Survey of Inmates in State and Federal Correctional
Facilities 1997; U.S. Bureau of the Census, Census 2000, Summary File 1, Table DP-1 Profile of
General Demographic Characteristics: 2000.

The weighted sample data in table A3 suggest a racial distribution of PIE inmate workers
markedly different from the overall U.S. State prison population, particularly a much higher
proportion of sampled PIE inmates categorized “white” than found in the U.S. prison population
as a whole, and tempting a conclusion that black inmates—for whatever reason—participate in
PIE much less frequently than their incarceration proportions suggest. However, preliminary
analysis of all sampled States’ race data does not indicate, and certainly does not prove,
overrepresentation of white inmates and under representation of black inmates in PIE. States’
tendency to include some persons of “other” races among whites no doubt contributes to some
small degree to an overestimate of whites in this research. But much more important, comparing
the 1998–2001 reported race distribution of PIE inmates in each individual sampled State with the
current (2003) reported race distribution of all inmates in each State indicates that, for all but one
State, black inmates are an insignificantly higher proportion of PIE participants than black
inmates in each State’s overall inmate population; in just one sampled State, black PIE inmates
appear to be a significantly lower share of PIE participants than their overall State proportion. If
anything, then, the sample data may suggest a slight, probably insignificant, overrepresentation
rather than under representation of black inmates in PIE, and the disproportionately lower
participation rate of black inmates in PIE shown in this research compared with all U.S. inmates
(Table A3) is almost certainly a function of a nonrepresentative sample with respect to race.
Demonstration of the overall racial proportions of PIE inmates, and conclusions about those

93

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

proportions, cannot be reliably drawn from these sample data and await further research
appropriate to the issue.

Education
The beneficiaries research was not designed to be representative with respect to PIE workers’
education, and Table A4 should not be interpreted as representatively portraying PIE inmate
education. The weighted sample data suggest that PIE inmate workers may be somewhat more
educated than the general State prison population, and, not surprisingly, less educated than the
overall U.S. population. Absent information on pre-prison employment and income, education
may also be viewed as a rough but direct indicator of legal earnings power. As usual, both the PIE
and the overall inmate data suggest that (a) many inmates are as well educated as the normal
population, and (b) nevertheless, State prison inmates and the PIE sample are disproportionately
less educated than the normal working-age U.S. population.
Although not entirely certain, and likely in some instances to be inconsistent among reporting
States, education reported by the sampled States is interpreted as highest education completed;
specifically, “12” or any of its variations is interpreted as having graduated from high school.
PIE-reported data do not specify whether “education” identifies year attended or year completed;
the 1997 inmate data and 2000 census data identify year completed.

Table A4. Comparison, Education (Percent)
Education
8th Grade or Less
9th—12th, No Diploma
High School Diploma/GED
Some College, No Bachelors
College Graduate, Post
Total
Median
Mean

PIE Inmates
12.0
41.8
35.4
9.2
1.7
100.0
11
11

State Inmate
Population 1997
13.0
48.8
24.9
11.0
2.4
100.0
10
10

U.S. Population
25+, 2000*
7.5
12.1
28.6
27.4
24.4
100.0
Some College

Notes: Sums may not equal totals because of independent rounding.
Sources: Bureau of Justice Statistics, Survey of Inmates in State and Federal Correctional Facilities 1997; U.S. Bureau of
the Census, Census 2000, Summary File 3, Matrices P37 and PCT25, retrieved July 22, 2003 from the U.S. Bureau of the
Census Web site: http://factfinder.census.gov.

94

Appendix A

Analysis
Marital Status/Parenthood
The research is unable to report reliably on the marital status or parenthood of the PIE inmate
population. No sampled State was able to report parenthood. Some States were unable to provide
marital status; where marital status was reported, however, the meaning and even the time period
of measurement (at arrest, at time of interview?) of reported status categories—such as “single”
and “separated” was sufficiently uncertain and ambiguous as to undermine any practical value of
the data obtained.

Offense
Because the sample is not known to be representative with respect to crime and sentence length,
differences in the tabulated crime and sentence length characteristics for the weighted sample of
PIE inmates from the overall population (Table A5) may reflect real differences from the normal
population of State prison inmates or may merely reflect characteristics of an atypical sampled
group in one or more States.
Respondent States were asked to provide the one most serious offense for which the PIE inmate
was serving during the survey period; States reported the most serious offense shown in the
inmate’s overall record, however determined.1
Weighted sample data show a PIE worker offense distribution disproportionately sentenced for
violent (particularly sexual assault) crimes and much less than proportionally sentenced for drug
offenses. Variations among individual sampled States in offense characteristics are occasionally
striking, suggesting that the overall weighted distribution of the sample may be significantly
affected by individual States. However, some sampled States exhibit much higher than U.S.
average—or their own State’s overall inmate population’s—participation by persons sentenced
for rape/sexual assaults. All sampled States exhibit relatively few PIE participants serving for
drug offenses, compared with the overall U.S. State inmate population and their own State’s
inmate populations.

1

It is conceivable that in some instances States reported an inmate’s current (2002) offense and sentence. However, it
is highly unlikely that either the number of such cases is significant or that either the offense or the sentence in those
instances differs notably from the offenses and sentences being served in the 1998-2001 survey year.

95

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

Table A5. Most Serious Offense, PIE Inmate Sample and
U.S. State Prison Population (Percent)
PIE Inmates

State Inmate
Population 1997

Violent
Murder
Manslaughter
Kidnap
Rape
Other Sexual Assault
Robbery
Assault
Other Violent

70.6
14.5
2.8
2.7
14.5
22.8
7.3
6.1
0

47.2
11.7
1.6
1.2
2.6
6.0
14.1
9.4
0.8

Property
Burglary
Larceny/Vehicle Theft
Arson/Fraud
Stolen Property/Other

19.8
8.7
5.5
4.3
1.3

22.0
10.7
6.0
3.2
2.1

Drug
Drug Possession
Drug Trafficking/Other

7.3
3.2
4.0

20.7
8.8
11.8

Public Order/Other

2.4

10.1

Offense

Note Where identifying a specific offense risked disclosing an individual PIE
inmate, offenses were grouped to prevent disclosure.
Source: Bureau of Justice Statistics, Survey of Inmates in State and Federal
Correctional Facilities 1997.

Sentence Length
Compared with the overall State prison inmate population, the weighted sample of PIE inmate
workers is serving longer sentences than the overall State prison population (Table A6). No
sampled PIE inmates were under sentence of death. Longer sentences undoubtedly to some extent
reflect the disproportionately more serious offenses for which PIE inmates are incarcerated.
Verification of the longer sentences and, if corroborated, other explanations, await further
research.

96

Appendix A

Table A6. Maximum Sentence, PIE Inmate Sample
and U.S. State Prison Population (Percent)
PIE Inmates

State Inmate
Population 1997

0–5 Years
6–10 Years
11–15 Years
16–25 Years
26–50 Years
51 + Years/Life

20.1
18.4
27.9
10.4
3.5
19.8

34.6
22.2
11.3
13.3
7.7
10.9

Median Sentence:

15

10.0

Maximum Sentence

Note: Sums may not equal totals because of independent rounding.
Source: Bureau of Justice Statistics, Survey of Inmates in State and
Federal Correctional Facilities 1997.

97

Appendix B. Introductory Materials
The following documents were provided (usually e-mailed) to sampled States and used to inform,
coordinate, and gain permission of participating jurisdictions.
Every invited State agreed to participate and provided signed concurrence to participate.
Protocols were prepared for submission to participating firms. However, with one exception, all
employer-model firms’ information was already in possession of the host States; the lone firm
also agreed to participate and provided signed agreement.

Introductory Protocol to Sampled States
The introductory protocol on the following was e-mailed to each State sampled for the research.
Its purpose is to introduce the research and key characteristics, request assistance, and provide
appropriate contacts.

.............
Introduction, Request for Help, and Research Outline
Identifying Beneficiaries of PIE Inmate Incomes
The George Washington University Center For Economic Research
Under Contract With the National Correctional Industries Association (NCIA)
IRB Number U080222ER
We need your help!
As part of the Prison Industry Enhancement Certification (PIE) program, the National
Correctional Industries Association (NCIA) has asked the Center for Economic Research
of the George Washington University (GWU) to identify in more detail the benefits of
PIE inmate incomes. The research is expected to help demonstrate the benefits of PIE.
Your Department and PIE firms have been statistically selected. Your participation is
voluntary. It is not for audit and is not part of ongoing PIE assessments. Your responses
are not being revealed to the NCIA. We think we can finish in a few days.

99

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

To help us maintain confidentiality, Please do not reveal to NCIA, to PIE assessment
teams, or to anyone that you are participating. Please direct comments or questions to
either –
Thomas Petersik, Ph.D. (703) 323-5272 (tkpetersik@yahoo.com), or
Anthony M. Yezer, Ph.D. (202) 994-6755 (yezer@gwu.edu)
The records for every PIE-participating inmate during a year to be specified are –
•
•
•
•
•

Basic demographic and criminal justice—e.g. age, sex, race, offense, sentence;
Work hours, gross wages, and Federal tax filing status for the year;
Firm costs and participant benefits not shown in PIE participants’ gross wages;
All payroll, PIE, mandatory and other deductions; and
Inmate accounts information for a month before and during PIE.

The research will likely involve your (1) Industries Director, (2) PIE Director, (3)
Records Officer, (4) Legal Counsel, and (5) Research Office, plus (6) representatives of
each firm. Please let us know if other staff should also participate.
GWU offers complete confidentiality to inmates; the University will not reveal inmate,
business, or State identifiers, or the year being examined. We have also instituted careful
confidentiality procedures for staff and participants during the research.
We should be able to do the work by phone, e-mail, facsimile, and mail, though visits
may help as well. We will identify to you by telephone the year to be studied and the PIE
firms that have been selected.
You should also be receiving now –
(1)
(2)
(3)
(4)
(5)

A complete list of the data elements to be sought via records;
A voluntary participation agreement form for return before we get data;
A proposed timetable for completing the research;
A State confidentiality outline; and
Instructions for coordinating with PIE businesses in your State.

Tom Petersik will call you again in a few hours to follow-up on the contact.
Thank you for your help.

.............

100

Appendix B

Voluntary Participation Agreement
Participation agreements were mailed or facsimile provided to GWU research staff; GWU
research staff, in turn, mailed or facsimile provided a signed response copy.

.............
Department of Corrections, Informed Consent
Identifying Beneficiaries of PIE Inmate Incomes
The George Washington University Center For Economic Research
Under Contract With the National Correctional Industries Association (NCIA)
IRB Number U080222ER
Principal Investigators:
Thomas Petersik, Ph.D. (703) 323-5272; tkpetersik@yahoo.com
Anthony M. Yezer, Ph.D. (202) 994-6755; yezer@gwu.edu
(To protect confidentiality, please do not contact NCIA. See “PIE Certification” on the
NCIA web page, http://www.nationalcia.org, for an NCIA Statement acknowledging and
encouraging participation in this research.
Your Prison Industry Enhancement (PIE) program has been selected for research on the
disposition of PIE inmate incomes. This consent form provides information you may
need in deciding whether your State will participate (We will invite firms separately).
You may contact Thomas Petersik or anyone shown above to answer your questions and
give further details. If you agree to participate, please sign the consent form. Next, please
mail the form, signed or unsigned, in the provided stamped return envelope.
Purpose: The George Washington University (GWU) Center For Economic Research,
under contract with the NCIA, is studying “who gets” PIE inmate income and other
benefits from inmate work in private sector firms. The results of the study will inform the
public and also tell policy makers how the goals of the PIE program are being met.
Procedures: We are asking your permission and assistance getting information from
payroll and associated records that you have, and from policies for PIE inmates working
for private PIE firms in your State:

101

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

•
•
•
•
•

Basic demographic and offense information for all PIE inmates in a recent year;
Gross wages (reported to IRS) and all PIE and other deductions for each inmate;
Firm matching or other contribution rates for PIE inmates;
Other benefits (health, retirement, or other) provided to PIE inmates; and
PIE inmate account information for a month before and during PIE work.

Possible Risks: Your State will not be identified. However, public discussion of overall
results could bring unwelcome as well as welcome attention.
Possible Benefits: Your State will not receive any direct benefit from this research.
However, the information you provide may help increase public acceptance of PIE
inmate-made products, inmate work, departments of corrections, and firms engaging
inmates, and improve inmate participation practices.
Costs: The research costs time providing information.
Compensation: Your State will not receive any payment.
Right to Withdraw From The Study: Your State’s participation is voluntary. You can
choose if your State will or will not take part in the research. You can stop at any time.
Privacy of Research Records: Your State will not be identified, and inmate, firm, and
State identifiers will be removed. Only GWU researchers and the GWU Committee on
Human Research will have access to identifying data. NCIA is not aware of your having
been asked to participate. Your records will be private unless you permit their release or
they are required by court order. Though unlikely, someone could identify your State or
firms from outside knowledge of special payroll features or deduction practices.
Questions? For project questions, please contact Tom Petersik. For questions about your
rights as a research participant, please call the GWU Assistant Vice President for Health
Research, Compliance and Technology Transfer—your representative—at 202-994-2995.
Consent to Participate: Please circle one of the choices below. If you agree to participate,
please also complete the information below.
I (Circle one) Consent / Do Not Consent to this State participating in this research.
Signature: By signing this consent form, you confirm that you have read this informed
consent form. You also agree that the study has been explained to you, your questions
have been answered, and your State will take part. You do not give up your legal rights if
you sign this informed consent form. Please retain a copy of this consent form.

102

Appendix B

______________________________________________
(Name and Title of Authorized Agent of the State)
______________________________________________
(Signature of Agent)
______________________________________________
(Date)
Researcher’s Statement (For completion before obtaining information): I confirm that
either I have explained the purpose, procedures, possible risks, and potential benefits of
this research, and have addressed any questions this State has raised.

Thomas W. Petersik, PhD
____________________________
(Signature)
____________________________
(Date)

.............

Record Layout
States were provided the below record layout, with data requested via automated datasets rather
than paper or other transmission methods. Most research data was obtained via automated files,
but information was also obtained via publicly available information on departmental Web sites
and via transcription from paper records.
Little difficulty occurred in obtaining basic inmate demographic, criminal justice, and PIE work
information. However, almost no information was readily available regarding prior employment
status or parenthood of minor children; nor was information readily available on deductions other
than for PIE.
Difficulties in obtaining optional demographic and expenditure data appear to be a combination
of lack of data, non-automated systems (meaning that data would need to be hand sought from
paper records, at best), and organizational separations (that is, records controlled by other offices

103

IDENTIFYING BENEFICIARIES OF PIE INMATE INCOMES

than PIE, with separate permissions and protocols, and with possibly different skills or software
facilities needed).

Items Requested:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.

104

Inmate ID
Numeric
Inmate Name
First & Last
State
PIE Firm Name or ID
Day, Month, Year of Birth
Sex
Race
Ethnic Origin
Marital Status at Arrest (Optional)
Parent of Child Born After 12/31/1979 Yes/No (Optional)
If yes, Birth Year of Youngest Child (Optional)
Highest Year Education Completed
Labor Force Status at Arrest
Most Serious Offense
Sentence, This Incarceration
PIE Worker—Annual Hours Worked
PIE Worker—Gross Wages Earned
PIE Firm—Social Security Contribution
PIE Firm—Medicare
PIE Firm—Workers’ Compensation
PIE Firm—Unemployment Compensation
PIE Firm—Other (specify)
PIE Firm—Health Care Contributions
PIE Firm—Retirement Contributions (not Social Security)
PIE Firm—Vacation/Annual Leave Contributions
PIE Firm—Other (specify)
PIE Firm—In-Kind Benefits (Banquets, Celebrations)
Corrections—Good Time or Other Credits
Corrections—Prison Industries Pay
Corrections—Housing or Other PIE-only Privilege
Corrections—Other (specify)
PIE Worker—Federal Income Tax Withheld
PIE Worker—State Income Tax Withheld
PIE Worker—Local Income Tax Withheld
PIE Worker—Social Security Withheld (OIASI + HI)
PIE Worker—Dependent Care Benefits
PIE Worker—Other Deduction (specify)
PIE Worker—PIE Board and Room

Appendix B

39.
40.
41.
42.
43.
44.

PIE Worker—PIE Victim Compensation
PIE Worker—PIE Family Support
PIE Worker—Non PIE Board and Room
PIE Worker—Non PIE Victim Restitution
PIE Worker—Non PIE Court Costs, Fines
PIE Worker—Non PIE Child Support

Timetable
The below proposed timetable was provided each invited State. Moreover, an attempt was made
to orally discuss and effect the proposed timetable. Particularly important components were (1)
organizing the core team of corrections staff, within and beyond PIE, to effect the research, and
(2) identifying and resolving issues quickly.
Although all required and some optional research objectives were met in every sampled State, the
proposed protocol quickly became almost entirely fiction. Teams were never identified or
assembled, and the in-State work appears to have been addressed primarily by PIE staff, with adhoc assistance by other corrections offices when needed, a technique that generally worked
relatively well, but on occasion likely meant missed data opportunities, confusion and
misunderstandings, delays, or inefficient (that is, reliant on paper rather than automated data) data
assembly, plus mailed transmission followed by tedious and error-prone hand re-entry of data.

.............
Timetable & Proposed Procedure
Identifying Beneficiaries of PIE Inmate Incomes
The George Washington University Center For Economic Research
Under Contract With the National Correctional Industries Association (NCIA)
IRB Number U080222ER
To Discuss, Please Contact the Principal Investigators:

Thomas Petersik, Ph.D. (703) 323-5272; tkpetersik@yahoo.com

.............

105