DETAINING
Ill

:::c

I-

PIRlllSIOIN
POLICY INITIATIVE
i

MAY 2016

ABOUT THE AUTHORS
Bernadette Rabuy is the Senior Policy Analyst at the Prison Policy
Initiative. Bernadette produced the first comprehensive national report on
the video visitation industry, Screening Out Family Time: The for-profit
video visitation industry in prisons and jails, finding that 74% of local jails
that adopt video visitation eliminate traditional in-person visits. Her
research has played a key role in protecting in-person family visits in jails
in Portland, Oregon and the state of Texas. In her other work with the
Prison Policy Initiative, Bernadette has worked to empower the criminal
justice reform movement with key but missing data through the annual
Mass Incarceration: The Whole Pie reports as well as last year's report,
Prisons of Poverty, which for the first time provided the pre-incarceration
incomes of women in state prison.
Daniel Kopf is a data scientist in California and writer at
Priceonomics who has been a member of our Young Professionals
Network since February 2015. He has previously written about bail and
has co-authored several exciting statistical reports with the Prison Policy
Initiative: Prisons of Poverty: Uncovering the pre-incarceration incomes of the
imprisoned, Separation by Bars and Miles: Visitation in state prisons, and The
Racial Geography of Mass Incarceration. He has a Masters in Economics
from the London School of Economics.

ABOUT THE PRISON POLICY INITIATIVE
The non-profit, non-partisan Prison Policy Initiative was founded in
2001 to challenge over-criminalization and mass incarceration through
research, advocacy, and organizing. We show how the United States’
excessive and unequal use of punishment and institutional control harms
individuals and undermines our communities and national well-being.
The Easthampton, Massachusetts-based organization is most famous
for its campaigns to end prison gerrymandering and bring fairness to the
prison and jail phone industry, but is increasingly gaining national
recognition for compiling and presenting up-to-date information about
the criminal justice system nationally and in each state.

FOR MORE INFORMATION
For more up-to-date information about mass incarceration, see http://
www.prisonpolicy.org/national/. To subscribe to the Prison Policy
Initiative email newsletter, go to: http://www.prisonpolicy.org/subscribe/.

PR ISON
POLICY INITIATIVE

Prison Policy Initiative
PO Box 127 Northampton MA 01061
http://www.prisonpolicy.org

i

Detaining the poor: New report finds that people detained pretrial
are too poor to afford money bail
FOR IMMEDIATE RELEASE: May 10, 2016
Contact: Bernadette Rabuy brabuy [at] prisonpolicy.org
Easthampton, MA — People in local jails are significantly poorer than
non-incarcerated people, and even poorer than people in prison, finds a
new report by the nonprofit Prison Policy Initiative.
Detaining the Poor: How money bail perpetuates an endless cycle of
poverty and jail time connects the large pretrial population in local jails to
the criminal justice system’s reliance on money bail. “I kept hearing that
80% of defendants are indigent, but I was curious if people in local jails
are even poorer than people in prison. To get a better picture of the role
that money bail plays in the large unconvicted jail population in the U.S.,
we focused specifically on people unable to meet bail. I expected people
unable to meet bail to be poor, but I was surprised that a majority fall
within the poorest third of the national income distribution,” said
Bernadette Rabuy, who, along with data scientist Dan Kopf, last year
published a similar report on the pre-incarceration incomes of people in
state prison.
The latest numbers from the Bureau of Justice Statistics (BJS) reveal
that median bail for felony defendants was $10,000. “Using another BJS
dataset, theSurvey of Inmates in Local Jails, we found that the typical
detained defendant would need to spend eight months’ income to cover
$10,000 in money bail,” explained Kopf.
Detaining the Poor’s release coincides with newly published research
by the Federal Reserve showing that many Americans are unable come up
with $400 in an emergency without borrowing money from others or
selling something. “If the average American cannot easily come up with
$400, it is clear that a system that requires $10,000 from the poorest
members of our society for pretrial release is a system set up to fail,”
explained Rabuy.
The report provides the pre-incarceration incomes of people in local
jails who had the opportunity to be released pretrial, but were unable to
meet the conditions of bail. The report further breaks down the incomes
of the detained population by race, ethnicity, and gender. Additionally, the
authors compare pre-incarceration incomes to the incomes of similarly
aged non-incarcerated Americans.
While the report focuses on the incomes of people who were detained
for their inability to meet bail, the authors recognize the scarcity of useful
information about the jail populations in this country, so they also provide
the pre-incarceration incomes of people in local jails generally in an
appendix.
The new report, Detaining the Poor: How money bail perpetuates an
endless cycle of poverty and jail time is available at: http://
www.prisonpolicy.org/reports/incomejails.html
The report is a collaboration between the Prison Policy Initiative and
Dan Kopf, a member of the organization’s Young Professionals Network
and co-author of last year’s report on the pre-incarceration incomes of
people in state prisons.
i

DETAINING THE POOR:  

How money bail perpetuates an endless cycle of poverty and jail time
A Prison Policy Initiative report

Press release .............................................................i

Report ......................................................................1

Recommendations
1. Eliminate the use of money bail ...............................4
2. Stop locking people up for failure to pay fines and
fees..........................................................................5
3. Reduce the number of arrests that lead to jail
bookings through increased use of citations and
diversion programs ..................................................5
4. Increase funding of indigent criminal defense ..........5
5. Eliminate all pay-to-stay programs ...........................6
6. Reduce the high costs of phone calls home from
prisons and jails and stop replacing in-person jail
visits with expensive video visitation .........................6

Methodology
A. Background .............................................................7
B. Data sources and process ........................................7
C. On definitions ..........................................................8

Acknowledgements ..................................................9
Appendix .................................................................9

Sidebars
Do defendants have to pay the whole bail bond
amount? Why do I often hear that only 10% is
required? ....................................................................3
What is the difference between a bail and a bond? ....4

i

DETAINING THE POOR:  

How money bail perpetuates an endless cycle of poverty and jail time
A Prison Policy Initiative report
By Bernadette Rabuy and Daniel Kopf
May 10, 2016
In addition to the 1.6 million people incarcerated in federal and state
prisons, there are 646,000 people locked up in more than 3,000 local jails
throughout the U.S. Seventy percent of these people in local jails are being
held pretrial1 — meaning they have not yet been convicted of a crime and
are legally presumed innocent.2 One reason that the unconvicted
population in the U.S. is so large is because our country largely has a
system of money bail,3 in which the constitutional principle of innocent
until proven guilty only really applies to the well off. With money bail, a
defendant is required to pay a certain amount of money as a pledged
guarantee he will attend future court hearings.4 If he is unable to come up
with the money either personally5 or through a commercial bail
bondsman,6 he can be incarcerated from his arrest until his case is resolved
or dismissed in court.78
While the jail population in the U.S. has grown substantially since the
1980s, the number of convicted people in jails has been flat for the last 15
years. Detention of the legally innocent has been consistently driving jail
growth, and the criminal justice reform discussion must include a
discussion of local jails and the need for pretrial detention reform. This
report will focus on one driver of pretrial detention: the inability to pay
what is typically $10,000 in money bail.9 Building off our July 2015

Figure 1. Since the 1980s, there has been a significant, nationwide
move away from courts allowing non-financial forms of pretrial
release (such as release on own recognizance) to money bail,
although this does vary substantially depending on jurisdiction. 8 This
chart illustrates the possible paths from arrest to pretrial detention.
Almost all defendants will have the opportunity to be released
pretrial if they meet certain conditions, and only a very small

1

number of defendants will be denied a bail bond, mainly because a
court finds that individual to be dangerous or a flight risk. The only
national data on pretrial detention that we are aware of comes from
the Bureau of Justice Statistics’ Felony Defendants in Large Urban
Counties series. Nationally, in 2009, 34% of defendants were
detained pretrial for the inability to post money bail. This report
focuses on this important population: those who are detained
pretrial because they could not afford money bail.

report on the pre-incarceration incomes of people in prison, this report
provides the pre-incarceration incomes of people in local jails who were
unable to post a bail bond. This report aims to give the public and
policymakers the foundation for a more informed discussion about
whether requiring thousands of dollars in bail bonds makes sense given the
widespread poverty of the people held in the criminal justice system and
the high fiscal10 and social costs of incarceration.
We find that most people who are unable to meet bail fall within the
poorest third of society.11 Using Bureau of Justice Statistics data, we find
that, in 2015 dollars, people in jail had a median annual income of
$15,109 prior to their incarceration, which is less than half (48%) of the
median for non-incarcerated people of similar ages.12 People in jail are
even poorer than people in prison13 and are drastically poorer than their
non-incarcerated counterparts.
People in jail unable to meet bail
(prior to incarceration)
Men
Women

Non-incarcerated people
Men
Women

All

$15,598

$11,071

 

$39,600

$22,704

Black

$11,275

$9,083

 

$31,284

$23,760

Hispanic

$17,449

$12,178

 

$27,720

$14,520

White

$18,283

$12,954

 

$43,560

$26,136

Figure 2. Median annual pre-incarceration incomes for people in local jails unable to
post a bail bond, ages 23-39, in 2015 dollars, by race/ethnicity and gender. The
incomes in bold fall below the Census Bureau poverty threshold. The median bail
bond amount nationally is almost a full year’s income for the typical person unable to
post a bail bond.
Men
Women
All
61%
51%
Black
64%
62%
Hispanic
37%
16%
White
58%
50%
Figure 3. Percentage difference between the median pre-incarceration annual incomes
for people in local jails unable to post a bail bond and non-incarcerated people, ages
23-39, in 2015 dollars, by race/ethnicity and gender.

Unsurprisingly, white men have the highest incomes before
incarceration while Black women have the lowest incomes before
incarceration. The difference for Black men is particularly dramatic. Black
men in jail have a pre-incarceration median income 64% lower than that
of their non-incarcerated counterparts.
Examining the median pre-incarceration incomes of people in jail
makes it clear that the system of money bail is set up so that it fails: the
ability to pay a bail bond is impossible for too many of the people
expected to pay it. In fact, the typical Black man, Black woman, and
Hispanic woman detained for failure to pay a bail bond were living below
the poverty line before incarceration. The income data reveals just how
unrealistic it is to expect defendants to be able to quickly patch together
$10,000, or a portion thereof, for a bail bond. The median bail bond
amount in this country represents eight months of income for the typical
detained defendant.

2

“ People in jail are even poorer
than people in prison and are
drastically poorer than their
non-incarcerated counterparts.

80%
Distribution
of annual pre-incarceration incomes for people in
local jails unable to meet bail, ages 23-39

60%

50%
44%

40%

37%

Percent of
people with
that income

30%
20%

0%

< $9,489

$9,489-31,664

19%

20%

$31,665+

< $9,489

Incarcerated people

$9,489-31,644

$31,665+

Non-incarcerated people

Source: Compiled by the Prison Policy Initiative from the Bureau of Justice Statistics Survey of Inmates in Local Jails, 2002
and the American Community Survey Integrated Public Use Microdata Series, 2002 for non-incarcerated comparison. Incomes
were adjusted for inflation to 2015 dollars. See methodology at http://www.prisonpolicy.org/reports/incomejails.html.

Figure 4. People in local jails unable to meet bail are concentrated at the lowest ends
of the national income distribution, especially in comparison to non-incarcerated
people. As this graph shows, 37% had no chance of being able to afford the typical
amount of money bail ($10,000) since their annual income is less than the median bail
amount.

Because a system of money bail allows income to be the determining
factor in whether someone can be released pretrial, our nation’s local jails
are incarcerating too many people who are likely to show up for their
court date and unlikely to be arrested for new criminal activity.14
Although, on paper, it is illegal to detain people for their poverty, such
detention is the reality in too many of our local jails. Our country now has
a two-track system of justice in which the cost of pretrial liberty is far
higher for poor people than for the well off.

“ The median bail bond amount
in this country represents eight
months of income for the
typical detained defendant.

DO DEFENDANTS HAVE TO PAY THE WHOLE AMOUNT?
WHY DO I OFTEN HEAR THAT ONLY 10% IS REQUIRED?
Courts almost always offer a defendant the option
to pay cash or surety for his bond amount. “Cash”
means the defendant would pay 100% of the bond
amount set, and pay it directly to the court prior to
release from jail. If he makes his court appearances,
the court must return the amount he paid. “Surety”
options allow defendants to pay a portion of the bail
bond amount as a nonrefundable fee to a bail
bondsman or agency. The fee is typically 10%, although
it can be more or less than that. Even if the defendant
shows up for all of his court dates, he will not get the
fee back.
The use of “surety” or a commercial bail bondsman
is the most common way that defendants meet the
conditions of money bail for two main reasons. First,
most defendants cannot afford the 100% bond amount
set by courts, and, second, most courts are not
equipped to handle the transaction of processing
“returns” of cash when people keep their court dates.

3

As a result, some courts prefer to use surety companies
for this transaction.
Few jurisdictions have tried to make it easier for
more defendants to be released pretrial through other
cash alternatives or deposit bonds. For example, in
Illinois, all defendants who have a bail bond amount
set are automatically eligible to pay a 10% cash
alternative to the court. (A similar practice in
Massachusetts has effectively eliminated the private bail
bond industry.) If the defendant shows up for court, the
money is refunded, except for a small administrative
fee.
In New York City, a third of cases had a cash
alternative in 2010, which most commonly offered a
50% discount on the full bail bond amount. One study
found that a minimum cash discount of 60% would
make the cash alternative in New York City competitive
with the use of commercial bail bondsmen or agencies.
At discounts below 60%, defendants were still likely to
use commercial bail bondsmen.

RECOMMENDATIONS
This report shines light on another injustice of the American criminal
justice system — the unnecessary and excessive detention of poor people
in our local jails. To truly make our local communities safer and ensure
that bail decisions are based on more than how much money one has,
states, local governments, and sheriffs should:
1. Eliminate the use of money bail
Too many jails are detaining people not because they are dangerous,
but because they are too poor to afford bail bonds. One study of felony
defendants nationwide found that an additional 25% percent of
defendants could be released pretrial without any increases to pretrial
crime. The study found that many counties could safely release older
defendants, defendants with clean records, and defendants charged with
fraud and public order offenses, all without threatening public safety.
At a time when the White House is condemning money bail as “a crude
way to screen pretrial defendants for their risk of flight or to the
community,” states and local governments should eliminate money bail.
Instead, courts can increase their use of non-financial forms of pretrial
release such as release on own recognizance, which is when a defendant
signs an agreement that he will appear in court as required and is not
required to pay any money for pretrial release.15 Another option is to use
unsecured bonds16 instead of money bail. Through unsecured bonds, a
defendant is not required to pay any money to be released pretrial, but he
will be liable to pay an agreed upon amount of money if he does not
appear for court. Unsecured bonds are as effective at achieving public
safety and court appearance as money bail and much more effective at
freeing up jail beds. States and local governments interested in eliminating
money bail can follow the lead of Kentucky and the District of Columbia.
Kentucky banned for-profit money bail 40 years ago, and a statewide
agency instead uses a risk assessment tool17 to determine who will be
released pretrial. In D.C., most defendants are released on own
recognizance, during which time they are supervised by the D.C. Pretrial
Services Agency.18 Both Kentucky and D.C. have remained successful at
ensuring defendants show up for court and avoid arrest for new criminal
activity.

WHAT IS THE DIFFERENCE BET WEEN A BAIL AND A BOND?
Bail refers to the process of releasing a defendant
from jail with conditions set to reasonably assure public
safety and that a defendant will show up for court. Note
that the word “bail” does not mean the amount of
money that a defendant must pay for pretrial release,
although the term is commonly, but incorrectly, used
this way.
A bond is an agreement. Therefore, a bail bond is
4

an agreement between a defendant and the court —
sometimes involving money and sometimes not (release
on own recognizance) — to maintain public safety and
assure that a defendant will show up for court.
For additional key terms and definitions about
pretrial release and detention, visit the Pretrial Justice
Institute’s Glossary of Terms and Phrases Relating to
Bail and the Pretrial Release or Detention Decision.

2. Stop locking people up for failure to pay fines and fees
As the criminal justice system and its associated costs have grown,
states and local governments such as Ferguson, Missouri have adopted a
misguided policy: charging defendants fines and fees to pad their
correctional, and even municipal, budgets.19 Because, as this report
shows, the people who are being charged these fines and fees are largely
poor, states and local governments unsurprisingly have difficulty collecting
these funds20 and, for failing to pay criminal justice debt, people can land
in jail. For example, in Rhode Island in 2007, 18% of defendants were
locked up for criminal justice debt.
States and local governments should stop locking people up for failure to
pay criminal justice debt that they cannot afford, a practice repeatedly
deemed unconstitutional by the U.S. Supreme Court.21 If states and local
governments decide that charging fines and fees is worth the effort, they
should consider ability to pay when assessing fines and fees and be flexible
by allowing payment plans, community service in lieu of payment, and
exemption waivers for poor defendants. For example, in 2011,
Washington State passed legislation permitting waivers of interest that is
accrued on criminal justice debt while a person is locked up. After
researching the net gains of fee collections, Leon County, Florida decided
to close its Collections Court and terminated thousands of outstanding
arrest warrants.
3. Reduce the number of arrests that lead to jail bookings through
increased use of citations and diversion programs
Despite falling crime rates, the likelihood of arrest has increased
modestly for violent and property crimes and dramatically for drug crimes
over the past three decades.22 More arrests hinder criminal justice reform
by increasing the number of people locked up in the U.S. By choosing to
lock up people who need mental health and substance use services and not
jail time, American jails have become de facto mental health institutions.
One of the best yet underused reforms available to our local criminal
justice systems is for police to reduce the number of arrests that lead to jail
bookings. Instead, police can cite23 people so that defendants can wait for
their court date at home without having to post money bail or risk losing
employment. Kentucky, Maryland, and D.C. have all increased their use
of citations. States and local governments can also follow the lead of
Seattle, which implemented Law Enforcement Assisted Diversion
(LEAD). Through LEAD, police officers connect people — who are often
battling chronic homelessness, substance use, and mental health challenges
— to social services rather than bringing them to jail. LEAD has been
effective at both reducing arrests and slowing the rate at which people
arrested for low-level crimes cycle through Seattle jails.
4. Increase funding of indigent criminal defense
Most defendants are too poor to afford a private attorney to represent
them. Further, while the Supreme Court has affirmed the constitutional
right to counsel at initial appearance,24 in reality, only 10 states and
D.C.25 provide counsel at initial appearance. A study of defendants in
Baltimore found that the failure to provide legal representation when bail
bonds are determined was a leading reason for lengthy pretrial detention.
Defendants who were represented had a median jail stay of two days while
5

unrepresented defendants had a much longer median jail stay of nine days.
Greater funding of indigent criminal defense is imperative to making sure
that incarceration is only used when necessary and that the sentence is
proportional to the offense. Increased funding could ensure both that the
right to counsel is the reality for even the poorest defendants and that
more defendants have the guidance of an attorney earlier in the legal
process, such as when the bail bond amount is set or reviewed. One recent
example is in San Francisco, where the public defender’s office has
assembled a Bail Unit. Comprised of two lawyers, two paralegals, and
interns, the Bail Unit works to contest bail bonds for nearly all of the
public defender’s clients.
5. Eliminate all pay-to-stay programs
Jails and prisons in forty-one states charge incarcerated people for
room and board through pay-to-stay programs. For example, Riverside
County, California requires incarcerated people to pay $142 per day for
their incarceration. Now that the data in this report can confirm that the
majority of people that fill our local jails are poor, states and local
governments should resist the temptation to create new forms of criminal
justice fees, such as increasingly common pay-to-stay programs.
Otherwise, states and local governments risk spending more on the
administrative costs of collection than the little money they are able to
chase down. In 2013, Riverside County had collected less than 1% of
what it hoped to generate through its pay-to-stay program.
If states and local governments insist on having a pay-to-stay program,
they should, at the very least, make sure to research the likelihood that a
program would be worth it before implementation. For example, when a
committee was formed in Massachusetts to consider whether introducing
a room and board fee in prisons and jails would feasibly increase revenue,
the committee concluded that the harms of additional fees outweighed the
benefit.
6. Reduce the high costs of phone calls home from prisons and jails
and stop replacing in-person jail visits with expensive video
visitation
Phone calls home from prisons and jails and increasingly common
remote video visits typically cost $1 per minute. The high prices of these
communications products act like a regressive tax, charging the people
who have the least the most to keep in touch. As a result, more than a
third of families of incarcerated people fall into debt to cover phone and
visitation costs. And it is these same family members who defendants
often turn to when trying to scrape together money for bail bonds26 or
other criminal justice fines and fees.
While the Federal Communications Commission has been working to
protect incarcerated people and their family members from these high
costs, a federal court has stayed parts of the FCC’s last round of
regulations. States and local governments should not wait around until the
lawsuit is resolved. Instead, they should follow the lead of many states
such as New York and, recently, Mississippi by immediately reducing the
rates in their phone and video visitation contracts. States and local
governments should also protect the in-person visitation rights of people
in local jails and resist the temptation to replace free in-person visitation
6

with expensive computer chats. Recognizing how video visitation is costprohibitive for many families, the state of Texas as well as Multnomah
County, Oregon and Alameda County, California have all protected inperson visitation rights for people in local jails.

METHODOLOGY
Background
This is not the first report to address the incomes of incarcerated
people. The Bureau of Justice Statistics (BJS) collects this data periodically
(most recently in 2002 with another survey scheduled to begin in 2018)
but does not routinely publish the results in a format that can be accessed
without statistical software. The BJS last published a complete analysis of
the survey results in 2004. Last year, we produced a report on the preincarceration incomes of people in state prison. This report focuses on the
pre-incarceration incomes of people in local jails, and, even more
specifically, on the incomes of people incarcerated in local jails who had
the opportunity to post bail, but could not meet it. This report included
both convicted (people who were detained for the entire pretrial period
and then sentenced to jail time) and unconvicted people because only
looking at the unconvicted population detained pretrial would have been
too small of a sample size.
This report was not intended to make the point that incarceration
causes poverty, although there is extensive research on that topic. Because
the Prison Policy Initiative is regularly asked about the role that poverty
plays in who ends up behind bars, this report is aimed at answering a
different question: are incarcerated people poorer than non-incarcerated
people? In particular, we often hear that 80% of defendants nationwide
are indigent, but we wanted to be able to put numbers to the problem in
the hopes of furthering the conversation on the growth of pretrial
detention and the urgent need for bail reform.
To be clear, this report relies on the Bureau of Justice Statistics survey
from 2002, which is both quite old and the newest available. While we
look forward to the Bureau of Justice Statistics administering a survey in
2018, it will take a few years before that data is published, and we know of
no reason or trend that would make relying on the 2002 survey less
reliable than the alternatives of using data that is even older or no data at
all.
Further research should look at the effects of educational attainment
and prior prison or jail time on the pre-incarceration incomes and identify
policies that could address those disparities.

Data sources and process
This report is the result of a collaboration between Bernadette Rabuy,
Senior Policy Analyst at the Prison Policy Initiative, and data scientist
Daniel Kopf, who joined our Young Professionals Network in February
2015.

7

Together, we studied the BJS Survey of Inmates in Local Jails, 2002
relying in particular on the questions listed below and then developing a
way to make the data comparable to non-incarcerated people.
• S7Q11c. Which category on this card represents your personal
monthly income from ALL sources for the month before your arrest?
• S1Q1a. Sex
• S1Q2a. What is your date of birth?
• S1Q3a. Are you of Spanish, Latino, or Hispanic origin?
• S1Q3c. Which of these categories describes your race?
• S3Q3a. At any time after your arrest for the … charge(s) was bail or
bond set?
• S3Q3c. Were you released on bail or bond?
The non-incarcerated data comes from the Census Bureau’s American
Community Survey (ACS), specifically from the Integrated Public Use
Microdata Series (IPUMS). We used data from 2002 both because this
was the same year as the incarcerated survey data, and because the ACS in
2002 included only people in households and did not include jails and
other group quarters.
Because income is correlated with age and because the incarcerated
population trends younger than the general U.S. population, we thought
it would be most accurate to compare people of similar ages. We limited
our study to the 25th and 75th percentiles of ages for incarcerated people
(ages 23-39), and we used the same age range for the non-incarcerated
population.
To make all of this data more accessible and useful, we converted all
data in two ways: We converted monthly incomes to annual incomes by
multiplying by 12, and we multiplied each income by 1.32 to adjust for
inflation from 2002 to 2015, as provided by the Bureau of Labor Statistics
CPI Inflation Calculator. (We chose to adjust to 2015 dollars instead of
2016 because the 2016 index value will change from month to month.
Because 2016 is not yet over, the 2016 index value is based only on the
latest monthly values.)
In addition, to provide an estimated median income for each
incarcerated race/ethnicity/gender group from the BJS “grouped
frequency” data, we followed these steps:
1. Take the distance between the smallest and largest number in the
group containing the median
2. Multiply this number by the following: ( ( (total data points/2) total data points in groups with lower numbers) / data points in
group containing median )
3. Add lowest number in group containing the median

On definitions
Note that throughout this report, the incomes for incarcerated people
are the incomes incarcerated people reported earning before their arrest,
not the incomes they earned through work programs behind bars. For
incarcerated people and non-incarcerated people, incomes include welfare
and other public assistance. For incarcerated people, incomes also include
illegal sources of income.
We use “Non-incarcerated” to refer to people in households, and
thereby exclude people in group quarters, including people in correctional
8

facilities, psychiatric hospitals, college/university housing, or residential
treatment facilities.
Our data on “Blacks” and “Whites,” relies on data for Non-Hispanic
Blacks and Non-Hispanic Whites. The federal government defines Black
and White as races while Hispanic is defined as an ethnicity (and,
therefore, it is possible to identify as both Hispanic and White or Hispanic
and Black). Our data for both incarcerated people and non-incarcerated
people allowed us to avoid overlap by separately talking about NonHispanic Whites, Non-Hispanic Blacks, and Hispanics.

AC KNOWLEDGMENTS
This report was made possible thanks to the generous contributions of
individuals across the country who support justice reform. Figure 1 and
the jail population by convicted status animated graph were made possible
by Elydah Joyce and Jacob Mitchell of the organization’s Young
Professionals Network, respectively. Bob Machuga created the covers, and
the rest of the Prison Policy Initiative staff helped the authors gather
research. The authors would like to thank the Pretrial Justice Institute
both for providing invaluable feedback on an early draft of the report and
the great work they do for pretrial justice. Any errors or omissions,
however, are the sole responsibility of the authors.

APPENDIX
The 2002 BJS survey asked incarcerated people what their personal
monthly income was the month before their arrest. First, the data in this
appendix is presented in monthly incomes and has not been adjusted for
inflation. Second, the data in this appendix is presented for people in local
jails in general. In this report, we focus on the people in local jails who
had the opportunity to be released pretrial but were unable to meet the
conditions of bail.
The following tables and graphs allow for comparisons between the
incomes of incarcerated people prior to incarceration and the incomes of
non-incarcerated people for each of the income categories that BJS
provides respondents in its Survey of Inmates in Local Jails. The graphs
also show that incarcerated people are dramatically concentrated at the
lower ends of the national income distribution.

9

All
Black
Hispanic
White

People in jails
(prior to incarceration)
Men
Women
$1,061
$671
$900
$568
$1,114
$709
$1,236
$813

 
 
 
 

Non-incarcerated people
Men
Women
$2,500
$1,433
$1,975
$1,500
$1,750
$917
$2,750
$1,650

Figure 5. Median monthly pre-incarceration incomes for people in local jails, in 2002
dollars, ages 23-39, by race/ethnicity and gender.

Figure 6. Distribution of monthly pre-incarceration incomes for men in local jails and
non-incarcerated men, 2002 dollars, ages 23-39

Figure 7. Distribution of monthly pre-incarceration incomes for men unable to meet
bail and non-incarcerated men, 2002 dollars, ages 23-39

10

Proportion of men
Proportion of men unable to meet
in local jails with
bail with that
that income
income
Proportion of non(prior to
(prior to
incarcerated men with
Income category incarceration)
incarceration)
that income
$0
4.37%
4.21%
3.22%
$1-99
2.35%
3.71%
1.13%
$100-199
3.48%
3.92%
1.06%
$200-299
5.04%
6.17%
0.95%
$300-399
3.75%
3.9%
0.96%
$400-499
4.99%
6.82%
1.26%
$500-599
7.79%
7.71%
2.02%
$600-799
6.44%
6.22%
2.60%
$800-999
8.81%
7.9%
3.22%
$1,000-1,199
9.78%
8.75%
4.45%
$1,200-1,499
11.70%
11.47%
5.88%
$1,500-1,999
9.72%
10.08%
11.08%
$2,000-2,499
8.30%
6.13%
11.18%
$2,500-4,999
8.66%
7.99%
36.15%
$5,000+
4.82%
5.01%
14.84%
Figure 8. Proportion of men in local jails (prior to incarceration), men unable to meet
bail (prior to incarceration), and non-incarcerated men that fall within a monthly
income category, 2002 dollars, ages 23-39

Figure 9. Distribution of monthly pre-incarceration incomes for women in local jails
and non-incarcerated women, 2002 dollars, ages 23-39. While most people in local
jails make less prior to incarceration than people on the outside, there is one
interesting anomaly in the data for women not present in the data for men. More nonincarcerated women report no income at all than incarcerated women prior to
incarceration. For both groups, the reported incomes include wages, welfare, and
other public assistance, but since these are individual surveys, they do not include
spousal income. It is likely that many of those non-incarcerated women with zero
reported income are receiving support from their spouses.

11

Figure 10. Distribution of monthly pre-incarceration incomes for women unable to
meet bail and non-incarcerated women, 2002 dollars, ages 23-39
Proportion of
Proportion of
women in local women unable to
jails with that meet bail with that
income
income
Proportion of non(prior to
(prior to
incarcerated women
incarceration)
incarceration)
with that income

Income category
$0
10.27%
7.13%
14.3%
$1-99
3.85%
4.29%
3.1%
$100-199
5.40%
6.38%
2.3%
$200-299
7.89%
8.57%
2.2%
$300-399
7.39%
6.37%
2.0%
$400-499
6.16%
7.12%
2.5%
$500-599
10.98%
5.45%
3.4%
$600-799
8.91%
9.45%
4.1%
$800-999
8.67%
7.65%
4.8%
$1,000-1,199
9.40%
7.15%
5.4%
$1,200-1,499
9.02%
6.81%
6.6%
$1,500-1,999
5.89%
4.99%
11.4%
$2,000-2,499
7.14%
7.35%
9.8%
$2,500-4,999
6.09%
7.13%
23%
$5,000+
3.21%
4.17%
5.2%
Figure 11. Proportion of women in local jails (prior to incarceration), women unable to
meet bail (prior to incarceration), and non-incarcerated women that fall within a
monthly income category, 2002 dollars.

12

FOOTNOTES
1 The research on the median or average length of pretrial detention will depend on

the jurisdiction. For example, in a study of New York City defendants, the median
length of pretrial detention was 5 days for nonfelony cases and 7 days for felonies in
2003-2004. See: Mary T. Phillips, A Decade of Bail Research in New York City
(New York, NY: New York City Criminal Justice Agency, Inc., 2012), p 109. In
another study of 153,407 defendants in Kentucky from 2009-2010, the average
length of pretrial detention was 35 days for felonies and 7 days for misdemeanors.
2 In some places, the unconvicted population is even larger. In San Francisco,

California, 85% of the local jail population is legally presumed innocent.
3 There are some important exceptions that show that money bail is not necessary to

make sure that defendants show up for court and avoid new crimes. For example, in
the District of Columbia, 80% of people charged with an offense are released on
nonfinancial bail bond options such as release on own recognizance, which is when a
defendant signs an agreement that he will appear in court as required and is not
required to pay any money for pretrial release. Even though D.C. rarely uses money
bail, the Pretrial Services Agency has a very high rate of pretrial success: 88% of
defendants successfully appear in court and avoid new offenses. See: Melissa Neal,
Bail Fail: Why the U.S. Should End the Practice of Using Money for Bail
(Washington, D.C.: Justice Policy Institute, 2012), p 40. Kentucky is another leader
in trying to reduce the use of money bail. In 1976, Kentucky made it illegal to profit
from bail and instead uses a risk assessment tool to determine who will be released
pretrial. Illinois, Oregon, and Wisconsin are the only other states that have banned
for-profit bail. See: Melissa Neal, Bail Fail: Why the U.S. Should End the Practice of
Using Money for Bail (Washington, D.C.: Justice Policy Institute, 2012), p 42.
4 If a defendant is able to come up with the money for his bail bond, he will be

released pretrial. Then, in theory, a majority of the bail bond amount will be returned
to him after his court date. For example, in New York State, a defendant gets the
entire bail bond amount if he is not convicted. If he is convicted, he will get the bail
bond amount back except for a 3% fee. See: Mary T. Phillips, A Decade of Bail
Research in New York City (New York, NY: New York City Criminal Justice Agency,
Inc., 2012), p 79. However, there is an important caveat. Because almost all
defendants nationwide use commercial bail bondsmen to meet money bail (a practice
that is rarer in New York City), it is extremely rare for a defendant to get any money
back if he does show up for court. He usually loses at least 10% of the money bail as a
nonrefundable fee to a bail bondsman or agency. See footnote 6.
5 It is common for defendants to look to their family and friends for help in being

able to afford a bail bond. In New York City in 2005, immediate family members
were the ones to post the bail bond amount in 41% of cases when they gathered the
money among themselves and 49% of cases when a commercial bail bondsman was
used. See: Mary T. Phillips, A Decade of Bail Research in New York City (New York,
NY: New York City Criminal Justice Agency, Inc., 2012), p 85.
6 This is the most common way that defendants are able to meet money bail. The use

of a commercial bail bondsman is also known as a surety bond. The defendant or his
or her family pays a commercial bail bondsman or agency what is around 10% of the
bail bond amount as a non-refundable fee. Sometimes the fee is higher, and often
collateral in the form of cash or property is involved as well. In exchange, the
commercial bail bondsman agrees to pay the court the full bail bond amount if the
defendant does not appear for his court date. Even if the defendant shows up for his
court date, he will not get the fee back from the commercial bail bondsman, although
collateral is usually refunded. See footnote 3 for the states that ban for-profit money
bail.
13

7 Some people who are detained for failure to post money bail will later have their

charges dropped or have the opportunity to instead receive services. A Bureau of
Justice Statistics study of felony defendants in the largest counties nationwide found
that 19% of detained defendants between 1990-2004 had their cases dismissed or
were acquitted. See: Thomas H. Cohen and Brian A. Reaves, Pretrial Release of
Felony Defendants in State Courts (Washington D.C.: Bureau of Justice Statistics,
2007) p 7. In another study of 1,900 defendants in ten Colorado counties, half of
defendants detained for the full pretrial period would later have their charges
dropped or return to the community through diversion or probation. Therefore,
these are people who are not just being detained while they are legally innocent, many
may not have committed the crime they are being charged with in the first place.
8 For example, in Dallas County, Texas, non-financial release (such as release on own

recognizance) is almost always only used after a defendant cannot first afford money
bail. See: Michael R. Jones, Unsecured Bonds: The As Effective and Most Efficient
Pretrial Release Option (Washington, D.C.: Pretrial Justice Institute, 2013) p 23.
9 The only national data we are aware of is for felony defendants in 2009, which

found that $10,000 was the median money bail. We are not sure how money bail
amounts have changed since 2009, but if money bail has kept pace with inflation, the
median money bail is $11,048 in 2015 dollars. The Bureau of Justice Statistics should
be releasing a new report on felony defendants in the largest counties soon, possibly
later this year. This report will include an updated figure for median money bail. In
one study in New York City, the median bail amount for both felonies and
misdemeanors was significantly lower: $2,000 in 2010 ($2,174 adjusted for inflation
to 2015 dollars). See: Mary T. Phillips, A Decade of Bail Research in New York City
(New York, NY: New York City Criminal Justice Agency, Inc., 2012), p 45. Setting
lower bail bond amounts could be one short-term option in a longer process to
reduce a local jail population, but its effects have been limited. For example, previous
research has found that even when a bail bond was set below $500, a majority of New
York City defendants still couldn’t afford it. See: Mary T. Phillips, <A Decade of Bail
Research in New York City (New York, NY: New York City Criminal Justice Agency,
Inc., 2012), p 51. Similarly, in Virginia, 92% of defendants in 2012 who were
detained for failure to post bail had been unable to pay a bail bond of $5,000 or less.
See: Kenneth Rose, A “New Norm” For Pretrial Justice in the Commonwealth of
Virginia (Richmond, VA: Virginia Department of Criminal Justice Services, 2013) p
19.
10 Pretrial detention costs our country billions of dollars each year. The Justice Policy

Institute estimates that poor defendants and their families pay the commercial bail
bond industry about $1.4 billion in a year. In a speech in 2011, former Attorney
General Eric Holder estimated that detaining the legally innocent costs our criminal
justice system roughly $9 billion taxpayer dollars.
11 Our analysis finds that over 60% of the people unable to post bail bonds fall within

the poorest third of society. Eighty percent fall within the bottom half.
12 In the unadjusted 2002 dollars, people in jail who were unable to post a bail bond

had a median annual income of $11,468 prior to their incarceration, which is 48% of
the median for non-incarcerated people of similar ages.

14

13 In Prisons of Poverty: Uncovering the pre-incarceration incomes of people in prison, we

found that, in 2014 dollars, people in prison had a median annual income of $19,185
prior to their incarceration. According to the Bureau of Labor Statistics’ CPI
inflation calculator, this is equivalent to $19,208 in 2015 dollars. Therefore, in 2015
dollars, the median annual income of a person in a local jail was, prior to their
incarceration, 79% of the median pre-incarceration income for a person in state
prison. One possible reason that people in local jails are poorer than people in prison
may be because some research has found that there is an even greater percentage of
people with mental illness in jails than in prisons.
14 Pretrial detention has lasting effects on the likelihood of conviction, the likelihood

of receiving a sentence of incarceration, and even sentence length. One controlled
study by the Laura and John Arnold Foundation found that defendants detained
until trial were more likely to be sentenced to jail and with longer sentences as well as
more likely to be sentenced to prison and with longer sentences. Another study of
New York City defendants included many different analyses but repeatedly came to
the same conclusion: pretrial detainees were more likely to be convicted, and if
convicted, more likely to be sentenced to incarceration, and if incarcerated, more
likely to have longer sentences. See: Mary T. Phillips, A Decade of Bail Research in
New York City (New York, NY: New York City Criminal Justice Agency, Inc., 2012),
p 127.
15 This includes both citations, which can be given by police officers instead of

arresting someone, and unsecured bonds, which is when a defendant does not have to
pay any money to the court to be released pretrial, but will be potentially liable to pay
the full bail bond amount if he does not show up for court. See pages 4-5 of the
Pretrial Justice Institute’s Glossary of Terms and Phrases Relating to Bail and the
Pretrial Release or Detention Decision.
16 Unsecured bonds are a type of release on own recognizance that includes a promise

of money if the defendant does not show up for court.
17 Jurisdictions use risk assessments at various stages of the criminal justice system: to

make pretrial release or detention decisions, to determine the appropriate level of
probation and parole supervision, and, recently, to decide the length of sentences. At
their most basic level, risk assessment tools are questionnaires that assign points to a
defendant or incarcerated person based on a number of factors -- from criminal
history to, sometimes, family background-- to determine the person’s risk. The tools
sometimes include interviews performed by a caseworker or probation officer while
other tools are scored solely from administrative records. The factors will also vary by
jurisdiction, and even what the tool is measuring may differ slightly. For example, a
pretrial risk assessment tool will likely focus mainly on the likelihood that a person
will show up for court and not engage in criminal activity during pretrial release,
while a risk assessment tool at sentencing will focus on the likelihood of future
crimes. Risk assessments have been met with some skepticism, with even former
Attorney General Eric Holder expressing concerns about whether risk assessments at
sentencing will increase the existing disparities in the criminal justice system. The
public and policymakers should feel empowered to ask hard questions about what
specific factors are being used in a jurisdiction’s particular risk assessment tool,
whether the tool has been validated for that specific jurisdiction, and whether the
tool is likely to increase disparities. For example, opponents of risk assessment have
been concerned about the use of factors that are unrelated to conduct and arguably
out of one’s control such as neighborhood or family members’ criminal history.

15

18 In D.C., supervision can include drug testing, reporting to a case manager, or

electronic surveillance. See: Melissa Neal, Bail Fail: Why the U.S. Should End the
Practice of Using Money for Bail (Washington, D.C.: Justice Policy Institute, 2012),
p 41. A 2013 study by the Laura and John Arnold Foundation found that defendants
who were supervised were more likely to show up for court than non-supervised
defendants. However, supervision can also be applied unnecessarily. To avoid
expanding the criminal justice system’s reach to people who are already likely to show
up for court without any supervision, restraint is necessary.
19 Fines and fees are technically different. Fines are monetary punishments for

offenses while fees are payments for anything from court activities to probation
supervision that are intended to support operational costs. See: Alicia Bannon, Mitali
Nagrecha, and Rebekah Diller, Criminal Justice Debt: A Barrier to Reentry (New
York, NY: Brennan Center for Justice, 2010).
20 Florida was able to collect less than 14% in felony fines and fees assessed while

Maryland was able to collect 17% in parole supervision fees. The result is too often
modern day debtors’ prisons in which defendants who cannot pay are being jailed.
21 Illegally locking people up for failure to pay fines can be costly for jurisdictions. For

example, a recent settlement will require Colorado Springs, Colorado to pay a total of
$103,000. Each person who was illegally locked up will receive $125 for each day
they were wrongly held.
22 Violent and property crime arrest rates have both increased by approximately 20%

since 1980. In this same time period, drug arrest rates have increased by over 90%.
See: Council of Economic Advisers, Economic Perspectives on Incarceration and the
Criminal Justice System (Washington, D.C.: Executive Office of the President of the
United States, 2016) p 13-14.
23 A citation is an order to appear before a judge on a date to defend oneself against a

stated charge. Using citations in lieu of arrests would reduce both pretrial detention
and jail populations in general. See page 6 of the Pretrial Justice Institute’s Glossary of
Terms and Phrases Relating to Bail and the Pretrial Release or Detention Decision.
24 Bail is almost always set at initial appearance, although it is sometimes set before

this. For more information on initial appearance, see the definition for “first
appearance”; in the Pretrial Justice Institute’s Glossary of Terms and Phrases Relating
to Bail and the Pretrial Release or Detention Decision.
25 The ten states that, as of a survey in 2008-2009, provide counsel at initial

appearance are: California, Connecticut, Delaware, Florida, Hawaii, Massachusetts,
Maine, North Dakota, Vermont, and Wisconsin. According to a Supreme Court case,
ten states do not provide representation at initial appearance: Alabama, Colorado,
Kansas, Michigan, Mississippi, New Hampshire, Oklahoma, South Carolina, Texas,
and Virginia. The practices of the other states vary from county to county. See:
Douglas Colbert, “Prosecution Without Representation” Buffalo Law Review Vol.
59:333 (2011).
26 See footnote 5.

16