HOW M A N Y A MER IC A NS ARE
UNNECESSARILY INCARCERATED?
Dr. James Austin and Lauren-Brooke Eisen
with James Cullen and Jonathan Frank
Preface by Inimai Chettiar

Brennan Center for Justice at New York University School of Law

about the brennan center for justice
The Brennan Center for Justice at NYU School of Law is a nonpartisan law and policy institute that
seeks to improve our systems of democracy and justice. We work to hold our political institutions and
laws accountable to the twin American ideals of democracy and equal justice for all. The Center’s work
ranges from voting rights to campaign finance reform, from ending mass incarceration to preserving
Constitutional protection in the fight against terrorism. Part think tank, part advocacy group, part cuttingedge communications hub, we start with rigorous research. We craft innovative policies. And we fight for
them — in Congress and the states, the courts, and in the court of public opinion.

ABOUT THE BRENNAN CENTER’S Justice Program
The Brennan Center’s Justice Program seeks to secure our nation’s promise of “equal justice for all” by
creating a rational, effective, and fair justice system. Its priority focus is to reduce mass incarceration while
keeping down crime. The program melds law, policy, and economics to produce new empirical analyses
and innovative policy solutions to advance this critical goal.

ABOUT THE BRENNAN CENTER’S PUBLICATIONS
Red cover | Research reports offer in-depth empirical findings.
Blue cover | Policy proposals offer innovative, concrete reform solutions.
White cover | White papers offer a compelling analysis of a pressing legal or policy issue.

© 2016. This paper is covered by the Creative Commons “Attribution-No Derivs-NonCommercial” license (see http://creativecommons.
org). It may be reproduced in its entirety as long as the Brennan Center for Justice at NYU School of Law is credited, a link to the Center’s
web pages is provided, and no charge is imposed. The paper may not be reproduced in part or in altered form, or if a fee is charged,
without the Center’s permission. Please let the Center know if you reprint.

ABOUT THE AUTHORS
Dr. James Austin is the President of the JFA Institute. He is the former director of the Institute on
Crime, Justice and Corrections at George Washington University in Washington, D.C., and former
Executive Vice President for the National Council on Crime and Delinquency. He has served as director
for several large U.S. Department of Justice-funded research and evaluation programs and has served as
the project director of the Bureau of Justice Assistance projects. Dr. Austin provides technical assistance
to state and local jails, probation, parole, and prison systems.
Over the last four decades, Dr. Austin has worked with the Federal Bureau of Prisons, 47 state prison
systems, 100 local jails, the District of Columbia, Puerto Rico, Virgin Islands, Australia, and Canada
to help officials find ways to reduce their incarcerated populations. He was an expert witness in the
landmark Supreme Court decision, Brown v. Plata, which mandated the state of California to reduce
its prison population. He has authored numerous publications, received the American Correctional
Association’s Peter P. Lejin’s Research Award, and the Western Society of Criminology Paul Tappin
Award for outstanding contributions in criminology. Dr. Austin holds a Ph.D. in sociology from the
University of California at Davis, and a master’s degree in sociology from DePaul University.
Lauren-Brooke Eisen is senior counsel in the Justice Program at the Brennan Center for Justice. With
expertise in criminal law and practice, Ms. Eisen spearheads program reports melding legal, policy, and
data analyses to produce new policy and legal recommendations. Previously, she was a Senior Program
Associate at the Vera Institute of Justice in the Center on Sentencing and Corrections. Ms. Eisen
also served as an Assistant District Attorney in New York City in the Sex Crimes and Special Victims
Bureau, Criminal Court Bureau, and Appeals Bureau where she prosecuted a wide variety of criminal
cases. She has worked as a journalist in Laredo, Texas, covering crime and justice; served as an adjunct
instructor at the John Jay College of Criminal Justice; and taught a seminar on mass incarceration at
Yale College. She holds an A.B. from Princeton University and a J.D. from the Georgetown University
Law Center.
James Cullen is a Research and Program Associate in the Justice Program at the Brennan Center for
Justice. As part of the program’s law and economics research team, he conducts economic, statistical,
policy, and legal research and analysis on issues related to mass incarceration. He holds a B.A. in
economics and political science from the University of Chicago.
Jonathan Frank was the Ford Foundation Legal Fellow at the Brennan Center, working in the Justice
Program from 2015 to 2016, where he focused on policy reforms to reduce mass incarceration. Prior
to the Brennan Center, he interned at the Federal Public Defender’s office in San Francisco and the
American Civil Liberties Union Criminal Law Reform Project in New York City, and represented
indigent defendants in the Stanford Criminal Defense Clinic. He holds a B.A. from Swarthmore
College and a J.D. from Stanford Law School.

ACKNOWLEDGEMENTS
The Brennan Center gratefully acknowledges Laura and John Arnold, Marguerite Casey Foundation,
Democracy Alliance Partners, Jason Flom, Ford Foundation, The Margaret and Daniel Loeb — Third
Point Foundation, Open Society Foundations, Public Welfare Foundation, Rockefeller Family Fund,
and William B. Wiener, Jr. Foundation for their support of the Justice Program.
The authors are grateful to Inimai Chettiar for guiding this report’s analysis, methodology, and writing;
Michael Waldman and John Kowal for their insights; Ames Grawert, Vivien Watts, Holly Harris, and Fritz
Schwartz for their assistance; and Brennan Center Board member Stephen Schulhofer for his insights.
The authors are especially appreciative to the author of this report’s foreword, Cornell Brooks.
They also thank Julia Bowling, Natasha Camhi, Brenna Christensen, Brittany Cohen, Jessica Eaglin,
Patrick Feeney, Nicole Fortier, Matthew Friedman, Nia Itoh, Morgan Lingar, Will Lockwood, Jason
O’Conner, Adureh Onyekwere, Eva Torres, and Bianca Tylek for their research assistance; Kate Robards
for her comments; Rebecca Autrey, Theresa Jefferson, Jim Lyons, Jeanine Plant-Chirlin, and Erik Opsal
for their communications assistance; and Beatriz Aldereguia, Nicole Austin-Hillery, Steven Hawkins,
Tanisha McClellan, Ngozi Ndulue, Kia Pearson, Mary Price, and Magdalene Zier for their assistance.

EXPERT REVIEWERS
The authors are indebted to the sentencing experts who carefully reviewed this report and its
methodology, and provided invaluable detailed feedback:
•	 J ohn Gleeson, former U.S. District  Judge, Eastern District of New York; Partner at Debevoise
& Plimpton LLP; and
•	 Carol Steiker, Henry J. Friendly Professor of Law at Harvard Law School.

TABLE OF CONTENTS
Foreword by Cornell William Brooks	
1
	
Preface by Inimai Chettiar	3
I.	

Current Prison Population 	

11

II.	

Time Served In Prison Today	

17

III.	

Ending Prison For Lower-Level Crimes	

21

IV.	

Reducing Time Served For Other Crimes	

35

V.	

Recommendations & Cost Savings	

42

Conclusion	46
Apendix A: Methodology	

47

Endnotes	57
	

FOREWORD
Cornell William Brooks
In 1963, the March on Washington marked a turning point in the long fight for civil rights for African
Americans. A century after President Lincoln issued the Emancipation Proclamation, hundreds of
thousands converged at his memorial to celebrate a century of liberation and to protest what Rev.
Dr. Martin Luther King, Jr. called “the manacles of segregation and the chains of discrimination.” In
the intervening fifty years, we have come a remarkable distance, but the shackles of systemic racism
continue to bind communities of color.
We stand on the frontlines in the fight to build a society free from racial discrimination. In 2015,
we honored the sacrifices of our forbearers and galvanized international attention to systemic
discrimination with a “Journey for Justice” from Selma, Ala. to Washington, D.C. While national
support for this effort provides hope the tide may be turning, it also belies a sad truth: Many of the
grave inequalities we fought decades ago still persist, more than fifty years after the Civil Rights Act.
The single greatest injustice that threatens our safety and hinders our progress? Mass incarceration.
People of color bear the brunt of our criminal justice system in disproportionate and devastating
numbers. This is in part because racial disparities exist at all stages of the system, which relies on
corrosive practices that harm people of color. Our communities have already suffered from historic
and systemic economic injustice and racially targeted criminal justice policies. These wounds have
not healed and have been aggravated by the staggering number of people trapped in prisons over
the past forty years. Today, an estimated 2.2 million people are locked inside jails and prisons.
African Americans make up roughly 13 percent of the U.S. population but 37 percent of the nation’s
prisoners.1 People with dreams and aspirations suffer in airtight cells of prison and poverty. But the
injustice does not end there. More than half of formerly incarcerated Americans are unemployed a
year after release. Communities of color are over policed, over-prosecuted, over-incarcerated and yet
underemployed.
If we do not take steps now, Americans of color will forever be relegated to a penal and permanent
underclass, and mass incarceration will continue to cage the economic growth of our communities.
We have reached a crisis point, and we need solutions. This groundbreaking report from the Brennan
Center for Justice offers a pathway to reduce our prison population and its tragic racial disparities. It
documents the number of people behind bars without rationale, and reveals the unnecessary trauma
this causes. It recommends real solutions that can help end over-incarceration. I urge lawmakers to
give deep consideration and deeper commitment to this report’s findings and recommendations.
This nation must continue to march forward, toward a day when all people are treated based not
on the color of their skin but on the content of their character, uncolored and un-stigmatized by a
criminal record. It is time that we end the plague of mass incarceration.
Mr. Brooks is the President and CEO of the National Association for the Advancement of Colored People.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 1

PREFACE
By Inimai Chettiar
United States incarceration statistics are well-known by now, but they still stun. Although the U.S. has
only 5 percent of the world’s population, it has nearly 25 percent of the world’s prisoners.2 There are
2.2 million people in prisons and jails in this country.3 With an incarceration rate of 693 per 100,000
people, the United States trails only the Seychelles in the frequency with which it deprives its residents
of liberty, and vastly outpaces that of Iran, Zimbabwe, and even notoriously punitive Singapore.4
This level of incarceration, hidden for so long in plain sight, has massive societal consequences. It
drives and reinforces deep seated racial inequity and disproportionately punishes African Americans. It
ruins people’s lives and breaks up families. Now it has become widely understood that this exorbitant
incarceration rate is not needed to keep the public safe. With the ascendancy of Donald Trump as
president, we hope the trans-partisan nationwide effort for criminal justice reform will continue. And
so we face a challenge as a nation: how can we reduce this prison population without endangering
public safety?

How We Got Here
In 1974, the country’s imprisonment rate was 102 per 100,000. By the turn of the century, it had
quadrupled, reaching 138 in 1980, 295 in 1990, and 470 in 2000. The surge continued. By 2007, the
imprisonment rate reached its peak of 506.5
Since then, the rate has decreased marginally to 471 prisoners per 100,000 Americans.6 While the
downward trend is welcome, a yearly declining rate of 1 percent is modest at best. At this pace, it would
take nearly 75 years to return to the 1985 incarceration rate of 200 per 100,000.7
Like other failed social experiments, mass incarceration was not a spasm without a cause. It was the
result of sustained policies beginning in the mid-1970s and campaigned on even earlier.
As early as 1964, Republican Presidential nominee Barry Goldwater promised “a government that
attends to its inherent responsibilities of maintaining . . . and enforcing law and order.”8 Four years
later, after riots in major cities including Chicago, Los Angeles, and Washington, D.C., and growing
protests against the Vietnam War, there was a palpable sense among many Americans that “order” had
to be restored. In 1968, Republican presidential candidate Richard Nixon appealed to what we would
later call the “silent majority.” In one campaign commercial, a series of still photos of angry protesters
and burning buildings appeared over a soundtrack of a snare drum and dissonant piano chords. “Let
us recognize that the first civil right of every American is to be free from domestic violence,” Nixon
intoned. “So I pledge to you, we shall have order in the United States.”9
To a large extent, what average Americans saw on their television screens squared with their own
experiences. From 1960 to 1980, violent crime soared 270 percent. It continued to increase, peaking

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 3

at 758 violent offenses per 100,000 people in 1991.10 African American and Latino communities bore
the brunt of this crime rise. By the late 1970s, people of color were crime victims at a rate 24 percent
higher than white Americans.11
In response, states and the federal government enacted a series of laws that dramatically lengthened
sentences for many crimes, and also created entirely new crimes.
One early example is the passage of the “Rockefeller Drug Laws” in 1973. Named after New York
Gov. Nelson Rockefeller, the laws imposed mandatory minimum 15-year terms for possession
of marijuana and other drugs.12 Michigan adopted similar sanctions, and other states began to
follow suit, setting mandatory sentences for various crimes. In the next decade, Washington state
adopted the nation’s first “truth-in-sentencing” law, which required inmates to serve at least 85
percent of their sentences before becoming eligible for parole. Then, 27 other states imposed
similar requirements. Three states — Idaho, New Hampshire, and Nevada — went even further,
requiring that prisoners serve 100 percent of their terms. Meanwhile, 14 states, including Florida,
Illinois, North Carolina, and Ohio abolished the ability of parole boards to release inmates.13
Similar events unfolded in the nation’s capital. In 1984, Congress created the U.S. Sentencing
Commission. Driven by concerns that judges retained too much discretion, resulting in disparate
treatment of defendants and overly lenient sentences, Congress charged the Commission with
crafting sentencing guidelines that limited the discretion of federal judges and resulted in stiffer
penalties.14 Then Congress abolished parole and enacted harsher sentences for federal crimes,
particularly drug offenses, over the next decade.15
By the end of the 20th century, yet another strict approach to sentencing had taken hold. This
time, it was “three-strikes-and-you’re-out” laws, which automatically imposed lengthy prison
sentences, up to and including life without parole, for those convicted of a third felony. Spurred
in part by the federal 1994 “Crime Bill,” half the states passed three strikes laws between 1993 and
1995 alone.16
There is disagreement over whether the devastating racial inequalities caused by these policies
were known and intentional, or inadvertent side effects. But it’s clear that our inheritance is an
incarceration rate five times higher than in the mid-1970s, with vast racial disparities.17

Effect on Crime
This history begs a central question. Did mass incarceration achieve its ostensible goal of reducing
crime? After all, it would be difficult to suggest cutting the prison population if doing so would
risk public safety.
Over the last four decades, crime plummeted dramatically. Today, the crime rate is half of what it
was at its peak in 1991. Despite recent troubling headlines about rising murder rates in Baltimore

4 | Brennan Center for Justice

and Chicago, crime nationwide is at historic lows. Violent crime today is about where it was in
1970. Property crime is at 1967 levels.18
It is tempting to look at this data and assume that mass incarceration caused this decline in crime.
But this is not the case. Rigorous social science research based on decades of data shows that
increased incarceration played an extremely limited role in the crime decline. It finds that social
and economic factors, and to some extent policing, drove this drop.19 Though this truth is counterintuitive, it is real.
Studies from the Brookings Institute’s Hamilton Project and the National Academy of Sciences
corroborate findings from the Brennan Center and leading economists: “When the incarceration
rate is high, the marginal crime reduction gains from further increases tend to be lower, because
the offender on the margin between incarceration and an alternative sanction tends to be less
serious. In other words, the crime fighting benefits of incarceration diminish with the scale of
the prison population.”20 Although there is some relationship between increased incarceration
and lower crime, at a certain point, locking up additional people is not an effective crime control
method, especially when imprisoning one person costs $31,000 a year.21
Another body of research shows that keeping people in prison for longer periods has similar
diminishing returns.22 For example, a 2007 National Bureau of Economic Research study found
that prison stays longer than 20 months had “close to no effect” on reducing commission of certain
crimes upon release.23 Other studies show that prison often has a “criminogenic” effect, meaning
that imprisonment can actually lead people to commit more crimes after release. Prisons provide
little rehabilitative programming and often release individuals back to communities without
proper support, leaving them vulnerable and likely to turn back to crime. This is one major reason
more than half of released prisoners are back behind bars within three years.24
Recent reforms enacted by states show that mass incarceration and crime are not inextricably linked.
Over the last decade, 27 states have reduced both imprisonment and crime together.25 From 1999 to
2012, New Jersey and New York reduced their prison populations by about 30 percent, while crime
fell faster than it did nationally. Texas decreased imprisonment and crime by more than 20 percent
during the same period. California, in part because of a court order, cut its prison population by 27
percent, and violence in the state also fell more than the national average.26
It has become clear that it is unnecessary to warehouse so many prisoners for such long periods of
time. Yet, last year, half of all offenders in federal prison were sentenced to more than 10 years.27
And in the states, lengths of prison stays have increased by 33 percent from 1993 to 2009.28
The current sentencing regime was largely a knee-jerk reaction to crime, not grounded in any
scientific rationale. While it may have seemed like a reasonable approach to protect the public, a
comprehensive examination of the data proves it is ineffective at that task.29 Worse yet, it is also
inequitable, placing a disproportionate burden on communities of color.30

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 5

Today’s public policy is dislocated from public reality. It is time to change our laws to reflect
and serve our country’s current public safety needs, rather than govern based on mistakes from
the past.

This Report’s Purpose
While mass incarceration has emerged as an urgent national issue to be addressed, the reforms currently
offered are dwarfed by the scale of the problem. The country needs bolder solutions.
How can we significantly cut the prison population while still keeping the country safe? This report
puts forth one answer to that question. Our path forward is not offered as the only answer or as an
absolute. Rather, it is meant to provide a starting point for a broader discussion about how the country
can rethink and revamp the outdated sentencing edifice of the last four decades.
This report is the product of three years of research conducted by one of the nation’s leading
criminologists, experienced criminal justice lawyers, and statistical researchers. First, we conducted an
in-depth examination of the federal and state criminal codes, as well as the convictions and sentences
of the nationwide prison population (1.46 million prisonersi serving time for 370 different crime
categories) to estimate how many people are currently incarcerated without a sufficient public safety
rationale. We find that alternatives to incarceration are more effective and just penalties for many
lower-level crimes. We also find that prison sentences can safely be shortened for a discrete set of more
serious crimes.
Second, based on these findings, we propose a new, alternative framework for sentencing grounded in
the science of public safety and rehabilitation.
Many have argued that regimented sentencing laws should be eliminated and replaced with broad
judicial discretion.31 Others counter that this would reinstate a system wherein judges are free to
deliver vastly divergent sentences for the same crime, potentially exacerbating racial disparities and
perpetuating the tradition of harsh sentences.32
This report proposes a new solution, building on these past proposals. We advocate that today’s
sentencing laws should change to provide default sentences that are proportional to the specific crime
i Th
 is number represents the total prisoners in this report’s dataset, which includes 94 percent of state and federal
prisoners. The total U.S. prison population is 1.56 million. We were unable to secure detailed data from New Mexico
and Vermont, and other sources did not cover every single state prisoner. For more information, see Appendix
A: Methodology. We originally intended to include an analysis of jail inmates. Due to difficulty in obtaining
detailed data on jail inmates, we were unable to complete that analysis and hope to do so in a subsequent report.
For the purpose of this report, we assumed that all current criminal laws remained in force. We only examined
sentencing changes. An analysis of whether some crimes may be best treated as civil infractions or become legal
conduct, thereby avoiding criminalization altogether, is a fruitful avenue for future research.

6 | Brennan Center for Justice

committed and in line with social science research, instead of based on conjecture. These defaults
should mandate sentences of alternatives to incarceration for lower-level crimes. For some other crimes
that warrant incarceration, they should mandate shorter sentences. Judges should have discretion to
depart from these defaults in special circumstances, such as a defendant’s criminal history, mental
health or addiction issues, or specifics of the crime committed. This approach is grounded in the
premise that the first principle of 21st century sentencing should be to protect public safety, and that
sentences should levy the most effective, proportional, and cost-efficient sanction to achieve that goal.
It aims to create more uniform sentences and reduce disparities, while preserving judicial discretion
when needed.
Our proposed sentencing defaults for each crime weigh four factors:
•	 Seriousness: Murder, for instance, should be treated as a far graver crime than writing
a bad check.
•	 Victim Impact: If a person has been harmed in the commission of a crime, especially
physically, weight toward a more serious sentence.
•	 Intent: If the actor knowingly and deliberately violated the law, a more severe sanction
may be appropriate.
•	 Recidivism: Those more likely to reoffend may need more intervention.
Our findings and recommendations, determined by applying the four factors above to the prison population,
are detailed below. (The rationale for these factors and our full methodology is described in Appendix A.)

Our Findings
As depicted in Figure 1, this report finds the following:
•	 Of the 1.46 million state and federal prisoners, an estimated 39 percent (approximately
576,000 people) are incarcerated with little public safety rationale. They could be more
appropriately sentenced to an alternative to prison or a shorter prison stay, with limited
impact on public safety. If these prisoners were released, it would result in cost savings of
nearly $20 billion per year, and almost $200 billion over 10 years. This sum is enough to
employ 270,000 new police officers, 360,000 probation officers, or 327,000 school teachers.
It is greater than the annual budgets of the United States Departments of Commerce and
Labor combined.33
•	 Alternatives to prison are likely more effective sentences for an estimated 364,000 lower-level
offenders — about 25 percent of the current prison population. Research shows that prison
does little to rehabilitate and can increase recidivism in such cases. Treatment, community

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 7

service, or probation are more effective. For example, of the nearly 66,000 prisoners whose
most severe crime is drug possession, the average sentence is over one year; these offenders
would be better sentenced to treatment or other alternatives.34
•	 An estimated 212,000 prisoners (14 percent of the total population) have already served
sufficiently long prison terms and could likely be released within the next year with little risk
to public safety. These prisoners are serving time for the more serious crimes that make up 58
percent of today’s prison population — aggravated assault, murder, nonviolent weapons offenses,
robbery, serious burglary, and serious drug trafficking.
•	

Approximately 79 percent of today’s prisoners suffer from either drug addiction or mental illness,
and 40 percent suffer from both.35 Alternative interventions such as treatment could be more
effective sanctions for many of these individuals.

Figure 1: Best Estimate of Americans Unnecessarily Incarcerated
1,600,000
1,400,000
1,463,000

364,000

1,200,000
1,000,000

212,000

800,000

Alternatives to Prison
More Appropriate
Appropriate Sentence
Already Served

600,000
887,000
400,000
200,000

Current Prison Population

Source: Brennan Center Analysis.36

8 | Brennan Center for Justice

Report Findings

Imprisonment More
Appropriate

Figure 2: Current and Recommended Prison Population, by Crime Category
Current Prison Population

Report Proposal

14%

22%

12%
14%

46%

59%

15%
18%

1.46 Million Prisoners

Violent

Violent

Property

Property

Drug

Drug

Other

887,000 Prisoners

Other

Source: Brennan Center Analysis.37

Recommendations
Based on these findings, this report issues the following recommendations to safely reduce the prison
population. As shown in Figures 1 and 2, these recommendations will decrease the total prison population
but ensure that those who have committed the most serious crimes remain behind bars. The majority of
prisoners remaining in the new system would be violent offenders (59 percent), up from less than half in
the current system (46 percent).ii
•	

Eliminate Prison for Lower-Level Crimes Barring Exceptional Circumstances: State
legislatures and Congress should change sentencing laws to mandate alternatives to prison as the
default sentences for certain lower-level crimes. These include drug possession, lesser burglary,
minor drug trafficking, minor fraud or forgery, minor theft, and simple assault — offenses
that now account for 25 percent of the prison population. Alternative sanctions — such as
community service, electronic monitoring, probation, restitution, or treatment — should be
the default for such crimes instead. Judges should have flexibility to depart and impose a prison
sentence if certain enumerated factors are present — for example, repeat serious offenses or
heinous circumstances of the crime.

ii Th
 is report separates crime into categories — drug, property, violent, other — consistent with the National
Corrections Reporting System, so that this analysis can be compared to other national datasets. Some advocates
have noted that the violent crime category may be overly broad, sweeping in crimes that do not actually include
perpetrate violence. Ideally, these categories would better reflect the acts committed. See generally Justice Policy
Inst., Defining Violence: Reducing Incarceration by Rethinking America’s Approach to Violence (2016), http://
www.justicepolicy.org/uploads/justicepolicy/documents/jpi_definingviolence_final_report.pdf.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 9

•	 Reduce Sentence Minimums and Maximums by Law: State and federal legislatures should
reduce the current minimums and maximums prison stays set by laws, or guidelines. These
ranges should be proportional to the crimes committed, with judges retaining discretion
to depart when appropriate. We recommend that legislators consider a 25 percent cut as a
starting point to determine how to reduce sentences for the six major crimes that make up
the bulk of the current prison population: aggravated assault, murder, nonviolent weapons
offense, robbery, serious burglary, and serious drug trafficking. Sentences would be shorter,
but still substantial. For example, the average inmate convicted of robbery now serves 4.2
years. A 25 percent cut would reduce the prison stay to 3.1 years. A similar analysis can be
applied to other crimes for which prison may be warranted to determine whether sentences
can be safely shortened.
•	

Retroactively Apply Reforms: Current inmates should be permitted to petition judges for
retroactive application of the two reforms above, on a case-by-case basis. This would allow for
safe release of prisoners whose sentences no longer serve a justifiable public safety purpose.

•	

Complementary Recommendations: Prosecutors should use their discretion to seek
alternatives to incarceration or shorter prison stays in line with the recommendations
of this report. Further, the nearly $200 billion in savings from implementing this report’s
recommendations can be reinvested in proven crime prevention tactics and in alternatives to
incarceration proven to reduce recidivism.

While the first steps many states have taken toward prison reform are welcome, they have not gone far
enough. It took roughly four decades to build mass incarceration. Yet, at current rates of decline, it will
take even longer to undo it.
This report provides evidence-based findings and puts forth one approach to rethink sentencing
that will reduce the disproportionate impact on communities of color, while maintaining hard-won
gains in public safety and saving cash-strapped states significant sums. Our goal is to jump-start a
conversation about how the country can implement specific reforms that are audacious enough to
truly end mass incarceration.
Inimai Chettiar is the director of the Justice Program at the Brennan Center for Justice. She directed the
research team conducting this report.

10 | Brennan Center for Justice

I. CURRENT PRISON POPULATION
From 1950 to 1972, the prison population grew by 18 percent, as shown in Figure 3. Then, from 1972
to 2009, it grew by 700 percent, reaching a peak of 1.6 million. This expansion was driven largely by
overly punitive policies enacted beginning in the 1970s and continuing through the 1990s, such as
higher new mandatory minimum sentencing laws, “truth-in-sentencing” laws, and “three-strike” laws
that imposed automatic life terms on repeat offenders, and an expansion of criminal codes.38 Since
2009, due to reforms enacted by state legislatures, the prison population has declined marginally, by 3
percent.39

Figure 3: State and Federal Prison Population (1950-2014)
1,800,000
1,600,000

Prison Population

1,400,000
1,200,000
1,000,000
800,000
600,000
400,000
200,000

19
50
19
52
19
54
19
56
19
58
19
60
19
62
19
64
19
66
19
68
19
70
19
72
19
74
19
76
19
78
19
80
19
82
19
84
19
86
19
88
19
90
19
92
19
94
19
96
19
98
20
00
20
02
20
04
20
06
20
08
20
10
20
12
20
14

0

Source: U.S. Bureau of Justice Statistics.40 (The 2014 BJS population is slightly higher than the authors’ total
prison population for the reasons explained in footnote i).

There are more mentally ill people in the nation’s prisons that in its mental hospitals.41 As shown in
Figure 4, the authors’ estimate of recent data indicates that 79 percent of today’s prisoners suffer from
either drug addiction or mental illness, and 40 percent suffer from both. As explained in the box
“Prisons: The New Mental Hospitals,” prisons are ill-equipped to respond to these challenges, and
usually simply warehouse sick people without treating them.42

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 11

Prisons also house significant numbers of juveniles and elderly people. Thirteen percent of prisoners are
under the age of 24, and 10 percent are over the age of 55.43 Research shows that human brains do not
reach full maturity until the mid-twenties.44 Research also indicates that prisoners over 50 are highly
unlikely to commit another crime if released.45

Figure 4: Substance Abuse & Mental Health Issues, State & Federal Prisoners (2014)

20.6%

24.8%

Substance Abuse Issue
Mental Health Issue

14.7%
39.9%

Both
None Documented

Source: Brennan Center estimate analyzing data from U.S. Bureau of Justice Statistics (2014).46

12 | Brennan Center for Justice

PRISONS: THE NEW MENTAL HOSPITALS
Pennsylvania Secretary of Corrections John Wetzel warns, “America is turning our correctional
system into a mental health system nationally, and . . . that’s a poor policy.”47 According to a
recent report, “The number of individuals with serious mental illness in prisons and jails now
exceeds the number in state psychiatric hospitals tenfold.”48 Worse, when those who suffer from
mental illness have completed their sentences, prison officials often release them with no services
or medical care. People whose illnesses may have worsened while detained, and who may no
longer have access to needed medications, confront a reentry process daunting for even those
well-prepared.49
Historians and advocates have noted the sharp rise in mentally ill prisoners following
“deinstitutionalization,” the shift away from inpatient mental health institutions to outpatient
care in the 1960s, compounded by policies in the 1980s.50 This shift left many former patients
with no care at all, with many ending up homeless.51 Between 1955 and 2012, the populations
in state and county mental hospitals dropped from approximately 559,000 to 35,000. During
that same period, the mentally ill population in jails and prisons expanded, reaching 356,268
people with severe mental illness. Today, about 15 percent of men and 30 percent of women
in prisons and jails have a severe mental illness such as schizophrenia or major depression,
compared to less than 5 percent of the general population.52
Along with the mentally ill, prisons have swept in thousands of people with drug addiction.
Our prisons have become ill-equipped hospitals. Such illnesses are extremely difficult to treat
in prison.53 Prisons do not have the best tools to treat mentally ill people, and even the bare
minimum treatment they receive in prison is more expensive and less effective than treatment in
the community or in mental health institutions.54
The rise of mass incarceration has exacerbated the problem, not only taxing limited mental
health treatment resources in prisons, but also exposing inmates with mental illness to traumatic
prison conditions. Mentally ill prisoners face the challenges of overcrowding, poor medical care,
isolation, violence, and sexual victimization. A recent study of the jail population showed that
suicide has been the leading cause of death since 2000, causing 34 percent of all deaths in 2013.55
Our prisons have become warehouses for people whom we don’t know how to help. Proven
alternative interventions, such as probation with intensive treatment, may yield better results for
these individuals.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 13

Two additional demographic trends are worth noting. First, the incarceration of women has accelerated
compared to past decades. The female imprisonment rate is six times higher than it was in 1980, as
shown in Figure 5. Second, the imprisonment rate of African-Americans is 5.5 times higher than that
of white Americans. As shown in Figure 6, though disparities have decreased marginally (by 1.8 percent
per year since 2000), it would take 100 years for the disparity to disappear at the current trend. 

Imprisonment Rate in 1980 Terms

Figure 5: Imprisonment by Gender (1980-2014)
7
6
5
4
3
2
Male Imprisonment Rate

1

Female Imprisonment Rate

14
20

12
20

10
20

20

08

06
20

02

04
20

20

00
20

98
19

96

94

19

19

92
19

90

88

19

19

19

86

84
19

82
19

19

80

0

Source: U.S. Bureau of Justice Statistics.56

Figure 6: Imprisonment by Race (2000-2014)
2,000
1,800

Imprisonment Rate

1,400
1,200
1,200

Black Imprisonment Rate

1,000

White Imprisonment Rate

800
600
400
200

Source: U.S. Bureau of Justice Statistics.57
14 | Brennan Center for Justice

14
20

13
20

12
20

11
20

10
20

09
20

08
20

07
20

06
20

05
20

04
20

03
20

02
20

01
20

20

00

0

And these booming incarceration rates have made the United States a distant outlier when compared
to other democracies.

Incarceration Rate per 100,000

Figure 7: Incarceration Rates for High Income Democracies (2016)

693
572

n

47

pa

en
ed

d

53

Sw

an

k
Fi

nl

ar

y

m

an

55

Ja

61

D

G

Sw

m

d

78

er

lan

er

Ita

ly

84

en

89

itz

um

ce

lgi

a

98

Be

an

re

103

Fr

da

107

Ko

na
So

ut

h

ain

114

Ca

l
ga

m

131

Sp

rtu

Po

do

d
lan

137

U

ni

ted

Ki

ng

n

Po

tio

da

en
m

m

co

144

Re

po

rt

Re

U

ni

ted

St

at

es

187

Source: World Prison Brief (2016) & Brennan Center Analysis.58

As Figure 7 shows, the United States has an incarceration rate (prisoners plus jail inmates) nearly four
times greater than Poland, the developed democracy with the second-highest rate. Even implementing
the recommendations of this report (explained in Sections III and IV below) would leave the country’s
incarceration rate more than double the rate of the democracy with the second highest incarceration
rate. This is not meant to imply that a rate of 572 prisoners per 100,000 people is an optimal rate of
incarceration. Rather, it merely depicts what the country’s incarceration rate would be after applying
this report’s recommendations.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 15

II. TIME SERVED IN PRISON TODAY
As the number of people in prison has increased, so has the time they spend there. The 1990s and
2000s saw states of all demographic and political compositions pass laws aimed at keeping people in
prison longer. These laws targeted both “front-end” policies, which increased the length of sentences
handed down (such as higher statutory sentence ranges, mandatory minimums, and “three-strikes”
laws), as well as “back-end” policies, which caused inmates to serve more time (such as requirements
that inmates serve 85 percent of their sentence behind bars, life without parole sentences, limits on the
ability to earn time toward early release, and restrictions on parole boards).59
As shown in Figure 8, between 1993 and 2009, the average prison stay for state inmates increased by
33 percent. While the increase in prison stays was most dramatic for violent and public order crimes,
prison stays also increased 18 percent for property crimes and 25 percent for drug crimes from 1993 to
2009.60 Following the same trend, the average stay for federal prisoners more than doubled from 1988
to 2012, rising from 1.5 to 3.1 years.61

Figure 8: Growth in Time Served By State Prisoners (1993-2009)
70%
62%

Percent Increase in Prison Stay

60%
50%
39%

40%
33%
30%

25%
18%

20%
10%
0%
Total

Violent

Property

Drug

Public Order

Source: National Corrections Reporting Program (The last comprehensive collection on lengths of stay was
conducted in 2009).62

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 17

Figure 9 provides a more detailed look, examining current average prison stays for six serious crimes that
make up 58 percent of the prison population.63 Notably, many of the laws and policies that increased
time served are not based on evidence showing that longer stays behind bars enhanced public safety.
Instead, they are the result of decades of severe sentencing laws, guidelines, and practices premised on
a collective “guess” that more punishment and more prison would reduce crime.64

Figure 9: Current Prison Stays for Major Crimes
14
12

11.7

Years

10
8
6
4.2
4

3.4
3.0

3.0
1.7

2
0
Murder
(165,000)

Robbery
(176,000)

Serious Drug
Trafficking
(165,000)

Weapons
(76,000)

Aggravated
Assault
(155,000)

Serious Burglary
(111,000)

Source: Brennan Center Analysis.65
Driven by advances in social science, states are increasingly turning toward risk assessment tools to
help decide how much time people should spend behind bars. These tools use data to predict whether
an individual has a sufficiently low likelihood of committing an additional crime to justify a shorter
sentence or an alternative to incarceration. They take into account various factors such as: criminal
history, drug and alcohol abuse, employment history and education, connection to others with criminal
histories, mental health, and financial well-being.
Some courts have implemented risk assessments to determine whether defendants should be held in jail
or released while waiting for trial; similarly, some parole boards use them to decide which prisoners to
release. States such as Kentucky and Virginia have implemented the former, while Arkansas and Nevada
have implemented the latter. More recently, states are applying risk assessments to guide sentencing
decisions.66 The first state to incorporate such an instrument in sentencing was Virginia in 1994. By

18 | Brennan Center for Justice

2004, the state implemented risk assessments statewide, requesting judges to consider the results in
individual sentencing decisions.67 Courts in at least 20 states have begun to experiment with using risk
assessments in some way during sentencing decisions.68
Some studies have shown these tools are effective at reducing incarceration while reducing the risk of
releasing people who may go on to commit new crimes.69 Other studies have shown that they produce
unjust racial disparities.70 Specifically, these tools rely on factors, such as educational and employment
history, that are affected more by structural inequities disproportionately affecting racial and ethnic
minorities instead of by individual factors. As these tools are implemented more widely, there will be
more evidence of their results. Because these instruments do not change existing sentencing laws, which
the authors believe are a root cause of overly long sentences, this report does not delve further into the
use of risk assessment in sentencing.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 19

III. ENDING PRISON FOR LOWER-LEVEL CRIMES
A. Research On Prison And Crime
The prison population exploded as crime rates began a dramatic decline. Today, the crime rate is half
of what it was at its height in 1991. Violent crime has fallen by 51 percent since 1991, and property
crime by 52 percent.71
One might conclude that the increase in incarceration at the same time as the crime decline indicates
that one trend is responsible for the other. But studies now show that incarceration has long since
passed the point where locking up additional prisoners would have a pronounced effect on reducing
crime. As shown in Figure 10, starting in about 2000, increased imprisonment had a limited effect on
the crime rate. The diminishing returns of incarceration are becoming more commonly accepted.72

Figure 10: Increased Imprisonment’s Effect on Crime (1980-2012)

Percent Decrease in Crime Rate Caused by
Percent Increase in Imprisonment

14%
12%
10%
8%
6%

Upper bound

4%

Best Estimate

2%

Lower bound

0%
-2%
-4%

12

10

20

08

20

06

20

04

20

02

20

00

20

98

20

96

19

94

19

92

19

90

19

88

19

86

19

84

19

82

19

19

19

80

-6%

Source: What Caused the Crime Decline? (Brennan Center 2015, analyzing data through 2012).73
What may account for this counterproductive effect of incarceration?
One body of research shows that prison may make some people more likely to commit crimes after
release. Criminologists call this the “criminogenic” effect of prison. It is particularly powerful on lowerlevel offenders. Once individuals enter prison, they are surrounded by other prisoners who have often
committed more serious or violent offenses. While in prison, they are generally not provided rehabilitative

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 21

programming, treatment, or any job or educational training to ensure success upon release. They are also cut
off from support networks and life and employment skills deteriorate.74 As a result, after release prisoners
often have trouble finding employment and reintegrating into society, often leading them to turn to crime.75
It is therefore unsurprising that the national recidivism rate for former prisoners hovers around 50 percent
— meaning that half are reincarcerated within three years of release.76
Studies on this effect abound. A September 2016 New York Times editorial, citing research published by the
Laura and John Arnold Foundation, concluded that unnecessary time spent behind bars “harms defendants
and may actually increase crime.”77 A 2002 study of more than 1,000 felony offenders in Kansas City,
Missouri, indicated that using prison instead of probation for lower-level drug offenders increased their
likelihood of committing crimes upon release.78 Similarly, a 2012 study of over 17,000 Florida inmates found
that prison, when compared to probation, resulted in higher recidivism for drug and property criminals.
Interestingly, it found that prison and probation were equally effective for violent offenders.79 That same year,
a study by criminologist William Bales of 79,000 prisoners found that those sentenced to prison were 10
percent more likely to reoffend than those put on “house arrest” with mandatory treatment.80
A second series of studies shows that alternatives to incarceration promote public safety more effectively
than incarceration, especially for lower-level offenders. In 1986, the Rand Corporation found that those
sentenced to probation were less likely to reoffend than those sentenced to prison. This difference was
particularly pronounced for property offenders. A 1999 report by the Centre for Criminal Justice Studies
showed that those who received probation or community service had lower recidivism rates than those who
served prison sentences. Reviewing nearly 100 studies on drug treatment and recidivism, a 2006 study on
violent and nonviolent offenders by the Washington State Institute of Public Policy concluded that, on
average, treatment-oriented probation reduced recidivism by 12 percent, while in-prison treatment reduced
recidivism by only 6 percent.81
Finally, other recent research shows no difference in the effect on recidivism reduction of incarceration
compared to alternative sanctions. A 2002 Vera Institute of Justice study found that offenders in New
York sentenced to alternatives to incarceration programs that provide counseling and treatment in lieu of
incarceration “were no more likely to be convicted of new offenses over the three years than the comparison
group of offenders who spent much more of that time incarcerated.”82 A 2009 analysis by criminologists of
55 empirical studies concluded that prison either increased reoffending or did not affect it, compared to
alternatives to prison.83 A 2011 follow-up study reached a more definite conclusion: “With some confidence,
we can conclude that, across all offenders, prisons do not have a specific deterrent effect. Custodial sentences
[jail and prison] do not reduce recidivism more than noncustodial sanctions.”84 A 2012 study in the Journal
of Quantitative Criminology found no benefit to incarceration compared with noncustodial alternatives
for violent and nonviolent offenders, declaring the result “quite consistent with an emerging body of work”
showing “little persuasive evidence that incarceration reduces future criminality.”85 It pointed to three studies
conducted from 2007 to 2011 on thousands of offenders in Chicago and Washington, D.C.86
This evidence conclusively weighs in favor of expanding the use of alternatives to incarceration as a
more effective crime reduction technique, especially for lower-level crimes.

22 | Brennan Center for Justice

EFFECTIVE ALTERNATIVES TO INCARCERATION
Not incarcerating people convicted of lower-level crimes does not mean they escape consequences.
Alternative sanctions exist that are cost-effective and better reduce recidivism. Notably, 75 percent
of crime victims prefer to hold offenders accountable through alternatives beyond incarceration.87
•	

 robation: Individuals are supervised by probation officers while continuing work or school.
P
The supervision aims to prevent probationers from engaging in criminal activity. Probation
conditions may include: anger management counseling, drug or alcohol treatment, drug
tests, or regular check-ins with an officer. Studies show that supervision reduces recidivism
by as much as 34 percent, at an average cost of $3.42 per day, less than one-twentieth of the
average cost of a day in prison ($79).88

•	

 lectronic Monitoring: Electronic devices, usually placed on the ankle, allow law enforcement
E
to track offenders’ whereabouts, ensuring compliance with location restrictions. Electronic
monitors have increased in the last decade from 50,000 to 125,000 active devices.89 Some
studies found that more research is needed to determine how well the technology ensures
returns to court, and others have contended that monitors can interfere with day-to-day
functioning of individuals.90 However, electronic monitoring is far less expensive than
incarceration and can reduce rearrest rates by as much as 25 percent.91

•	

Treatment: This includes drug or alcohol abuse treatment, mental health treatment, and
other rehabilitative programs. Treatment costs less than incarceration and reduces recidivism
by addressing the core problem (the illness) instead of the symptom (the crime). A 2013
Vera Report showed that mental health treatment is less expensive and more effective than
incarceration.92 Successful drug treatment programs include Drug Treatment Alternatives to
Prison (“DTAP”) in New York, New York, Law Enforcement Assisted Diversion in Seattle,
Washington, and Vermont’s Cognitive Self-Change Program, reducing rearrests by between
21 and 58 percent compared to prison.93 Although treatment programs require upfront costs,
they offer long-term savings from averted incarceration.

•	

 ommunity Service: This includes required work that improves the community, such
C
as cleaning a park, stocking a food bank, or working with a youth group. By connecting
offenders to their community, service may also be rehabilitative.

•	

 ines and Restitution: Fines go to the government, whereas restitution compensates victims.
F
Of course, payments must be calibrated to a defendant’s income and ability to pay.94 Restitution
may be especially appropriate for property crimes. In some countries, like Germany, where 79
percent of people who commit crimes are sanctioned with fines assessed according to ability
to pay, fines are used as stand-alone sanctions and the country has a low crime rate.95

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 23

B. Analysis
To determine whether incarceration is an effective sanction for specific crimes, this report weighed
four factors: seriousness of the crime, victim impact, intent, and likelihood of recidivism (defined as
reimprisonment for a new crime within three years). These factors reflect a social science and popular
consensus on why we incarcerate — to deliver punishment, promote public safety, and rehabilitate.
They also lean heavily in favor of protecting public safety. (See Appendix A: Methodology for more
explanation of the rationale for these factors.)
Using data on 1.46 million inmates from 48 state prisons and all federal prisons, the authors analyzed
why inmates were behind bars, and for how long.96 The authors applied the four-factor framework in
a comprehensive, crime-by-crime analysis of the 370 crimes for which all prisoners were convicted.
A detailed summary of this report’s recommendations can be found in Table 1. These recommendations
are conservative. When the circumstances are ambiguous, they err on the side of protecting public
safety. For a large subset of crimes, the report concludes that prison is likely not the most effective
intervention. In these cases, public safety may be equally served by using an alternative to incarceration.
These categories of crimes include those that do not result in serious harm to a victim or substantial
destruction of property; where malicious intent may not have been present; and/or where prison may
not serve as the most effective penalty to reduce recidivism.
Figure 11 depicts the lower-level crimes for which the authors’ analysis finds an alternative to
incarceration to be the more effective sanction:
•	 Approximately 25 percent of the national prisoner population (364,000 prisoners) would
benefit from alternatives to prison based on the four-factor analysis. (See Table 1 for more
explanation and Section III. C for examples.)
•	 Ninety-five percent of these crimes are non-violent. The only crime included that is categorized
as “violent” is simple assault, which typically involves only a verbal threat against another
person. If bodily contact is involved, a higher crime, such as aggravated assault, can be charged.
•	 The category of “other minor crimes” comprises a number of lower-level offenses, including
gambling, offenses against decency, and traffic infractions. These crimes are not violent and
usually involve little to no direct harm to others. If an offender perpetrates violence during
these crimes, they become more serious crimes that can merit incarceration.

24 | Brennan Center for Justice

Figure 11: Lower-Level Crimes Likely Warranting Alternatives to Prison
Crime Category
Drug Possession
Minor Trafficking of Serious Drugs
Minor Larceny
Other Minor Property Crimes
Minor Fraud/Forgery
Unlawful Entry and Reentry to the Country
Simple Assault
Minor Trafficking of Marijuana
Lesser Burglary
Other Minor Drug Offenses
Other Minor Crimes
(i.e. gambling, prostitution, first DUI offense)
TOTAL

Number of
Prisoners Today

Percent of Total
for Alternatives

66,000
62,000
45,000
25,000
23,000
22,000
18,000
18,000
15,000
4,000

18.1%
17.0%
12.4%
6.9%
6.3%
6.0%
4.9%
4.9%
4.1%
1.1%

66,000

18.1%

364,000

100%

Drug Possession (18.1%)
18.1%

18.1%

Minor Trafficking of Serious Drugs (17.0%)
Minor Larceny (12.4%)

1.1%

Other Minor Property Crimes (6.9%)

4.1%

Minor Fraud/Forgery (6.3%)
17.0%

4.9%

Unlawful Entry or Reentry to the Country (6.0%)
Simple Assault (4.9%)

4.9%

Minor Trafficking of Marijuana (4.9%)

6.0%
6.3%

12.4%
6.9%

Lesser Burglary (4.1%)
Other Drug (1.1.%)
Other Minor Crimes (18.1%)

Source: Brennan Center Analysis.97

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 25

Figure 12: Estimated Percentage of Americans Unnecessarily Incarcerated, by Crime Type

Estimated Percent of Unnecessarily Incarcerated

50%

47.2%

45%

41.1%

40%

40.6%

35%
30%
25%
20%
15%
10%
5%

3.1%

0%
Drug

Other

Property

Violent

Source: Brennan Center Analysis.98	
Figure 12 depicts the percent of prisoners that the authors found were incarcerated without a compelling
public safety reason according to the four-factor analysis. This includes almost half of drug offenders
and 40 percent of those convicted of property or other crimes.

C. Examples
These findings emerged from a careful examination of each category of crime. A summary of the
authors’ analysis can be found in Table 1 and is further explained in the Methodology. Examples are
provided below.
1. Trafficking of Serious Drugs (227,000 prisoners, 15.5 percent of national prison population)
In the original data collected by the authors, eight crime categories, including trafficking of serious
drugs, had broad definitions that encompassed prisoners convicted of serious activities that likely merit
incarceration and those convicted of less serious versions where alternative forms of punishment are
likely more appropriate.99 For these crime categories, the authors first used the four factors to determine
what conduct constituted a “serious” and a “minor” definition of the crime. They then used more
detailed data from a representative sample of states to estimate how many prisoners nationwide fell into
each definition. The analysis for serious trafficking and minor trafficking is below.

26 | Brennan Center for Justice

Serious Trafficking of Serious Drugs (165,000 prisoners, 72.7percent of drug trafficking prisoners, 11.3
percent of all prisoners)100
In these crimes, individuals play a managerial or high-level role in producing, importing, selling, or
otherwise supplying illegal substances drugs — other than marijuana. These include cocaine, heroin,
and methamphetamine.
•	 Seriousness: Those with higher-level roles do more to facilitate the spread of harmful drugs,
and are more likely to use violence or fraud to further their organization.
•	 Victim Impact: Serious trafficking often entails physical violence, threats, or fraud, in addition
to facilitating harmful drug use.
•	 Intent: These crimes generally require knowing or intentional action.
•	 Recidivism: The recidivism rate of prisoners released after serving time for drug trafficking is
moderate relative to other crimes. According to data from the U.S. Bureau of Justice Statistics
(BJS), nearly 25 percent were reimprisoned for a new crime within three years.101
All four factors point to incarceration as the recommended default sanction for serious trafficking.
Minor Trafficking of Serious Drugs (62,000 prisoners, 27.3 percent of drug trafficking prisoners, 4.2 percent
of all prisoners)
In these crimes, offenders play a low-level role in trafficking, such as couriers, drug “mules,” or street
dealers selling small quantities.
•	 Seriousness: Those with low-level roles contribute little to the operation and are replaceable.
They are usually motivated to commit the crime due to poverty or drug addiction.
•	 Victim Impact: Individuals play a smaller role with regard to harmful drug use than serious
drug traffickers.
•	 Intent: These crimes generally require knowing or intentional action.
•	 Recidivism: The recidivism rate is moderate relative to other crimes. Nearly 25 percent were
reimprisoned for a new crime within three years.102
Two of the four factors for minor trafficking indicate the public safety threat presented is low and that a
sanction other than incarceration may be an appropriate default. While the recidivism rate is moderate,
many of these offenders are impoverished or suffering from addiction, which is more appropriately
handled through social services and treatment.
2. Robbery (176,000 prisoners, 12.0 percent of national prison population)
Robbery is defined as theft of property by force or intimidation, in the presence of the victim.
•	 Seriousness: This is a serious crime, entailing use of force or intimidation to subdue a victim
and take his or her property.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 27

•	 Victim Impact: This crime can result in both physical and psychological trauma to victims.
•	 Intent: Intent is required to commit this crime.
•	 Recidivism: Recidivism data from BJS indicates that 25 percent of these offenders were
reimprisoned for a new crime within three years.103
While not all robberies entail physical harm, the trauma and potential for physical harm or death
may merit incarceration. Because the authors chose to err on the side of public safety even when
circumstances are questionable, this report recommends incarceration as the default sanction.
3. Murder (165,000 prisoners, 11.3 percent of national prison population)
Murder is the intentional killing of another person, or killing of a person during a felony.
•	 Seriousness: Murder is the most serious crime.
•	 Victim Impact: It not only irrevocably ends the lives of victims, it causes trauma and suffering
for the friends and family of the person killed.
•	 Intent: Murder is planned or intentional, or committed in the course of a planned or intentional
felony crime. (A killing committed under sufficient provocation, or resulting from recklessness,
is defined as manslaughter.)
•	 Recidivism: The recidivism rate is relatively low. 11 percent of those released after a murder
conviction were reimprisoned for a new crime within three years.104
Three of the four factors point to incarceration as the recommended sanction. While the recidivism rate
is relatively low, the gravity of this crime merits incarceration as a default.
4. Burglary (126,000 prisoners, 8.6 percent of national prison population)
Burglary is entering a residence or occupied structure without permission, with intent to commit any
crime inside. The crime committed inside could be completed or merely intended, including possibly
arson, damage, loitering, or property theft. (Notably, if violence is committed or intended, the crime
would escalate to attempted assault or use of deadly force).
Burglary is one of the eight categories that lumped together serious conduct along with other less
serious conduct. The median victim loss amount for this crime is $280.105 Of course, some burglaries
result in more, and others in less loss. The authors split this category into their own definitions of
“serious burglary” and “lesser burglary” as discussed below.
Serious Burglary (111,000 prisoners, 88.1 percent of burglary prisoners, 7.6 percent of all prisoners)
Serious burglary is defined as a burglary that involves either: a deadly weapon, intent to commit a
serious or violent felony, or entering a home or non-residential structure when occupants are present.

28 | Brennan Center for Justice

•	 Seriousness: These burglaries involve a violation of the home or other safe space such as an
office, regardless of whether a resident is present or not. They may involve a weapon. Due to
the sanctity of the home in common law and in society, the crime is considered serious.
•	 Victim Impact: People feel a particularly acute fear of having their homes violated and their
families threatened. These burglaries may cause trauma as well as significant property loss.
•	 Intent: The crime requires, at the very least, an intention to inflict property damage upon others.
•	 Recidivism: Recidivism data from BJS indicates that 31 percent of burglary offenders were
reimprisoned for another crime within three years.106
Although many would argue that if no violence was perpetrated incarceration is not appropriate, the
authors erred on the side of public safety even in equivocal circumstances. Given the societal value
placed on the sanctity of the home, they recommend incarceration as the default sanction.
Lesser Burglary (15,000 prisoners, 11.9 percent of burglary prisoners, 1 percent of all prisoners)
“Lesser burglaries” are burglaries of abandoned or unoccupied structures when occupants are not present.
•	 Seriousness: These burglaries occur on abandoned or unoccupied structures, but may result in
property loss or damage. They do not involve homes, residential buildings, or office buildings.
•	 Victim Impact: These crimes do not involve contact with people directly, and therefore may
cause less trauma.
•	 Intent: Intent to inflict property damage on others is required.
•	 Recidivism: Recidivism data from BJS indicates that 31 percent of burglary offenders were
reimprisoned for another crime within three years.107
Three of the four factors indicate the public safety threat these prisoners present is low and that a default
sanction other than incarceration may be appropriate. While the recidivism rate is relatively high,
rehabilitation may be more likely achieved with an alternative sanction given the potential criminogenic
effect of prison for these types of lower-level crimes.
5. Drug Possession (66,000 prisoners, 4.5 percent of national prison population)
Drug possession involves having one or more specifically enumerated controlled substances, such as
cocaine, heroin, methamphetamine, or marijuana.
•	 Seriousness: These are low-level crimes committed by people who often suffer from substance
abuse problems. They generally possess these illegal drugs for their own personal use not for
sale. Any sale or distribution (or intent to do so) escalates to a different crime captured under
the drug trafficking category.
•	 Victim Impact: The primary harm is self-inflicted. However, purchase of an illegal substance provides
at least marginal support for a criminal enterprise. Offenders also sometimes commit violence or
other crimes while under the influence, or commit property crimes to support their drug purchases.
But if these additional crimes are committed, they can be charged as separate crimes.
•	 Intent: The intent is usually to possess the drugs for personal use, not to harm another.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 29

•	 Recidivism: Recidivism data from BJS indicates that 24 percent were imprisoned for another
crime within three years.108
Three of the four factors indicate the public safety threat these prisoners present is low and that an
alternative to incarceration may be appropriate. While the recidivism rate is moderate, a review of many
studies on drug treatment and recidivism concluded that on average, treatment reduced recidivism
more than prison.109 For these reasons, the authors recommend a default sanction of alternatives to
incarceration, such as treatment or probation.

DEPARTING FROM RECOMMENDED SANCTIONS
This report recommends that laws set default sentences for crimes of either imprisonment or
alternatives to incarceration, as detailed in Table 1. This is one way to restore proportionality
to sentencing while ensuring that those who commit similar offenses are treated similarly. But
these defaults are not absolute. Judges should retain discretion to make upward or downward
departures in special circumstances such as:
•	 Mental Illness, Drug Addiction, or Other Vulnerability: When an offender suffers
from mental illness, substance addiction, or history of victimization, an alternative
sanction may be merited. In these cases, a defendant is more likely to be rehabilitated,
refrain from recidivating, and become productive in society if offered treatment or
another intervention instead of imprisonment. If the offender is a child or an elderly
person, a departure may also be warranted.
•	 Pleading Down: Where an initial charge is for a more serious crime, but a defendant
admits to, or pleads, to a lesser crime and it is clear that the defendant committed the
underlying conduct, an upward departure may be appropriate.
•	 Criminal History: When a defendant has committed the same or other crimes in
the past, imprisonment may be merited even for a lower-level crime.
•	 Aggravating Circumstances: Where specific circumstances, such as possession of
gun or particularly egregious behavior, make a crime more serious than is evident
from the crime itself, an upward departure may be warranted.
•	 Remorse: A remorseful defendant, who may be less likely to commit another crime
and more receptive to rehabilitation, may merit a lesser sanction.

30 | Brennan Center for Justice

Prison
Recommended

Average Years
Served Now

Recommended
Years Served

176,000 12.0%

No

Yes

4.2

3.1

Theft by force or intimidation, in the presence of the
victim. Can result in severe physical and psychological
trauma. Incarceration is the recommended default
sanction, but the authors believe time served can be
safely reduced; an approximate 25 percent cut should be
considered.

Murder

165,000 11.3%

No

Yes

11.7

8.8

The intentional killing of another, or killing a person
during a felony. Incarceration is recommended, but the
authors believe time served can be safely reduced; an
approximate 25 percent cut should be considered.

Aggravated
Assault

155,000 10.6%

No

Yes

3.0

2.3

Knowingly or extremely recklessly causing serious injury
to another. Because it causes severe physical injury,
incarceration is the recommended default sanction, but
the authors believe time served can be safely reduced; an
approximate 25 percent cut should be considered.110

Rape

50,000

3.4%

No

Yes

7.4

N/A**

Sexual intercourse forced on a non-consenting victim.
Because it causes severe physical and psychological
trauma, incarceration is recommended.

Manslaughter

47,000

3.2%

No

Yes

5.6

N/A**

Unplanned killing. Incarceration is recommended.

Lewd Act
with a Child

26,000

1.8%

No

Yes

3.6

N/A**

Unlawful sexual act committed with a child. Because this
results in physical and psychological trauma to children,
incarceration is recommended.

Simple
Assault

18,000

1.2%

Yes

No

1.4

N/A

Knowingly or recklessly causing minor injury to another,
or causing fear of injury. Because injuries are often minor,
and the rate of reoffending is low, alternatives such as
community service or probation are the recommended
default sanction.

Serious
Statutory
Rape

3,000*

0.2%*

No

Yes

3.3

N/A**

Defined as sex with someone much younger or where
consent is ambiguous. Because this entails an abuse of
power and likely psychological harm, incarceration is
recommended.

Other
Statutory
Rape

3,000*

0.2%*

Yes

No

3.3

N/A

Defined as sex with someone younger but close in age
with clear consent. This could include a high school
senior having intercourse with another senior or a junior,
with clear consent. Alternatives such as counseling
or probation are the recommended default sanction,
especially if both parties are young adults.

Other Serious
Violent
Crimes

29,000

2.0%

No

Yes

N/A

N/A**

Examples include: assault on police, child abuse, and
kidnapping. Because these crimes may cause severe
physical and psychological trauma or may impede law
enforcement’s work, incarceration is the recommended
default sanction.

3.1%

96.9%

N/A

N/A

N/A

% of Total

Robbery

Crime
Category

# of Prisoners

Alternatives
Recommended

Table 1: Findings and Recommendations, State and Federal Prisoners

Explanation

Violent

Total Violent

672,000 45.9%

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 31

Alternatives
Recommended

Prison
Recommended

Average Years
Served Now

Recommended
Years Served

No

Yes

1.7*

1.3

Defined as entering a residence or occupied structure
without permission, with intent to commit a crime
inside, involving a deadly weapon or intent to commit
a serious or violent felony – regardless of whether
occupants are present. Given the sanctity of the home
reflected in American law and culture, incarceration
is recommended as the default sanction. The authors
believe time served can be safely reduced; an approximate
25 percent cut should be considered.

Minor
Larceny

45,000*

3.1%*

Yes

No

1.1*

N/A

Taking another’s property worth less than $10,000
without consent. Because it results only in property loss,
alternatives such as fines, restitution, and community
service are the recommended default sanction.

Minor Fraud
or Forgery

23,000*

1.6%*

Yes

No

1.1*

N/A

A knowing misrepresentation causing financial or material
loss less than $10,000. Because it results only in property
loss, alternatives such as fines, restitution, and community
service are the recommended default sanction.

Serious Fraud
or Forgery

19,000*

1.3%*

No

Yes

1.1*

N/A**

Defined as a knowing misrepresentation causing financial
or material loss of $10,000 or more. This includes “white
collar” crimes like securities fraud, money laundering, or tax
fraud. Because fraud of such significant sums (20 percent
of average annual household income) may substantially
diminish a victim’s wealth and cause mental anguish,
incarceration is the recommended default sanction.

Serious
Larceny

19,000*

1.3%*

No

Yes

1.2*

N/A**

Defined as theft of $10,000 or more. Because larceny of
over $10,000 (20 percent of average annual household
income) may substantially diminish a victim’s wealth and
cause mental anguish, incarceration is the recommended
default sanction.

Lesser
Burglary

15,000*

1.0%*

Yes

No

1.7*

N/A

Defined as entering an abandoned structure with intent
to commit any crime inside. The crime committed could
be completed or merely intended. Because these burglaries
are less likely to result in a physical altercation and do
not involve occupants, alternatives such as probation,
restitution, or community service are the recommended
default sanction.

Arson

6,000

0.4%

No

Yes

2.6

N/A**

Setting a fire to purposely destroy another’s property.
Because arson causes severe property damage, and
can cause severe injury or death, incarceration is the
recommended default sanction.

Other Serious
Property
Crimes

3,000

0.2%

No

Yes

N/A

N/A**

Examples include blackmail and embezzlement. Because
these crimes diminish victims’ wealth, cause mental
anguish, or threaten the integrity of political and
economic institutions, incarceration is the recommended
default sanction.

Other Minor
Property

25,000

1.7%

Yes

No

N/A

N/A

Examples include receiving stolen property and
trespassing. Because these crimes result only in minor
property loss, alternatives such as fines, restitution, and
community service are the recommended default sanction.

40.6%

59.4%

N/A

N/A

N/A

% of Total

111,000* 7.6%*

# of Prisoners

Serious
Burglary

Crime
Category

Explanation

Property

Total
Property

266,000 18.2%

32 | Brennan Center for Justice

Alternatives
Recommended

Prison
Recommended

Average Years
Served Now

Recommended
Years Served

Yes

3.4*

2.6

Defined as having a managerial or high-level role in
producing, importing, selling, or supplying illegal
substances other than marijuana (e.g. cocaine, heroin,
and methamphetamine). These offenders contribute
more to the distribution of harmful substances, and
are more likely to commit violence to further the
organization’s interest. Incarceration is the recommended
default sanction, but the authors believe time served can
be safely reduced; an approximate 25 percent cut should
be considered.

Drug
Possession

66,000

4.5%

Yes

No

1.1

N/A

Because many users suffer from addiction or mental
health problems, and this crime generally inflicts harm
primarily on the offender, alternatives such as treatment
or probation are the recommended default sanction.

Minor
Trafficking of
Serious Drugs

62,000*

4.2%*

Yes

No

3.4*

N/A

Defined as lower-level roles in trafficking, such as street
dealers of small amounts or acting as a courier. Because
these offenders do less to facilitate the spread of harmful
drugs, and are often motivated by addiction or poverty,
alternatives such as probation, treatment, or fines are the
recommended default sanction.

Minor
Trafficking of
Marijuana

18,000*

1.2%*

Yes

No

1.3*

N/A

Defined as lower-level roles in selling marijuana, such
as street deals of small amounts or acting as a courier.
Research shows that marijuana is less harmful than other
more serious drugs, and popular sentiment has moved
toward favoring legalizing it.111 Given these two factors,
alternatives such as probation, treatment, or fines are the
recommended default sanction.

Serious
Trafficking of
Marijuana

3,000*

0.2%*

No

Yes

1.3*

N/A**

Defined as having a managerial or high-level role in
producing, importing, selling, or supplying marijuana.
Because these offenders facilitate widespread drug trade
for their own financial gain, and may have motive
to commit fraud and violence, incarceration is the
recommended default sanction.

Other Minor
Drug Offenses

4,000

0.3%

Yes

No

N/A

N/A

Includes crimes such as possessing drug paraphernalia.
Because they do not involve large scale trafficking,
alternatives such as probation, fines, or treatment are the
recommended default sanction.

47.2%

52.8%

N/A

N/A

N/A

% of Total

No

# of Prisoners

Serious
165,000* 11.3%*
Trafficking of
Serious Drugs

Crime
Category

Explanation

DRUG

Total Drug

318,000 21.7%

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 33

Alternatives
Recommended

Prison
Recommended

Average Years
Served Now

Recommended
Years Served

5.2%

No

Yes

3.0

2.3

Includes weapons possession, negligent discharge, and
other weapons uses that are not directed at a person or
aimed at hurting a person. Use of a deadly weapon is a
separate crime category. These crimes facilitate violence
by expanding access to and hindering effective regulation
of deadly weapons. Incarceration is recommended, but
the authors believe time served can be safely reduced; an
approximate 25 percent cut should be considered.

DUI —
Multiple

27,000*

1.8%*

No

Yes

1.0*

N/A**

“Driving under the influence” involves driving a vehicle
while beyond the legally permitted level of intoxication
with one or more prior conviction for the same. DUI
offenses pose a risk of severe injury or death to motorists
or pedestrians. Because multiple DUI offenses involve
intentional or extremely reckless disregard for the safety of
others, incarceration is the recommended default sanction.

Unlawful
Entry and
Reentry to the
Country

22,000*

1.5%*

Yes

No

1.8*

N/A

In most cases there has been no additional crime
committed (if there is, it can be charged as a separate
crime). This crime generally automatically results in the
sanction of deportation under existing immigration law,
making incarceration in federal prison unnecessary.iii

Immigration
Smuggling

1,000*

0.1%*

No

Yes

1.8*

N/A**

Facilitation of illegal immigration of others. Because it
facilitates widespread violation of the law, places people
at risk of death or injury, and is associated with fraud,
violence, or smuggling of sex workers, incarceration is
recommended.

Other Serious
Crimes

18,000

1.2%

No

Yes

N/A

N/A

Examples include escape from custody, racketeering, and
bribery. Because these crimes not only cause physical or
psychological harm, but impede the orderly functioning
of prisons and other important institutions, incarceration
is the recommended default sanction.

Other Minor
Crimes

63,000

4.3%

Yes

No

N/A

N/A

Examples include prostitution, gambling, obscenity, traffic
offenses, and first time DUIs.112 These crimes involve
behavior deemed immoral but may not necessarily cause
physical injury to others; others involve risk of minor
injury to others. Alternatives such as community service or
fines are the recommended default sanction.

% of Total

76,000

# of Prisoners

Non-Violent
Weapon
Offenses

Crime
Category

Explanation

other

Total Other

207,000 14.1%

41.1%

58.9%

N/A

N/A

N/A

Total
Prisoners
(authors
dataset)

1,463,000 100%

24.9%

75.1%

N/A

N/A

By reducing prison stays for the six crimes noted
above, an additional 14.5 percent of inmates (212,000
people) could be safely released.

* Estimates.113 ** As explained in Section IV, the authors did not conduct a prison stay analysis for these crimes;
however, they believe these stays could likely be safely reduced.

Source: Brennan Center Analysis.114
iii Th
 is report takes no position on immigration policy. This is merely meant to note that undocumented people
already convicted of immigration felonies will be placed in the immigration system if not incarcerated.

34 | Brennan Center for Justice

IV. REDUCING TIME SERVED FOR OTHER CRIMES
The recommendations in Section III would reduce the nation’s prison population from 1.46 million to
about 1.1 million. This section turns to a thoughtful approach to reduce sentence lengths for offenders
whose crimes may still warrant prison.

A. Research On Lengths Of Stay & Recidivism
In Section III, this report analyzed existing research that showed that increasing incarceration has
diminishing returns on crime reduction. The same can be said of the relationship between the length
of time spent in prison and crime.
One body of research demonstrates that longer sentences do not reduce recidivism more than shorter
sentences. For example, a 2013 paper by economist David Abrams reviewed a range of studies examining
the effects of sentencing and parole in three cities (Chicago, Las Vegas, and Washington, D.C.), and
two states (California and Georgia). He concluded that “these studies seem to find fairly consistent
evidence of specific deterrence for low sentence ranges, but not for longer ones.”115 An analysis of over
440,000 prisoners by psychologist Paul Gendreau in 2002 also found significant diminishing returns.
Further, he concluded that after 12 months, prison stays caused higher recidivism.116
A similar case study was conducted in 2007 by economist Ilyana Kuziemko for the National Bureau
of Economic Research (“NBER”).117 That analysis focused on 80,000 prisoners in Georgia. She chose
this locale because the state’s inmate population was representative of national prison population
demographics, including percentage of violent and non-violent offenders and labor-market conditions
upon release from prison. She also found conditions ripe for a natural experiment on sentence lengths.
The Georgia Board of Parole assigned inmates “scores” that determined parole eligibility based on risk
of recidivism – the lower the score, the higher the presumed risk of reoffending upon release. Scores
were based on age, past record, and other factors believed to correlate with recidivism. However, release
decisions were made based on which risk “tier” an inmate’s score placed him in: a score of nine might
place an inmate in the medium-risk tier, but a score of eight would correspond to high risk. This
allowed Kuziemko to test the effect of prison on inmates who received similar risk scores, but were
treated differently.118 She was then able to assess whether lengths of time spent in prison resulted in
different rates of recidivism. She chose to assess the recidivism risks of those imprisoned for less than
20 months (1.7 years) versus those imprisoned for longer. She found that for prisoners serving less than
20 months, an extra month behind bars moderately reduced reoffending. However, she found “an extra
month ha[d] close to no effect on those serving more” than 20 months. Kuziemko concluded that her
results strongly suggest that longer prison terms have diminishing returns to sentence lengths beyond
20 months.119
Other studies indicate that as individuals get older, they tend to commit fewer crimes, providing more
evidence that longer sentences may not be as effective as assumed. Long sentences often hold people in
prison after they “age out of crime.” In 1999, economist Steven Levitt found that people between the
ages of 15 and 24 were most likely to commit crimes.120 Similarly, sociologists Charles Tittle, David

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 35

Ward, and Harold Grasmick demonstrated in 2003 that 18 to 24 year olds were more likely to commit
crimes than those who were older.121 A study by sociologist Robert Sampson conducted that same year
showed almost the exact same result in a longitudinal study, finding people over the age of 55 are 10
times less likely to commit crimes than those aged 23, even for individuals who had committed crimes
earlier on in life.122 Despite this evidence, the proportion of elderly prisoners tripled from 1993 to
2009.123 This is a direct result of the lengthy sentences of the past four decades.
A second body of research finds little to no relationship between lengths of stay and recidivism. In
its most recent multi-state study on recidivism published in 2002, the Justice Department examined
data from 15 states reporting lengths of stays and recidivism rates from prisoners released. It found
that recidivism rates did not differ significantly among those released after serving 6 months or less
compared to those serving sentences all the way up to 30 months in prison.124 A 2009 analysis by
criminologists Daniel Nagin, Francis Cullen, and Cheryl Jonson reviewed 19 major studies and found
that some research indicated that longer stays slightly decreased reoffending, while other research found
that longer stays slightly increased reoffending.125 A study that same year of juveniles convicted of
serious crimes in Philadelphia, Pa., and Phoenix, Ariz., found that longer stays — increasing from
three to 13 months — did not decrease future rearrest rates.126 A 2010 evaluation by political scientist
Donald Green and law professor Daniel Winik analyzed drug offenders in Washington, D.C., who
were sentenced by more strict judges versus more lenient ones. They also did not find a relationship
between longer incarceration and recidivism. In fact, they found that an extra month of imprisonment
raised recidivism rates by about 1 percent, but this result was not statistically significant.127
	
Further, a third group of studies indicates that the longer one stays in prison, the more likely he or
she is to reoffend upon release. In 2016, criminologist Jason Rydberg examined incarceration length’s
impact on recidivism rates in four states — Michigan, Missouri, New York, and Utah. He found that
longer length of stays between caused higher recidivism rates especially for some public order and
property crimes.128 University of Michigan economist Michael Mueller-Smith also recently studied
the effect of prison stays on recidivism rates in Harris County, Tex. He found that each additional year
that a prisoner served time behind bars made them 4 to 7 percent more likely to reoffend within threemonths after release. Further, he noted that this “should be of particular concern since it rapidly reverses
any cost savings from crime prevented.”129
Similarly, studies indicate that long prison sentences have little or no impact on reducing the criminal
behavior of the public at large. This theory, what criminologists call “general deterrence,” argues that
humans are rational beings who consider the punishment for their behavior before acting. Studies,
however, find this theory does not hold true, as most people consider immediate circumstances and
emotions instead of longer term legal consequences when acting or reacting. A seminal study by the
National Academy of Sciences in 2014 conducted a comprehensive analysis of more than a dozen
leading studies on general deterrence, and concluded that “the evidence on the deterrent effect of
sentence length suggests that the relationship between crime rate and sentence length” has “diminishing
deterrent returns” at best.130 Other studies find similar results. For example, in a 2002 study in Kentucky
and North Carolina, economist David Anderson interviewed males imprisoned for felonies about their
prior knowledge of the penalties they would face. He found that only 22 percent knew beforehand

36 | Brennan Center for Justice

what the punishment would be for their crime while more than half did not know or even consider
the punishment. When it came to deadly crimes, 55 percent of the men responded that they did not
think about the punishment at all.131 In 2003, economists Steven Raphael and Jens Ludwig evaluated
the effect of a policy implemented in 1997 in Richmond, Va., that increased prison sentences for
gun crimes by prosecuting them as federal crimes. They found no deterrent effect associated with the
sentence increases.132 In 2009, economists David Lee and Justin McCrary examined young people in
Florida about to reach their 18th birthday. Since the likelihood of receiving longer prison sentences
increases when individuals turn 18, theoretically fewer people should be willing to commit crimes
just after turning 18 compared to a month before. Lee and McCrary found a very small effect that
was not statistically significant, concluding that deterrence effects were minor or nonexistent.133
Finally, much research shows that certainty of punishment — as opposed to severity — more
effectively reduces future criminal activity. In recent years, researchers have focused on evidence that
“swift and certain” punishment using non-prison alternatives to respond to violations of probation
conditions are more effective than waiting for multiple violations and then revoking the probationer
to prison. The theory is that the threat of immediate punishment deters criminal conduct more than
the threat of more severe punishment occurring at some point in the future. Though this research
was conducted on probationers, it supports the notion that severity of punishment on its own does
not drive crime reduction.134
In sum, social science evidence indicates that in the worst case scenario, longer lengths of stay produce
higher recidivism rates, while the best case scenario points to diminishing returns of incarceration
on public safety. It also provides compelling evidence of the possibility that there is no relationship
at all between long lengths of stay and recidivism rates. After decades of using long prison stays as a
response to crime, these studies strongly encourage a need to rethink this approach.

B. Options For Reducing Prison Stays
The authors attempted to search for a way to bring sentence lengths into line with current research and to
make them more proportional to the crime committed. The authors recognize that their approach below
may seem arbitrary. However, they believe that their research-driven approach will result in more effective
sentences than the current arbitrary and overly punitive sentencing regime. They hope their recommendation
will serve as a starting point for sentence reductions, and not a call for uniform application.
To provide a starting point, the authors tested different options. They considered options of cutting
prison stays by 10 percent, 25 percent, and 50 percent, which they labeled as cautious, moderate, and
assertive, respectively.
They tested these options on the six major crimes that make up 58 percent of the national prison
population: aggravated assault, murder, nonviolent weapons crimes, robbery, serious burglary, and serious
drug trafficking.135 The authors chose these six crimes because they make up a significant portion of the
prison population, and because their lengths of stays increased significantly between 1993 and 2009.136

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 37

Table 2 depicts the effects of these reductions on the prison population, and Table 3 shows the cost
savings for each option:
•	

Cautious: A 10 percent reduction would shave off an average of about six months for these
prisoners. This modest step would reduce the prison population by 85,000 (5.8 percent of the
current prison population) and save $2.7 billion per year.

•	

Moderate: A 25 percent reduction would cut the prison population by 212,000 (14.5 percent
of the current prison population), saving almost $6.6 billion annually and reducing sentences
on average by about 15 months. It would save $66 billion over 10 years.

•	

Assertive: Leading criminal justice advocacy groups such as the ACLU, Beyond the Dream,
#Cut50, Ella Baker Center, #FreeAmerica, and JustLeadershipUSA support cutting the
national prison population in half.137 A similar approach could be applied to sentence lengths.
It would reduce the prison population by 424,000 (28.9 percent) and cut sentence lengths by
one year and four months, on average. It may risk short-term increases in recidivism, but the
amount saved would be staggering — $13.3 billion annually and $133 billion over 10 years.

Table 2: Proposed Reductions in Prison Stays for Major Crimes

Current
Prisoners

Percent of
Total Prison
Population

Current
Average
Prison Stay
(Years)

10%
Reduction

25%
Reduction

Robbery

176,000

12.0%

4.2

3.8

3.2

2.1

Serious Drug
Trafficking

165,000

11.3%

3.4*

3.1

2.6

1.7

Murder

165,000

11.3%

11.7

10.6

8.8

5.9

Aggravated Assault

155,000

10.6%

3

2.7

2.3

1.5

Serious Burglary

111,000

7.6%

1.7*

1.5

1.3

0.9

Nonviolent
Weapons Offenses

76,000

5.2%

3

2.7

2.3

1.5

Crime Category

*Estimates.138 Source: Brennan Center Analysis.139

38 | Brennan Center for Justice

50%
Reduction

Table 3: Annual Cost Savings from Reduced Prison Stays
Crime Category

10% Reduction

25% Reduction

50% Reduction

Robbery

$550 million

$1.38 billion

$2.75 billion

Serious Drug Trafficking

$520 million

$1.29 billion

$2.58 billion

Murder

$520 million

$1.29 billion

$2.58 billion

Aggravated Assault

$490 million

$1.21 billion

$2.43 billion

Serious Burglary

$350 million

$870 million

$1.74 billion

Nonviolent Weapons Offenses

$240 million

$590 million

$1.19 billion

Total

$2.7 billion

$6.6 billion

$13.3 billion

Source: Brennan Center Analysis.140
Though it is difficult to categorically determine precise optimal sentence lengths, the authors attempted
to find a way to safely roll back sentences to more proportional lengths.
To see which of these options seemed most effective, the authors applied their four-factor analysis
to these six crimes, with an increased emphasis on recidivism due to the heightened public safety
concern invoked by these more serious crimes. This analysis led to an average sentence reduction of
26.6 percent. To shed light on the authors’ analysis, two examples are illustrative:
•	 Murder: The average current prison stay for federal and state murder is 11.7 years.141 These
offenders also serve an average of roughly 1 to 3 years of additional incarceration in county
jails before their prison stays. In Broward County, Florida, those charged with murder served
an average of 2.8 years in jail before they were transferred to prison.142 This crime is one of
the most serious offenses, which weighs in favor of a long prison sentence. Murder is an
intentional crime where the victim impact is lethal and produces life-altering trauma for
family members, which also weighs toward a long sentence. However, the recidivism rate
is lower than average — 11 percent of those convicted of murder are reimprisoned for a
new crime within three years of release.143 Because the NBER study specifically found that
sentences less than 20 months did help bring down recidivism rates, the authors caution
against reducing prison stays to less than 20 months for the most serious violent crimes.
However, given that there are diminishing returns to long prison stays generally, the authors
recommend that prison stays for murder should be reduced somewhat. A 25 percent cut
would result in a reduction of three years, leaving an 8.8 year sentence — which still keeps
inmates behind bars for almost a decade.
•	 Serious Burglary: On average, offenders spend 1.7 years in prison, along with an additional
roughly 3 to 5 months in county jail. In Broward County, those charged with burglary served
an average of 1.1 years in jail before being transferred to prison.144 Burglary involves serious
property damage and alarm to residents. It is intentional, and the victim impact varies from
extensive property damage to emotional trauma, although it does not involve violence. (If

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 39

any physical harm to a victim is involved, the crime would escalate to assault, for example.) The
recidivism rate is below average, 31 percent of all burglary offenders were reimprisoned for a new
crime within three years of release.145 Given that there are diminishing returns to long prison stays
generally and that this crime is not violent, this report recommends considering cuts in stays. A
25 percent cut would result in a four-month reduction, leaving a 1.3 year sentence.
The authors also looked to two additional points of reference. First, as mentioned above, while
cutting sentence lengths by 50 percent may appear to many as a radical step, a powerful national
coalition is calling for a similar approach.
Second, prominent criminologists and other experts have called for a return to the shorter prison
stays that existed before the prison population exploded. They argue that the low crime rates
of the 1970s indicate that crime can remain low, even with shorter prison stays. Todd Clear,
criminologist and former Rutgers University Dean of the School of Criminal Justice, and James
Austin, co-author of this report, argued in a 2009 paper, “An in-depth examination of sentence
lengths and time served is called for, it might begin with the presumption that terms be cut back
to what they were circa 1975 when the imprisonment binge began.”146 Others have implied that
even returning to the sentencing regime of the early 1990s, before the imposition of truth-insentencing and mandatory minimums policies and the narrowing of parole eligibility, would make
a significant impact in reducing prison populations.147 To return to the length of stay in 1990,
today’s average prison stays would need to be cut by almost 40 percent.148
A 25 percent cut in sentence lengths for these six crimes seems more modest in comparison.
Based on the four-factor analysis, as well as these additional reference points, the authors conclude
that a 25 percent cut on average for these six crimes seems moderate and should serve as a starting
point for consideration. This recommendation — which would shave about 15 months from
prison sentences for these crimes — ensures that sanctions for serious crimes involve significant
prison time, are at levels that deter recidivism while protecting public safety, and achieve significant
savings.149 There are, of course, myriad different methods to project more proportional lengths of stay.

40 | Brennan Center for Justice

Table 4: Current Prison Population, by Recommendation
 Category

Number of Prisoners

Alternatives to Prison Appropriate — Lower Level Crimes
Prison Appropriate and Recommended Term Already Served — Six Major
Crimes
Prison Appropriate and Recommended Term Not Yet Served — Six Major
Crimes
Prison Appropriate and Length of Stay Analysis Can Be Applied — Other
Crimes
Total Prison Population

364,000
212,0000
636,000 
251,000
1,463,000

Source: Brennan Center Analysis.150
As shown in Table 4, if this recommendation were retroactively applied to prisoners currently serving
time for these six crimes, 212,000 inmates would have already served their reduced sentences and could
be released within the next year. Under this hypothetical, another 636,000 current prisoners would
continue to serve time until they reached the recommended length of stay.
Further, given today’s lengthy sentences, the authors believe a similar analysis can be applied to determine
a safe but shorter length of stay for the remaining 17.2 percent of the national prison population
(251,000 prisoners) for which the authors recommend imprisonment as the default sanction. Some of
these crimes may warrant more or less than a 25 percent cut depending on the crime.
Of course, appropriate recidivism reduction, job training, education, and reentry programs should
be implemented before releasing any prisoners to ensure proper reintegration into society without
increasing recidivism.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 41

V. RECOMMENDATIONS & COST SAVINGS
A. Recommendations
Whether viewed through a lens of justice, fairness, public safety, or cost, the U.S. prison system
unnecessarily warehouses hundreds of thousands of people. This report calls for the following changes
to sentencing laws as opposed to guidelines or practices, which can too easily be reversed.
•	

Eliminate Prison for Lower-Level Crimes Barring Exceptional Circumstances: State
legislatures and Congress should amend sentencing laws to mandate alternatives to prison
as the default penalty for lower-level crimes. These offenses include: drug possession, lesser
burglary, minor drug trafficking, minor fraud and forgery, minor theft, simple assault, and
other crimes such as prostitution and public disorder. Alternatives include: community service;
drug, alcohol, and mental health treatment; electronic monitoring; fines calibrated to ability
to pay; and probation. In exceptional circumstances, such as repeated past criminal activity,
judges should have discretion to depart from the defaults.

•	

Reduce Maximum and Minimum Sentences Set by Law: Lawmakers should consider
reducing sentence maximums and minimums defined in criminal statutes and sentencing
guidelines. A 25 percent cut can be considered as a starting point for the six major crimes
that make up 58 percent of the prison population: aggravated assault, murder, nonviolent
weapons offenses, robbery, serious burglary, and serious drug trafficking.151 Sentencing ranges
for other crimes that warrant prison should similarly be reviewed and likely recalibrated based
on research showing diminishing returns to sentence lengths.

•	

Retroactively Apply Reforms if Safe: Lawmakers should allow current inmates to petition
judges for retroactive application of these reforms. Judges can apply an analysis similar to the
four-factor test in this report and grant petitions if risks to public safety are minimal.

•	

Additional Recommendations:
o	 Prosecutors should use their discretion to seek alternatives to incarceration or shorter
prison stays in line with this report’s recommendations.
o	 Lawmakers should seek to repeal “three-strikes” and “truth-in-sentencing” laws, which
arbitrarily increase time-served, and expand parole eligibility.
o	 Lawmakers should reinvest savings from imprisonment reduction into policies proven to
reduce incarceration and crime, including: alternatives to incarceration (especially drug and
mental health treatment), community policing, and reentry programs. Although treatment
programs require upfront costs, they offer long-term cost efficiencies through savings from
averted incarceration. Attention should also be paid to the more than 500,000 Americans
employed in the nation’s prisons, for whom downsizing may mean a loss of livelihoods.
These experienced workers could be retrained to work in alternative to prison programs.

42 | Brennan Center for Justice

Sentencing Reform Success in States
Several states have started changing their sentencing laws in the direction recommended by
this report. Below are some examples.
•	

Florida: In 2009, Florida eliminated prison sentences for some low-level, nonviolent
felonies, such as lower-level theft and insurance fraud. These individuals may now be
sentenced to probation or community supervision with requirements that they participate
in community service, educational classes, training programs, or other recidivism
reduction programs.152 By 2012, Florida was able to shut down 11 correctional facilities
due to this and other reforms.153 From 2009 to 2014, the state’s imprisonment rate fell
9 percent while violent and property crime fell by more than 10 percent.154

•	

Mississippi: In 2008, the state repealed its truth-in-sentencing law for offenses involving
possession of a “small quantity of drugs,” reducing the time-served requirement to
25 percent of the original sentence.155 After this and other reforms, the state’s rate of
incarceration fell by 17 percent between 2008 and 2014, while violent crime fell by
nearly 10 percent.

•	

South Carolina: In 2010, the state eliminated mandatory minimums for drug possession,
and expanded the use of probation instead of prison for drug possession.156 Between 2010
and 2014, South Carolina saw its imprisonment rate fall by 13 percent along with a 17
percent decline in violent crime and an 11 percent decline in property crime.

•	

Texas: In 2007, state legislators increased the availability of drug addiction and mental
health treatment instead of prison for lower-level offenders including nonviolent drug
offenses.157 In 2011, the Texas legislature passed a bill allowing drug offenders to reduce
their prison sentence lengths by completing educational programs.158 From 2011 to
2014, the state reduced crime by 14 percent with an accompanying 8 percent reduction
in the incarceration rate.159

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 43

B. Cost Savings
Ending prison as the default punishment for lower-level crimes would reduce the current prison
population by 364,000 people. Reducing time served for the six major crimes by 25 percent would allow
an additional 212,000 prisoners to be released within the next year. These two recommendations, taken
together, would shrink today’s prison population of 1.46 million people by almost 580,000, and would
also curtail the major drivers of mass incarceration going forward.
The average cost to house an inmate is about $31,000 per year. Therefore, these two recommendations
would yield $18.1 billion in savings annually and $181 billion over 10 years, assuming the number of
people convicted of such crimes remained constant.160
Figure 13 depicts some of these analogues. This savings is the equivalent of:
•	

Hiring and training 270,000 new police officers.161

•	

Employing 360,000 probation officers.162

•	

Hiring 327,000 new teachers.163

•	

Sending two million young people to college.164

•	

The annual budgets of Los Angeles and Chicago, combined.165

•	

Nearly the combined budgets of the U.S. Departments of Commerce and Labor.166

•	

The annual budget of NASA.167

•	

The annual budget of Connecticut.168

All these dollars could be saved with little effect on public safety.

44 | Brennan Center for Justice

Figure 13: Cost Savings from Reduced Imprisonment
25

Billions of Dollars

20

15

10

5

0
Cost Savings from
Reduced Imprisionment

Chicago and
Los Angeles
Budget

Dept. of Commerce
and Dept. of Labor

270,000 New
Police Officers

360,000 New
Probation Officers

Source: Brennan Center Analysis of Agency Budgets (2016).169

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 45

CONCLUSION
America’s experiment in mass incarceration has failed. Not only does using prison as a knee-jerk
reaction to crime devastate families and communities, but many of today’s overly punitive prison
sentences produce little public safety benefits. Many people in prison today are serving sentences for
lower-level crimes, for which an alternative sanction would prove more just, equitable, less costly, and
more effective at reducing recidivism. Additionally, many prisoners could serve less time in prison with
similar public safety results.
This report’s findings and recommendations are intended to offer the country one practical and effective
approach to end mass incarceration while preserving the public safety gains of the last two decades. The
authors hope that it serves as a catalyst for discussion for much needed change to criminal punishment
in the United States.

46 | Brennan Center for Justice

APPENDIX A: METHODOLOGY
I. DATA SOURCES
The authors began research for this report in 2013, at which time 2012 data on state and federal prisoners
was the most recent available. Notably, Dr. Austin and his co-investigators relied on data obtained from
the National Corrections Reporting Program, a restricted source that cannot be distributed to outside
parties.170 Therefore the dataset for this report cannot be shared.
The authors initially intended to include jail data, but due to difficulty securing representative jail
samples, they ultimately could not include jail populations in this report. They hope to conduct a
similar analysis on jails in the future, if data allows.
Ultimately, the authors analyzed state inmate data from 48 states plus federal prisons, totaling 1.46
million prisoners, representing 94 percent of the national prisoner population.

A. State Prison Data
This report includes data from 48 states:
•	

National Corrections Reporting Program (NCRP) Data: The NCRP, run by the Justice
Department, collects data every year on individual prisoners in state prisons. In 2012,
47 states reported some data to the NCRP.171 But only 42 of those states reported data
containing information on prison populations and lengths of prison stay, sorted by crime
(into 370 crime categories) — the format needed for this report. Notably, NCRP data
does not capture information about every single inmate in state prison so some individuals
are missing from the dataset.

•	

Individual State Data: The authors secured data on 2012 prison populations directly from
six additional states: Arizona, Arkansas, Connecticut, Hawaii, Louisiana, and Virginia.172

•	

Missing Data: The only states unable to fulfill any part of the request were New Mexico and
Vermont. New Mexico organizes its data by criminal offense, but not individual offense,
therefore over-counting all incarcerated people convicted of multiple offenses. Vermont’s
Department of Corrections does not organize its data by criminal offense. Therefore data
from these states could not be inputted into the authors’ dataset. Prisoners in New Mexico
and Vermont make up just 0.5 percent of the total, national prison population, and around
0.7 percent of the state prison population.173

B. Federal Prison Data
Data on federal prisoners were collected from the Federal Justice Statistics Program (FJSP), which
describes criminal cases in the federal justice system from arrests through corrections. It lists 33 federal

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 47

crime categories. Data in the FJSP is collected from the U.S. Marshals Service, Drug Enforcement
Administration, Executive Office for U.S. Attorneys, Administrative Office of the U.S. Courts, U.S.
Sentencing Commission, and the Federal Bureau of Prisons.174 The data on length of prison stays details
the mean time served for prisoners released in 2012, sorted by crime category.
The FJSP differs from the NCRP data in two main ways. First, the FJSP data uses slightly different
crime categories than the NCRP. Second, the FJSP year runs from October 1st to September 30th,
while the NCRP year runs from January 1st to December 31st.

C. Rounding
Because prison populations fluctuate, the authors rounded all statistics on the number of prisoners.
Totals listed in this report are the sums of the unrounded figures, rounded to the nearest thousand or
single percent.

D. Definitions
Some sources did not include definitions of crimes. In those circumstances, the authors relied on
definitions from the American Law Institute’s Model Penal Code, Black’s Law Dictionary, and other
leading casebooks on modern criminal law.175

II. ANALYSIS TO IDENTIFY LOWER-LEVEL CRIMES
To determine the appropriate default sanction, whether incarceration or an alternative, the authors
devised a four-part test. They then applied the test to each crime category in their dataset.

A. Crime Categories
The NCRP dataset contained 160 crime categories. Individual states reported between 16 and 57 crime
categories each. The FJSP reported 33 crime categories. Added together, the authors analyzed 370
crime categories.
For presentation purposes, the authors grouped these categories into the larger categories (listed in Table
1) when they entailed the same or similar conduct. For example, the overarching category of serious
larceny includes grand larceny, attempted larceny, and conspiracy to commit larceny from the NCRP
data, crimes labeled “theft” from individual state data, and crimes labeled larceny from the federal data.
Notably, this report uses the labels drug, property, violent, and other consistent with the NCRP
and FJSP, so that this analysis can be compared to other national datasets and studies. The violent
crime category in these sources may be overly broad, sweeping in crimes that do not actually include
perpetrated violence.176 Ideally, these categories would be more reflective of the acts committed.

48 | Brennan Center for Justice

B. Four-Factor Analysis
The authors devised a four-factor analysis that relies heavily on public safety and applied it to each of the
370 crime categories. Their analysis weighed the following factors: seriousness of crime, victim impact,
state of mind, and likelihood of recidivism. These factors were chosen based on a comprehensive review of
scholarly and popular consensus on why we incarcerate — to promote public safety and deliver deserved
punishment. Each of the first three factors speaks to both of these rationales. The final factor, recidivism,
focuses solely on public safety.
Some crimes, such as attempted murder, are serious but may not physically injure victims, while others,
like some regulatory offenses, are serious and potentially harmful, but lack culpable intent. Consideration
of all four factors allowed the authors to determine when incarceration adequately reflected a concern
for the maintenance of public safety, and when it unnecessarily burdened individuals, budgets, and
communities with little public safety benefit.
While these factors were generally weighed equally, on occasion one or more factors predominated. For
example, attempted murder had an exceptionally low recidivism rate, and by definition the victim was not
killed. However, because it required intent to commit the most serious, irreversible crime, the first and third
factors were given greater weight, leading to a recommendation of incarceration as the default sanction.
The authors labeled crimes as “lower-level” when the analysis led to a conclusion that the crime did not
merit incarceration as the default sanction. Among these lower-level offenses were crimes that did not
result in serious harm to a victim or substantial destruction of property, for which malicious intent may
not have been present, and/or for which prison was not effective at reducing recidivism.
These recommendations are meant to instill default guidelines into sentencing laws, not provide absolute
sanctions. As noted in the report, these defaults should be adjusted up or down depending on individual
circumstances of each case and offender.
Further description of the four factors is below.
1. Seriousness
This factor reflects the intuitive idea that a crime that causes, or has the potential to cause, great harm to
victims and to society may merit more severe means of prevention and incapacitation. The broad metric
of “seriousness” ensures consideration of both real and potential harm, and both tangible and intangible
harm. For example, an attempted murder in which the victim does not know of the attempt may not
harm the victim, although it has great potential to do so. Moreover, a mugging may not cause serious
bodily or material harm, but it can cause fear, mental anguish, and property loss.177
This factor takes into consideration the retribution theory of punishment (by relying on the argument that
a more serious crime deserves a more serious punishment) while using proportionality to cabin its effect
(to ensure that sentences are not more severe than necessary).

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 49

2. Victim Impact
This factor reflects the idea that actual, as opposed to potential harm, deserves special consideration.
Bodily harm is the most common distinction in popular discussion of crime, differentiated by violent and
non-violent harm. The special significance of bodily harm is both intuitive and deeply rooted in American
law. Criminal law in every state reserves the most serious punishments for infliction of bodily harm.178
Today, criminologists and the public also recognize the significance of psychological harm.179 Some
crimes that involve no physical contact at all, such as burglary of someone’s unoccupied home, may still
cause significant fear and mental anguish.
3. State of Mind
A culpable mental state, also called “mens rea” or intent, is a core tenet of criminal liability in American
law. The Model Penal Code, produced by the American Law Institute, forms the basis of many state
criminal codes and requires that all crimes have some mental state element — at a minimum negligence
(carelessness, or a lack of “reasonable care”), but usually requiring knowledge or intent.180 This means
that the individual must know they are committing the criminal act or have the will or desire to commit
the act to be held guilty of that crime. The idea that mental state affects culpability is reflected in the
different categories of killing in almost every jurisdiction. All states have determined that a planned
murder deserves a more severe punishment than a negligent killing, although the harm to the victim
may be the same.181
4. Recidivism
Concern for reducing recidivism reflects part of the public safety rationale for incarceration — the idea
being that a sanction for a crime will deter the individual from future criminal conduct, often called
“specific deterrence,” thereby making society safer. (The other public safety rational is incapacitation
— meaning that as long as the offender is kept out of society and in prison, they won’t have the
opportunity to commit another crime.) It is important to consider recidivism because over 95 percent
of today’s prisoners will be released at some point in the future.182 Recidivism is most commonly
defined as the rearrest, reconviction, or reincarceration of a former offender within a specific time
frame.183 Recidivism also encompasses the idea of rehabilitation, as lower recidivism rates should ideally
reflect more rehabilitative programs that deter future crime. It also takes into account the theory of
incapacitation (the idea that a prisoner will not commit crimes while he or she is in prison) as part of
its analysis of public safety impact.
In this report, the authors use reincarceration within three years as the standard for recidivism as this
report’s recommendations focus on incarceration.184
When determining recidivism rates for the 370 crime categories, the authors relied on data and research
from four seminal federal government studies on nationwide recidivism: one of state prisoners released
in 1983 (published in 1990); one of state prisoners released in 1994 (published in 2002); one of state
prisoners released in 2005 (published in 2014); and one of federal prisoners released in 2005 (published

50 | Brennan Center for Justice

in 2016).185 Generally, these studies examined rearrest, reconviction, and reimprisonment rates for
released prisoners over three and five years.
In conducting the recidivism analysis for the vast majority of crimes, the authors relied on the 2002
study of offenders released in 1994 because it was the most recent available analysis providing recidivism
data specifically broken down by the crime for which the released person served time.186 The more
recent studies examined recidivism rates for crimes combined into the overall categories of violent,
property, drug, and other. The authors compared the overall recidivism rates reported by each study
and found a high level of consistency over a period of more than 20 years for rearrest, reconviction, and
reimprisonment within these overall categories. Given this long-term consistency, the authors believe it
was prudent to use the 1994 data, although a more recent source broken down by specific crime would
have been preferable.
5. Other Factors
There are myriad factors that one could consider when determining sentences. The authors examined
some additional factors and ultimately decided not to include or to implicitly include them in their
analysis for the reasons below.
•	

Incapacitation: Incapacitation focuses on the value of preventing individuals who have
committed crimes from committing additional crimes by removing them from society through
imprisonment.187 This theory could be used to lock up individuals for the longest period of
time possible — on the rationale that the longer someone is removed from society, the longer
their capacity for committing crimes remains zero. The authors believe there should be a
limiting principle to this theory because otherwise it could be used to justify unnecessarily long
punishment, which has produced our current regime of an overreliance on incarceration. For
this reason, the authors considered incapacitation in two ways. First, as part of the public safety
impact consideration under the recidivism factor. Second, this report suggests that incapacitation
should be proportional to the crime committed, linked to seriousness of the crime and victim
impact. If one has committed a more serious crime, there is a rationale to incapacitate that
individual for a longer period of time (compared to a lower-level offender).

•	

General deterrence: This theory hinges on the idea that the threat of incarceration for long
periods prevents future crimes by deterring individuals from committing crimes by instilling a
fear of punishment. Two underlying assumptions call into question the veracity of this concept.
First, general deterrence depends on an individual’s knowledge of criminal penalties. Second, it
requires the individual to instantaneously weigh their potential criminal behavior against those
penalties when acting or reacting in often volatile and emotional situations.188 As explained
in Section IV of this report, a large body of research indicates that general deterrence is not
effective at preventing crime. Therefore the authors did not include this in their four-factor
analysis. They believe that their factor considering the individual offender’s recidivism postprison captures the majority of the relevant public safety effects of sentences.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 51

•	 Retribution: This theory of punishment focuses on the moral wrong committed by the
offender. It is punishment inflicted primarily as vengeance for a wrong or criminal act. It
does not consider the positive or negative effects of punishment on the individual or society,
but is grounded in the idea that punishment is deserved regardless of its future effects.189
This theory is, for example, codified in the Old Testament, which calls for “an eye for an
eye, a tooth for a tooth, and a life for a life.”190 The authors incorporated retribution in their
analysis in two ways. First, they did so through their seriousness factor. This factor takes into
consideration the retribution theory of punishment (by relying on the argument that a more
serious crime deserves a more serious punishment) while using proportionality to cabin its
effect (to ensure that sentences are not more severe than necessary). Second, the impact on
victims and offender’s intent both include aspects of the morality of the offense.
•	 Individual Liberty and Effects on Individuals and Communities: One reason to reduce
excessive sentences is to prevent the unnecessary deprivation of individual liberty and other
collateral consequences to individuals and their families.191 With this in mind, in making
their decisions, the authors implicitly weighed their four factors alongside an individual’s
right not to be deprived of liberty without justification and the country’s and communities’
interest in reducing over-incarceration.

C. Estimates For Crimes With Broad Definitions
Eight NCRP crime categories describe conduct that varied widely in gravity, from less serious to
more serious types. These crimes are: burglary, driving under the influence, embezzlement, fraud
and forgery, grand larceny, statutory rape, trafficking of marijuana, and trafficking of serious drugs
(substances other than marijuana). The NCRP does not provide detailed data on the specifics of
crimes committed under these categories.
In order to appropriately apply the four-factor analysis, the authors first devised a definition for
lower-level conduct compared to more serious conduct for each of these eight categories, and split
them into a “minor” and “serious” versions, respectively. They then estimated how much of the
national prison population fell into each category by using representative data from individual states
that broke down their prison populations into more detailed crime categories. These states include:
Arizona, Arkansas, California, Connecticut, Florida, Kentucky, and Louisiana. Through this analysis,
they were able to devise the estimates of the national prison population incarcerated for each degree
of offense. Estimates produced using this analysis are noted with an asterisk in Table 1.
The analysis for two offenses — burglary and trafficking of serious drugs — is shown in Section III
of this report as examples.

52 | Brennan Center for Justice

III. ANALYSIS FOR REDUCTIONS IN PRISON STAYS
A. Calculating Current Average Prison Stay
The NCRP contains average prison stays for 42 states. The federal data from the FJSP reports sentence
lengths, but not length of prison stay. Most of the data collected from individual states did not include
prison stay lengths. Therefore, the authors relied on the NCRP for prison stay lengths as it provided the
needed detail on sentence lengths and was nationally representative.
The authors combined length of stay data for individual crimes to create an average length of stay for
the larger crime categories in Table 1. The authors weighted these averages by population, adjusting for
the number of people in each crime category, to ensure accurate representation.
For the eight crimes with broad definitions noted above, the average length of stay in Table 1 represents
the length of stay for the broad category since data for the more specific categories created by the
authors was not available.

B. Options For Recommended Reductions
Because it is difficult to ascertain an exact, ideal sentence length for a crime, the authors picked three
benchmark options to test percentage cuts (10, 25, and 50 percent) to sentence lengths in order to
provide policymakers with different options to consider as starting points.
Instead of testing the benchmark options on all 370 crime categories, the authors opted to test them
on the six crimes that comprised 58 percent of the national prison population — aggravated assault,
murder, nonviolent weapons crimes, robbery, serious burglary, and serious drug trafficking — and
comprised 75 percent of the prison population for which the authors recommended incarceration as
the default sanction.192
The authors then sought to determine which of these options would most appropriately bring prison
stay lengths to be more proportional to the crimes committed, while still protecting public safety. They
applied the four-factor analysis used in Section III, with an increased emphasis on recidivism due to the
heightened public safety concern for these more serious crimes. The four-factor analysis led, on average,
to a 26.6 percent sentence reduction.
The authors concluded that a 25 percent reduction in length of stay closely resembled their calculated
average reduction and could be achieved without producing harm to public safety.
The authors recognize that this methodology can be viewed as somewhat arbitrary. The purpose of this
analysis is not to put forth absolute sentence lengths. Instead, it is to put forth options for policymakers
to consider, starting with one the authors believe is moderate and effective.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 53

The authors believe a similar analysis should be applied to the remaining 17 percent of the national
prison population (251,000 prisoners) for whom the authors recommend incarceration as the default
sanction to determine whether current sentence lengths are appropriate or could be safely cut. This
could further reduce the prison population.

C. Estimating Effect Of Prison Stay Reductions On The Prison Population
The authors then estimated the impact of these prison stay reduction recommendations on the current
prison population to determine how many prisoners convicted of the six crimes have already served the
recommended prison terms.
To estimate this number, the authors used a model that relies on a “steady state” assumption. Under
this assumption, the same number of people would be admitted and released each year for the same
crimes so that the prison population stays flat. For example, the number of people in prison for drug
trafficking, robbery, and theft would stay constant with the same length of stay year after year. In
reality, incremental changes each year cause admissions, releases, and lengths of stay to vary. But for
the purposes of estimating the number of prisoners that can be released, the authors assumed that
the prison population equaled the average prison stay multiplied by the admissions per year. This is a
common assumption employed when making such estimates.
To estimate admissions, the authors divided the current prison population by length of stay for the six
crimes. The authors multiplied this admissions estimate by the new recommended lengths of stay for
each of the six crimes to estimate the reduced prison population.

IV. METHODOLOGICAL LIMITATIONS
Limitations include:
•	 This report’s analysis is the authors’ best attempt at creating a more evidence-based sentencing
scheme. They recognize that modifications and alternate approaches could be made to improve
these recommendations, and they encourage such research and proposals.
•	 This report makes recommendations for default sanctions. They do not take into consideration
individual factors. For that reason, these defaults should be adjusted up or down according
to individualized factors, as listed in the report. The authors believe that an equitable system
of sentencing should treat like offenders alike. For that reason, they erred on the side of
recommendations based on offenses while giving judges the discretion to depart.
•	 There are many factors that one could consider when devising sentencing laws. The authors
chose to focus on the factors they believe were the most relevant to public safety and most in
line with social science research on punishment. Of course, others could select different factors,
arriving at different conclusions.

54 | Brennan Center for Justice

•	 The authors relied on 2012 data because that was the most recent data available when the
authors began their research. The prison population has changed slightly since then; for this
reason, the authors rounded and approximated data on prisoners. Despite any fluctuations, the
overarching findings and recommendations continue to be valid.
•	 The authors’ dataset is confidential and cannot be shared as it includes NCRP data, which is
restricted from general dissemination and cannot be distributed to outside parties.
•	 The authors estimated the number of prisoners for the eight overly broad crime categories
based on data from sample states, rather than an exact accounting of the number of people
incarcerated for each version. This could slightly skew the data; however, the overarching
recommendations continue to be valid.
•	 This report relies primarily on recidivism data from the 1994 study referenced above, because it
was the most recently available data that provided a breakdown by specific crime. Consistency
in recidivism data over time suggests that the data remains accurate. However, a more recent
source of data may change the recommendations slightly.
•	 The authors recognize that their methodology to cut lengths of stays can be viewed as somewhat
arbitrary. The purpose of this analysis is not to put forth absolute sentence lengths. Instead, it
is to put forth options for policymakers to consider, starting with one that the authors believe
is modest and effective for the six major crimes that make up a substantial percentage of the
prison population.
•	 The authors recognize that the 50 states and the federal government each have their own unique
criminal and sentencing laws and enforcement policies. However, for presentation purposes and
to tally the number of people incarcerated with limited public safety rationale, they presented
a national level analysis in this report. This presentation minimizes individual differences
between these jurisdictions. However, the authors’ data set preserves these differences.
•	 The data in this report does not indicate whether the incarceration term is a result of a plea
bargain or a conviction after trial. However, around 95 percent of criminal convictions in this
country result from plea bargains.193 This is significant for two reasons. First, the prosecution
might have been very likely to win the case, but in the interest of the victim or efficiency,
may have wanted to avoid a full trial and negotiated a lesser charge with a less serious prison
sentence. Or alternatively, the defendant may have felt pressure to agree to a plea deal to a lesser
crime even if not guilty because of the threat of mandatory minimums sentences attached to
the charges a defendant may face after conviction at trial. Because the authors searched for a
more fair and proportional method for sentence lengths, they made their recommendations
based on the actual offense for which the offender was convicted, while providing judges the
discretion to depart in such cases.

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 55

ENDNOTES
1	
See U.S. Census Bureau, Quick Facts, United States, https://www.census.gov/quickfacts/table/
PST045215/00; E. Ann Carson, Bureau of Justice Statistics, Prisoners in 2014 15 (2015), http://
www.bjs.gov/content/pub/pdf/p14.pdf.
2	
See Michelle Ye Hee Lee, Does The United States Really Have 5 Percent Of The World’s Population And
One Quarter Of The World’s Prisoners?, Wash. Post, Apr. 30, 2015, https://www.washingtonpost.com/
news/fact-checker/wp/2015/04/30/does-the-united-states-really-have-five-percent-of-worlds-populationand-one-quarter-of-the-worlds-prisoners (finding that the United States contains 4.4% of the global
population, and 22% of global prisoners); see also Roy Walmsley, World Prison Brief, World Prison
Population List 5, 14 (11th ed. 2015), http://www.prisonstudies.org/sites/default/files/resources/
downloads/world_prison_population_list_11th_edition.pdf (placing the United States prison population
at 2,217,000, or 21.4% of the global prison population of 10,357,134) and U.S. Census Bureau, U.S.
World & Population Clock, https://www.census.gov/popclock/world (stating that, as of mid-2016, the
United States contained approximately 4.4% of the world’s population).
3	
See Danielle Kaeble et al., Bureau of Justice Statistics, Correctional Populations in the
United States 2014 2 tbl.1 (2016), http://www.bjs.gov/content/pub/pdf/cpus14.pdf; World Prison
Brief, Highest to Lowest – Prison Population Total, http://www.prisonstudies.org/highest-tolowest/prison-population-total/trackback?field_region_taxonomy_tid=All.
4	
See World Prison Brief, Highest to Lowest – Prison Population Total, http://www.prisonstudies.
org/highest-to-lowest/prison-population-total/trackback?field_region_taxonomy_tid=All (ranking
countries by incarceration rate).
5	The imprisonment rate includes only prison. The incarceration rate includes jail and prison. See Chet
Bowie, Bureau of Justice Statistics, Prisoners 1925-1981 2 tbl.1 (1982), http://www.bjs.gov/content/
pub/pdf/p2581.pdf and Bureau of Justice Statistics, Corrections Statistics Analysis Tool (CSAT)
– Prisoners, http://www.bjs.gov/index.cfm?ty=nps (follow “Quick Tables” hyperlink).
6	
See Chet Bowie, Bureau of Justice Statistics, Prisoners 1925-1981 2 tbl.1 (1982), http://www.bjs.
gov/content/pub/pdf/p2581.pdf and Bureau of Justice Statistics, Corrections Statistics Analysis
Tool (CSAT) – Prisoners, http://www.bjs.gov/index.cfm?ty=nps (follow “Quick Tables” hyperlink).
7	The incarceration rate leveled off in 2008, and then began to decline. See Bureau of Justice Statistics,
Corrections Statistics Analysis Tool (CSAT) – Prisoners, http://www.bjs.gov/index.cfm?ty=nps
(follow “Quick Tables” hyperlink). The average annual decline between 2008 and 2014 was about 1.15%.
See id. If that trend continued, the incarceration rate would fall below 200 by 2089.
See Barry Goldwater, Goldwater’s 1964 Acceptance Speech (Jul. 16, 1964), in Wash. Post, http://www.
8	
washingtonpost.com/wp-srv/politics/daily/may98/goldwaterspeech.htm.
9	
See 1968: Nixon vs. Humphrey vs. Wallace, Museum of the Moving Image: The Living Room Candidate
(2012), http://www.livingroomcandidate.org/commercials/1968 (follow “The First Civil Right” hyperlink
in the “Republican” row; under “Click on thumbnail to view video”).
10	
See United States Dep’t of Justice, Federal Bureau of Investigation, Uniform Crime Reporting
Statistics, State-by-State and National Crime Estimates by Year(s), http://www.ucrdatatool.gov/
Search/Crime/State/RunCrimeStatebyState.cfm (providing crime statistics from 1960 to 2013).
11	
See Bureau of Justice Statistics, Criminal Victimization in the U.S.: 1973-79 Trends 43 tbl.17
(1981), http://www.bjs.gov/content/pub/pdf/cvus7379.pdf (data for 1979).
12	
See Jim Parsons et al., Vera Inst. of Justice, End of an Era: The Impact of Drug Law Reform in
New York City 5 (2015), http://www.vera.org/sites/default/files/resources/downloads/drug-law-reformnew-york-city-summary-01.pdf (laying out the history of the Rockefeller drug laws).

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 57

13	
See Paula M. Ditton & James Wilson, Bureau of Justice Statistics, Truth in Sentencing in State
Prisons 2 tbls.1, 3 (1999), http://bjs.gov/content/pub/pdf/tssp.pdf.
14	

28 U.S.C. § 991.

15	
See., e.g., Anti-Drug Abuse Act of 1986, Pub. L. No. 99–570, 100 Stat. 3207 (1986).
16	
See Violent Crime Control and Law Enforcement Act of 1994, H.R. 3355, 103d Cong. (1994) (enacted);
Nat’l Conf. of State Legislatures, Principles of Effective State Sentencing and Corrections
Policy (2011), http://www.ncsl.org/research/civil-and-criminal-justice/principles-of-sentencing-andcorrections-policy.aspx (see “mandatory minimum sentences”). (This bill is commonly known as the
“Crime Bill.”)
17	
See Bureau of Justice Statistics, Prisoners 1925-1981 2 tbl.1 (1982), http://www.bjs.gov/content/
pub/pdf/p2581.pdf (providing data through 1978) see also Bureau of Justice Statistics, Corrections
Statistics Analysis Tool (CSAT) – Prisoners, http://www.bjs.gov/index.cfm?ty=nps (follow “Quick
Tables” hyperlink; then follow “1978-2014” hyperlink under “Prisoners under the jurisdiction of state or
federal correctional authorities, December 31, 1978-2014” to download data table).
18	
See United States Dep’t of Justice, Federal Bureau of Investigation, Uniform Crime Reporting
Statistics, State-by-State and National Crime Estimates by Year(s), http://www.ucrdatatool.gov/
Search/Crime/State/RunCrimeStatebyState.cfm (providing crime statistics from 1960 to 2013) and
United States Dep’t of Justice, Federal Bureau of Investigation, Uniform Crime Reporting
Statistics, Crime in the United States, 2015 (2016), https://ucr.fbi.gov/crime-in-the-u.s/2015/crimein-the-u.s.-2015/home (providing data through 2015). For this calculation, the overall crime rate is the
sum of the property and violent crime rates.
19	
See Oliver Roeder et al., Brennan Ctr. for Justice, What Caused the Crime Decline (2015),
https://www.brennancenter.org/publication/what-caused-crime-decline (including a detailed review of
other existing research on the topic).
20	
See Nat’l Research Council, The Growth of Incarceration in the United States: Exploring
Causes and Consequences 138 (Jeremy Travis, Bruce Western, & Steve Redburn eds., 2014);
Steven Raphael & Michael Stoll, The Hamilton Project, A New Approach to Reducing
Incarceration While Maintaining Low Rates of Crime 9 (2014), http://brook.gs/1E6xzGl; Oliver
Roeder et al., Brennan Ctr. for Justice, What Caused the Crime Decline (2015), https://www.
brennancenter.org/publication/what-caused-crime-decline (including a detailed review of other existing
research on the topic).
See Christian Henrichson & Ruth Delaney, Vera Inst. of Justice, The Price of Prisons: What
21	
Incarceration Costs Taxpayers 10 (2012), http://www.vera.org/sites/default/files/resources/downloads/
price-of-prisons-updated-version-021914.pdf.
22	
See Section IV.
23	Ilyana Kuziemko, Going off Parole: How the Elimination of Discretionary Prison Release Affects the Social Cost
of Crime 21 (Nat’l Bureau of Econ Research, Working Paper No. 13380, 2007), http://www.nber.org/
papers/w13380.
24	
See Matthew R. Durose et al., Bureau of Justice Statistics, Recidivism of Prisoners Released
in 30 States in 2005: Patterns from 2005 to 2010 1 (2014), http://www.bjs.gov/content/pub/pdf/
rprts05p0510.pdf (providing recidivism rates for those who are returned to prison with or without a new
prison sentence).
25	
See Lauren-Brooke Eisen & James Cullen, Brennan Ctr. for Justice, Update: Change in
State Imprisonment Rates 1 (2016), https://www.brennancenter.org/sites/default/files/analysis/
UpdateChangesinStateImprisonment.pdf.

58 | Brennan Center for Justice

26	For imprisonment rates, see Bureau of Justice Statistics, Corrections Statistics Analysis Tool
(CSAT) – Prisoners, http://www.bjs.gov/index.cfm?ty=nps (follow “Quick Tables” hyperlink). For
crime rates, see United States Dep’t of Justice, Federal Bureau of Investigation, Uniform Crime
Reporting Statistics Reports, State-by-State and National Crime Estimates by Year(s), http://
www.ucrdatatool.gov/Search/Crime/State/StatebyState.cfm (providing crime statistics from 1960 to
2013). For the Supreme Court case affecting California imprisonment rates, see Brown v. Plata, 131 S. Ct.
1910, 1923 (2011).
27	
U.S. Sentencing Comm’n, Quick Facts: Federal Offenders in Prison – January 2015 2 (2015),
http://www.ussc.gov/sites/default/files/pdf/research-and-publications/quick-facts/Quick-Facts_BOP.pdf.
28	
Thomas P. Bonczar et al., Bureau of Justice Statistics, Nat’l Corrections Reporting Program:
Sentence Length of State Prisoners, by Offense, Admission Type, Sex, and Race (2011), http://
www.bjs.gov/index.cfm?ty=pbdetail&iid=2056 (follow the “1993” and “2009” hyperlinks to download
data tables for 1993 and 2009; then compare mean time served in prison in Table 9 (“NCRP9309.csv”
and “ncrp0909.csv” for 1993 and 2009, respectively)).
29	
See Matthew R. Durose et al., Bureau of Justice Statistics, Recidivism of Prisoners Released
in 30 States in 2005: Patterns from 2005 to 2010 1 (2014), http://www.bjs.gov/content/pub/pdf/
rprts05p0510.pdf (describing high rates of recidivism, around 50%, for former prisoners).
30	
See generally Jessica Eaglin & Danyelle Solomon, Brennan Ctr. for Justice, Reducing Racial and
Ethnic Disparities in Jails (2013), https://www.brennancenter.org/publication/reducing-racial-andethnic-disparities-jails-recommendations-local-practice.
31	

S ee Abe Cho, Lowering Sentences for Illegal Immigrants: Why Judges Should Have Discretion to Vary from
the Guidelines Based on Fast-Track Sentencing Disparities, 43 SSRN Electronic J. 4 (2009), https://
www.researchgate.net/publication/228294416_Lowering_Sentences_for_Illegal_Immigrants_Why_
Judges_Should_Have_Discretion_to_Vary_from_the_Guidelines_Based_on_Fast-Track_Sentencing_
Disparities; Daniel J. Freed, Federal Sentencing in the Wake of Guidelines: Unacceptable Limits on the
Discretion of Sentences (Yale Law School Faculty Scholarship Series, Paper No. 2050, 1992), http://
digitalcommons.law.yale.edu/cgi/viewcontent.cgi?article=3083&context=fss_papers;
William
W.
Schwarzer, Judicial Discretion in Sentencing, 3 Fed. Sent’g Rep. 339 (1991), http://repository.uchastings.
edu/cgi/viewcontent.cgi?article=2183&context=faculty_scholarship.

See Charles W. Ostrom et al., Nat’l Inst. of Justice, Judges & Discrimination: Assessing the Theory
32	
& Practice of Criminal Sentencing (2003), https://www.ncjrs.gov/pdffiles1/nij/grants/204024.pdf;
David Mustard, Racial, Ethnic, and Gender Disparities in Sentencing: Evidence from the U.S. Federal Courts,
44 J. Law & Econ. 285 (2001), https://www.researchgate.net/publication/24101050_Racial_Ethnic_
and_Gender_Disparities_in_Sentencing_Evidence_from_the_US_Federal_Courts; Mirko Bagaric,
Consistency and Fairness in Sentencing, 2 Cal. Crim. L. Rev. 1 (2000), http://scholarship.law.berkeley.
edu/bjcl/vol2/iss1/1; Abigail Payne, Does Inter-Judge Disparity Really Matter? An Analysis of the Effects of
Sentencing Reforms in Three Federal District Courts, 17 Int’l Rev. of Law & Econ. 337 (1997) (suggesting
that the disparities resulting from judicial discretion may be better than the increase to incarceration
caused by determinate sentencing).
33	

See Section V.

34	

See Table 1.

35	

See analysis in Section I.

36	This report’s findings and analysis, as explained in greater detail in Appendix A: Methodology, build on
data describing the number of prisoners incarcerated nationwide, and the average lengths of prison stays.
For data on federal prisoners, see Mark Motivans, Bureau of Justice Statistics, Federal Justice
Statistics, 2012 – Statistical Tables 37 tbl.7.9, 39 tbl.7.11 (2015), http://www.bjs.gov/content/pub/
pdf/fjs12st.pdf (listing, respectively, the number of persons incarcerated in the federal system by offense,
and the mean time served by offense).

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 59

Data on prison populations and average lengths of stay for 42 states were collected from the National
Corrections Reporting Program. This dataset is confidential, and the authors obtained access under a
restricted use agreement. Accordingly, this data cannot be shared publicly. See Nat’l Archive of Crim.
J. Data, National Corrections Reporting Program Resource Guide, https://www.icpsr.umich.edu/
icpsrweb/content/NACJD/guides/ncrp.html.
Data from 4 other states were collected directly from government sources. See Va. Dep’t of Corrections,
State Responsible Offender Population Trends FY2008 – FY2012 (2013), https://vadoc.virginia.
gov/about/facts/research/new-statsum/offenderpopulationtrends_fy08-fy12.pdf; Email from Mike Dolny,
Research Unit, Arizona Dep’t of Corrections, to the authors (Jul. 8, 2015) (on file with the authors); Email
from Mary Lansing, Associate Research Analyst, Ct. Dep’t of Corrections, to the authors (Jan. 21, 2106)
(on file with the authors); Email from George R. King, Ph.D., Research Statistician, Hawaii Dep’t of Pub.
Safety, to the authors (Jul. 8 2015) (on file with the authors).
The authors also collected data from Arkansas and Louisiana, but were unable to obtain data on prisoners
and lengths of stay from Vermont and New Mexico. However, Vermont and New Mexico prisoners
combined make less than 1 percent of the national prison population, making this report’s data substantially
complete. See E. Ann Carson, Bureau of Justice Statistics, Prisoners in 2014 3 tbl. 2 (2015), http://
www.bjs.gov/content/pub/pdf/p14.pdf.
This summary is hereinafter referred to below as the “Authors’ Analysis.” More information can be found
in Appendix A: Methodology.
37	For a description of the data sources underlying this report’s analysis, see endnote above describing Authors’
Analysis.
38	
See generally Nat’l Research Council, The Growth of Incarceration in the United States:
Exploring Causes and Consequences (Jeremy Travis, Bruce Western, & Steve Redburn eds., 2014).
39	
See Lauren-Brooke Eisen & Inimai Chettiar, Brennan Ctr. for Justice, The Reverse Mass
Incarceration Act 10 (2015), https://www.brennancenter.org/publication/reverse-mass-incarcerationact (summarizing and detailing BJS statistics).
40	
See Bureau of Justice Statistics, Prisoners 1925-1981 2 tbl.1 (1982), http://www.bjs.gov/content/
pub/pdf/p2581.pdf (providing data through 1978) and Bureau of Justice Statistics, Corrections
Statistics Analysis Tool (CSAT) – Prisoners, http://www.bjs.gov/index.cfm?ty=nps (follow “Quick
Tables” hyperlink; then follow “1978-2014” hyperlink under “Prisoners under the jurisdiction of state or
federal correctional authorities, December 31, 1978-2014” to download data table).
41	Stephanie Mencimer, There are 10 Times More Mentally Ill People Behind Bars Than in State Hospitals,
Mother Jones, Apr. 8, 2014, http://www.motherjones.com/mojo/2014/04/record-numbers-mentallyill-prisons-and-jails.
42	
See David Cloud, Vera Institute of Justice, On Life Support: Public Health in the Age of Mass
Incarceration 10 (2014), http://vera.org/pubs/public-health-mass-incarceration (citing Jennifer C.
Karberg & Dorris J. James, Bureau of Justice Statistics, Substance Dependence, Abuse, and
Treatment of Jail Inmates (2002); Seena Fazel et al., Substance Abuse and Dependence in Prisoners:
a Systematic Review, 101 Addiction 181, 181-189 (2006); Bridget F. Grant et al., Prevalence and Cooccurrence of Substance Use Disorders and Independent Mood and Anxiety Disorders: Results From the National
Epidemiologic Survey on Alcohol and Related Conditions, 61 Archives of General Psychiatry 807, 807816 (2004)).
E. Ann Carson & William J. Sabol, Bureau of Justice Statistics, Aging of the State Prison
43	
Population, 1993-2013 1-2 (2016), http://www.bjs.gov/content/pub/pdf/aspp9313.pdf.
44	
MIT Young Adult Development Project, Brain Changes (2008), http://hrweb.mit.edu/worklife/
youngadult/brain.html.

60 | Brennan Center for Justice

45	Medical expenses also make these inmates far more expensive than younger inmates. See KiDeuk Kim &
Bryce Peterson, Urban Inst., Aging Behind Bars: Trends and Implications of Graying Prisoners
in the Federal Prison System 4-6 (2014), http://www.urban.org/research/publication/aging-behindbars-trends-and-implications-graying-prisoners-federal-prison-system/view/full_report.
46	The authors calculated these estimates by taking the percentage of the state and federal prison populations
experiencing mental health and substance dependence issues from 2006 (the last year for which data was
available), assuming those ratios have remained unchanged, and applying those percentages to the 2014
state and federal prison populations. See E. Ann Carson, Bureau of Justice Statistics, Prisoners in
2014 5 tbl.3 (2015), http://www.bjs.gov/content/pub/pdf/p14.pdf; Doris J. James & Lauren E. Glaze,
Bureau of Justice Statistics, Mental Health Problems of Prison and Jail Inmates 5 tbl.5 (2006),
http://www.bjs.gov/content/pub/pdf/mhppji.pdf.
47	Rich Lord, Pennsylvania Prison System Develops Separate Housing For Mentally Ill Inmates, Pittsburgh
Post-Gazette, Jan. 6, 2015, http://www.post-gazette.com/news/state/2015/01/06/Pennsylvania-prisonsystem-develops-separate-housing-for-mentally-ill-inmates/stories/201501060042
48	
E. Fuller Torrey, M.D. et al., Treatment Advocacy Ctr. & Nat’l Sheriff ’s Ass’n, The Treatment of
Persons with Mental Illness in Prisons and Jails: A State Survey 6 (2014). http://www.tacreports.
org/storage/documents/treatment-behind-bars/treatment-behind-bars.pdf.
49	
See generally KiDuek Kim, Urban Inst., The Processing & Treatment of Mentally Ill Persons
in the Criminal Justice System (2015), http://www.urban.org/sites/default/files/alfresco/publicationpdfs/2000173-The-Processing-and-Treatment-of-Mentally-Ill-Persons-in-the-Criminal-Justice-System.
pdf; see also Ram Subramanian & Alison Shames, Vera Inst. of Justice, Sentencing and Prison
Practices in Germany and the Netherlands: Implications for the United States 11-13 (2013),
http://www.vera.org/sites/default/files/resources/downloads/european-american-prison-report-v3.pdf
(noting the number of jail inmates with mental health problems, and the damage done to them by
incarceration).
50	E. Fuller Torrey, American Psychosis: How the Federal Government Destroyed the Mental
Illness Treatment System 88 (Oxford Univ. Press ed., 2014).
51	
See, e.g., Darrell Steinberg et al., Stanford Law School Three Strikes Project, When Did Prisons
Become Acceptable Mental Healthcare Facilities? 5-6 (2015), http://law.stanford.edu/wp-content/
uploads/sites/default/files/child-page/632655/doc/slspublic/Report_v12.pdf (describing this process in
California); Treatment Advocacy Ctr., The Treatment of Persons with Mental Illness in Prisons
and Jails: A State Survey 101 (2014), http://www.tacreports.org/storage/documents/treatment-behindbars/treatment-behind-bars.pdf.
52	

 reatment Advocacy Ctr., The Treatment of Persons with Mental Illness in Prisons and Jails:
T
A State Survey 6, 101 (2014) http://www.tacreports.org/storage/documents/treatment-behind-bars/
treatment-behind-bars.pdf; David Cloud, Vera Inst. of Justice, On Life Support: Public Health in
the Age of Mass Incarceration 7 (2014), http://vera.org/pubs/public-health-mass-incarceration.

53	
See Figure 4.
54	
David Cloud, Vera Inst. of Justice, On Life Support: Public Health in the Age of Mass
Incarceration 12-14 (2014), http://vera.org/pubs/public-health-mass-incarceration.
55	
Margaret Noonan et al., Bureau of Justice Statistics, Mortality in Local Jails and State Prisons,
2000-2013 – Statistical Tables 1 (2015), http://www.bjs.gov/content/pub/pdf/mljsp0013st.pdf.
56	
See Bureau of Justice Statistics, Corrections Statistics Analysis Tool (CSAT) – Prisoners,
Bureau of Justice Statistics, http://www.bjs.gov/index.cfm?ty=nps (follow “Quick Tables” hyperlink).
57	
Paul Guerino et al., Bureau of Justice Statistics, Prisoners in 2010 27 appx. tbl.14 (2011) http://
www.bjs.gov/content/pub/pdf/p10.pdf; E. Ann Carson & William J. Sabol, Bureau of Justice
Statistics, Prisoners in 2011 8 tbl.8 (2012), http://www.bjs.gov/content/pub/pdf/p11.pdf; E. Ann

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 61

Carson & Daniella Golinelli, Bureau of Justice Statistics, Prisoners in 2012: Trends in
Admissions & Releases, 1991-2012 25 tbl.18 (2013), http://www.bjs.gov/content/pub/pdf/p12tar9112.
pdf; E. Ann Carson, Bureau of Justice Statistics, Prisoners in 2013 9 tbl.8 (2014), http://www.bjs.
gov/content/pub/pdf/p13.pdf; E. Ann Carson, Bureau of Justice Statistics, Prisoners in 2014 15
tbl.10 (2015), http://www.bjs.gov/content/pub/pdf/p14.pdf.
58	An “incarceration rate” includes both prison and jail inmates. For the imprisonment rate under this
report’s proposal, see Table 1. The authors then added in the current jail population to calculate the overall
incarceration rate under the report’s proposal. For the current incarceration rates of the United States
and other countries, see World Prison Brief, Highest to Lowest – Prison Population Total, http://www.
prisonstudies.org/highest-to-lowest/prison-population-total/trackback?field_region_taxonomy_tid=All
(listing most recent estimates of the incarceration rate in each country). This source is routinely updated
with new information; rates in this figure were drawn from the site as it appeared in September 2016. The
incarceration rate for the United Kingdom is the weighted average of England & Wales, Northern Ireland,
and Scotland.
59	
See generally Executive Office of the President, Council of Economic Advisers, Economic
Perspectives on Incarceration & the Criminal Justice System 15-20 (2016), https://www.
whitehouse.gov/sites/default/files/page/files/20160423_cea_incarceration_criminal_justice.pdf (finding
that rate of convictions, sentencing options, and length-of-stay increases all contributed to the growth of
mass incarceration); Pew Ctr. on the States, Time Served: The High Cost, Low Return of Longer
Prison Terms (2012), http://www.pewtrusts.org/~/media/assets/2012/06/6/time_served_report.pdf
(analyzing the growth in prison stays at state levels).
60	“Length of stay” is commonly used to describe the amount of time a person actually spends in prison,
which often differs from the person’s sentence length because low-risk inmates can frequently earn release
before serving their full sentences. For the data underlying this analysis, see Figure 8.
61	
Pew Ctr. on the States, Prison Time Surges for federal Inmates 1 (2015), http://www.pewtrusts.
org/~/media/assets/2015/11/prison_time_surges_for_federal_inmates.pdf.
62	
Thomas P. Bonczar et al., Bureau of Justice Statistics, Nat’l Corrections Reporting Program:
Sentence Length of State Prisoners, by Offense, Admission Type, Sex, and race (2011), http://
www.bjs.gov/index.cfm?ty=pbdetail&iid=2056 (follow the “1993” and “2009” hyperlinks to download
data tables for 1993 and 2009; then compare mean times served in prison by offense according to Table 9
(“NCRP9309.csv” and “ncrp0909.csv” for 1993 and 2009, respectively)).
63	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
64	

See Hon. Jed S. Rakoff, Mass Incarceration: the Silence of the Judges, N.Y. Rev. of Books (May 21, 2015),
http://www.nybooks.com/articles/2015/05/21/mass-incarceration-silence-judges (“[T]he supposition on
which our mass incarceration is premised — namely, that it materially reduces crime — is, at best, a
hunch. Yet the price we pay for acting on this hunch is enormous.”).


65	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
66	Sarah Desmarais & Jay Singh, Risk Assessment Instruments Validated and Implemented in Correctional
Settings in the United States: An Empirical Guide, Council of State Governments Justice Center
(2013), https://csgjusticecenter.org/wp-content/uploads/2014/07/Risk-Instruments-Guide.pdf.
67	Mark H. Bergstrom & Richard Kern, A View from the Field: Practitioners’ Response to Actuarial Sentencing:
An “Unsettled” Proposition, Criminology Pub. Pol’y (forthcoming 2011).
68	Sonja B. Starr, Evidence-Based Sentencing and the Scientific Rationalization of Discrimination, 66 Stanford
L.J.803, 809 (2014).

62 | Brennan Center for Justice

69	
PEW Ctr. on the States, Risk/Needs Assessment 101: Science Reveals New Tools to Manage
Offenders 1 (2011), http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2011/
pewriskassessmentbriefpdf.pdf.
70	James T. McCafferty, Unjust Disparities? The Impact of Race on Juvenile Risk Assessment
Outcomes, 25 Criminal Justice Policy Review, (2016); http://cjp.sagepub.com/content/
early/2016/02/25/0887403416634163.abstract.
71	
See United States Dep’t of Justice, Federal Bureau of Investigation, Uniform Crime Reporting
Statistics, State-by-State and National Crime Estimates by Year(s), http://www.ucrdatatool.gov/
Search/Crime/State/RunCrimeStatebyState.cfm (providing crime statistics from 1960 to 2013) and
United States Dep’t of Justice, Federal Bureau of Investigation, Uniform Crime Reporting
Statistics, Crime in the United States, 2015 (2016), https://ucr.fbi.gov/crime-in-the-u.s/2015/crimein-the-u.s.-2015/home (providing data through 2015). For this calculation, the overall crime rate is the
sum of the property and violent crime rates.
72	
See, e.g., Steven Raphael & Michael Stoll, The Hamilton Project, A New Approach to Reducing
Incarceration While Maintaining Low Rates of Crime 9-12 (2005), http://www.hamiltonproject.
org/assets/legacy/files/downloads_and_links/v5_THP_RaphaelStoll_DiscPaper.pdf.
73	
Oliver Roeder et al., What Caused the Crime Decline?, Brennan Ctr. for Justice 23 (2015),
https://www.brennancenter.org/publication/what-caused-crime-decline.
74	
See, e.g., Joan Petersilia & Susan Turner, Rand Corp., Prison versus Probation in California:
Implications for Crime and Offender Recidivism vii-viii (1986), https://www.rand.org/content/
dam/rand/pubs/reports/2007/R3323.pdf (noting that prison markedly increased recidivism for persons
convicted of property crimes); Gerald Gaes, Unintended Consequences: Experimental Evidence for the
Criminogenic Effect of Prison Security Level Placement on Post-Release Recidivism, 5 J. Experimental
Criminology 139 (2009), http://link.springer.com/article/10.1007/s11292-009-9070-z; Daniel S.
Nagin & G. Matthew Snodgrass, The Effect of Incarceration on Re-Offending: Evidence from a Natural
Experiment in Pennsylvania, 29 J. Quantitative Criminology 601, 642 (2012).
75	
See, e.g., Devah Pager, The Mark of a Criminal Record, 108 Am. J. Soc. 937 (2003), http://scholar.
harvard.edu/files/pager/files/pager_ajs.pdf; Bruce Western & Becky Pettit, Pew Charitable Trusts,
Collateral Costs: Incarceration’s Effect on Economic Mobility (2010), http://www.pewtrusts.
org/~/media/legacy/uploadedfiles/pcs_assets/2010/collateralcosts1pdf.pdf.
See Matthew R. Durose et al., Bureau of Justice Statistics, Recidivism of Prisoners Released in
76	
30 States in 2005: Patterns from 2005 to 2010 1 fig.1 (2014), http://www.bjs.gov/content/pub/pdf/
rprts05p0510.pdf; Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of
Prisoners Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf.
77	
See Editorial Board, A Formula to Make Bail More Fair, N.Y. Times, Sept. 16, 2016, http://www.nytimes.
com/2016/09/17/opinion/a-formula-to-make-bail-more-fair.html?_r=1.
78	Cassia Spohn & David Holleran, The Effect of Imprisonment on Recidivism Rates of Felony Offenders: A Focus
on Drug Offenders, 40 Criminology 329, 347 fig.1 (2002).
79	
See Daniel Mears et al., Gender Differences in the Effects of Prison on Recidivism, 40 J. Crim. Just. 370 (2012).
80	William D. Bales & Alex R. Piquero, Assessing the Impact of Imprisonment on Recidivism, 8 J. Experimental
Criminology 71, 76 n.4, 96 & tbl. 9 (2012).
81	
Joan Petersilia & Susan Turner, RAND Corp., Prison versus Probation in California:
Implications for Crime and Offender Recidivism vii-viii (1986), https://www.rand.org/content/dam/
rand/pubs/reports/2007/R3323.pdf; Paul Gendreau et al., Public Works and Government Services
Canada, The Effects of Prison Sentences on Recidivism (1999), http://www.prisonpolicy.org/scans/
e199912.htm (“The results were as follows: [relative to shorter sentences and alternative sanctions,] prison

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 63

produced slight increases in recidivism.”); Wash. State Inst. for Pub. Policy, Evidence-Based Adult
Corrections Programs: What Works and What Does Not 3 (2006), http://www.wsipp.wa.gov/
ReportFile/924.
82	Rachel Porter et al., Balancing Punishment and Treatment: Alternatives to Incarceration in New York City, 24
Fed. Sent’g Rep. 26, 26-29 (2011) (republishing 2002 Vera Institute investigation).
83	

See Daniel S. Nagin et al., Imprisonment and Reoffending, 38 Crime & Just. 115, 178 (2009).

84	Francis T. Cullen et al., Prisons Do Not Reduce Recidivism: The High Cost of Ignoring Science, 91 Prison J.
60S (2011).
85	Daniel S. Nagin & G. Matthew Snodgrass, The Effect of Incarceration on Re-Offending: Evidence from a
Natural Experiment in Pennsylvania, 29 J. Quantitative Criminology 601, 631-633 (2012), http://
repository.cmu.edu/cgi/viewcontent.cgi?article=1407&context=heinzworks.
86	Daniel S. Nagin & G. Matthew Snodgrass, The Effect of Incarceration on Re-Offending: Evidence from
a Natural Experiment in Pennsylvania, 29 J. Quantitative Criminology 601, 607 (2012), http://
repository.cmu.edu/cgi/viewcontent.cgi?article=1407&context=heinzworks (citing Donald P. Green &
Daniel Winik, Using Random Judge Assignments to Estimate the Effects of Incarceration and Probation on
Recidivism Among Drug Offenders, 48 Criminology 357-387 (2010); Danton Berube & Donald P. Green,
The Effects of Sentencing on Recidivism: Results from a Natural Experiment (2007) (unpublished paper) (on
file with Yale University); Charles E. Loeffler, Estimating the Effects of Imprisonment on the Life-Course
(2011) (unpublished Ph.D. dissertation, Harvard University)).
87	
Alliance for Safety & Justice, Crime Survivors Speak: The First-Ever National Survey of
Victims’ Views on Safety and Justice 20 (2016), https://www.allianceforsafetyandjustice.org/wpcontent/uploads/documents/Crime%20Survivors%20Speak%20Report.pdf.
88	For information on recidivism, see Joan Petersilia & Susan Turner, RAND Corp., Prison versus
Probation in California: Implications for Crime and Offender Recidivism vii-viii (1986), https://
www.rand.org/content/dam/rand/pubs/reports/2007/R3323.pdf (finding the percent of probationers
incarcerated within two-years of release to be 34 percent lower than the percent of prisoners incarcerated in
the same time period); Paul Gendreau et al., Public Works and Government Services Canada, The
Effects of Prison Sentences on Recidivism (1999) (“The results were as follows: [relative to shorter
sentences and alternative sanctions] prison produced slight increases in recidivism”). For information on
cost, see Pew Ctr. on the States, One in 31: The Long Reach of American Corrections 13 (2009),
http://www.pewtrusts.org/~/media/assets/2009/03/02/pspp_1in31_report_final_web_32609.pdf; Joan
Petersilia, Probation in the United States, Crime and Just. 149, 186 (1997) (“[I]n programs where offenders
received both surveillance (e.g. drug tests) and participated in relevant treatment, recidivism was reduced
20-30 percent.”).
See PEW Ctr. on the States, Risk/Needs Assessment 101: Science Reveals New Tools to Manage
89	
Offenders 1 (2011), http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2011/
pewriskassessmentbriefpdf.pdf.
90	
See PEW Charitable Trusts, Use of Electronic Offender-Tracking Devices Expands Sharply 1 (2016),
http://www.pewtrusts.org/~/media/assets/2016/10/use_of_electronic_offender_tracking_devices_expands_
sharply.pdf; James Kilgore, E-Carceration: The Problematic World of Being On an Electric Monitor, Alternet,
Oct. 20, 2016 http://www.alternet.org/human-rights/electronic-monitoring-restrictive-and-wrong.
91	
See John K. Roman et al., District of Columbia Crime Policy Inst., The Costs and Benefits of
Electronic Monitoring for Washington, D.C. 6 (2012) (analyzing seven evaluations of electronic
monitoring and finding that those under monitoring had an average rearrest rate 23.5 percent lower
than standard probationers and parolees), http://www.urban.org/sites/default/files/alfresco/publicationpdfs/412678-The-Costs-and-Benefits-of-Electronic-Monitoring-for-Washington-D-C-.PDF.

64 | Brennan Center for Justice

92	See David Cloud & Chelsea Davis, The Vera Inst. for Justice, Treatment Alternatives to Incarceration for
People with Mental Health Needs in the Criminal Justice System: The Cost-Savings Implications (2013),
http://archive.vera.org/sites/default/files/resources/downloads/treatment-alternatives-to-incarceration.pdf.
93	

 n evaluation by the National Center on Addiction and Substance Abuse found that participation in the
A
DTAP program reduced rearrest rates by 26 percent. Nat’l Ctr. on Addiction & Substance Abuse at
Columbia Univ., Crossing the Bridge: An Evaluation of the Drug Treatment Alternative-toPrison (DTAP) Program 6 (2003). An independent evaluation of Seattle’s Law Enforcement Assisted
Diversion found that participants had a 58 percent lower chance of arrest than a control group. Susan E.
Collins et al., Univ. of Wash. LEAD Evaluation Team, LEAD Program Evaluation: Recidivism
Report 3 (2015), http://static1.1.sqspcdn.com/static/f/1185392/26121870/1428513375150/LEAD_
EVALUATION_4-7-15.pdf?token=xTmZUgmkb4HTY3bnogXuOI8Uows%3D. Vermont’s cognitive
self-change program, targeting violent offenders with treatment in and out of prison, reduced the rate of
new criminal charges by 21 percent relative to a comparison group. Kris R. Henning & B. Christopher
Frueh, Cognitive-Behavioral Treatment of Incarcerated Offenders: An Evaluation of the Vermont Department
of Corrections’ Cognitive Self-Change Program, 23 Crim. J. & Behavior 522, 534 (1996).

94	
See Alicia Bannon et al., Brennan Ctr. For Justice, Criminal Justice Debt: A Barrier To Reentry
7-10 (2010).
95	
See Ram Subramanian & Alison Shames, Vera Inst. of Justice, Sentencing and Prison Practices
in Germany and the Netherlands: Implications for the United States 8-9 (2013), http://www.
vera.org/sites/default/files/resources/downloads/european-american-prison-report-v3.pdf (describing the
role of fines in the German justice system); United Nations Office on Drugs & Crime, Global Study
on Homicide (2013), http://www.unodc.org/gsh/en/data.html (from the landing page, select “Homicide
counts and rates, time series 2000-2012”) (showing a German 2011 homicide rate of 0.8 per 100,000,
compared to 4.7 in the United States). Homicide rates are a valid comparison because the definition of the
crime is unlikely to vary internationally.
96	Data from New Mexico and Vermont were not included in this report’s analysis. For a description of the
data sources underlying this report’s analysis, see endnote above describing Authors’ Analysis.
97	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
98	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
99	The other such crimes included driving under the influence, drug trafficking non-marijuana, drug
trafficking marijuana, fraud and forgery, grand larceny, and statutory rape.
100	For references in this section to the number of prisoners incarcerated for a specific offense, see Table
1. For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
101	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994
data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
102	
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 65

data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.
103	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994
data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.
104	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994
data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.
105	
Nat’l Archive of Crim. J. Data, National Corrections Reporting Program Resource Guide,
(2002) https://www.icpsr.umich.edu/icpsrweb/content/NACJD/guides/ncrp.html. The report lists the
cost to victim as $280 dollars which would be $375 dollars in today’s terms according to the Bureau of
Labor Statistics tool at http://data.bls.gov/cgi-bin/cpicalc.pl?cost1=280&year1=2002&year2=2016.
106	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994
data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
107	
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994
data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
108	
Released in 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf. While BJS has collected
recidivism data more recently, the more recent data is not broken down as narrowly by crime as the 1994
data, and is thus not as useful in this report’s crime by crime analysis. Additionally, the more recent data
is only for rearrest, not for reimprisonment. This and subsequent balancing tests use data from Table 9
of the 1994 report, stating the percentage of people who returned to prison within three years of release.
Comparing the 1994 data to more recent data sets not broken down by crime shows a high degree of
consistency over time.

66 | Brennan Center for Justice

109	
See, e.g., Wash. State Inst. for Pub. Policy, Evidence-Based Adult Corrections Programs: What
Works and What Does Not 3 (2006), http://www.wsipp.wa.gov/ReportFile/924.
110	 This category also includes some crimes of a sexual nature.
111	Christopher Ingraham, Americans Are More Bullish Than Ever on Legal Weed, Wash. Post, Oct. 26,
2016, https://www.washingtonpost.com/news/wonk/wp/2016/10/26/americans-are-more-bullish-thanever-on-legal-weed/ (citing recent polling data indicating that 64 percent of Americans support making
marijuana legal).
112	The estimated total number of first time DUI offenders in prison nationally is 290. Due to rounding, this
number rounded to zero and is therefore not included in the total numbers.
113	For most crimes, the authors’ analysis yielded a straightforward recommendation for whether incarceration
or an alternative was most appropriate. However, in the authors’ dataset, eight crimes — including burglary,
driving under the influence, drug trafficking, embezzlement, fraud and forgery, larceny, and statutory
rape — had broad definitions that encompass both serious versions (more appropriately punished by
incarceration) and less serious versions (more appropriately punished by alternatives to incarceration). For
these eight crimes, the authors first used the four factors to determine what conduct constituted “serious”
and “minor” versions of the crime, and then used more detailed data from a representative sample of states
to estimate how many prisoners nationwide fell into each category. For more details, see Appendix A:
Methodology.
114	This table summarizes all of the report’s recommendations. For a description of the data sources underlying
this report’s analysis, see endnote above describing Authors’ Analysis.
115	David Abrams, The Imprisoner’s Dilemma: A Cost Benefit Approach to Incarceration, 98 Iowa L.J. 907,
936 (2013).
116	
Paul Gendreau et al., The Effect of Prison Sentences & Intermediate Sanctions on Recidivism:
General Effects & Individual Differences (2002), http://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/
ffcts-prsn-sntncs/index-eng.aspx.
117	Ilyana Kuziemko, Going off Parole: How the Elimination of Discretionary Prison Release Affects the Social Cost
of Crime 21 (Nat’l Bureau of Econ Research, Working Paper No. 13380, 2007), http://www.nber.org/
papers/w13380.
118	 Id. at 19-22.
119	
Id. at 20-22 & n.28 (calling for further study of this effect). A modified version of this working paper was
published in 2013. See Ilyana Kuziemko, How Should Inmates Be Released From Prison? An Assessment of
Parole Versus Fixed Sentence Regimes, 128 Q.J. Econ. 371 (2013).
120	
See Steven D. Levitt, The Limited Role of Changing Age Structure in Explaining Aggregate Crime Rates, 37
Criminology 581, 592 (1999).
121	
See Charles R. Tittle et al., Gender, Age, and Crime/Deviance: A Challenge to Self-Control Theory, 40 J.
Research In Crime And Delinquency 426, 443 (2003).
122	Robert J. Sampson, Life-Course Desisters? Trajectories of Crime Among Delinquent Boys Followed to the Age of
70, 41 American Society of Criminology, 301, 312 (2003).
123	Bureau of Justice Statistics, Aging of the State Prison Population (1993-2009) 1, https://www.bjs.gov/
index.cfm?ty=pbdetail&iid=5602.
124	
Patrick Langan & David Levin, Bureau of Justice Statistics, Recidivism of Prisoners Released in
1994 11 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf.
125	Daniel S. Nagin et al., Imprisonment and Reoffending, 38 Crime & Justice 115, 169-175 (2009).
126	Thomas A. Loughran et al., Estimating a Dose-Response Relationship Between Length of Stay and Recidivism
in Serious Juvenile Offenders, 47 Criminology 699, 723-26 (2009).

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 67

127	Donald Green & Daniel Winik, Using Random Judge Assignment to Estimate the Effects of Incarceration and
Probation on Recidivism, 48 Criminology 357, 374 tbl.5, 380-81 (2010).
128	 S ee Jason Rydberg & Kyleigh Clark, Variation in the Incarceration Length Recidivism Dose-Response
Relationship, 125 Journal of Criminal Justice (2016), https://www.academia.edu/27391540/Variation_in_
the_Incarceration_Length-Recidivism_Dose-Response_Relationship.
129	Michael Mueller-Smith, The Criminal and Labor Market Impacts of Incarceration 25 (2015) (working
paper), http://sites.lsa.umich.edu/mgms/wp-content/uploads/sites/283/2015/09/incar.pdf.
130	Jeremy Travis et al., Nat’l Research Council, The Growth of Incarceration in the United States: Exploring
Causes and Consequences 139, 154 (2014). See also, Daniel S. Nagin, Deterrence in the Twenty-first Century:
A Review of the Evidence, 42 Crime and Justice: Crime and Justice in America 1975-2025 199, 231 (2013).
131	David Anderson, The Deterrence Hypothesis and Picking Pockets at the Pickpocket’s Hanging, 4 Amer. L. &
Econ. Rev. 295 (2002), http://aler.oxfordjournals.org/content/4/2/295.short.
132	Steven Raphael & Jens Ludwig, Prison Sentence Enhancements: The Case of Project Exile, in Evaluating
Gun Policy: Effects on Crime and Violence 252 (2003), http://home.uchicago.edu/ludwigj/papers/Exile_
chapter_2003.pdf.
133	David Lee. Justin McCrary, The Deterrence Effect of Prison: Dynamic Theory and Evidence, Princeton
University, July 2009, http://eml.berkeley.edu//~jmccrary/lee_and_mccrary2009.pdf
134	
See, e.g., Angela Hawken & Mark Kleiman, Managing Drug Involved Probationers with Swift
and Certain Sanctions: Evaluating Hawaii’s HOPE 64 tbl.A2 (2009), https://www.ncjrs.gov/
pdffiles1/nij/grants/229023.pdf (finding lower new arrest and revocations rates for participants in the
Hawaii Opportunity Probation with Enforcement program compared to other probationers).
135	For a description of the data sources underlying this report’s analysis, see endnote above describing Authors’
Analysis.
136	For data showing the increase in average length of stay for these offenses, see Bureau of Justice Statistics,
Nat’l Corrections Reporting Program: Sentence Length of State Prisoners, by Offense,
Admission Type, Sex, and Race (2011), http://www.bjs.gov/index.cfm?ty=pbdetail&iid=2056 (follow
the “1993” and “2009” hyperlinks to download data tables for 1993 and 2009; then compare mean time
served in prison in Table 9 (“NCRP9309.csv” and “ncrp0909.csv” for 1993 and 2009, respectively)).
See Let’s Cut Our Prison Population in Half, ACLU, https://action.aclu.org/secure/end-mass-incarceration;
137	
What We Do, #Cut50, http://www.cut50.org/mission (scroll down to “About Us”); Ella Baker Center on the
Move, Ella Baker Ctr. for Human Rights (Dec. 12, 2012), http://ellabakercenter.org/blog/2012/12/
ella-baker-center-on-the-move; The #FreeAmerica Pledge, #FreeAmerica, http://letsfreeamerica.com/
pledge; About Us, JustLeadershipUSA, https://www.justleadershipusa.org/about-us (scroll down to
“Mission Statement”).
138	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
139	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
140	To calculate savings, the authors multiplied the average cost of one year of imprisonment ($31,286), as
calculated in Christian Henrichson & Ruth delaney, Vera Inst. of Justice, The Price of Prisons:
What Incarceration Costs Taxpayers 9 (2012), http://www.vera.org/sites/default/files/resources/
downloads/price-of-prisons-updated-version-021914.pdf, by the number of people who would not
be incarcerated if the proposed reductions were fully implemented. The authors would prefer to base
calculations on long-term marginal costs of incarceration, but these costs were difficult to identify, due to
a lack of data.

68 | Brennan Center for Justice

141	For average lengths of stay discussed in this section, see Table 1.
142	James F. Austin, Evaluation of Broward Jail Population: Current Trends and Recommended Options (Oct.
2016) (unpublished manuscript) (on file with authors). The national estimate is a rough estimate by lead
author James Austin based on working directly with 100 jails and 47 state prison systems to help reduce
their populations.
143	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism Of Prisoners
Released In 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf.
144	James F. Austin, Evaluation of Broward Jail Population: Current Trends and Recommended Options (Oct.
2016) (unpublished manuscript) (on file with authors). The national estimate is a rough estimate by lead
author James Austin based on working directly with 100 jails and 47 state prison systems to help reduce
their populations.
145	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism Of Prisoners
Released In 1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf.
146	
James Austin et al., Unlocking America: Why and How to Reduce America’s Prison Population
19 (The JFA Inst. 2009); see also James Austin & Todd R. Clear, Reducing Mass Incarceration: Implications
of the Iron Law of Prison Populations, 3 Harv. L. & Pol’y Rev. 307, 311(2009).
147	
See e.g., Ctr. on Budget and Policy Priorities, Changing Priorities: State Criminal Justice Reforms
and Investments in Education (2014) (recommending that policymakers repeal truth-in-sentencing
requirements and mandatory minimum sentences), http://www.cbpp.org/research/changing-priorities-statecriminal-justice-reforms-and-investments-in-education; Jeremy Travis, Bruce Western, and Steve Redburn,
The Growth of Incarceration in the United States: Exploring Causes and Consequences (2014)
(suggesting states modify truth-in-sentencing laws, scale back “three strike enhancements,” relax mandatory
minimum sentencing laws, expand good time credit to increase parole eligibility, prevent reimprisonment for
parole violations, and retroactively apply sentencing guidelines changes).
148	Erica Goode, Average Prison Stay Grew 36 Percent in Two Decades, N.Y. Times, June 6, 2012, at A12, http://
www.nytimes.com/2012/06/06/us/average-prison-stay-grew-36-percent-in-two-decades.html?_r=1.
149	The 15-month estimate represents the weighted average of sentence reductions described in Table 2.
150	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
151	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
152	
See S.B. 1722, 1st Reg. Sess. (Fla. 2009), http://laws.flrules.org/files/Ch_2009-063.pdf.
153	Press Release, Florida Dep’t of Corrections, Florida Department of Corrections to Close Prisons, Work
Camps (Jan. 12, 2012), http://www.dc.state.fl.us/secretary/press/2012/01-12PrisonClosings.html.
154	For crime rates through 2012 reported in this section, see United States Dep’t of Justice, Federal
Bureau of Investigation, Uniform Crime Reporting Statistics, State-by-State and National
Crime Estimates by Year(s) (1985-2012), http://www.ucrdatatool.gov/Search/Crime/State/StatebyState.
cfm. For 2014 crime rates, see United States Dep’t of Justice, Federal Bureau of Investigation,
Crime in the United States, 2014 tbl.5 (2015), https://ucr.fbi.gov/crime-in-the-u.s/2014/crime-inthe-u.s.-2014/tables/table-5. Rates of incarceration are for sentenced prisoners under state control. See
William J. Sabol et al., Bureau of Justice Statistics, Prisoners in 2008 30 appx. tbl.10 (2009),
http://www.bjs.gov/content/pub/pdf/p08.pdf; Heather C. West & William J. Sabol, Bureau of
Justice Statistics, Prisoners in 2009 24, appx. tbl.9 (2010), http://www.bjs.gov/content/pub/pdf/p09.
pdf; Paul Guerino et al., Bureau of Justice Statistics, Prisoners in 2010 22 appx. tbl.9 (2011)
http://www.bjs.gov/content/pub/pdf/p10.pdf; E. Ann Carson & William J. Sabol, Bureau of Justice

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 69

Statistics, Prisoners in 2011 8 tbl.6 (2012), http://www.bjs.gov/content/pub/pdf/p11.pdf; E. Ann
Carson, Bureau of Justice Statistics, Prisoners in 2014 8 tbl.6 (2015), http://www.bjs.gov/content/
pub/pdf/p14.pdf.
155	S.B. 2136, Reg. Sess. (Miss. 2008), http://billstatus.ls.state.ms.us/documents/2008/html/SB/2100-2199/
SB2136CS.htm.
156	
ACLU, Smart Reform is Possible: States Reducing Incarceration Rates and Costs while
Protecting Communities 36 (2011), https://www.aclu.org/files/assets/smartreformispossible.pdf.
157	
Marc Levin, Tex. Pub. Pol’y Found., Adult Corrections Reform: Lower Crime, Lower Costs 1
(2011), http://www.rightoncrime.com/wp-content/uploads/2011/09/Texas-Model-Adult.pdf.
158	H.B. 1205, 82d Leg. (Tex. 2011); H.B. 2649, 82d Leg. (Tex. 2011); see also Lauren-Brooke Eisen
& Juliene James, Vera Inst. of Justice, Reallocating Justice Resources: A Review of 2011
State Sentencing Trends 26 (2012) http://archive.vera.org/sites/default/files/resources/downloads/
reallocating-justice-resources.pdf (surveying Texas house bills on earned credit programs).
159	
Bureau of Justice Statistics, Corrections Statistical Analysis Tool (CSAT) – Prisoners, http://
www.bjs.gov/index.cfm?ty=nps.
160	To calculate savings, the authors multiplied the average cost of one year of imprisonment ($31,286), as
calculated in Christian Henrichson & Ruth Delaney, Vera Inst. of Justice, The Price of Prisons:
What Incarceration costs Taxpayers 9 (2012), http://www.veraF.org/sites/default/files/resources/
downloads/price-of-prisons-updated-version-021914.pdf, by the number of people who would not
be incarcerated if the proposed reductions were fully implemented. The authors would prefer to base
calculations on long-term marginal costs of incarceration, but these costs were difficult to identify, due to
a lack of data.
161	Using BJS data on the average starting salary ($40,500) and training cost ($16,000) for police officers, and
adjusting for inflation, the authors found that the savings could cover the training and salaries for 368,000
new officers for ten years. This number is not adjusted to include salary increases or additional training.
See Brian A. Reaves, Bureau of Justice Statistics, Local Police Departments, 2007 6 (2010), http://
www.bjs.gov/content/pub/pdf/lpd07.pdf; Brian A. Reaves, Bureau of Justice Statistics, State and
Local Law Enforcement Training Academies, 2006 5 (2009), http://www.bjs.gov/content/pub/pdf/
slleta06.pdf; Bureau of Labor Statistics, Consumer Price Index Inflation Calculator, http://
www.bls.gov/data/inflation_calculator.htm.
162	The number of probation officers who could be employed is calculated using the median salary for probation
officers, as reported by BLS in 2015 ($49,360), adjusted for inflation. See Bureau of Labor Statistics,
Occupational Outlook Handbook, 2016-17 Edition, Probation Officers and Correctional
Treatment Specialists, http://www.bls.gov/ooh/community-and-social-service/probation-officers-andcorrectional-treatment-specialists.htm; Bureau of Labor Statistics, Consumer Price Index Inflation
Calculator, http://www.bls.gov/data/inflation_calculator.htm.
163	Using data from the National Center for Educations Statistics on the average starting salary ($55,418) for
high school teachers and elementary school teachers the authors found that the savings could cover the
salaries for 327,000 new teachers. All figures are in 2012 dollars. This number is not adjusted to include
salary increases, healthcare or additional training. See National Center for Educations Statistics,
Estimated Average Salary of Teachers in Public Elementary and Secondary Schools, tbl. 211.50,
(2014) https://nces.ed.gov/programs/digest/d14/tables/dt14_211.50.asp
164	According to the College Board, the average in-state public college tuition is $9,410 per year. Average
Published Undergraduate Sector, 2015-2016, The College Board, https://trends.collegeboard.org/
college-pricing/figures-tables/average-published-undergraduate-charges-sector-2015-16.
165	The 2015 budget of Los Angeles was $8.58 billion. City of Los Angeles, Budget Summary, FY 2015-16
2 (2016), http://cao.lacity.org/budget/summary/2015-16BudgetSummaryBooklet.pdf. The 2015 budget

70 | Brennan Center for Justice

of Chicago was $9.32 billion. City of Chicago, 2016 Budget Overview 21 (2016), http://www.
cityofchicago.org/content/dam/city/depts/obm/supp_info/2016Budget/2016BudgetOverviewCoC.pdf.
166	The requested budget for the Department of Labor for FY 2017 is $12.8 billion. Dep’t of Labor, FY
2017 Department of Labor in Brief 7 (2016), https://www.dol.gov/sites/default/files/documents/
general/budget/FY2017BIB_0.pdf. The President’s budget allotted $9.75 billion for the Department of
Commerce for FY 2017. Dep’t of Commerce, Fact Sheet: FY 2017 U.S. Department of Commerce
Budget (2016), https://www.commerce.gov/news/fact-sheets/2016/02/fact-sheet-fy-2017-us-departmentcommerce-budget.
167	The projected budget of NASA for FY 2017 is $19 billion. NASA, NASA FY 2017 Budget Request,
http://www.nasa.gov/sites/default/files/atoms/files/fy_2017_nasa_agency_fact_sheet.pdf.
168	The budget of Connecticut in FY 2017 is $20.4 billion. Ct. Office of Fiscal Analysis, Budget
Summary, FY 16-FY 17 Budget 2 (June 30, 2015), https://www.cga.ct.gov/ofa/Documents/year/
HLT/2015HLT-20150630_Budget%20Summary%20of%20the%20FY%2016%20-%20FY%20
17%20Budget.pdf.
169	
See endnotes above for sources.
170	
See Nat’l Archive of Crim. J. Data, National Corrections Reporting Program Resource Guide,
https://www.icpsr.umich.edu/icpsrweb/content/NACJD/guides/ncrp.html.
171	
See Bureau of Justice Statistics, Data Collection: National Corrections Reporting Program
(NCRP), http://www.bjs.gov/index.cfm?ty=dcdetail&iid=268.
172	
See Va. Dep’t of Corrections, State Responsible Offender Population Trends FY2008 – FY2012
(2013),
https://vadoc.virginia.gov/about/facts/research/new-statsum/offenderpopulationtrends_fy08fy12.pdf; Email from Mike Dolny, Research Unit, Arizona Dep’t of Corrections, to the authors (Jul.
8, 2015) (on file with the authors); Email from Mary Lansing, Associate Research Analyst, Ct. Dep’t of
Corrections, to the authors (Jan. 21, 2106) (on file with the authors); Email from George R. King, Ph.D.,
Research Statistician, Hawaii Dep’t of Pub. Safety, to the authors (Jul. 8 2015) (on file with the authors).
173	E. Ann Carson, Bureau of Justice Statistics, Prisoners in 2014 3 tbl.2 (2015), http://www.bjs.gov/
content/pub/pdf/p14.pdf.
174	
Mark Motivans, Bureau of Justice Statistics, Federal Justice Statistics, 2012 – Statistical Tables
37 tbl.7.9, 39 tbl.7.11 (2015), http://www.bjs.gov/content/pub/pdf/fjs12st.pdf (listing, respectively, the
number of persons incarcerated in the federal system by offense, and the mean time served by offense).
175	
See, e.g., Model Penal Code (Am. Law Inst., Official Draft 1962); Black’s Law Dictionary (10th ed.
2014); Wayne R. LaFave, Criminal Law (5th ed. 2010).
176	
See generally Justice Policy Inst., Defining Violence: Reducing Incarceration by Rethinking
America’s Approach to Violence (2016), http://www.justicepolicy.org/uploads/justicepolicy/
documents/jpi_definingviolence_final_report.pdf.
177	
See Paul H. Robinson et al., Extralegal Punishment Factors: A Study of Forgiveness, Hardship, GoodDeeds, Apology, Remorse, and Other Such Discretionary Factors in Assessing Criminal Punishment (Faculty
Scholarship Paper 353, 2012), http://scholarship.law.upenn.edu/cgi/viewcontent.cgi?article=1352&cont
ext=faculty_scholarship.
178	
See, e.g., Model Penal Code § 211.1 (Am. Law Inst., Official Draft 1962) (distinguishing “simple” and
“aggravated” assault based on degree of harm suffered).
179	
Bureau of Justice Statistics, Socio-Emotional Impact of Violent Crime (Sept. 2014), http://www.
bjs.gov/content/pub/pdf/sivc.pdf.
180	
Model Penal Code § 2.02 (Am. Law Inst., Official Draft 1962).

HOW MANY AMERICANS ARE UNNECESSARILY INCARCERATED? | 71

181	New York state offers a representative example. Compare N.Y. Penal Law 125.25 (2016) (proscribing
homicide committed “with intent to cause the death of another person,” and designating it a Class A-I
felony) with N.Y. Penal Law 125.10 (2016) (designating criminally negligent homicide a Class E, or less
serious, felony).
182	
Timothy Hughes & Doris James Wilson, Bureau of Justice Statistics, Reentry Trends in the U.S.
(2016), http://www.bjs.gov/content/reentry/reentry.cfm.
183	
See Matthew R. Durose et al., Bureau of Justice Statistics, Recidivism of Prisoners Released
in 30 States in 2005: Patterns from 2005 to 2010 1 (2014), http://www.bjs.gov/content/pub/pdf/
rprts05p0510.pdf.
184	
See Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners
Released in 1994 1 (2002) (providing data on different measures of recidivism), http://www.bjs.gov/
content/pub/pdf/rpr94.pdf (stating three-year reincarceration rates).
185	
Allen J. Beck & Bernard E. Shipley, Bureau of Justice Statistics, Recidivism of Prisoners Released
in 1983 (1990), http://www.bjs.gov/content/pub/pdf/rpr83.pdf; Patrick A. Langan & David J. Levin,
Bureau of Justice Statistics, Recidivism of Prisoners Released in 1994 (2002), http://www.bjs.
gov/content/pub/pdf/rpr94.pdf; Matthew R. Durose et al., Recidivism of Prisoners Released in
2005: Patterns from 2005 to 2010 (2014), http://www.bjs.gov/content/pub/pdf/rprts05p0510.pdf;
Kim Steven Hunt, et al., United States Sentencing Commission, Recidivism Among Federal
Offenders: A Comprehensive Overview (2016), http://www.ussc.gov/sites/default/files/pdf/researchand-publications/research-publications/2016/recidivism_overview.pdf.
186	
Patrick A. Langan & David J. Levin, Bureau of Justice Statistics, Recidivism of Prisoners Released in
1994 8 tbl.9 (2002), http://www.bjs.gov/content/pub/pdf/rpr94.pdf (stating three-year reincarceration rates).
187	
See Cyndi Banks, Criminal Justice Ethics: Theory and Practice 107, 117 (2004).
188	
Michel Foucault, Discipline and Punish: The Birth of the Prison 9 (Alan Sheridan trans., Vintage
Books, 2d ed. 1995) (1977).
189	Richard S. Frase, A More Perfect System: Twenty-Five Years of Guidelines Sentencing Reform Purposes, 58
(Stan. L. Rev.) 67, 73 (2005) (emphasis added).
190	Deuteronomy 19:21.
191	
See e.g. American Civil Liberties Union, Racial Disparities in Criminal Justice, available at https://
www.aclu.org/issues/mass-incarceration/racial-disparities-criminal-justice. See also Richard S. Frase,
Limiting Excessive Prison Sentences Under Federal and State Constitutions, 11 Journal of Constitutional
Law 1, 39-72 (2008), http://scholarship.law.upenn.edu/cgi/viewcontent.cgi?article=1142&context=jcl.
192	For a description of the data sources underlying this report’s analysis, see endnote above describing
Authors’ Analysis.
193	
See Lindsey Devers, Bureau of Justice Assistance, Plea and Charge Bargaining: Research
Summary 1 (2011); see also Hon. Jed S. Rakoff, Why Innocent People Plead Guilty, N.Y. Rev. of Books (Nov.
20, 2014), http://www.nybooks.com/articles/2014/11/20/why-innocent-people-plead-guilty/ (“In 2013, while
8 percent of all federal criminal charges were dismissed . . . more than 97 percent of the remainder were resolved
through plea bargains, and fewer than 3 percent went to trial.”).

72 | Brennan Center for Justice

STAY CONNECTED TO THE BRENNAN CENTER
Visit our website at www.brennancenter.org.
Sign up for our electronic newsletters at www.brennancenter.org/signup.

Latest News | Up-to-the-minute information on our work, publications, events, and more.
Election 2016 Newsletter | Latest developments, state updates, new research, and media roundup.
Justice Update | Snapshot of our justice work and latest developments in the field.
Redistricting Round-Up | Analysis of current legal battles and legislative efforts.
Fair Courts | Comprehensive news roundup spotlighting judges and the courts.
Liberty & National Security | Updates on privacy, government oversight, and accountability.
Twitter | www.twitter.com/BrennanCenter
Facebook | www.facebook.com/BrennanCenter
Instagram | www.instagram.com/BrennanCenter

NEW AND FORTHCOMING
BRENNAN CENTER PUBLICATIONS
Crime in 2016: A Preliminary Analysis
Matthew Friedman, Ames Grawert, and James Cullen
What Caused The Crime Decline?
Dr. Oliver Roeder, Lauren-Brooke Eisen, and Julia Bowling
The Reverse Mass Incarceration Act
Lauren-Brook Eisen and Inimai Chettiar
The New Era of Secret Law
Elizabeth (Liza) Goitein
Secret Spending in the States
Chisun Lee, Katherine Valde, Benjamin T. Brickner, and Douglas Keith
Judicial Selection for the 21st Century
John Kowal
The Fight to Vote
Michael Waldman
Democracy & Justice: Collected Writings, vol. IX
Brennan Center for Justice
For more information, please visit www.brennancenter.org

120 Broadway, Suite 1750
New York, NY 10271