By Peter Wagner and Bernadette Rabuy
January 25, 2017
The cost of imprisonment — including who benefits and who pays — is a major part of the national discussion around criminal justice policy. But prisons and jails are just one piece of the criminal justice system and the amount of media and policy attention that the various players get is not necessarily proportional to their influence.
In this first-of-its-kind report, we find that the system of mass incarceration costs the government and families of justice-involved people at least $182 billion every year. In this report:
Our goal with this report is to give a hint as to how the criminal justice system works by identifying some of the key stakeholders and quantifying their “stake” in the status quo. Our visualization shows how wide and how deep mass incarceration and over-criminalization have spread into our economy. We find:
This report and infographic are a first step toward better understanding who benefits from mass incarceration and who might be resistant to reform. We have no doubt that we missed some costs, and we did not include some costs because they are relatively small in the big picture or are currently unknowable. But, by following the money, one can see that private prison corporations aren’t the only ones who benefit from mass incarceration.
Some of the lesser-known major players in the system of mass incarceration and over-criminalization are:
A graphic like this shows the relative economic cost of different parts of mass incarceration, but it can also obscure the fact that we don’t have a single monolithic system. Instead, we have a federal system, 50 state systems, and thousands of local government systems. Sometimes these systems work together, although often they do not; and looking at just the national picture can obscure the importance of state and local policy decisions. For example, while state government spending makes up the majority (57%) of corrections costs, local governments make up almost a third (32%).4 Local governments are largely enforcing state law, and local discretionary arrest and bail policies can have tremendous influence on both the state budget and justice outcomes. For example, more than half ($13.6 billion) of the cost of running local jails is spent detaining people who have not been convicted.5
To be sure, there are ideological as well as economic reasons for mass incarceration and over-criminalization. But at this moment, when crime is near record lows and there is increasing attention to the role of privatization in the justice system, we need a far more expansive view of how our criminal justice system works, whom it hurts, and whom it really serves.
If we are to make our society safer and stronger, we’ll need to be making far smarter investments than we are today.
While this report asks a very simple question about the financial costs of mass incarceration, a comprehensive answer has never existed before because the government doesn’t collect or organize these statistics in one place. Like our report, Mass Incarceration: The Whole Pie, which looked at who is behind bars and why, this report aggregates economic data that have never been put together before to offer a big picture view of who pays for and who benefits from mass incarceration.
Before we explain our sources and methodology, it might be helpful to explain our goals and how this report’s finding that mass incarceration consumes at least $182 billion each year is different than the two more commonly cited Bureau of Justice Statistics figures:
This report’s goal is to identify the total cost of mass incarceration and the comparative investment that each part of that system has in the status quo. We wanted to take a holistic view without also exaggerating our findings by including tens of billions of dollars in policing and court expenses that have little to do with the enforcement of criminal law.
In general, this report includes a number of firsts, including the identification of sufficient data points to develop national estimates where none existed before. But our approach does have a few weaknesses that other researchers building on this work should be aware of:
What follows is a description of the data sources and assumptions used for each part of the infographic. Each circle’s area is in proportion with the value being represented. We welcome ideas on newer data sources, more precise estimates, and information on costs that we missed.
The Bureau of Justice Statistics reports that the combined total of federal, state and local expenditures on the judicial and legal system was $57.9 billion in 2012.11 Because these figures include both criminal and civil law aspects of the court system, we reduced this figure by world renowned criminologist Nils Christie’s estimate that 50% of court expenditures were criminal law related. With rounding, our figure is $29.0 billion. 12
While our calculations are aimed at excluding costs associated with the civil justice system, it is important to recognize that mass incarceration has substantial economic impacts on civil courts. First, a material portion of the federal courts’ workload consists of habeas corpus petitions, motions to vacate sentences, complaints involving prison conditions and cases concerning the civil rights of incarcerated people. These are all reported as civil matters, even though they are driven by the system of mass criminalization. Data from the federal court system shows that such matters constituted 20% of district court civil filings and 48% of appellate court civil filings in the year ending in March 2015.13 Second, courts’ criminal caseloads can “squeeze out” civil cases.14 Since 1986, workloads for federal judges have increased, driven by the increase in criminal cases. Unlike civil cases, criminal defendants have a right to a speedy trial, which means that as criminal workload increases, judges are sometimes forced to prioritize criminal cases, resulting in a slower pace for civil matters and imposing opportunity costs on civil litigants. Presumably this dynamic is even more acute in state courts, which handle more criminal cases than the federal system.
Judicial and legal expenditures include expenses for prosecutors ($5.8 billion in 2007)15 and for indigent defense ($4.5 billion in 2008).16 The indigent defense expenses include both public defender agencies and private counsel appointed by the courts; but it would not include the currently unknown number of billions paid by individuals and families to private defense and appellate attorneys.
And notably, judicial and legal costs do not include monetary payments paid by governments when it is sued in a judicial or administrative proceeding.
The Bureau of Justice Statistics reports in Justice Expenditure And Employment Extracts, 2012 that the combined total of federal, state and local expenditures on policing was $126.4 billion in 2012. Unfortunately and significantly, these figures do not include separate costs for the criminal and civil components of police work, so we used Nils Christie’s estimate that 50% of police expenditures are related to the enforcement of the criminal law to reduce the Bureau of Justice Statistics figure by half.17 We also note a Cincinnati, Ohio study, which found that police officers spend 33% of their time patrolling and 17% on crime calls. There is, obviously, an urgent need for more precise data on police expenditures for projects like this report and, more importantly, to help state and local governments make useful, comparative policy decisions about what are and what are not appropriate duties for law enforcement.
In addition, note that because in most states, sheriffs’ departments are multifunctional agencies providing police protection, judicial, or correctional services, sheriffs’ expenditures are prorated by function to corrections, policing, and judicial and legal.18
Our $4.5 billion figure reflects the net assets of the Department of Justice and Treasury forfeiture funds in fiscal year 2014 as reported on page 10 of the Institute for Justice’s 2015 report, Policing for Profit: The Abuse of Civil Forfeiture, 2nd Edition. According to the Institute for Justice, net assets are a more stable metric than annual deposits into the forfeiture funds (over $5 billion) because net assets measure the funds remaining after the government pays various obligations like payments to victims. Notably, this figure is an undercount of the total because it does not include state forfeiture revenue. According to the Institute of Justice, “Unfortunately, deriving similar totals at the state level is impossible because most states require little to no public reporting of forfeiture activity. However, of the  states from which the Institute for Justice was able to obtain usable data, the totals are” more than $254 million. (page 11)
There is likely some overlap between the $4.5 billion net assets of the Department of Justice and Treasury forfeiture funds and the expenditures reported for the policing and judicial and legal systems. We didn’t adjust the policing and judicial and legal system figures because it is likely impossible to figure out how much overlap there is and to determine how the money is distributed among the various government agencies. The federal government and states vary in what percentage of the seized property can be kept by the law enforcement agency that seized the property.
While this report adjusts policing and judicial and legal costs to focus on the criminal parts of these systems and exclude the civil parts, we include civil asset forfeiture because civil asset forfeiture is a mechanism by which law enforcement agencies can seize and retain property on the suspicion that the property is connected to a crime.
In its report, For Better or For Profit: How the Bail Bonding Industry Stands in the Way of Fair and Effective Pretrial Justice, the Justice Policy Institute uses a figure of $14 billion in bail bonds written every year, and cites its source as an email from Dennis Bartlett from the American Bail Coalition (which lobbies on behalf of the bail bondsman industry) in footnote 2 of the Executive Summary of the report. Since defendants and their family members typically pay 10% to commercial bail bond agencies as a nonrefundable fee, this comes out to $1.4 billion actually paid by the families. To learn more about the high costs of money bail in the U.S., see our report Detaining the Poor: How money bail perpetuates an endless cycle of poverty and jail time.
For our calculations of the total value of commissary expenditures, see Stephen Raher, Paging anti-trust lawyers: Prison commissary giants prepare to merge, July 5, 2016. We assigned incarcerated people’s commissary purchases to a section on money spent by families precisely because incarcerated people don’t make very much money. Many people confined in jails don’t work, and four state prison systems19 don’t pay at all. And the states that pay for work aren’t much better. Looking at just the states that paid incarcerated people for non-industry work in 2001, the average minimum wage per day was 93 cents. Therefore, the majority of the money spent on commissary comes from the families of incarcerated people and not from the incarcerated people themselves.
As with the telephone industry (below), a portion of this cost to the families feeds back into the $81 billion correctional budgets via the corrupt commission system — where private companies provide commissary services for correctional facilities but share a percentage of the revenue with the government. In this way, it’s possible we are counting these dollars twice in our total estimate, although our methodology of ignoring costs under a billion dollars more than makes up for any double counting in this respect.
Lee Petro, pro bono counsel for the Wright Petitioners analyzed the 2015 financial reports of 13 prison telephone companies, including the nation’s largest, which were required to submit that information to Alabama state regulators. For one large company, CenturyLink, he used the 2014 figure because an accounting change at that company included its other, non-prison related, businesses in the 2015 figure, but he expects the telephone figure for CenturyLink to be generally unchanged from 2014 to 2015. Other companies not regulated by the state of Alabama are not included, so this is a slight under-estimate. In some locations these costs are physically paid by incarcerated people and in some cases by the families, but regardless of who makes the payment, the source of the funds is almost entirely the families.
A large portion of this cost to the families feeds back into the $81 billion correctional budgets via the corrupt commission system — where contracts are awarded not on the basis of the best price and service to the consumer but on the size of the revenue that is paid to the government authority that awards the monopoly contract. In this way, it’s possible we are counting these dollars twice in our total estimate, although our methodology of ignoring costs under a billion dollars more than makes up for any double counting in this respect.
The Bureau of Justice Statistics reports that the combined total of federal, state and local expenditures on corrections — which includes prisons, jails, juvenile facilities, probation and parole, and immigration detention was $80.7 billion in 2012. Within this cost, we provide more detail:
Because construction costs can be highly variable, we averaged several recent years when it was possible to do so:
|Level of government||Annual Expenditure in billions||Years averaged||Sourcing/notes|
|Local||1.4||2005-2011||Local Government Corrections Expenditures, FY 2005-2011, table 3|
|State||1.8||2002-2010||State Corrections Expenditures, FY 1982-2010 appendix table 2|
|Federal||0.1||2014||Bureau of Prisons 2014 budget request p. 3 (although note, this may include some maintenance.)|
Many prisons are built with borrowed money. When governments borrow money, they usually issue bonds that pay interest to investors (typically institutional investors such as pension funds). The principal amount of such bonds is already included in the construction-cost segment of our graphic. Because there is no available data on the interest expense associated with prison-specific bonds, we calculated that figure as follows. In recent decades, municipal bonds have paid interest for an average period of 18 years, according to the Securities Industry and Financial Markets Association (SIFMA) [Excel]. Accordingly, to estimate the total interest expense in any given year, one must calculate the principal amount of bonds issued during the preceding 18 years. We looked at prison bonds from 1998 through 2015. For 2012 through 2015, we relied on SIFMA’s Municipal Bond Credit Report series, which reports the principal amount of all municipal bonds issued for construction of state and local correctional facilities. For years prior to 2012, we obtained state prison-bond figures from the National Association of State Budget Officers’ State Expenditure Report series (NASBO reports only state bonds; thus, we estimated local government bonds by calculating the ratio of state-to-local correctional facility bonds in 2012-15, and extrapolating local bond issuances for 1998-2011 based on the same ratio). The aforementioned calculations resulted in an estimated principal amount of $47.4 billion between 1998 and 2015. Unfortunately there is no way to know what interest rate the different bonds were paying, so we calculated interest by conservatively assuming an average rate of 4%, although actual rates for prison bonds are likely higher.
Food is another large factor where the data is hard to access. The Bureau of Justice Statistics reports in State Prison Expenditures, 2001 (table 5) that the per capita food cost was $955 per year — $2.62 per day. Multiplying this figure by 2.3 million incarcerated people produces a total annual expenditure of $2.197 billion. Because our visualization separates out private prisons, we multiplied the annual figure by the number of people in private prisons, and concluded that $125 million was too small to include in the graph. The far larger number of people confined in publicly run prisons, however, sums to $2.071 billion and is labeled on the data visualization.
The Bureau of Justice Statistics reports in State Prison Expenditures, 2001 (table 5) that the per capita utility cost was $795 per year. Multiplying this figure by 2.3 million incarcerated people produces a total expenditure of $1.8 billion. Because our visualization separates out private prisons, we multiplied the annual figure by the number of people in private prisons, and concluded that $104 million was too small to include in the infographic. The far larger number of people confined in publicly run prisons, however, sums to $1.7 billion and is labeled on the data visualization. Given that energy costs have been rising faster than inflation, this could be much higher. Utility costs should not be ignored because, as Ruth Gilmore explains in Golden Gulag: Prisons, Surplus, Crisis and Opposition in Globalizing California, prisons (and large jails) are essentially small cities with the water, heating and electrical demands of — as you would expect once you begin to see the facilities that way — small cities.
To illustrate both the scale of the private prison industry and the critical fact that this industry works under contract for government agencies — rather than arresting, prosecuting, convicting and incarcerating people on its own — we displayed these companies as a subset of the public corrections system.
This industry is dominated by two large publicly traded companies — CoreCivic (which until recently was called Corrections Corporation of America (CCA)) and The GEO Group — as well as one small private company, Management & Training Corp (MTC). We relied on the public annual reports of the two large companies, and estimated MTC’s figures using records from a decade-old public record request.
In sum, we estimate that the three companies received $3.9 billion in revenue from mass incarceration and immigration detention and made $0.37 billion in profits ($374 million).
The figure for private corrections is one area where we are including civil and criminal costs. While these immigration detention facilities, many of which are private,22 and the confinement there are technically civil, in reality, they are quite like prisons. People in federal prison for criminal convictions of violating federal immigration laws and people detained civilly in local jails under contract with U.S. Immigration and Customs Enforcement are included in the public corrections costs.
For further reading on some of the topics not fully explored in this report:
All Prison Policy Initiative reports are collaborative endeavors, but this report was particularly collaborative. It required multiple in-depth investigations to answer questions that had never successfully been answered before. The authors are indebted to James Kilgore, Alex Friedmann, Ruth Wilson Gilmore, John Pfaff, Chris Sturr and Bruce Reilly who all offered invaluable feedback on earlier drafts of this report. Aleks Kajstura, Lauren Powers, Wendy Sawyer, Alison Walsh, and Emily Widra helped with the research; and Stephen Raher of our Young Professionals Network provided groundbreaking research on the commissary industry, the money transfer industry, and helped us understand several other key topics. Bob Machuga came up with the initial visual design allowing us to depict how the money of mass incarceration flows. Any errors or omissions in the final report, however, are the sole responsibility of the authors. We also thank the MacArthur Foundation for their support and each of our individual donors who invests in the Prison Policy Initiative’s work and who gives us the resources and the flexibility to quickly turn our insights into new movement resources.
The non-profit, non-partisan Prison Policy Initiative has been shining a fresh light on how our criminal justice system really works since 2001. Through research, innovative media work, and intersectional organizing, our staff members shape national reform campaigns from our office in Western Massachusetts.
Peter Wagner is an attorney and the Executive Director of the Prison Policy Initiative. He co-founded the Prison Policy Initiative in 2001 in order to spark a national discussion about the negative side effects of mass incarceration. His research and advocacy on the issue of prison gerrymandering have led four states and more than 200 local governments to end prison gerrymandering.
Some of his most recent work include exposing the entire mass incarceration pie, uncovering that prisons are disproportionately built in White areas, and working with Josh Begley to put each state’s overuse of incarceration into the international context.
He is @PWPolicy on Twitter.
Bernadette Rabuy is the Senior Policy Analyst at the Prison Policy Initiative. Bernadette produced the first comprehensive national report on the video visitation industry, Screening Out Family Time: The for-profit video visitation industry in prisons and jails, finding that 74% of local jails that adopt video visitation eliminate traditional in-person visits. Her research has played a key role in protecting in-person family visits in jails in Portland, Oregon and the state of Texas. In her other work with the Prison Policy Initiative, Bernadette has worked to empower the criminal justice reform movement with key but missing data through the annual Mass Incarceration: The Whole Pie reports and, most recently, Detaining the Poor: How money bail perpetuates an endless cycle of poverty and jail time.
Bernadette is on Twitter at @BRabuy
Except for private prison profits ($0.37 billion), we excluded costs of less than $1 billion from our infographic. ↩
See the following for examples of prison and jail population declines not being followed by proportional staffing reductions: California, New York State and New York City. New York State law even gives the guard union what amounts to veto power over prison closures. ↩
Alaska, Delaware, Hawaii, Idaho, Maine, Montana, Nebraska, New Hampshire, New Mexico, North Dakota, Rhode Island, South Dakota, Vermont, West Virginia, and Wyoming each has a population smaller than 2.3 million people. ↩
The federal government is responsible for only a small portion (10%) of correctional expenditures. These percentages are based on the direct expenditures provided in jeeus1201.csv from Bureau of Justice Statistics, Justice Expenditures and Employment Extracts, 2012 - Preliminary. ↩
In order to calculate the cost of pretrial detention, we first needed to calculate the cost of running local jails. The Bureau of Justice Statistics reports that local government corrections spending in 2012 was $26,397,777,000, but this figure includes probation. See jeeus1201.csv at Bureau of Justice Statistics in Justice Expenditures and Employment Extracts, 2012 - Preliminary. Thus, we needed to subtract the cost of probation from the local government corrections spending. To calculate the annual probation costs, we used the probation costs per year per person ($1,250) reported by the Pew Center on the States’ Public Safety Performance Project on page 12 of its March 2009 report, One in 31: The Long Reach of American Corrections. Pew got this cost from surveying 33 states in 2008. We multiplied $1,250 by the number of people on probation in 2015 reported by the Bureau of Justice Statistics in Table 1 of Probation and Parole in the United States, 2015: 3,789,800. Therefore, we found that the annual cost of probation is $4,737,250,000. Now that we had figures for local government corrections spending and probation, we subtracted the cost of probation from the cost of local government corrections spending, and we got $21,660,527,000 as the annual cost of running local jails. Finally, we calculated the cost of pretrial detention by multiplying the percentage of people pretrial in local jails in 2014 (62.8%) from Table 3 of Bureau of Justice Statistics, Jails in Midyear 2014 by local spending on jails to arrive at $13.6 billion. Another available figure for the cost of detaining people pretrial is roughly $9 billion, an estimate provided by former Attorney General in a speech at the National Symposium on Pretrial Justice in 2011, but we were unable to locate the source for this figure. We used the unadjusted percentage for the pretrial population, which includes people held in local jails for other authorities like Immigration and Customs Enforcement. We did not adjust for people who are on probation and also on parole or incarcerated. ↩
See the corrections data, which is in the file jeeus1201.csv from the Bureau of Justice Statistics, Justice Expenditure And Employment Extracts, 2012 based on data from the Census Bureau’s Annual Survey of State and Local Government Finances and Annual Survey of Public Employment and Payroll. ↩
See the total justice system data, which is in the file jeeus1201.csv of the Bureau of Justice Statistics, Justice Expenditure and Employment Extracts 2012. ↩
See the groundbreaking 2015 report Who Pays: The True Cost of Incarceration on Families, produced by Ella Baker Center for Human Rights, Forward Together and Research Action Design, which surveyed 1,080 formerly incarcerated people and their family members to find that families paid an average of $13,607 in court-related costs. Given the high costs and pervasiveness of fines and fees, the harm to impoverished families can hardly be seen as an unintended consequence of reasonable policies. Also see the persuasive argument that these costs amount to a system of “seizure” that taxes poor families of the incarcerated “to subsidize the carceral state.” Katzenstein, Mary Fainsod, and Maureen R Waller. “Taxing the Poor: Incarceration, Poverty Governance, and the Seizure of Family Resources.” Perspectives on Politics 13.03 (2015): 639. ↩
See Gov’t Accounting Standards Bd, Statement No. 10, “Accounting and Financial Reporting for Risk Financing and Related Insurance Issues” (Nov. 1989) PP 13, 64-72. ↩
For a comprehensive review of statutes controlling where telephone revenue goes, see Exhibit H in the Reply Comments of the Wright Petitioners to the FCC on April 22, 2013, and for an example of how these funds restricted to “Inmate Welfare” are often misused, see Dear Butte County: You Can’t Fleece the Inmate Welfare Fund to Pay for a New Jail. ↩
See Bureau of Justice Statistics, Justice Expenditures and Employment Extracts, 2012 - Preliminary, jeeus1201.csv file. ↩
Nils Christie, Crime Control as Industry, 3rd edition (2000), page 141 fn 4, which Christie calls a low estimate. But like Christie, we think it better to under-estimate than over-estimate. We could identify only one other possible way to identify the portion of judicial and legal expenditures that were criminal law in nature: the National Center for State Courts’ Court Statistics Project finds that, in 2013, state court caseloads were 54% traffic, 21% criminal, 18% civil, 6% domestic relations, and 1.4% juvenile. See: R. LaFountain, S. Strickland, R. Schauffler, K. Holt, and K. Lewis, Examining the Work of State Courts: An Overview of 2013 State Court Caseloads (National Center for State Courts 2015). But given that the data on court caseloads by type includes traffic offenses and other types of cases that would require far less time, attention, and resources than criminal cases, it did not seem possible to base our estimate on the court caseload statistics. ↩
See Administrative Office of U.S. Courts, Federal Judicial Caseload Statistics: 2015 tables C-3 and B-7. ↩
See Patricia W. Hatamayer Moore, The Civil Caseload of the Federal District Courts, 2015 U. of Ill. L. Rev. 1177, 1187-91 (2015). ↩
See page 2 of Steven W. Perry and Duren Banks, Prosecutors in State Courts, 2007 - Statistical Tables (Washington, D.C.: Bureau of Justice Statistics, 2011). ↩
See page 76 of Holly R. Stevens, Colleen E. Sheppard, Robert Spangenberg, Aimee Wickman, and Jon B. Gould, State, County and Local Expenditures for Indigent Defense Services Fiscal Year 2008 (Fairfax, VA: The Spangenberg Project for the American Bar Association, 2010). ↩
Nils Christie, Crime Control as Industry, 3rd edition (2000), page 138. See also fn 3 on page 141 where he calls this a low estimate, but like Christie we think it better to under- rather than over-estimate. ↩
See the definitions of terms and concepts for the Bureau of Justice Statistics’ report, Justice Expenditure and Employment Extracts, 2012 - Preliminary. ↩
Alabama, Florida, Georgia, and Texas ↩
The “Governmental Employment” section of the Justice Expenditure and Employment Extracts Definitions of Terms and Concepts says that contractors and their employees are excluded from the employees included in the Justice Expenditure and Employment Extracts series. ↩
The combined capacity of Management & Training Corp’s facilities is 31,962, which is 36% of CoreCivic’s 88,142. Applying that estimate to the dollar value of the contracts, we estimate that Management & Training Corp had $0.6 billion in contracts and $80 million in profits. (Had we based our estimates on GEO Group and the fact that Management & Training Corp’s contracts have 46% of the capacity of GEO Group’s, we would have estimated $0.56 billion in contracts and $63 million in profits.) ↩
Sixty-five percent of immigration detention facilities are run by the private prison industry under contract with U.S. Immigration and Customs Enforcement. See page 6 of Homeland Security Advisory Council, Report on the Subcommittee on Privatized Immigration Detention Facilities (Washington, D.C.: U.S. Department of Homeland Security, 2016). ↩