Skip navigation

Philadelphia Prison System Reincarceration Recidivism Report by Paul Heroux 2008

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
REINCARCERATION IN THE
PHILADELPHIA PRISON SYSTEM
ALL DISCHARGED INMATES BETWEEN
2000 AND 2006

Report Author:
Paul Heroux, MS, MSc

2

This report does not necessarily reflect the views of
the Philadelphia Prison System or the City of Philadelphia. All comments herein are that of the author.

Questions about the methodology, qualifications and/or limitations of what can or can’t be inferred from this report
should directed to:
Paul Heroux at
PaulHeroux@aol.com
All other questions should be directed to the Philadelphia
Prison System or the City of Philadelphia.

This report was commissioned by then Commissioner Leon A King, Esq
in September 2007 and was completed under the supervision and direction
of Commissioner Louis Giorla in May 2008.
Data used to begin and complete this report were obtained from the PPS
MIS Department in 2007 and 2008.

3

About the Report Author:
Paul Heroux worked for the Philadelphia Prison System between 2006 and 2008 as
the Special Assistant to the Commissioner for then Commissioner Leon King. He
conducted analysis and provided research support on various PPS issues. Prior to
working for the PPS, Heroux worked as a military analyst in a national security
think-tank. Currently, he is the Director of the Research and Planning Division at
the Massachusetts Department of Correction. Heroux earned his Master’s in Criminology from the University of Pennsylvania, and he earned a second Master’s degree from the London School of Economics.

4

CONTENTS
1. SUMMARY OF FINDINGS

PAGE 6

2. INTRODUCTION

PAGE 7

3. MEASURING REINCARCERATION

PAGE 10

4. TREATMENT PROGRAM EFFECTIVENSS

PAGE 12

5. GENERAL FINDINGS IN REINCARCERATION

PAGE 15

6. REINCARCERATION: 2000-2006

PAGE 22

7. OVERCROWDING AND REINCARCERATION

PAGE 27

8. REINCARCERATION AND AGE

PAGE 30

9. REINCARCERATION AND INMATE LENGTH OF STAY

PAGE 37

10. REINCARCERATION AND EDUCATION

PAGE 43

11. REINCARCERATION AND VIOLATION OF PAROLE

PAGE 47

12. WHO IS NOT REINCARCERATED?

PAGE 58

13. NOTE FROM THE AUTHOR

PAGE 63

14. WORKS CITED

PAGE 64

15. CREDITS

PAGE 66

5

SUMMARY OF FINDINGS
•

Reincarceration rates decreased between 2000 and 2006. The general rate of reincarceration at one year fluctuated between 30 and 45%, with an average around 35%
between 2000 and 2006 (Page 23). Recidivism in Philadelphia is not 70% (page 63).

•

The percentage of inmates discharged who had their first incarceration in the PPS increased between 2001 and 2007. This finding suggests two things: 1) that reincarceration is not the driving force behind the PPS population crisis, and 2) the PPS
population was populated by a growing percentage first-time inmates (Page 27).

•

Increased new admissions and increased length of stay are the driving force behind
the PPS population crisis (see pages 29 and 37 for LOS, and 29 for admissions).

•

As the number of career incarcerations increases in the inmate population, the percentage of inmates with serious mental illness (SMI) also increases with inmates who
have higher numbers of career admissions—the City reincarcerates the seriously
mentally ill at high rates (Page 21).

•

The largest percentage of reincarceration occurs in the first year of release and
within that first year, most reincarceration occurs within the first six months of release.

•

At any given time, just over 20% of inmates in the PPS are incarcerated on a most
serious charge that is related to violence; while just less than 80% of the inmates in
the PPS are incarcerated on a most serious charge not related to violence.

•

About 48% of inmates are incarcerated for drug related charges (Page 16).

•

Drug offenses are the most prevalent reason for inmate reincarceration, followed by
violence related charges (Page 18-19). Robbery was the leading cause within violence
related reasons.

•

Drug offenders reincarcerated for distribution-related offenses outnumber possession related offenses by a ratio of approximately 4:1 to 8:1, depending on the year.

•

The average offender age is 31 years and the most frequently occurring age is 20 or
21 years old, depending on the year. Female desistance begins after about 45 years of
age. Females, however, do not have a specific most frequently occurring age and are
represented evenly from their early 20s until their mid 40s (Pages 30-36).

•

The length of incarceration seems to have no relationship with deterrence – longer
stays do not seem to decrease reincarceration (Pages 37-42).

•

Most inmates have only 1 to 3 readmissions in the PPS; less than 7% of inmates
have 10 or more admissions (Page 20).

•

Race is not a strong predictor of who will be reincarcerated on VOP (Page 52).

•

“Other legal violation” offenders (marginal and non-violent) are not reincarcerated
at the same rate as more serious offenders (Page 59).

Paul Heroux

6

Reincarceration in Philadelphia

INTRODUCTION
Recidivism rates are among the most
frequently cited figures in corrections. However,
recidivism is one of the most difficult to measures in all of criminal justice. There are several
definitions of recidivism, including if an offender:
1)

Reoffended but was not caught

2)

Was rearrested

3)

Was reincarcerated

4)

Was reconvicted but not reincarcerated

5)

Was reconvicted and sentenced

This report will not make use of the
term recidivism but will make use of the term reincarcerated. This report will examine inmates reincarcerated in the PPS for a new crime or a violation of parole.
Empirical research and experience indicate that incarcerating individuals temporarily
incapacitates offenders, but generally speaking,
incarceration does not deter criminal activity.
Considering that 46% of inmates are released
within 2 weeks of admission (Goldkamp, 2006)
and more than 99% of the Philadelphia Prison
System’s admitted inmates will eventually be released back to the community after a stay in the
PPS, the rate at which inmates return to the PPS
is more than just a statistic. Knowing which
types of offenders are returning to the PPS, for
what offenses, where they are coming from, and
the amount of time from their release to reincarceration are all important when allocating reentry
resources and planning.
Incarceration serves five general functions (Bartol, 2002):
1.

punishment – punishing the offender for the crime

2.

incapacitation – keeping the offender from offending
for a specified period of time

3.

specific deterrence – to attempt to reduce the likelihood of criminal reoffending

4.

general deterrence – to discourage others from engaging in crime

5.

retribution – to serve as a source of justice for the
victim of the criminal’s bad acts

Paul Heroux

“If we know where and how
long it is before offenders return to the PPS, the City can
appropriately allocate its resources to reducing the risks of
reoffending.”

Reincarceration in Philadelphia 7

1,600

Violent Crimes
per 100,000
Citizens

Fig a.

10,000
8,000
6,000
4,000
2,000
0

1,500
1,400
1,300
2005

2004

2003

2002

2001

2000

1995

1,200
1990

PPS Average
Daily
Population

PPS Population and Philadelphia Violent Crime
Rate

Year
PPS ADP

Violent Crime Rate Per 100,000

As seen in the fig above, the PPS population has increased at a
consistent rate from 1990 to 2005, while the violent crime rate
increases and decreases regardless of the number of inmates in the
PPS. As such, it seems that the violent crime rate is unaffected any
threat of, or deterrent effect of incarceration. This pattern makes
sense since many violent crimes are impulsive with no premeditation; deterrence is unlikely to work for these cases.
In addition, after accounting for impulsivity, research shows that
certainty of being caught is more of an effective deterrent than the
threat of being caught. Put another way, certainty of punishment
places an upper limit of the severity of punishment.
The rise in the Prison population is not caused by an increase in
the City’s population as there was a decrease in the City population
concurrent with an increase in the PPS population (see fig. b below).

2006

2004

2002

2000

1998

1996

1994

Philadelphi
a
Population

1,600,000
1,550,000
1,500,000
1,450,000
1,400,000
1,350,000
1992

10,000
8,000
6,000
4,000
2,000
0
1990

PPS ADP

Philadelphia Citizen Population and the PPS
Average Daily Population

Fig b

Year
Average Daily Population (ADP)
Philadelphia Citizen Population
The Increase in the PPS population is not due to an increase I the
City’s population; the population of the City decreased as the
population of the PPS increased between 1990 and 2007. Philadelphia has more inmates per citizen than other major US cities. (see
Fig ‘c’ on page 9).This information was originally provided by the
author in a report dated 9 November 2007 from the PPS Commissioner Leon King to then Mayor-Elect Michael Nutter.

Paul Heroux

Of these five general functions, not one
actually addresses an effective and empirically
proven long-term criminal risk reduction strategy. Incapacitation stops the criminal behavior
only while he or she is incarcerated. One estimate finds that a 10% increase in incarceration
yields an estimated 1.6-3.1% (cited from Travis,
2005) reduction in crime on the street – not a
cost effective means of reducing crime. Philadelphia is consistent with national trends and has
not seen a decrease in violent crimes with any
increases in incarceration (See fig. a). Incarceration is necessary, but the notion that we can simply lock-up criminals and therefore have less
crime is simplistic and does not address the
many causes of crime.
In addition, incapacitation as a crime
strategy has its limits. If the upper limits of incarceration are extended, such as in Truth in
Sentencing laws, Three Strikes Laws and 25
Years to Life, incapacitation still has not been
proven through empirical research to produce
the desired effects in reducing long-term crime;
costs for such long sentences reduce limited
public finds for other, better crime reduction
means (Petersilia, 2003). The threat of short or
long punishment, including jail and other sanctions does not deter career criminals - incarceration is actually a right-of-passage for some. Since
the likelihood of getting caught for crime is so
low (only a fraction of reported crimes actually
end in an arrest, much less for a conviction),
general deterrence is not an effective strategy.
Retribution for a crime sometimes makes the
victim feel better but does nothing to reduce the
criminogenic risk factors after an offender has
served his or her time.
The solution, therefore, is not to simply
warehouse inmates in jail or prison, but to proactively and comprehensively prepare exoffenders for crime-free reentry using rigorously
evaluated evidence-based interventions aimed at
changing the criminal mindset and behavior of
the offender. Proactive, comprehensive reentry
support is not a liberal ideology; reducing the
risks of criminal reoffending upon release is a
moral obligation of government.

8

Reincarceration in Philadelphia

Even if the PPS can reduce the reincarceration rate, the overall crime rate should not necessarily
be expected to go down. The PPS is able to reduce
reincarceration. Reducing reincarceration may or may
not, however, result in a decrease in the overall crime
rate. That is not to say that efforts should not be
taken to reduce reincarceration, clearly reducing reincarceration improves the quality of life for many. If
the crime on the street grows more quickly than a
decrease in reincarceration, we will see an increase in
crime rates. If decreases in reincarceration outpace
increases in crime then we may see a decrease in
crime rates. Because of this complexity, as a general
rule, criminologists do not equate a decrease in reincarceration with a decrease in crime rates. A decrease
in reincarceration is only less criminal offending,
clearly a good thing, but it might not be enough to
decrease increasing crime rates. In addition, this report includes victimless crimes and parole violations.
Both of these are reason for reincarceration, but are
not typically what the public thinks of when it thinks
of ‘crime.’

“There are many factors
that can increase reincarceration including the lack
of housing, the lack of
jobs, and drug addiction, .”

There are many factors that can increase reincarceration. These include, but are not limited to:

newly learned criminal endeavors while incarcerated

•

Fig c.

new legislation that seeks to reincarcerate parolees for technical violations, which can also be understood as reincarceration for violations that are not necessarily crime

The contents of this report has significance
because once it is known where and how long it is
before offenders return to the PPS, the City can appropriately allocate its resources to reducing the risks
of reoffending. If the City knows which types of offenders are returning for drug offenses, violence, or
property crimes, to name a few, it can better address
treatment efforts before an inmate is released from
the PPS.
The City of Philadelphia needs to know who
is reoffending in terms of age, zip code, race, ethnicity and gender, at what rate, and for what crimes. It
needs to know what relationship exists between the

Paul Heroux

9

40%
30%
20%
10%
0%
2007

•

2000-2007 - Percent of Inmates' First-Time
Admission to the PPS

2006

the lack of job opportunities

2005

•

2004

old criminal habits not eliminated

2003

•

2002

the lack of adequate family or community support

2001

•

2000

the lack of housing upon release

Percent of

•

Year
Fig. b shows that the percentage of released inmates who had their
first admission in the PPS. Between 2001 and 2007, Philadelphia
incarcerated more first time inmates than the year before. This finding
dispels the notion that the Prison’s growing population is caused by
inmates returning to the PPS – Reincarceration is not really a driving
factor in the Prison’s overcrowding. Inmate reincarceration contributed, but inmate reincarceration was actually decreasing between 2001
and 2006 (see page 26, bottom chart).

Reincarceration in Philadelphia

Number of Inmates

Inmate Average Daily Admissions and
Discharges
105
100
95
90
85
80

90 89

100 99

Fig d.
102
99

FY07

FY08

93 91

inmates’ length of stay and the amount of time
elapsed before inmates return after release. How
many inmates return? What direction has reincarceration taken in the past seven years? These and
more questions will be answered in this report. But
first, the City must understand what reincarceration
is, and what limitations restrict the measuring of reincarceration.
MEASURING REINCARCERATION

FY05

FY06

Average Daily ADMISSIONS

Average Daily RELEASES

The PPS has experienced a 14.4% increase in inmate admissions since
FY2005. In FY2005, the PPS admitted 90 inmates on an average day,
while it discharged 89, a difference of one. However, in the current
fiscal year, FY08, the PPS has been admitting an average per day of
102 inmates but discharging only 99, a difference of three. We are
admitting more inmates but discharging fewer, and as we saw from
Fig. 5 those that are staying are staying longer.
Three additional daily admissions does not sound like a lot, but over
the course of one year that amounts to an increase of 1,095 inmates.

“Reincarceration is not the
main factor in the Prison’s
overcrowding.”

As stated, this report does not make use of recidivism.
Instead, this report examines reincarceration; a former
inmate who is returned to the Philadelphia Prison
System, within one year of release, for a new crime
and/or a violation of parole (VOP).
Despite Prison in its name, the PPS is a
county jail that houses both pre-trial inmates (about
60-70%) and sentenced inmates (about 30-40%).
This report does not distinguish between re-detained
and re-sentenced. If an individual is a former inmate
at the Philadelphia Prison System and is subsequently reincarcerated for a new crime or a VOP,
that individual is included in these calculations.
As with any definition of reincarceration, our
definition is not without problems. For example, if a
former inmate was re-detained but eventually released due to ‘dropped charges’ this inmate is included in this analysis as a reincarcerated offender. If
an inmate was previously sentenced for a charge and
re-detained on a new charge but has not yet been
found guilty, this is also included here. This is a very
conservative definition. However, with our operational definition clearly identified as reincarceration for
a new crime or a VOP, there should be no confusion
concerning what is being measured and what can be
inferred and concluded from the data.
Other problems with almost any definition
include:

Paul Heroux

•

An offender might commit a crime but never be caught.

•

An ex-offender might violate parole and subsequently
reincarcerated, but a new crime never actually occurred.

•

An offender might commit crime, be caught, but post bail
before being reincarcerated in the PPS.

•

An offender might move to another city, county, state or
country and engage in criminal activity in a new location.

10

Reincarceration in Philadelphia

•

An offender might be killed while engaging in a criminal
activity, or is currently incarcerated elsewhere.

•

An offender might be rearrested and use an alias; (the PPS
has certain safeguards to identify these offenders but it is
possible that a small number go undetected).

This report measures the more clearly defined reincarceration since it is virtually impossible to
measure recidivism flawlessly, where all crime committed by a former inmate is known, reported, and
included in a measure of recidivism. Despite limitations, due to the enormous amount of data retrieved
by the PPS Management Information Services (the
Prison’s information technology department) as initially recorded by the PPS Classification, Movement
and Registration Division (CMR), there is a substantial amount that can be determined from electronic
records kept by the Philadelphia Prison System.
This report considers every inmate discharged from the PPS between 1 January 2000 and
31 December 2006. This amounts to 232,638 inmate
discharges. Of these 232,683 discharges, 45,613 remained in government custody. This leaves 187,025
discharges, of which, 96,330 inmates are represented.
Many of these inmates posted bail and were then reincarcerated after being found guilty and sentenced
to the PPS.
Between 2000 and 2006, the PPS processed
between 31,853 and 34,467 inmate discharges each
year, which represented 24,243 to 28,765 different
inmates in a year. In order to properly measure reincarceration, a designated time period must elapse; in
this case it is at least one year. As such, this report
does not consider any inmates released in 2007.
Variables included in this report are each inmate’s age, sex, race, marital status, given home zip
code at arrest (if any), length of stay in the PPS, time
duration before returning to the PPS, most serious
criminal charge, if the initial detention was the result
of a parole or probation violation, reason for release
from the PPS, number of total career incarcerations
per inmate, number of incarcerations in a given year
and the occurrence of serious mental illness. Total
yearly counts will be inconsistent due to blank spots
in computer data. At no juncture does this analysis
discuss missing data, which is minimal in MIS.

Paul Heroux

11

“Proactive, comprehensive
reentry support is not a liberal ideology... reducing the
risks of criminal reoffending upon release is a moral
obligation.”

Inmate Average Length of Stay (LOS) per
Year
Fig e.

100
80

74.1

76.4

80.1

87.8

88.9

89.7

90.5

91.3

60
40
20
0
2000 2001 2002

2003 2004

2005 2006 2007

The Length of Stay (LOS) for inmates has been increasing.
In years past, the LOS was as low as 74 days; now it is about
91 days.
If 36,000 inmate admissions stay an average of 74 days, the
average daily population at the Prison will be 7299 inmates.
However, if 36,000 inmates stay 91 days, as is presently the
case, the average daily population at the Prison will be 8975
inmates. This shows the powerful impact of the increased
LOS.

Reincarceration in Philadelphia

TREATMENT PROGRAM EFFECTIVENESS

“45,613 inmate discharges
remained in government
custody leaving 187,025 discharges, of which, 96,330
inmates were represented.”

This report does not measure the effectiveness of any PPS treatment program, nor can this report be used to make any conclusions about PPS
treatment programs. As of the writing of this report,
the PPS was not measuring whether or not its treatment programs are effective in reducing reincarceration. To do so would require either the preferred
method, a randomized experiment with a control
group, or the less preferred but still effective method,
a matched pair statistical analysis. To do the former
would involve the cooperation of the courts and expert researchers. To do the latter would involve enormous amounts of pre- and post-release information
and expert academic researchers, neither of which the
PPS has. Either method compares a treatment group
to an similar control group.
The standard in measuring treatment programs is described in Sherman’s book Evidence-Based
Crime Prevention, and in MacKenzie’s book What
Works in Corrections. In these publications, the authors
describe a 1-5 level methodology hierarchy with the
higher levels offering more confidence in results
while lower levels don’t offer any confidence. These
levels are as follows.
No confidence in research results
Level 1: This level indicates some correlation
between the program and measure(s) of recidivism.
Usually there was no comparison group. Studies in
this category should be judged to be so low in scientific rigor that they are not used to assess the effectiveness of the correctional program.
Level 2: These studies indicate some association between the program and recidivism but were
severely limited because many alternative explanations could not be ruled out, given the research design. Frequently in the correctional evaluations, these
studies used dropouts and no variables were included
in the statistical analysis to control for initial difference between groups.
Confidence Levels in research results
Level 3: These contained a comparison between two or more groups, one receiving the program and one without the program. The assignment
of inmates and the statistical analysis assured reasonable similarity between the treated group and the

Paul Heroux

12

Reincarceration in Philadelphia

comparison group(s). Double blind, and controls for
placebo are often considered.
Level 4: This level indicates a comparison between a program group and one or more control
groups with controls for other factors or a nonequivalent comparison group that is only slightly different
from the program group. Double blind, and controls
for placebo are considered
Level 5: This level is considered the “gold
standard” because studies in this level employ random
assignment and analysis of comparable programs and
comparison groups, including controls for attrition.
Double blind, and controls for placebo are considered.
“Level 1-2” analysis does not account for
other variables that may be responsible for any observed change in criminal behavior. For the PPS to do
“Level 3-5” measurements, it would need to make use
of experienced university level researchers. Generally
speaking, university professors in criminology, sociology and psychology departments, to name a few, have
the experience and skill set needed to properly measure a treatment program beyond the level of simple
correlation and suggestive associations.

“This report does not
measure the effectiveness of any PPS treatment program, nor can
this report be used to
make any conclusions
about PPS treatment
programs.”

The PPS does not measure programs, however
the PPS has hired an intelligent, caring and well intentioned staff dedicated to reducing reincarceration and
facilitating prisoner re-entry in Philadelphia. While the
PPS does not measure the outcomes of its treatment
programs, it does make use of evidence-based programs such as therapeutic communities for drug offenders, cognitive behavioral therapy for violence and
sexually violent offenders, and GED and vocational
training programs.
It would be ideal to measure treatment efforts
but the fact that the PPS does not is not necessarily
reason for alarm. For example, a similar situation can
be drawn to medical doctors’ use of prescription
drugs: they do not empirically measure the outcomes
of each patient they serve, but they do make use of
already proven medications that they proscribe. Generally speaking, it is the same principle with the Philadelphia Prison System’s Restorative and Transitional
Services (RTS) department, or more commonly
thought of as the PPS ‘treatment staff.’
As stated, to measure treatment programs accurately, a randomized experiment or matched pairing
is necessary, but not sufficient. A well planned study
of the effectiveness of treatment programs would have

Paul Heroux

13

Reincarceration in Philadelphia

“...to measure treatment
programs accurately, a
randomized experiment
or matched pairing is
necessary…”

Paul Heroux

to take into consideration that both a control group
and treatment group received the same level of aftercare upon release from the PPS. Another variable that
must be considered are laws that reincarcerate parolees
for technical violations.
Another point on measuring treatment programs includes the proper identification of outputs and
outcomes. The number of inmates served is an
‘output.’ the change in behavior is an ‘outcome.’ The
only way to determine an ‘outcome’ is to do a study
that makes use of a control group and a treatment
group.
Treatment programs aimed at reducing reincarceration can be measured, but treatment programs have
not been assessed in this report. In addition, there is
virtually nothing that can be implicated about the effectiveness of treatment programs by comparing the rates
of reincarceration in the PPS to other jurisdictions. Extreme care should go into measuring the effectiveness
of treatment programs. Incorrect measurement could
lead to a false sense of security and jeopardize public
safety, and taxpayers could waste millions of dollars on
ineffective programs.

14

Reincarceration in Philadelphia

GENERAL FINDINGS ON
REINCARCERATION
This section includes information on offenses discharges and readmitted to the PPS, the
relationship between SMI and Reincarceration, and the number of career admissions for
PPS inmates in 2006.

Paul Heroux

15

Reincarceration in Philadelphia

2006 Cohort - Incarcerated Inmates' Most Serious Charge

Sexual Violence
Related
3%

Violence Related
33%
Weapons Related
1%

Sex Crime Related
1%
Property Crime
Related
8%
Other Legal
Violation
6%

Drug Related
48%

The pie chart above illustrates the individual inmates most serious reason for admission to PPS in 2006
including the number of inmates and the percent. The below pie chart illustrates the individual inmates
most serious reason for reincarceration to the PPS in 2006, including the number of inmates and percent.
It is of note that incarceration related to sexual violence decreased from 3.4% to 1.65%. One of the interesting findings in this study was the sexually violent offenders more often than not are reincarcerated for
non-sexual violence offense, whereas drug, property and violent offenders are reincarcerated more often
than not for the same charge type. These pie charts include inmate duplicates, which are inmates reincarcerated multiple times.

2006 Cohort - Reincarcerated Inmates' Most Serious Charge

Sexual Violence
Violence Related
Related
33%
2%
Sex Crime Related
1%
Property Crime
Related
8%

Weapons Related
1%

Other Legal
Violation
3%

Paul Heroux

Drug Related
52%

16

Reincarceration in Philadelphia

2006 Cohort - Average Number of Incarcerations per Person by Charge
6
Average Number of Incarcerations

5.08
5

4.43

4.48

4.23

4

3.57
2.99

3

2.10
2

1.81

1
0
Drug
Related

Violence
Related

Sexual
Violence
Related

Property
Crime
Related

Sex Crime
Related

Other
Legal
Violation

Weapons
Related

Total
Average

Charge Category
The average number of incarcerations per person in 2006 by charge is listed in fig 12 above.
Drug related reincarcerations are the most frequent reason for an offender’s return to the PPS, followed by
violence related charges.
Not seen in the figure above is that drug offenders tend to return to the PPS for drug related charges,
whereas ‘violence’ related offenders tend to return for both ‘violence’ and ‘drug related’ charges.
Sexual violence offenders return to the PPS at a very high rate, however, these offenders are not returning
for sexual violence related crimes. The majority of the time they are returning for non-sexual violence related offenses. This is consistent with research b Western and Clear that a sex offender record poses more
barriers to reentry than other offenders face. The risk of re-offense is not as high and the severity is very
significant.
This bar graph includes inmate duplicates.

Paul Heroux

17

Reincarceration in Philadelphia

2006 Cohort - Reincarcerated Inmates: Less than 1 Year
6,000

Number of Inmates

5,000
4,000
3,000
2,000
1,000
0
Drug
Related

Violence
Related

Property
Crime
Related

1 to 90 days

Other
Legal
Violation

91 to 180 days

Sexual
Violence
Related

Sex Crime Weapons
Related
Related

181 to 270 days

N/A

271 to 365 days

In the graph above, all inmates reincarcerated in the PPS within the first year of their release are listed by reincarceration interval and most serious charge (often inmates have multiple charges so this analysis makes use of the
most serious charge.) This chart includes inmate duplicates—inmates who are reincarcerated more than once.
From this, we learn that the largest number of inmates returning to the PPS within the first year are incarcerated
on drug related charges; this holds true all the way down to the quarter year intervals. The leading drug related offenses are distribution related. However, distribution related drug offenses outweigh possession related drug offenses by
a factor of 4 to 1. There are two possible explanations for this: 1) more distributors get caught than possessors;
and 2) the legal level that constitutes distribution is so low that many offenders who are caught possessing with an
intention of personal use are charged with distribution.
However, we also learn that violence is the second leading reason that inmates are reincarcerated in the PPS. Not
seen in the chart above, the leading violence charge is robbery.
In-prison treatment efforts should seek to reduce drug possession and distribution related attitudes and behaviors,
and re-entry efforts need to continue to focus on decreasing offenders’ addictive behaviors (Travis, 2005).
We also learn than anger management is very important to decrease the occurrence of violence (Mackenzie, 2006).
However, robbery, often an instrumental crime, could be decreased by targeting its causes, i.e. lack of job opportunities and dire socioeconomic conditions. The City should invest substantial resources in either the FY 09, and if it
is too late, the FY10 for these recommendations in this report.

Paul Heroux

18

Reincarceration in Philadelphia

2006 Cohort - Reincarcerated Female Inmates: Less than 1 Year
800

Number of Female Inmates

700
600
500
400
300
200
100
0
Drug
Related

Violence
Related

Property
Crime
Related

1 to 90 days

Other
Legal
Violation

91 to 180 days

Sexual
Violence
Related

Sex Crime
Related

181 to 270 days

Weapons
Related

N/A

271 to 365 days

Above, we examine 2006 data for the reincarceration reason for female offenders returning within the first year.
Female offenders have returned to the PPS for violence related charges at a greater rate than have male offenders. Females reincarcerated for drug related offenses within the first 90 days is about equal to females returning
for violence related offenses. This chart includes inmate duplicates—inmates who are reincarcerated more than
once.
Unlike males, females returning for violence related offenses are not returning predominately for robbery related
charges; instead they are returning for assault related offenses. Violence for females seems to be less the product
of instrumental crime, and more the result of other factors.
Drug treatment related to both use and distribution needs to be a high priority for female offenders.
Anger management should also be considered as a necessary component for successful female offender reentry.

Paul Heroux

19

Reincarceration in Philadelphia

2006 Cohort - Career Incarcerations
7,000

Number of Inmates

6,000
5,000
4,000
3,000
2,000
1,000
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 30 34
Number of Career Incarcerations
The number of career incarcerations for each inmate discharged in 2006 that have been recorded by the
PPS Lock and Track System since 1989 are plotted above.
From this we learn that:
• 78% of the inmates have 2 or more incarcerations.
• 62% of the inmates have 3 or more incarcerations.
• 48% of the inmates have 4 or more incarcerations.
• 37% of the inmates have 5 or more incarcerations.
• 28% of the inmates have 6 or more incarcerations.
• 10% of the inmates have 10 or more incarcerations.
• 5% of the inmates have 12 or more incarcerations.
• Less than 1% of the inmates have 18 or more incarcerations.
Upon investigation of the individuals with 10 or more incarcerations, serious mental illness is a frequent
characteristic (next page).
Qualifications: It cannot be concluded from the above figure that 78% of the inmates who enter the PPS
recidivate. This is a snapshot of all discharged inmate from one year and the number of admissions in
their criminal history. What is not included is the number of inmates who did not come back to the PPS
that are admitted and discharged to the PPS each year of the years above that incarcerations occurred. If
these former inmates were included, the percentages above would be drastically lower. In addition, this is
a look at “reincarceration”, as such, many inmates who posted bail and then were released, were then subsequently reincarcerated after being sentenced on the same initial charge. This represents only the total
number of reincarcerations, not necessarily a reflection of the crime rate or recidivism.

Paul Heroux

20

Reincarceration in Philadelphia

January 2008 Daily Population Snap Shot Seriously Mentally Ill and Career Incarcerations

1
1

2

8

17

1

1

2

3

2

1

6

16

25

44

49

48

46

89

94

103

66

132

157

441

148

72

556

154

199

739

138

95

848

188

214

1,032

321

1,139

185

60%

22

80%
1,399

40%
2

9

11

13

15

17

1

7

5

5

6

3

13

1

2

26

0%

29

20%
101

Percent SMI or Not SMI

100%

19

21

23

25

33

Career Admissions
SMI

Not SMI

Above, using more recent data from January 2008, we find that as the number of career incarcerations increases, the percentage of
inmates with serious mental illness (SMI) increases. The implication behind this finding is that our local criminal justice system is
using incarceration to address the consequences of unregulated serious mental illness. There are more appropriate alternative sanctions that could be used to decrease the likelihood of reincarceration including day reporting, probation and mental health courts.
Alternative sanctions, such as the aforementioned, have been found to be more effective in reducing reoffending than incarceration
(Mackenzie, 2006), and are also overwhelmingly more appropriate to address the needs of the seriously mentally ill than incarceration. Below, we see the charges associated with SMI inmates. “Sex Crime Related” offenses is represented by about 80% female
prostitutes. In addition, the length of stay for SMI inmates is about 142 days whereas the LOS for non-SMI inmates is about 90
days, which is about 58% more expensive in terms of inmate bed costs alone; this does not take into consideration the enormous
costs that are needed to staff mental health professionals and provide psychotropic medication. Note: after 13 admissions, the
statistical accuracy begins to decrease when we have fewer and fewer inmates who have more than 13 career admissions. Typically,
researchers need at least 30 people to make any confident conclusions about data. This, however, does not detract from the finding
that as the number of incarcerations increases, we find a greater percent of seriously mentally ill inmates. In addition, this January
2008 population snapshot identifies inmates who had or currently have a SMI; this is not to say that all identified inmates are currently suffering from a SMI.
Seriously Mentally Ill and Charge

20

59

25%

Chil

ted
Rel a

l ate d
d Re

e
Cri m
Se x

e Re

ted

lenc

Re la

l Vi o

s
pon

lat ed

io n

te d
Rel a

at
Vi ol
ega l

rim e
rt y C

21

a
Se xu

Wea

er L
O th

pe
Pro

d
e la te

la ted

R
nce

g Re

e
Vi o l

D ru

SMI

6

68

51

152

282

564

386

0%

Paul Heroux

241

448

551

917

2,153

50%

64

75%
2,820

Percent SMI or Not SMI

100%

Not SMI

Reincarceration in Philadelphia

REINCARCERATION:
2000-2006
This report will examine data on inmate discharges from 2000-2006.
In general, reincarceration decreased between 2000 and 2006.
The reincarceration rate in Philadelphia is about 35% per year.

Paul Heroux

22

Reincarceration in Philadelphia

2000-2006 Inmate Reincarcerations Within the First Year of Release With Duplicates
14,000
Number of Inmate
Reincarcerations

12,000
10,000
8,000
6,000
4,000
2,000
0
2000

2001

2002

2003

2004

2005

2006

Year
1 to 90 Days

91 to 180 Days

181 to 270 Days

271 to 365 Days

Above, we see that reincarceration appears to have decreased between 2001 and 2003 and then increased
between 2004 and 2006. However, these are raw counts that do not take into consideration the number of
total admissions. The above figures include duplicates, which are inmates who return to the PPS more than
once. Below, the figure shows the percent of inmates who are reincarcerated after release from the PPS,
factoring out the duplicates.

Percent Of Inmates
Reincarcerated Within the First
Year of Release

2000-2006 Rate of Reincarceration for All Crimes Without Duplicates
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
2000

2001

2002

2003

2004

2005

2006

Year
Paul Heroux

23

Reincarceration in Philadelphia

2000-2006 Inmate Drug Related Reincarcerations Within the First Year
of Release - With Duplicates
7,000
Number of Inmate
Reincarcerations

6,000
5,000
4,000
3,000
2,000
1,000
0
2000

2001

2002

2003

2004

2005

2006

Year
1 to 90 Days

91 to 180 Days

181 to 270 Days

271 to 365 Days

The figure above shows the number of inmate drug related reincarcerations by quarter. The figure below
shows that between 2004 and 2006, drug related reincarcerations were lower than they were in the 4 years
prior.
The City should continue to allocate resources to the PPS to decrease drug related offending. If drug related
reincarceration can be decreased, the large number of inmates who do not return to the PPS relieves resources
to address other, more dangerous offenders.

Percent Of Inmates
Reincarcerated Within the First
Year of Release

2000-2006 Rate of Reincarceration for Drug Related Crimes Without Duplicates
60%
50%
40%
30%
20%
10%
0%
2000

2001

2002

2003

2004

2005

2006

Year
Paul Heroux

24

Reincarceration in Philadelphia

2000-2006 Inmate Violence Related Reincarcerations Within the First
Year of Release - With Duplicates
4,500
Number of Inmate
Reincarcerations

4,000
3,500
3,000
2,500
2,000
1,500
1,000
500
0
2000

2001

2002

2003

2004

2005

2006

Year
1 to 90 Days

91 to 180 Days

181 to 270 Days

271 to 365 Days

Seen above, from 2001 to 2004 there was a decrease reincarceration related to violence. However, in 2005,
the PPS started to see a significant increase in reincarceration due to violence. However, seen below, between 2001 and 2006, the percentage of reincarcerated offenders due to violence decreased as a percent of
total violence related discharges. So if crime in the City is increasing, it is not due to recidivist, it is due to first
time offenders. Not seen in here: robbery is the leading cause of violence at about 40% of the violence
charges, followed by aggravated assault at about 25% of the violence charges.

Percent Of Inmates
Reincarcerated Within the First
Year of Release

2000-2006 Percent of Reincarcerated Inmates Discharged for
Violence - Without Duplicates
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
2000

2001

2002

2003

2004

2005

2006

Year
Paul Heroux

25

Reincarceration in Philadelphia

Number of Inmate
Reincarcerations

2000-2006 Inmate Sex Violence Related Reincarcerations Within the
First Year of Release - With Duplicates
500
450
400
350
300
250
200
150
100
50
0
2000

2001

2002

2003

2004

2005

2006

Year
1 to 90 Days

91 to 180 Days

181 to 270 Days

271 to 365 Days

The graph above shows that there has been a leveling of sexual violence reincarcerations between 2002 and
2006. In treating sexual violence, cognitive-behavioral therapy/relapse prevention and chemical castration/
psychotherapy have been shown to be effective in reducing sexual violence recidivism (Mackenzie, 2006).

Number of Inmate
Reincarcerations

2000-2006 Inmate Property Crimes Related Reincarcerations
Within the First Year of Release - With Duplicates
800
700
600
500
400
300
200
100
0
2000

2001

2002

2003

2004

2005

2006

Year
1 to 90 Days

91 to 180 Days

181 to 270 Days

271 to 365 Days

Above is a summary of all reincarcerated inmates who returned to the PPS within one year for a property crime.
Since 2004 property crime has increased nearly 50%. Making use of ‘hot-spot’ policing techniques and improved
lighting in high crime areas can decrease crime in specific areas (Sherman, 2002). In addition, improving the racial and economic disparity in the City is an effective means of decreasing crime (Pager, 2007; Western 2006).

Paul Heroux

26

Reincarceration in Philadelphia

OVERCROWDING AND
REINCARCERATION
In completing this report, one of the most timely and significant findings is that reincarceration is not a cause of the Philadelphia Prison System’s current overcrowding crisis.

Paul Heroux

27

Reincarceration in Philadelphia

2000-2007 - Percent of Inmates' First-Time Admission to the PPS
35%
30%
Percent of

25%
20%
15%
10%
5%
2007

2006

2005

2004

2003

2002

2001

2000

0%

Year

The above chart shows that the percentage of released inmates who had their first admission in the
PPS. Between 2001 and 2007, Philadelphia incarcerated more first time inmates than the year before.
This finding dispels the notion that the Prison’s growing population is caused by inmates returning to
the PPS – Reincarceration is not really a driving factor in the Prison’s overcrowding. Inmate reincarceration contributed, but inmate reincarceration was actually decreasing between 2001 and 2006 (see
page 24, bottom chart). Note that the above chart cannot be inverted to find recidivism rates. In
other words, in 2007, it cannot be concluded from this chart that 71% is our reincarceration number.
Each number represents one year’s worth of first time admissions, while reincarcerations extend back
over many years. As such, the two would not be a uniform comparison.
The primary factors in driving the PPS’ population are length of stay and increased admissions. Next
page.
Population

Inmates

Citizens per Inmate

Cook County (Chicago)

5.3 million

9,800

540

Los Angeles County

9.9 million

21,000

471

New York City

8 million

14,000

571

Philadelphia County

1.4 million

9,100

153

Above we see that Philadelphia has more inmates per citizen than any other of the major cities in the
United States, but still does not have a lower crime rate. This suggests that an alternative to incarceration crime strategy is necessary. This information was originally provided by the author in a report
dated 9 November 2007 from the PPS Commissioner Leon King to then Mayor-Elect Michael

Paul Heroux

28

Reincarceration in Philadelphia

Inmate Average Daily Admissions and
Discharges
Number of Inmates

The PPS has experienced a
14.4% increase in inmate admis105
102
100 99
99
sions since FY2005. In FY2005,
100
the PPS admitted 90 inmates on
93
an average day, while it dis95
91
90 89
charged 89, a difference of one.
90
However, in the current fiscal
year, FY08, the PPS has been
85
admitting an average per day of
80
102 inmates but discharging
FY05
FY06
FY07
FY08
only 99, a difference of three.
We are admitting more inmates
Average Daily ADMISSIONS Average Daily RELEASES
but discharging fewer, and as we
can see in the fig below, inmates
are staying are staying longer. This information was originally provided by the author in a report dated 9 November 2007 from the PPS Commissioner Leon King to then Mayor-Elect Michael Nutter.
Three additional daily admissions does not sound like a lot, but over the course of one year that amounts to an
increase of 1,095 inmates.
The Length of Stay (LOS) for inmates has been increasing. In years past, the LOS was as low as 74 days; now it
is about 91 days.
If 36,000 inmate admissions stay an average of 74 days, the average daily population at the Prison will be 7299
inmates. However, if 36,000 inmates stay 91 days, as is presently the case, the average daily population at the
Prison will be 8975 inmates. This shows the powerful impact of the increased LOS. For additional LOS data, see
page 37.

Inmate Average Length of Stay (LOS) per Year
100
90
80
70
60
50
40
30
20
10
0

74.1

76.4

80.1

87.8

88.9

89.7

90.5 91.3

2000 2001 2002 2003 2004 2005 2006 2007

Paul Heroux

29

By way of comparison, Maricopa County
(Phoenix, AZ) has an average LOS of 26.8
days. Maricopa County/Phoenix was the
sixth largest City/County in the United
States, but recently moved one spot up in
the rankings, replacing Philadelphia as the
5th largest City/Count in the US.
It cannot be said that the decrease in reincarceration is due to an increase in LOS.
Starting on page 37, this notion is examined
and these analyses show that an inmates
individual LOS is unrelated to reincarceration. This information was originally provided in a report dated 9 November 2007
from the PPS Commissioner Leon King to
then Mayor-Elect Michael Nutter.
Reincarceration in Philadelphia

REINCARCERATION AND AGE

Paul Heroux

30

Reincarceration in Philadelphia

2006 Cohort - First Year Reincarcerations by Age and Sex

Number of Inmate Reincarcerations

1,400
1,200
1,000
800
600
400
200
0
13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77
Age
Female

Male

The most frequently occurring age for an inmate is 20 years old. Females offend at consistent rates throughout
their 20s, 30s but start to desist in their 40s. This chart includes inmate duplicates—inmates who are reincarcerated more than once.

Paul Heroux

31

Reincarceration in Philadelphia

2006 Cohort - First Year Reincarcerations by Age and Race/Hispanic Ethnicity

Number of Inmate Reincarcerations

1,200
1,000
800
600
400
200
0
13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77
Age
AFRICAN AMERICAN

HISPANIC

WHITE

Black offenders are reincarcerated at the highest rate in their early 20s, whereas white offenders are reincarcerated at the highest rate in their mid-20s. Black offenders desist in their offending before white males, but
both groups tend to continue a high rate of reincarceration until about their 40s. This chart includes inmate
duplicates—inmates who are reincarcerated more than once.

Paul Heroux

32

Reincarceration in Philadelphia

2006 Cohort - First Year Reincarcerations by Age and Sex

Number of Inmate Reincarcerations

400
350
300
250
200
150
100
50
0
15 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64
Age
1 to 90 days

91 to 180 days

181 to 270 days

271 to 365 days

Most offenders reincarcerated within the first year of release are 20 years old. This chart includes
inmate duplicates—inmates who are reincarcerated more than once.

Paul Heroux

33

Reincarceration in Philadelphia

2006 Cohort - Drug Related First Year Reincarcerations by Age and Year Quarters

Number of Inmate Reincarcerations

800
700
600
500
400
300
200
100
0
14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 72 75 73 75 77
Age
1 to 90 days

91 to 180 days

181 to 270 days

271 to 365 days

The age distribution above indicates that drug related charges are primarily activity associated with young
offenders. This chart includes inmate duplicates—inmates who are reincarcerated more than once.

Paul Heroux

34

Reincarceration in Philadelphia

2006 Cohort - Viollence Related First Year Reincarcerations by Age and Sex

Number of Inmate Reincarcerations

300
250
200
150
100
50
0
14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 63 65 73
Age
1 to 90 days

91 to 180 days

181 to 270 days

271 to 365 days

The graph above shows the frequency of first-year reincarcerated inmates by age. Violence tends to most frequently occur in males in their early twenties, and then starts to taper off by mid to late 20s. The rate of occurrence remains fairly
stable until about the mid 40s, and then gradually starts to decline again. This chart includes inmate duplicates—inmates
who are reincarcerated more than once.

Paul Heroux

35

Reincarceration in Philadelphia

40

35
30
25

20
15
10
5
0

2006 Sex Offenders Reincarcerated with in the First Year

15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 67 69 72
Age

Average Age of Group (Red Line)

Sex offenders tend to return to the PPS as older adults rather than younger adults. This could be for several reasons: 1) Older Caucasian men tend to
be the most violent sexual cohort; or 2) Sex offenders reoffend after release from a long stay at the state level department of corrections. Sex offender
reentry efforts are important for a number of reasons, including risk of reoffending as well as risk of committing new crimes resulting from being institutionalized or the unique reentry issues faced by released sex offenders.

Reincarceration in Philadelphia

36

Paul Heroux

Number of Inmates (Blue Bars)

REINCARCERATION AND
INMATE LENGTH OF STAY

There is a widespread belief that increasing the length of stay (LOS) will deter offenders from reoffending. Using the 2006 cohort, this deterrent notion is examined.
The 2006 cohort is used because it offers the most up to date information on reincarceration within the first year after release. Not seen in this report, is an examination
of 2000-2005 data which also yields consistent results when compared to the findings
in 2006. These findings strongly suggest that extended LOS does not reduce recidivism. A more scientific analysis using risk assessment tools is necessary to definitively
prove this.

Paul Heroux

37

Reincarceration in Philadelphia

Average Length of Stay by Charge
2000 & 2006 Cohorts
140
117

120

LOS in Days

100
80

91

87

84
69

66

60

84
65

58 61
44

42
40

30 30

29

23

20
0

39%

26%

38%

5%

0%

45%

91%

29%

Sexual
Violence
Related

Violence
Related

Drug
Related

Property
Crime
Related

Sex Crime
Related

Other
Legal
Violation

Weapons
Related

Grand
Total

2000 Cohort

2006 Cohort

Above, the average length of stay (LOS) is given for each charge category for 2000 and 2006. The percent given between the two years is the percent increase from 2000 to 2006. This chart and all of these
LOS charts do not make use of duplicates.
For the sake of ease in reading the graph, 2001-2005 are not included. While, two points don’t make a
trend, if 2001-2005 were included, a clear upward trend would be apparent. This means the LOS has
been consistently increasing. Between 2000 and 2006 the increase was 29%. This finding has significance
in terms of the PPS population. As a hypothetical admissions count, but using real LOS figures, if the
PPS admitted 30,000 inmates in 2000, and admitted 30,000 in 2006, based on LOS alone, the average
daily population for each year would be 5,342 inmates in 2000 but 6,904 in 2006. LOS has contributed
significantly to the PPS’s growing population.
A longer LOS can occur for any number of reasons. For example, an increased case load at the District
Attorney’s and the Public Defender’s offices, or legislation that increases the LOS as a penalty for committing crime. The City needs invest additional money and resources into the Court system to addresses
this problem. Solving the legal stalemate over funding the Courts may prove critical. Please refer to the
Goldkamp report submitted to the City of Philadelphia, and then Commissioner King’s report to then
Mayor Elect Michael Nutter on 9 November 2007 for important recommendations in reducing the
prison population. Both reports note that while the caseload for the First Judicial District increased over
the last few years, there has been no commensurate increase in resources. This increased caseload is inmate length of stay and is a major factor causing overcrowding in the Philadelphia Prison System

Paul Heroux

38

Reincarceration in Philadelphia

2006 Cohort - Reincarceration in the First Year and Length of Stay in the PPS
100%
90%
80%
70%
60%
50%
40%
30%
167

128

263

559

517

675

1,009

5,270

20%
10%
0%

a) 0-49
days

b) 50-99
days

c) 100-149 d) 150-199 e) 200-299 f) 300-399 g) 400-499
days
days
days
days
days

h) 500+
days

Reincarcerated in one year or less
The graph above examines the LOS and its effects on offender reincarceration. The average rate of reincarceration in the first year is between 35 and 40%. First year reincarceration was unaffected by whether inmates stayed between 0-49 days or for 500+ days; the rate of reincarceration did not decrease as the LOS
increased. As such, we cannot say that a longer LOS decreases reincarceration in Philadelphia County. Note
that this analysis does not consider inmate risk of reincarceration, which is necessary to determine with confidence that LOS is in fact unrelated to deterrence.
A question often arises about why LOS would not deter. The answer is fairly straightforward. Inmates often
become institutionalized whereby they become dependent on the secure housing, meals and healthcare that
are Constitutionally guaranteed while incarcerated; for some, this may abrogate any deterrent effects incarceration may offer. But more importantly, significant numbers of inmates face barriers to reentry that include difficulties finding work, lack of educational and opportunities to gain the skills for work, lack of
housing options, and isolation from family (Clear, 2007).
It could also be said that long incarceration LOS does deter future offending, however the aforementioned
iatrogenic effects of incarceration undermine any deterrent effect. Work, education, housing, and family are
all important factors that assist an ex-inmate and decrease the likelihood of reincarceration.

Paul Heroux

39

Reincarceration in Philadelphia

2006 Cohort - Drug Related Reincarceration in the First Year and Length of Stay in the
PPS
100%
90%
80%
70%
60%
50%
40%
30%
94

71

160

306

b) 50-99
days

302

472

a) 0-49
days

347

2,804

20%
10%
0%

c) 100-149 d) 150-199 e) 200-299 f) 300-399 g) 400-499
days
days
days
days
days

h) 500+
days

Reincarcerated in one year or less
Within the first year of release, the rate of reincarceration for offenders who stay in the PPS for 500+ days is
within the range of normal fluctuation as individuals staying in the PPS for 0-49 days. Below, using 2000 cohort
data, which can examine the effect of LOS beyond the first year of release, reincarceration increases just less
than 10% compared to a 0-49 days LOS. If there is any deterrent effect of incarceration, it seems to evaporate
with each successive year after release. This analysis, however, does not consider inmate risk of reincarceration,
which is necessary to determine with confidence that LOS is, in fact, unrelated to deterrence.

Paul Heroux

40

Reincarceration in Philadelphia

2006 Cohort - Violence Related Reincarceration in the First Year and Length of Stay in
the PPS
100%

80%

60%

40%

58

45

89

194

b) 50-99
days

164

407

a) 0-49
days

246

1,611

20%

0%
c) 100-149 d) 150-199 e) 200-299 f) 300-399 g) 400-499
days
days
days
days
days

h) 500+
days

Reincarcerated in one year or less
The effect length of stay in the PPS has on the rate of sexual violence related reincarcerations is seen
above. There is no consistent relationship between length of stay (LOS) and inmate reincarceration.
This analysis, however, does not consider inmate risk of reincarceration, which is necessary to determine
with confidence that LOS is, in fact, unrelated to deterrence.
Compared to drug offenders or violent offenders who are predominately young black males, sex offenders are represented with less disparity than and come from different socioeconomic classes, races and
ages. The numerous factors that effect sex offender reentry are likely responsible for the observed fluctuations in reincarcerations rates.

Paul Heroux

41

Reincarceration in Philadelphia

2006 Cohort - Sexual Violence Related Reincarceration in the First Year and Length of
Stay in the PPS
100%

80%

60%

40%

23

10

35

40

b) 50-99
days

c) 100-149 d) 150-199 e) 200-299 f) 300-399 g) 400-499
days
days
days
days
days

125

20%

7

8

10

0%
a) 0-49
days

h) 500+
days

Reincarcerated in one year or less
Longer length of stay is not correlated with decreases in reincarceration for inmates incarcerated on a sexual
violence related charge. As with other charges, institutionalization and/or barriers to reentry may account
for the increase in reincarceration (Petersilia, 2003). This analysis, however, does not consider inmate risk of
reincarceration, which is necessary to determine with confidence that LOS is, in fact, unrelated to deterrence.

Paul Heroux

42

Reincarceration in Philadelphia

REINCARCERATION AND
EDUCATION LEVEL

The following analysis will examine reincarceration from the year 2006 which contains
the cohort that is the most up-to-date data that can be used for this analysis.

Paul Heroux

43

Reincarceration in Philadelphia

2006 Cohort - Education Level and Reincarceration in the First Year
100%
90%
80%
70%
60%
50%
40%
30%

91

399

3,545

4,352

89

20%
10%
0%

1-6 Grade

7-11 Grade

12th Grade

Some College

College Degree

1 year or less
The chart above shows that as an inmates level of education increases, the percent of first year reincarcerations decreases. The chart below shows the same pattern, but for females only.
2006 Cohort - Female Education Level and Reincarceration in the First Year
100%

80%

60%

40%

559

94

7-11 Grade

12th Grade

Some College

20

703

20

20%

0%
1-6 Grade

College Degree

1 year or less

Paul Heroux

44

Reincarceration in Philadelphia

2006 Cohort - Black Education Level and Reincarceration in the First
Year

25

2,982

2,609

272

63

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

1-6 Grade

7-11 Grade

12th Grade

Some College

College
Degree

1 year or less

The chart above shows that for black inmates, the percent of reincarcerations decreases as educational attainment
increases.
2006 Cohort - Hispanic Education Level and Reincarceration in the First
Year

598

255

26

1-6 Grade

7-11 Grade

12th Grade

Some College

2

47

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

College
Degree

1 year or less

The chart above and below shows that for Hispanic and white inmates, the percent of reincarcerations decreases
as educational attainment increases.
2006 Cohort - White Education Level and Reincarceration in the First
Year

95

7-11 Grade

12th Grade

Some College

23

666

1-6 Grade

740

17

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

College
Degree

1 year or less

Paul Heroux

45

Reincarceration in Philadelphia

2006 Cohort - Percent of Inmates Reincarcerated within the First Year by Charge and
by Education Level
100%
80%
60%
40%
20%
0%

1-6 Grade

7-11 Grade
27

29

3

Violence Related

26

1866

1580

184

Sexual Violence Related

5

115

124

10

1

39

39

5

3

373

383

71

27

Weapons Related

Sex Crime Related
Property Crime Related

7

12th Grade

Some College

College Degree
42

Other Legal Violation

1

116

149

30

12

Drug Related

79

3257

2413

226

31

The chart above shows the percentile representation of charge by education level. Of particular note is the increasing drug related reincarceration for inmates with lower levels of education. Also, the high percent of reincarceration
for property crime and other legal violations associated with inmates with college degrees. Below, is the percentile
representation of admissions to the PPS by charge category and educational level. The same salient patterns occur
with admitted inmates as with readmitted/reincarcerated inmates.
2006 Cohort - Percent of Inmates Incarcerated within the First Year by Charge and by
Education Level
100%
80%
60%
40%
20%
0%

1-6 Grade

Weapons Related

Paul Heroux

7-11 Grade

12th Grade

Some College

97

150

18

College Degree
1

460

105
11

Violence Related

81

4278

3649

Sexual Violence Related

11

299

338

36

Sex Crime Related

1

92

89

10

7

Property Crime Related

22

887

1013

188

67

Other Legal Violation

22

506

762

171

76

Drug Related

171

6532

5197

582

116

46

Reincarceration in Philadelphia

REINCARCERATION AND
VIOLATION OF PAROLE

The following analysis will examine reincarceration from the year 2000 cohort of released inmates, and then from the 2006 cohort. The 2000 cohort is used because VOP
occurs at a greater frequency after the second year following release than general reincarceration. The 2006 Cohort is used because it is the most up-to-date data that can be
used in this analysis. This analysis does not include inmate duplicates—inmates who
are reincarcerated more than once.

Paul Heroux

47

Reincarceration in Philadelphia

2000 Cohort - Percent of Inmates Reincarcerated and When the VOP Occurred after
Release
100%

80%
6626

1961

958
655

60%

448

381
309

83

40%

20%

0%
First year

Second
Year

Third Year

Fourth
Year

Parole Violator

Fifth Year Sixth Year

Seventh
Year

Eighth
Year

Not a Violation of Parole

The second year after release has the greatest percentage of parole violation reincarcerations relative to
the number of total reincarcerations for that year.
Above is the number of inmate discharges from year 2000 that resulted in reincarceration. Most discharges result in reincarceration within the first year, and the greatest number of parole violations occur
within the first year as well. However, (using the graph on the opposite page) the second year has the
greatest percentage of parole violation reincarcerations relative to the number of total reincarcerations
for that year.
On average, about 26% of the discharges that result in a reincarceration within the PPS are from violation of parole. These reincarcerations are not new crimes but stem from technical violations of parole
including but not limited to not holding a job, missing appointments with a parole officer, traveling out
of jurisdiction, failing a drug test, which is not a crime, whereas possession or distribution of drugs is a
new crime. If an individual on parole committed a new crime, such as being caught possessing or distributing drugs, this individual is not identified here as a VOP; he or she would be identified as having
committed a new crime.

Paul Heroux

48

Reincarceration in Philadelphia

2006 Cohort - VOP and Lenght of Stay
100%
90%
80%
70%
60%

901

563

411

426

199

100-149
days

150-199
days

200-299
days

300-399
days

90

130

5,335

50%
40%
30%
20%
10%
0%
0-49 days 50-99 days

Violation of Parole

400-499
days

500+ days

New Crime

Looking at all reincarcerations from the 2006 discharge cohort, there is a strong correlation between an
inmates length of stay (LOS) and the prevalence of a violation of parole (VOP). This analysis, however,
does not consider inmate risk of reincarceration, which is necessary to determine with confidence that
LOS is in-fact unrelated to deterrence.
As inmates stay longer in the PPS, the prevalence of a VOP increases. The VOP reincarcerations are a
subset of the inmates who reoffend. While the difference between reincarceration based on 0-49 days and
500+ days was about 10% in general, the increase for VOP is more dramatic increasing from about 20%
to about 45%.
This could suggest that 1) a longer LOS increases the probability of an inmate violating parole due to institutionalization and/or the barriers to reentry that returning offenders face, or 2) that inmates who stay
shorter periods of time are less likely to violate parole due to personal characteristics.

Paul Heroux

49

Reincarceration in Philadelphia

2006 Cohort - VOP Reincarcerated and Most Serious Charge Category
100%
90%
80%
70%

166

4,190

225

638

64

Drug
Related

Other
Legal
Violation

Property
Crime
Related

Sex Crime
Related

60%

2,713

42

8,038

Violence
Related

Weapons
Related

Grand
Total

50%
40%
30%
20%
10%
0%
Sexual
Violence
Related

Violation of Parole

New Crime

Looking at all reincarcerations from the 2006 discharge cohort, the percent of inmates faced with violation of parole varies by charge type. Non-violent Sex Crimes have the lowest amount of parole violations
and Sexual Violence Related has the highest, which could be read another way: sexually violent offenders
have the lowest rate of new criminal offenses.
With regards to drug related VOP, the practice of screening for drugs in parolees is partly responsible for
the high rate of VOP.

Paul Heroux

50

Reincarceration in Philadelphia

2006 Cohort - VOP and Age Range
100%

80%

4

11

33

20,139

100

363

752

928

950

995

1,411

2,157

60%
349

40%

20%

0%
15 to 20 to 25 to 30 to 35 to 40 to 45 to 50 to 55 to 60 to 65 to 70 to Grand
19
24
29
34
39
44
49
54
59
64
69
74 Total
Violation of Parole

New Crime

Looking at all reincarcerations from the 2006 discharge cohort, violation of parole tends to remain
constant with age. Males and females tend to violate parole at about the same rate in each age category (page 52).
This does not follow the same pattern as seen in general desistance (page 27). In general, desistance
begins in the early 20s for males. VOP occurs in a relatively stable manner up to about the mid 40s.
Therefore, it would be reasonable to assume that offenders who are returning to the PPS as they get
older are returning for VOP at a far greater rate than new crimes are occurring.

Paul Heroux

51

Reincarceration in Philadelphia

2000 Cohort - VOP and Race
100%
90%
80%
70%
35

5,677

885

1,438

8,035

ASIAN

AFRICAN
AMERICAN

HISPANIC

WHITE

Grand Total

60%
50%
40%
30%
20%
10%
0%

Violation of Parole

New Crime

Looking at reincarcerations within the first year from discharges in the 2006, Asians, blacks, whites and
Hispanics all violate parole at about the same rate.

Paul Heroux

52

Reincarceration in Philadelphia

2006 Cohort - VOP and Marital Status
100%
90%
80%
70%
60%

939

19,383

3,175

151

50%
40%
30%
20%
10%
0%
DIVORCED

MARRIED

SINGLE

Violation of Parole

WIDOW(ER)

New Crime

Of all reincarcerated inmates within the first year of release from the 2006 Cohort, widowed
offenders had the lowest rate of VOP while divorced and single offenders had about the same
levels.

Paul Heroux

53

Reincarceration in Philadelphia

2006 Cohort - Percent of Discharges Reincarcerated by Zip Code

70

92
75

107
104
110

190
169
144

191
177

214
192
181

306
306
200

278
305

346
348
348

438
389

517
418
428

621
543
493

799
703

80%

782
647

100%

60%

40%

20%

Reincarcerated

1915 4

1911 9
1912 6
1911 4

1914 9
1913 5
1912 5
1915 1
1913 6
1911 1
1915 0
1914 7
1912 3
1913 0

1913 8
1914 2
1912 2

1913 9
1914 6
1914 4
1912 0
1910 0
1914 5
1914 1
1913 1
1914 8
1910 4

1912 4
1912 1
1913 3

1914 0
1913 4
1914 3
1913 2

0%

Not Reincarcerated

Looking at all reincarcerations from the 2006 discharge cohort, seen above are the highest rate
and most frequently occurring zip codes associated with inmate reincarceration. Note that the
above ZIP codes have the highest rate of reincarceration, however, what is not seen are the ZIP
codes that do not have a high rate of reincarceration. As such, we see that almost all of these are
above 40%, where as the normal rate of reincarceration within the first year is between 35-40%.
2006 Cohort - Percent of Discharges Reincarcerated by Zip Code
800

Seen to the left, are the
raw counts of inmates
who are reincarcerated by
zip code. It is the same
data as above but in a different chart format.

700
600
500
400
300
200
100

19 154

19 114

19 126

19 119

19 130

19 123

19 147

19 150

19 111

19 136

19 151

19 125

19 149
19 135

19 122

19 142

19 138

19 104

19 148

19 131

19 141

19 145

19 100

19 120

19 144

19 146

19 139

19 133

19 121

19 124

19 132

19 143

19 134

19 140

0

Reincarcerated

Paul Heroux

54

Reincarceration in Philadelphia

2006 Cohort - VOP and Inmates' Admission Zip Code
100%

56

50

54

57

54
70

65

75

83

96

102

97

105

111

142

123

148

149

159

191

195

217

214

216

60%

231

246

251

309

316

317

362

431

453

470

80%

40%

20%

Violation of Parole

1911 4
1915 4

1915 0
1912 6
1911 9

1914 7
1913 0

1912 3
1911 1

1913 6
1913 5
1915 1

1914 9
1912 5

1914 2
1912 2

1913 1
1913 8

1912 0
1914 1
1910 4

1910 0
1914 8

1914 4
1914 5

1913 9
1914 6

1912 4
1912 1

1914 3
1913 2
1913 3

1913 4
1914 0

0%

New Crime

Looking at all reincarcerations from the 2000 discharge cohort, above, of the zip codes that have the highest frequency of offender reincarceration, VOP is responsible for about 38% of inmates who return to the PPS. In general, VOP only accounts for about 30% of reincarceration.
Reentry resources need to be allocated to the zip codes where the most reentry is occurring . Within those
zip codes, efforts to reduce the occurrence of technical violations should be a top priority. Inmates who
are incarcerated on a VOP consume considerable resources from the County and do so without committing a new crime.

Paul Heroux

55

Reincarceration in Philadelphia

2006 Cohort - VOP and Sex
100%

75%
4,129

19,824

Female

Male

50%

25%

0%

Violation of Parole

New Crime

Above: Looking at all reincarcerations from the 2006 discharge cohort, there is no functional difference between males and females that violate parole.
Below: Looking at all reincarcerations from the 2006 discharge cohort, excluding the VOPs, there is a difference between males and females who don’t violate parole (males commit slightly more new crimes).

2006 Cohort - Non-VOP and Reincarceration
100%
90%
80%
70%

2,669

12,116

Female

Male

60%
50%
40%
30%
20%
10%
0%

Reincarcerated
Paul Heroux

56

Not Reincarcerated
Reincarceration in Philadelphia

Paul Heroux

57

Reincarceration in Philadelphia

WHO IS NOT
REINCARCERATED?
This chapter examines who has not been reincarcerated in the PPS. It is more difficult to
determine who is going to be reincarcerated than who is not.

Paul Heroux

58

Reincarceration in Philadelphia

2006 Cohort - Not Reincarcerated by Most Serious Charge Category
100%
80%
60%
217

4,724

437

102

6,188

20%

1,316

1,330

40%
0%

ted

ed

Rel...

elated
nce R

ence

ela
ons R
Weap

Viole

l Viol
Sex ua

ed
Relat
rime
Sex C

Relat

lation
a l Vio
r L eg

elated

r im e
er ty C
Pro p

Othe

R
Dr ug

Not Reincarcerated
Within the first year of release, offenders incarcerated for “other legal violation”, which are usually, but not
always, non-violent/marginal offenses, the occurrence of reincarceration is 20%. (See pg 60).

2006 Cohort - Not Reincarcerated by Age Category
100%
80%
60%

6

6 5 to

32

62

6 0 to

3 5 to

187

531

1,998

3 0 to

1,596

1,862

2 5 to

1,929

2,447

2,457

568

20%

1,053

40%

0%
7 5 to

7 0 to

5 5 to

5 0 to

4 5 to

4 0 to

2 0 to

1 5 to

79

74

69

64

59

54

49

44

39

34

29

24

19

Not Reincarcerated
Within the first year of release, as offenders get older, and after 25-29 years old, the occurrence of reincarceration is less than 50%.
Paul Heroux

59

Reincarceration in Philadelphia

2006 Cohort - Not Reincarcerated by Length of Stay in the PPS
100%
80%
60%
40%
845

309

49 da
100-1

224

1,027

50-9 9

531

1,566

0-49

808

9,475

20%
0%

500+
da ys

ys

ys

99 da
300-3

ys

99 da
400-4

99 da
200-2

ys

ys

99 da
150-1

da ys

days

Not Reincarcerated
Within the first year of release, there is no relationship between inmate length of stay and
not being reincarcerated in the PPS.

2006 Cohort - Not Reincarcerated by Race

2,804

TE
WHI

OTH
ER

48

A
CAN

AN IC
HISP

AFRI

1,638

10,136

ASIA
N

108

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

ICAN
MER

Not Reincarcerated
Within the first year of release, with the exception of Asians, there is no relationship between inmate race and
not being reincarcerated in the PPS.
Paul Heroux

60

Reincarceration in Philadelphia

2006 Cohort - Not Reincarcerated by Number of Career Admissions

10,473

3 3 to

2 5 to

1 9 to

1 6 to

1 3 to

1 0 to

7 to 9

4 to 6

1 to 3

2 8 to

1

2 2 to

6

5

44

114

331

997

2,818

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

36

30

27

24

21

18

15

12

Not Reincarcerated
The above chart shows the number of career admissions by reincarceration within the
first year of release. For individuals with 1-3 career admissions, the percent of individuals not reincarcerated is about 75%. For individuals with 4-6 admissions, the percent of
inmates reincarcerated is about 40%. Consistent with other research findings (Clear,
2007; Kurlychek, 2006; Western, 2006; Travis, 2005; Petersilia, 2003) this charts indicates that increased incarceration is associated with facilitating a criminogenic cycle of
institutionalization and difficulty with re-entry.
Other analysis directed by the author while at the PPS found that marginal offenders
have a lower rate of recidivism, are not violent, and have contributed to the PPS average daily population by about 600 inmates each year. This is particularly important in
light of current population overcrowding. If marginal offenders are addressed via alternative sanctions (mental health courts, day reporting, probation or fines, to name a
few), the average daily population would be about 600 inmates lower, the costs to the
City could be 50-75% less per inmate, and for many offenders, alternative sanctions are
more effective in reducing recidivism than is incarceration.
The City’s need to invest in mental health courts, drug courts, day reporting and other
alternatives to incarceration is immediate and should be addressed in the fY09 budget,
otherwise its contribution to reincarceration and overcrowding will continue until the
next budget cycle.

Paul Heroux

61

Reincarceration in Philadelphia

GENERAL CHARGE
Other Legal Violation

2006: Data CHARGE DESCRIPTION
VIOLATION OF JUDICIAL CODE

Total
632 LOIT & PROWLING AT NIGHT TIME

4

CRIMINAL MISCHIEF

201 ACCIDENTS-DAMAGE ATTN VEH/PROP

3

POSSESSING INSTRUMENT OF CRIME

167 BAD CHECKS

3

ENDANGERING WELFARE OF CHILD

70

CRUELTY TO ANIMALS

3
3
3

69

FALSE ALARMS

FUGITIVE FROM ANOTHER STATE

61

IMPERSONATING A PUBLIC SERVANT

CRIMINAL TRESPASS

56

INSURANCE FRAUD

3

OBST HWY & OTHER PUB PASSAGE

42

POSSESSING INSTRUMENT OF CRIME

3

CORRUPTION OF MINORS

38

TRADEMARK COUNTERF-CONSP

3

FORGERY

34

COPYING; RECORDING DEVICES

2

RECKLESSLY ENDANGERING

27

CREDIT CARDS

2

SUMMARY VIO OF MV CODE

26

CRIMINAL TRESSPASS

2

TRADEMARK COUNTERF

16

FAILURE TO DISPERSE - ORDERED

2

15

FALSE EVIDENCE OF IDENTITY

2

RESISTING ARREST

15

FLIGHT TO AVOID PROSECUTION

2

ESCAPE

14

OBST ADMIN-LAW OR GOV FUNCTION

2

14

ADJUDICATED DELIQUENT FOR:

1

CAUSING OR RISKING CATASTROPHE

12

ALT OR OBL MARKS OF IDENTIF

CONTRABAND
VIOLATION PUBLIC WELFARE CODE

12 ALTERED/FRG/CTFT DOC & PLATES
ATTEMPTED ARSON AND RELATED OF12 FENSES
12 CONTEMPT OF GENERAL ASSEMBLY

1

SCATTERING RUBBISH

10

CORRUPT ORGANIZATIONS

1

ACCIDENT INV DEATH/PERS INJURY

9

DEALING IN UNLAWFUL ACTIVITIES

1

HINDING APPREH OR PROSECUTION

9

DRIVING WHILE SUSP OR REVOKED

1

ACCIDENT-DEATH-NO LICEN

7

HOMICIDE BY VEHICLE

1

DEFAULT IN REQUIRED APPEARANCE

7

INCITING TO RIOT

1

HARASSMENT

7

LOITERING

1

INTERF W/CUSTODY OF CHILDREN

7

LOTTERIES

1

UNAUTH USE OF AUTO & VEHICLES

7

PERJURY

1

UNLAWFUL USE OF COMPUTER

7

PUBLIC DRUNKENNESS

1

ETHNIC INTIMIDATION

5

SECURING EXEC OF DOC BY DECEPT

1

INVOLUNTARY MANSLAUGHTER

5

FLEEING TO ELUDE POLICE OFF

4

TAMPERING WITH PUB REC OR INFO
VIO UNEMPL COMPENSATIONS LAWS

1
1

ARSON AND RELATED OFFENSES

DISORDERLY CONDUCT

CRIMINAL CONSPIRACY

PUBLIC ASSISTANCE ACT

1
1
1

Throughout this report, drug related charges, violence, related charges, sexual and sexual
violence related charges and property related charges are describe. However, at times, we
also see “other legal violation.” Seen above is what creates the ‘Other Legal Violation’ category.

Paul Heroux

62

Reincarceration in Philadelphia

NOTE FROM THE AUTHOR
70%?
This report shows that reincarceration in the PPS is around 35% within the first year of release.
In the 9 November 2007 report titled “Prison Population Report and Recommendations to The Mayor Elect
of Philadelphia Michael A. Nutter” from Commissioner Leon King to then Mayor Elect, Michael
Nutter, this author supplied information for that report. On page 39, it states that “Eventually,
more than 70% of inmate discharges return to the PPS”, and then on page 41, figure 14 shows the
number of inmate reincarcerations from the first year after release to the seventh year after release.
Under the methods used, an inmate discharge might be a reincarceration for the same crime without even violating parole or committing a new crime as well as several other non-recidivism related
reasons for reincarceration such as routine transfers, and reincarceration after posting bail and then
being later sentenced on the same initial charge (this would be two incarcerations for one crime).
The methods used to arrive at this “more than 70%” number do not represent a recidivism finding, or even exclusively a reincarceration for a new crime/and or VOP. As such that 70% figure is a meaningless and misleading figure to quote as a rate of recidivism for the Philadelphia Prison System, and any belief to
the contrary is in error.
In addition, this same ‘70% number’ was released as “a work in progress” on 14 April 2008 by the
PPS. While this report only makes use of reincarceration at one year, and does so for accounting
purposes as well as accuracy of data, best estimates at reincarceration for a new crime or a violation of parole at 7 years places the actual figure between 49% and 59%. As such this particular
number (70% or 72%)that has been cited many times by several parties should not be taken as a
true or accurate recidivism or reincarceration number.
The City of Philadelphia and/or the Philadelphia Prison System does not have a recidivism rate
that is anywhere near 70%.

Paul Heroux

63

Reincarceration in Philadelphia

WORKS CITED

Paul Heroux

64

Reincarceration in Philadelphia

WORKS CITED:
•
•
•
•
•
•
•
•
•
•
•

Paul Heroux

Barol, Curt (2001) Criminal Behavior
Clear, Todd (2007) Imprisoning Communities
Goldkamp, John (2006) Report to the City of Philadelphia
King, Leon, (2007) Prison Population Report and Recommendations to The Mayor Elect of Philadelphia Michael A. Nutter
Megan Kurlychek (2006) Criminology and Public Policy Vol. 5 No. 3
MacKenzie, Doris (2006) What Works In Corrections
Pager, Devah (2007) Marked: Race, Crime, and Finding Work in an Era of Mass Incarceration
Petersilia, Joan (2003) When Prisoners Come Home
Sherman, Lawrence, et al (2002) Evidence-Based Crime Prevention
Travis, Jeremy (2005) But They All Come Back
Western, Bruce (2006) Punishment and Inequality in America

65

Reincarceration in Philadelphia

IN COMPLETEING THIS REPORT
SPECIAL THANKS MUST BE GIVEN TO:

CHERYL MORRISON
WITHOUT HER HELP, THIS PROJECT AND MANY OTHERS
WOULD NOT HAVE BEEN POSSIBLE. SHE IS AN EXCELLENT TEAMMATE.

Paul Heroux

66

Reincarceration in Philadelphia

And special thanks to:
Reed Domer-Shank, MS
Bruce Herdman, PhD
Rebecca Katz, MPP
Leon A. King Esq.
Kate Roach, MS, MSW
For their help in reviewing this document
&
Lt Ernie Webb
For his help in understanding release data

And to

Commissioner Louis Giorla
for making sure that this project was supported in its
completion, and his patience in the completion of this
project.

Paul Heroux

67

Reincarceration in Philadelphia

68