RESEARCH
REPORT
April 2008

Massachusetts Recidivism
Study: A Closer Look at
Releases and Returns to
Prison
Rhiana Kohl
Hollie Matthews Hoover
Susan M. McDonald
Amy L. Solomon

URBAN INSTITUTE

Justice Policy Center

THE URBAN INSTITUTE
Justice Policy Center
2100 M Street NW
Washington, DC 20037
www.urban.org

© April 2008. The Urban Institute.
This project was supported by Grant #2005-DB-BX-0014, awarded by the Bureau of Justice
Assistance, Office of Justice Programs, U.S. Department of Justice, through the Executive Office
of Public Safety and Security (EOPSS) Office of Grants and Research. The views expressed in
this report are those of the authors and do not necessarily represent the official position or
policies of the Urban Institute, the Massachusetts Department of Correction, the U.S. Department
of Justice, or EOPSS.

About the Authors
Rhiana Kohl, Ph.D., co-principal investigator for the Massachusetts Recidivism Study, is the
Executive Director of Strategic Planning and Research for the Massachusetts Department of
Correction. In that capacity, she oversees research, information technology, performance
measures, and grant management. Dr. Kohl has been integrally involved in numerous studies and
publications examining prison reentry, female offenders, inmate population trends, and
recidivism.
Hollie Matthews Hoover is a senior research analyst with the Massachusetts Department of
Correction, where she is responsible for providing statistical information on the inmate
population. Hollie is the primary contact for recidivism data for the Massachusetts Department of
Correction.
Susan M. McDonald is a senior research associate with the Massachusetts Department of
Correction, where she is involved in projects related to prisoner reentry, recidivism, and the
Prison Rape Elimination Act (PREA).
Amy L. Solomon is a senior research associate at the Urban Institute, where she works to link
the research activities of the Justice Policy Center to the policy and practice arenas. She directs
projects relating to reentry from local jails, community supervision, and innovative reentry
practices at the neighborhood level. Ms. Solomon served as co-principal investigator of the
Massachusetts Prisoner Recidivism Study.

Massachusetts DOC and Urban Institute Collaboration
This report was produced as part of the Massachusetts Prisoner Recidivism Study, a
collaborative project conducted by the Justice Policy Center of the Urban Institute and the
Massachusetts Department of Correction (DOC). A companion report within the same study was
also produced: “Reincarcerated: The Experiences of Men Returning to Massachusetts Prisons.”

i

Acknowledgments
The authors would like to acknowledge the support of James R. Bender, Deputy Commissioner
of the Massachusetts Department of Correction (DOC), and the insights from former
Commissioner Kathleen M. Dennehy, which were instrumental in initiating and completing this
project. We also appreciate the letter of support for the study by Massachusetts Parole Board
Chair Maureen Walsh; access to Parole’s SPIRIT database was very helpful in ensuring data
quality for this endeavor.
Sarah Lawrence, former Director of Research and Policy for the Executive Office of Public
Safety and Security (EOPSS), was invaluable in her role, keeping the project well-paced and
completed.
We are grateful to our colleagues at the Urban Institute and DOC Research and Planning
Division who contributed to this project. In particular, we would like to thank Lisa Brooks, Jenny
Osborne, and Jay Reid, project staff at the Urban Institute and co-authors of the companion
report of the Massachusetts Recidivism Study, Reincarcerated: Experiences of Men Returning to
Massachusetts Prisons. Throughout the development of this report, they provided key feedback.
Additionally, we appreciate the time Christy Visher, also at the Urban Institute, devoted to this
study.
At the DOC Research and Planning Division, Pamela McLaughlin and Christopher Calia’s
assistance with reviewing criminal history data as well as Emily Parson’s numerous reviews of
report documentation. Without continued patience and logistical support from the Executive
Assistant for the DOC Office of Strategic Planning and Research, Sharon Savoie, this project
would not have been completed or as enjoyable to work on. Previous contributions to this study
from Margaux Reinoso, former Executive Assistant, are also very much appreciated.
Lisa Jackson, in her former capacity as the first Director of the Reentry Unit for the
Massachusetts DOC, was part of the original concept to conduct a study interviewing incoming
recidivists. Her input and continued support has been greatly appreciated. The authors would
also like to extend their appreciation for work conducted by Brian Kearnan, DOC Deputy
Director of Program Services, and Spectrum Health Services staff.

ii

Contents
Executive Summary ........................................................................................................ 1
Introduction and Background. Massachusetts Recidivism Study .................................... 3
Section 1. Recidivism Definition and Rates: Massachusetts Department of Correction .. 7
Section 2. Characteristics of the Release Cohort and Recidivists................................... 9
Race and Hispanic Ethnicity ...................................................................................................9
Marital Status ........................................................................................................................10
Age........................................................................................................................................10
Demographic Characteristics: Recidivists versus Nonrecidivists..........................................11

Section 3. Offense Types and Time Served.................................................................. 13
Offense Types.......................................................................................................................13
Drug Offenses with “Mandatory Minimum” Sentences..........................................................15
Time Served..........................................................................................................................15

Section 4. Criminal Histories ......................................................................................... 17
History of Juvenile and Adult Arraignments ..........................................................................18
History of Adult Convictions ..................................................................................................18

Section 5. Recidivism by Release Types and Return Data ........................................... 21
Recidivism by Type of Release (Parole vs. Expiration of Sentence) ....................................21
Time from Release to Reincarceration..................................................................................22
Recidivism by Type of Return ...............................................................................................23
Security Level of Releases....................................................................................................25

Section 6. In-Prison Reentry Preparation ...................................................................... 27
Participation in Transition Planning Workshop......................................................................27

Summary of Findings and Discussion ........................................................................... 31

Tables and Figures
Figure 1. Recidivism Trends, 1995–2002 .....................................................................................7
Table 1. Recidivism Rates by Race ..............................................................................................9
Table 2. Recidivism Rates by Age at Release............................................................................10
Table 3. Demographic Information for Recidivists vs. Nonrecidivists .........................................11
Figure 2. Offense Type..... ..........................................................................................................13
Figure 3. Violent vs. Nonviolent Offense…………………………………………………………….13
Table 4. Recidivism Rates by Offense Types .............................................................................14
Table 5. Recidivism Rate for Drug Offenders by Mandatory Offense .........................................15
Table 6. Criminal History of Technical Violators, New Crime Offenders, Recidivists, and
Nonrecidivists..............................................................................................................................17
Table 7. Recidivism by Juvenile Arraignment History .................................................................18
Table 8. Recidivism by Age of First Arraignment in Adult Court .................................................19
Table 9. Recidivism by Adult Drug Arraignment History .............................................................19
Table 10. Recidivism by Number of Convictions in Adult Court..................................................20
Table 11. Recidivism Rates by Offense Type and Type of Release...........................................22
Table 12. Time until Return for Recidivists .................................................................................22
Table 13. Type of Return by Year for Male Recidivists Released in 2002..................................23
Table 14. Type of Return by Release Type for Male Recidivists Released During 2002 ...........24
Table 15. Recidivism Rate for Security Level of Releasing Institution........................................25
Table 16. Transition Workshop Participation by Release Type ..................................................27

Sidebars
Selection and Limitations of Release Cohort ................................................................................4
Definition of Terms........................................................................................................................5
Offender-Reported Postrelease Statistics...................................................................................28

EXECUTIVE SUMMARY
The Massachusetts Recidivism Study reflects a collaborative effort between the Urban Institute’s
Justice Policy Center and the Massachusetts Department of Correction (DOC) to closely examine
factors contributing to recidivism, with three interrelated study components: an analysis of DOC
administrative data and recidivism, parole officer focus groups, and interviews of recidivists as
they return to prison. This report provides the findings of the first component while the
companion report, “Reincarcerated: The Experiences of Men Returning to Massachusetts
Prisons,” provides results from the focus groups and survey of recidivists.

Key Findings
•

The overall three-year reincarceration recidivism rate among the 1,786 males released from
the DOC in 2002 was 39 percent; with technical parole violators excluded, the recidivism
rate dropped to 35 percent.

•

Recidivists, on average, were younger, served shorter-prison terms, and were more likely to
be unmarried. Additionally, blacks recidivated at a significantly higher rate than other races.

•

Property offenders had the highest recidivism rates (57 percent) with similar patterns among
inmates paroled and those released because their sentences expired. Nonviolent offenders,
consisting primarily of property and drug offenders, recidivated at a significantly higher rate
(43 percent) than violent offenders (36 percent).

•

Of the 426 inmates in the cohort serving a sentence for a drug offense, 227 (53 percent) were
serving a drug sentence associated with a mandatory minimum term; recidivism rates among
those serving a mandatory minimum drug offense were significantly lower (29 percent)
compared with the rate for nonmandatory drug offenders (46 percent).

•

On average, recidivists became involved in the criminal justice system at an earlier age and
had criminal histories with more juvenile and adult arraignments, convictions, and prior adult
incarcerations.

•

Sixty-five percent of the release cohort was released to the street via expiration of sentence;
the remaining 35 percent were paroled to the street.

•

The recidivism rate (45 percent) for inmates paroled to the street was significantly higher
than the rate (36 percent) of inmates released to the street via expiration of sentence.

•

Among the 623 inmates paroled, 29 percent were revoked and returned to prison for either a
technical violation of their parole conditions or for committing a new offense. When
technical violations were excluded from the calculation, the parolee return rate decreased to
10 percent.

1

•

Inmates released for the first time had lower recidivism rates than those who had a prior
parole or probation violation (and had been “rereleased”).

•

Almost half (47 percent) of inmates who recidivated did so within one year of being released;
by 18 months after release, 67 percent of those who recidivated had returned to prison.

•

The majority (58 percent) of first-year recidivists were returned to prison on a new sentence
(38 percent county and 20 percent state), 40 percent were returned for violations of parole
(28 percent for a technical violation and 12 percent for a new crime arrest), and 2 percent for
violations of a probation term.

•

Higher recidivism rates corresponded to release from higher security levels; half of all
inmates in the cohort released from a maximum security facility recidivated.

•

Participants in the Transition Workshop recidivated at a higher rate (43 percent) than those
who did not participate (35 percent), but also represented a disproportionate number of
inmates released from facilities with higher security levels.

2

INTRODUCTION AND BACKGROUND.

MASSACHUSETTS RECIDIVISM STUDY
In Massachusetts state prisons, recidivism statistics show a slight reduction in rates over the past
few years.1 At the same time, though, admissions, custody levels, and overcrowding rates are on
the rise. From 2004 to 2007, the criminally sentenced custody population at the Massachusetts
Department of Correction (DOC) increased 10 percent, by 897 inmates. Along with the increase
in the custody population, the number of criminally2 sentenced admissions increased by 10
percent during the same time frame. The number of parole violators has also increased and
represents approximately 10 percent of admissions among all criminally sentenced inmates over
the past four years, second only to new court commitments.
This study was originally based on prison reentry and research staff asking the question, “What
happens to inmates when they get out of prison that impacts their ability to stay out of prison?”
Furthermore, “How can we improve positive behavior and increase the chances of successful
reentry into the community by considering inmates’ previous incarceration experiences?” In our
partnership with the Urban Institute’s Justice Policy Center, we looked more closely at
recidivism in three ways: interviewing recidivists as they were returned to prison, conducting
focus groups with parole officers, and conducting statistical analyses of administrative data
comparing recidivists versus nonrecidivists. The companion report, “Reincarcerated: The
Experiences of Men Returning to Massachusetts Prisons,” covers the first two points; this report
examines the third.
The overall study was designed to improve our understanding of how the experiences of
incarceration, transition to community, and living in the community influence recidivism. This
report focuses on findings from analysis of the 2002 male release data. Traditionally, the DOC
produces annual recidivism reports.3 This report expands on those annual recidivism reports by
including additional administrative data in relation to an inmate’s risk to recidivate, including
race, Hispanic ethnicity, age, offense type, type of release from prison, type of return to prison,
security level of releasing institution, time served, criminal history, and participation in a DOC
transition workshop. This report also looks closely at recidivism data in relation to parole, while
the companion report explores the process of parole revocations by parole officers. By
comparing recidivists with nonrecidivists, we gain a better understanding of the known static
variables. This component of the larger study was also intended to provide a conceptual
framework to our findings in the survey of recidivists provided in the companion report. The
analysis of administrative data provides a basis of comparison for the sample of recidivists
interviewed for the study.
The sidebar on offender-reported postrelease statistics summarizes comparable findings from the
Massachusetts DOC postrelease survey with those from the recidivism survey conducted for this

1

Kohl, Rhiana and Hollie Matthews Hoover (2007). Recidivism Overview. Milford, MA: Massachusetts Department of
Correction.

2

Compared to other admission types, specifically those awaiting trial detainees and civil commitments.

3

Massachusetts DOC recidivism reports are available on the Internet at http://www.mass.gov/doc.

3

study. Beginning in 2003, the DOC piloted a comprehensive postrelease survey designed to
capture the progress of offenders in the community after leaving prison. All (male and female)
criminally sentenced inmates released from DOC custody via expiration of their sentence or
paroled “to the street” were deemed eligible for the survey. In that effort, attempts to telephone
all eligible inmates were conducted at 30 days, 6 months, one year, and 18 months following
release. Approximately, 10 to 12 percent of those eligible were reached, and almost all who were
asked to participate agreed to do so.
A central focus of the Massachusetts Recidivism Study has been to improve our understanding of
the inmate’s experience in his or her community after being released from the DOC, particularly,
but not limited to, factors leading to his or her return to prison. This study also presents selected
findings from the DOC postrelease pilot survey that can be compared with the recidivism survey
to enrich the findings from the perspective of released inmates who were not recidivists when
they participated in the postrelease survey.
Selection and Limitations of Release Cohort
The 2002 release cohort was selected for this study for several reasons. The Massachusetts Recidivism
Study was initiated in 2006. At the time, 2002 releases were the most current recidivism cohort to include
all three years after release, which allowed the time4 needed to collect, analyze, and report recidivism
data. Another reason for using the 2002 release cohort was because the DOC had previously collected
criminal histories for all 2002 released inmates, which, due to the nature of the court activity record
information (CARI) files, had to be done manually and was only completed for that year’s cohort of
releases. Additionally, the Massachusetts DOC created a reentry unit in 2001 dedicated to coordinating
and improving prison reentry, resulting in the expansion of transition programming and other initiatives in
2002.5
During 2002, there was less reentry programming compared with what has since been implemented, and
programming may have become available to many inmates in the recidivism survey6 who were released
after 2002. There were, however, major limitations regarding information on incarceration experiences
among the 2002 release cohort. Statistics reflecting whether or not an inmate had a probation term to
serve after being released from the DOC were not available. Data (2004 to the present) indicate
anywhere from 10 to 12 percent of inmates released to the street via parole from the DOC also had a
probation term to serve after release. Additionally, another 20 to 26 percent of inmates released without
parole had probation terms to serve, such that probation supervision was the only postrelease criminal
justice supervision in the community. In other words, more inmates released from state prisons in
Massachusetts had probation than parole supervision. Approximately 33 to 40 percent of inmates
released to the street were under no criminal justice postrelease supervision.

4
In addition to the three years after a calendar year of inmates released has passed, at least one additional year is needed for all
the court information to be entered into the CARI files and the time needed to extract, analyze and report the data.
5

Unfortunately, program participation data and other key indicators reflecting inmates’ incarceration experiences were largely
inconsistent or unavailable among the records of inmates released in 2002, one drawback of using this release cohort.

6

Inmates surveyed for the companion portion of the Massachusetts Recidivism Study consisted of 178 male recidivists who had
previously been released from the DOC between 2003 and 2006.

4

Definition of Terms
CARI Files. The court activity record information (CARI) file provides criminal history information starting
with each arraignment. The Massachusetts Board of Probation (BOP) maintains the CARI file on the
Massachusetts criminal justice information system (CJIS).
Earned Good Time (Massachusetts General Law: M.G.L. c. 127, § 129D). An inmate may receive a
maximum of 7.5 days a month deduction from sentence for satisfactory performance in an approved
employment, educational, or vocational training program or activity. An inmate may earn good time unless
the law under which the inmate is committed specifically prohibits deductions for earned good time. Any
deductions from sentence earned, once granted, cannot be forfeited for disciplinary infractions.
Expiration of Sentence. Inmates who serve out their sentence (less any time deducted for earned good
time or statutory good time).
“From and After” Sentence. A sentence (which can be a probation term) which does not take effect until
the completion of another previously imposed sentence is a consecutive or “from and after” sentence. A
consecutive sentence commences upon discharge of the earlier sentence
Governing Offense. The offense associated with the longest maximum discharge date, when there are
multiple offenses per inmate.
Mandatory Minimum Sentence. A minimum sentence term required by statute (Massachusetts General
Law) for inmates convicted of specific crimes, including but not limited to numerous drug offenses and
crimes against a person.
Nonviolent Offense. Any offense that falls under the categories of property, drug, or “other.”
Parole Regional Reentry Center. Eight centers throughout Massachusetts operated by the Massachusetts
Parole Board servicing all individuals under parole supervision and any inmate returning to the
community.
Parole Violation. An act or failure to act by a parolee that does not conform to the conditions of parole
related to having been arrested for a new offense.
Provisional Revocation. A provisional revocation is the withdrawal of a decision to parole an inmate on a
provisional or temporary basis. Only parole board members have the authority to provisionally revoke
parole.
Technical Violation. A violation of parole conditions that does not necessarily constitute grounds for
revoking parole or being arrested for a new offense.
Time Served. For the purposes of this report, time served is calculated using the offenders incarceration
and release dates. For inmates who were serving a state prison sentence, jail credits are included in the
calculation.
Violent Offense. Any offense that falls under crimes against the person or a sex offense category.
Wrap-ups. A term for inmates released via expiration of sentence, having completed their sentence with
no time left or without any period of parole remaining

5

SECTION 1.

RECIDIVISM DEFINITION AND RATES: MASSACHUSETTS
DEPARTMENT OF CORRECTION
The Massachusetts DOC defines a recidivist as any inmate7 released from the DOC in a given
year who is reincarcerated8 within three years of his or her release to the community for either a
new sentence or a technical violation. In 2002, the cohort of inmates “released to the street”
totaled 2,524 inmates; 1,786 were males (71 percent) and 738 were females. The data and the
following discussion focus on the male cohort. This report provides and compares findings on
male recidivists and nonrecidivists within three years of being released in 2002.
This report provides macro-level information on what happens to all criminally sentenced male
prisoners released from the DOC in 2002, while the companion report, “Reincarcerated: The
Experiences of Men Returning to Massachusetts Prisons,” provides a more detailed description
of what led to reincarceration for a subset of individuals. Three-year recidivism rates for the
Massachusetts DOC date back to 1995.9 Overall, three-year recidivism rates for males dropped
from approximately 44 percent in 1995 and 1996 to approximately 39 percent for subsequent
years through 200210. One year postrelease recidivism rates, as reflected in figure 1, remained
relatively consistent across these years, fluctuating slightly between 18 and 22 percent and
suggesting a slight overall decline.
Figure 1. Recidivism Trends, 1995–2002
Recidivism Rates for Male Releases to the Street from a
Massachusetts DOC Facility 1995 - 2002
50%

44%

44%

43%

42%

40%
30%
20%

21%

22%

20%

40%

39%

20%

20%

39%

18%

18%

10%
0%
1995

1996

1997

1998
3 Year Rate

1999

2000*

2001

2002

1 Year Rate

Note: 2000 data are not yet available.

7

This includes all criminally sentenced inmates released from a DOC facility (state and county sentenced).

8

To a Massachusetts state, county, or federal correctional facility.

9

The Massachusetts DOC expanded the recidivism definition from one to three years postrelease starting with the 1995 release
cohort. Resources and time needed to collect and analyze recidivism data are taxing, even after progress has been made to collect
data electronically versus manually.
10

After completing the recidivism report based on 1999 releases, it was determined to skip ahead to 2002 releases and work
backward. As a result recidivism data based on the 2000 release population has not yet been completed.

7

SECTION 2.

CHARACTERISTICS OF THE RELEASE COHORT AND
RECIDIVISTS
Race and Hispanic Ethnicity
When analyzing the characteristics of the release cohort, statistical testing11 was used to
determine if there was a statistically significant difference between recidivists and nonrecidivists.
A significant difference was found between these two groups in race, marital status, and age;
however, there was no significant difference for Hispanic ethnicity. Black and Native
American/Alaskan inmates recidivated within three years at disproportionately higher rates than
Caucasian and Asian inmates. Twenty-five percent of males released reported a Hispanic
ethnicity12; the recidivism rate among male Hispanics was 41 percent, which differed minimally
from non-Hispanics.
Table 1. Recidivism Rates by Race

2002 Male Release Cohort Recidivism Data

Race

Number
of
releases

Percent of
releases

Number
of
recidivists

Recidivism
rate (%)

African American

531

30%

234

44%

Native
American/Alaskan

13

1%

5

38%

1,233

69%

461

37%

Asian

9

1%

0

0%

Total

1,786

100%

700

39%

Caucasian

11

Pearson’s Chi-Square test was used to determine statistical significance (p < 0.05).

12

Hispanic ethnicity is included separately from race by the DOC, consistent with the U.S. Office of Management and Budget
(OMB) statistical policy.

9

Marital Status
Most (68 percent) of the men in the cohort reported themselves as “single.” The marital status of
recidivists was significantly different than that of nonrecidivists. Recidivists were more likely to
report being single (74 percent) than nonrecidivists (63 percent), and they were less likely than
nonrecidivists to report being married (12 percent and 15 percent, respectively). Further, inmates
reporting a marital status of single had a recidivism rate of 43 percent, compared to a recidivism
rate of 30 percent for married male inmates. This finding was consistent with other studies
identifying male married inmates as having a lower risk of reoffending.

Age
Age at incarceration yielded statistically significant results as well. On average, inmates were 30
years old when they were incarcerated, ranging from 15 to 75 years old. Significantly more
recidivists were under age 35 when they were originally incarcerated. Since the average time
served among DOC males is between four and five years, an inmate’s average age increases by
about five years from the time he is incarcerated until he is released. Although the average age at
release for this cohort was 35 years, a disproportionate amount of recidivists were under age 35
at release and when they were reincarcerated. This finding, that significantly more recidivists
were younger than 35 when incarcerated, when released, and when reincarcerated supports other
research that younger inmates have higher risks of reoffending.
Table 2: Recidivism Rates by Age at Release

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

35 years or older

853

283

33%

Younger than 35 years

933

417

45%

1,786

700

39%

Age at time of release

Total

10

Demographic Characteristics: Recidivists versus Nonrecidivists
A disproportionate number of recidivists in the 2002 cohort were black, single, and younger
when incarcerated, released, and reincarcerated compared with those who did not recidivate to
prison. Table 3 shows further comparisons between the two groups based on age, race, ethnicity,
and marital status.

Table 3. Demographic Information for Recidivists vs. Nonrecidivists

2002 Male Release Cohort Recidivism Data
Characteristic (selfreported)

Recidivists

Nonrecidivists

(n = 700)

(n = 1,086)

Average age at incarceration

28 years

31 years

Average age at release

33 years

36 years

Caucasian

66%

71%

Black

33%

27%

American Indian/Alaska Native

1%

1%

Hispanic

26%

24%

Never been married

74%

63%

Married

12%

17%

Divorced/Separated

13%

16%

11

SECTION 3.

OFFENSE TYPES AND TIME SERVED
Offense Types
The largest proportion of offenders among the male release cohort were those serving time for a
crime against a person (42 percent), followed by drug (24 percent), property (16 percent), sex (11
percent) and other (7 percent) offense types. Among the 1,786 males released in 2002, 54 percent
had been incarcerated for a violent offense (figure 3). The remaining offense types consisted of
nonviolent governing offenses.13

Figure 2: Offense Type

Figure 3: Violent vs. Nonviolent Offense

Offense Type for 2002 Male
Release Cohort

Property
16%

Violent vs. Nonviolent Offenses for
2002 Male Release Cohort

Other
7%
Person
42%

Violent
Offense
54%

Sex
11%

Nonviolent
Offense
46%

Drug
24%

13

Nonviolent offense types include those for drug, property, or “other” offenses. The “other” offense category includes those
related to motor vehicle crimes, such as operating under the influence, weapons possession, and public disorder. The data
relevant to offense types are based on the “governing offense,” which is an inmate’s offense with the longest maximum sentence.

13

The finding, as shown in table 4, that those serving time for property crimes had the highest
recidivism rate (57 percent) was statistically significant and consistent with findings from
previous years (in Massachusetts as well as national studies). Further analysis of property-crime
recidivists in terms of their history of adult convictions showed property offenders who
recidivated had many more prior adult convictions, 41 on average, compared with nonrecidivists,
who averaged 26 convictions.

Table 4. Recidivism Rates by Offense Types

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

Property

281

159

57%

Person

761

302

40%

Drug

426

158

37%

Other

119

38

32%

Sex

199

43

22%

1,786

700

39%

Offense type

Total

14

Drug Offenses with “Mandatory Minimum” Sentences
Of the 426 inmates in the release cohort serving time for a drug offense, 227 (53 percent) were
serving a drug sentence with a mandatory minimum term. Over the past 30 years, there have
been many changes in sentencing legislation and corrections policies nationally and at the state
level. Massachusetts enacted statutes requiring mandatory minimum sentencing on specific
crimes, including numerous drug offenses. Such statutes not only impact the minimum term of a
given sentence but may include restrictions associated with that minimum term, such as
participation in work-release or community-based programs.14 Offenders serving nonmandatory
drug sentences recidivated at significantly higher rates compared with offenders serving
mandatory drug sentences (table 5).

Table 5. Recidivism Rate for Drug Offenders by Mandatory Offense

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

Not a Mandatory Drug Offender

199

92

46%

Mandatory Drug Offender

227

66

29%

Total

426

158

37%

Mandatory offense

Note: Of the drug offenders released during 2002, mandatory offenders had the lower rate.

Time Served
The average time served calculated for the 2002 male release cohort was 50 months or just over
four years.15 Less time incarcerated was negatively associated with recidivism rates; in other
words, as time served increased, recidivism rates decreased. This trend generally held true; as
inmates who served six months or less had a recidivism rate of 48 percent, inmates who served
between six months and five years had a rate of 43 percent, and inmates who served more than
five years had a rate of 30 percent.

14

See Rhiana Kohl, Statutory Restrictions on Inmate Placement and Other Sentencing Related Statistics. Milford, MA:
Massachusetts Department of Correction, 2007.
15

This calculation reflects the time an inmate was incarcerated after conviction, including jail credits or a complex sentence at
the time of release.

15

SECTION 4.

CRIMINAL HISTORIES
Recidivists in the cohort had a more extensive criminal history than nonrecidivists. On average,
recidivists started their criminal careers at a younger age. Similarly, recidivists’ alcohol and
substance abuse arraignments occurred at younger ages than nonrecidivists. Table 6 shows
average ages and totals for criminal history data for recidivists and nonrecidivists. This table also
highlights recidivists broken out by those who returned for a “new crime” versus “technical
violation” of parole.
The analysis of the 2002 release cohort was conducted, in part, to present a framework on which
to base the findings from the Massachusetts recidivism survey.16 The last column in table 6
provides results in the analyses of criminal history data for all inmates (recidivists) who
participated in the survey. An assessment of these findings demonstrates that the criminal
histories of inmates who returned to the DOC and participated in the survey, generally, had
criminal histories similar to the 2002 release cohort of recidivists.
Table 6. Criminal History of Technical Violators, New Crime Offenders, Recidivists, and
Nonrecidivists
2002 Release cohort by
recidivism

2002 Release cohort
recidivists by return type
Technical
violators
(n = 115)

Recidivism
survey
participants
(n = 178)

Nonrecidivists
(n = 1086)

Recidivists
(n = 700)

New crime
offenders
(n = 585)

Age at first arraignment

21

19

19

19

19

Number of juvenile
arraignment charges

3

6

6

4

6

Age at first adult drugrelated arraignment

23

22

22

21

23

Age at first adult alcoholrelated arraignment

26

22

22

22

21

Number of prior adult
arraignment charges

31

47

48

39

47

Number of prior adult
incarcerations

3

5

5

4

4

Criminal history

Note: Multiple arraignments can represent one arrest incident.

16

As reported in “Reincarcerated: The Experiences of Men Returning to Massachusetts Prisons.”

17

In Massachusetts, police statistics are not currently maintained in the same database with court or
DOC (state correction) data. Thus, in lieu of traditional arrest data, arraignment statistics were
included in this report.17 An analysis of juvenile and adult criminal histories revealed most
offenders had at least one prior incarceration and many had considerable juvenile records. The
mean number of juvenile arraignments was close to four and close to two for juvenile
convictions. On average, this cohort had over 37 prior adult arraignments and 18 adult
convictions. Most of those who served prison time did so in Massachusetts county facilities.

History of Juvenile and Adult Arraignments
Table 7 shows that 58 percent of inmates in the cohort had no juvenile arraignments, and 42
percent had at least one juvenile arraignment. Offenders who had at least one juvenile
arraignment18 recidivated at a disproportionately higher rate, making this variable a strong risk
factor for reoffending.
Table 7. Recidivism by Juvenile Arraignment History

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

No juvenile arraignments

1029

326

32%

At least one juvenile arraignment

757

374

49%

Total

1786

700

39%

Juvenile arraignment history

The average age in this cohort for first adult arraignment was 20 years old. The average age at
first arraignment for a drug or alcohol-related charge was 23 years old, with 68 percent having a
history of at least one drug charge and 37 percent a history of at least one alcohol charge.

History of Adult Convictions
In terms of actual adult convictions, the average age at the first adult conviction for offenders in
the cohort was 22 years old. Fifty-seven percent of the cohort had a history of at least one drug
conviction. The average age of first drug or alcohol-related conviction was between 24 and 25
years old. Recidivism rates were inversely and significantly correlated to an inmate’s age at his
first adult arraignment or conviction. The younger he was when first arraigned or convicted in an
adult court, the higher his rate of recidivism among the 2002 release cohort.

17

The vast majority of arrests resulted in a court arraignment within a very brief time from arrest, usually one day, making
“arraignments” a suitable replacement for arrest data.
18

Inmates arraigned in a juvenile court or arraigned in an adult court at age 17 or younger are included.

18

Table 8. Recidivism by Age of First Arraignment in Adult Court

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

17 yrs of age or less

842

410

49%

18 yrs of age

295

109

37%

19 - 22 yrs of age

309

98

32%

23 yrs of age or more

305

74

24%

1,751

691

39%

Age of first arraignment in adult court

Total

Note: There were 35 inmates with missing age at arraignment information.

History of a drug arraignment was also statistically significant for recidivism. With 67% percent
of the release cohort having a criminal history with a drug arraignment, this was a significant
factor related to recidivism rates. Table 9 illustrates how 45 percent of offenders with a history of
drug arraignments recidivated as opposed to only 28 percent for offenders with no drug
arraignments.
Table 9. Recidivism by Adult Drug Arraignment History

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

562

159

28%

At least one adult drug arraignment

1,195

534

45%

Total

1,757

693

39%

Adult drug arraignment history
No adult drug arraignments

Note: There were 31 inmates with missing drug arraignment history data.

19

As table 10 shows, almost all (98 percent) of the male release cohort had a history of at least one
adult conviction. Analysis of the data on prior incarcerations19 in a state, county, or federal prison
facility indicated approximately 72 percent of the cohort had been previously incarcerated prior
to the prison term they were released from in 2002. Among the 700 recidivists, 85 percent had a
prior incarceration, compared with 65 percent of the nonrecidivists, which was a statistically
significant difference between the two groups. Having a history of a prior incarceration was less
pronounced among those in the cohort who were paroled (67 percent), compared with those
released via expiration of sentence (75 percent). In sum, most men released from the DOC in
2002 were, at least conceptually, already recidivists at the start of that incarceration term.
Table 10. Recidivism by Number of Convictions in Adult Court

2002 Male Release Cohort Recidivism Data
Number of convictions in adult
court

Number of
releases

Number of
recidivists

Recidivism
rate (%)

6 or less

457

105

23%

7–13

452

154

34%

14–23

385

172

45%

24 or more

462

262

57%

1,756

693

39%

Total

19

Prior incarceration data were not available for 41 inmates in the cohort.

20

SECTION 5.

RECIDIVISM BY RELEASE TYPES AND RETURN DATA
Recidivism by Type of Release (Parole vs. Expiration of Sentence)
A major focus of this project was to consider the impact of inmates paroled versus those released
via expiration of sentence (“wrapped up”). The majority (65 percent) of those released served
their prison term and wrapped up, compared with 35 percent who were paroled. The recidivism
rate among parolees was 45 percent, compared with 36 percent for those released via expiration
of sentence. Among the 623 inmates paroled, 29 percent were returned to prison for a parole
violation.20 When technical violators are excluded from the calculations, the percentage of those
paroled who returned as parole violators decreased to 10 percent. The total recidivism rate for the
male cohort decreased from 39 to 35 percent when excluding technical violators.
Recidivism rates are based on each “release to the street” in a given calendar year. Each year,
some inmates paroled and revoked to prison are rereleased from the same sentence. For 19
percent of the 2002 cohort (n = 335), this was not their first release for the same sentence. The
recidivism rate among those having their “first release” was 38 percent, compared with 44
percent among those for which it was not their first release.
To further examine the differences between inmates paroled versus those who “wrap up,” crosstabulations were calculated in relation to offense types (table 11). Parolees of all offense types
had a higher recidivism rate than inmates who “wrapped-up,” with the exception of inmates
serving an offense in the “other” category. Recidivism rates among parolees serving a violent
(person or sex) or drug offense were significantly higher than those released via expiration of
sentence for the same offense types. This may be due, in part, to the close supervision of these
known violent offenders. This close supervision could result in more technical violations, such as
failing a drug test, rather than the offender committing new crimes.

20

Inmates being returned to prison on a violation of parole with a new offense or a technical violation are returned on the
sentence they were serving prior to their release.

21

Table 11. Recidivism Rates by Offense Type and Type of Release

Type of Release
Parole to street

Expiration of sentence

Number of
releases

Number of
recidivists

Recidivism
rate (%)

Number of
releases

Number of
recidivists

Recidivism
rate (%)

Person

277

125

45%

484

177

37%

Sex

57

14

25%

142

29

20%

Property

93

54

58%

188

105

56%

Drug

160

78

49%

266

80

30%

Other

36

11

31%

83

27

33%

Total

623

282

45%

1,163

418

36%

Offense type

Time from Release to Reincarceration
Most inmates who recidivated did so within the first 18 months of being released. Table 12
shows almost half (47 percent) of the recidivists in the sample returned to prison within one year
of release. After six more months, two-thirds of the recidivists were back in prison.
Table 12. Time until Return for Recidivists

2002 Male Release Cohort Recidivism Data
Number of
recidivists

Percent
returned (%)

< 30 days

12

2%

2%

1 to < 6 Months

129

18%

20%

6 to < 12 Months

186

27%

47%

12 to < 18 Months

144

21%

67%

18 to < 24 Months

109

16%

83%

24 to < 30 Months

59

8%

91%

30 to 36 Months

61

9%

100%

Time until return

Cumulative
total (%)

22

Recidivism by Type of Return
The majority (58 percent) of first-year recidivists were returned for a new sentence21 (38 percent
county and 20 percent state). An additional 40 percent of male recidivists were returned for
violations of parole, 28 percent on a technical violation, and 12 percent for a new crime arrest.
Parole violation rates were highest in the first year after release.
Table 13. Type of Return by Year for Male Recidivists Released in 2002
Year of Return for Recidivists
Type of
return

First Year

Second Year

Third Year

Three Year Total

Number
of
recidivists

Percent
returned
(%)

Number
of
recidivists

Percent
returned
(%)

Number
of
recidivists

Percent
returned
(%)

Number
of
recidivists

Percent
returned
(%)

Technical
parole
violator

93

28%

17

7%

3

2%

113

16%

Parole
violator new
offense

41

12%

17

7%

7

6%

65

9%

County
sentence

125

38%

165

66%

84

69%

374

53%

State
sentence

65

20%

43

17%

25

21%

133

19%

Technical
probation
violator

0

0%

1

0%

1

1%

2

0%

Probation
violator new
offense

6

2%

6

2%

1

1%

13

2%

330

100%

249

100%

121

100%

700

100%

Total

21

This includes inmates reincarcerated due to receiving a new state or county criminal sentence.

23

While criminally sentenced inmates are released “to the street” from the DOC either via
expiration of sentence or parole, they can be returned to prison for a variety of reasons.22 Table
14 reflects the different “type of returns” among recidivists based on either being released via
expiration of sentence or paroled. Almost half (43 percent) of the 700 recidivists who wrapped
up were reincarcerated for a new county sentence. The second largest subgroup by type of return
was parolees revoked for technical violations, followed closely by inmates wrapping up and
returned to prison for a new state sentence.
Table 14. Type of Return by Release Type for Male Recidivists Released During 2002

Type of Release
Parole

Expiration of Sentence

Number

Percent of return
type

Number

Percent of return
type

113

40%

0

0%

Parole violator, new offense

65

23%

0

0%

County sentence

70

25%

304

73%

State sentence

26

9%

107

26%

Technical probation violator

1

0%

1

0%

Probation violator, new offense

7

2%

6

1%

282

100%

418

100%

Type of return
Technical parole violator

Total

22

Only those paroled on a given sentence can be revoked by parole on that sentence. Inmates released with a probation term to
serve as part of a “from and after” sentence can be revoked on a probation violation back to prison based on the sentence from
which they were originally released. It is possible for an inmate to be under both parole and probation supervision.

24

Security Level of Releases
The year 2002 represented a time of transition at the Massachusetts DOC, when some prison
facilities were closed while others changed their mission in terms of housing by gender or
security levels. Most of the 21 DOC facilities housing males in 2002 were medium security.
Among 2002 males released, 11 percent were released from a maximum security facility, 58
percent from a medium security facility, and the remaining 31 percent from a
minimum/prerelease facility.
Seven DOC facilities accounted for 61 percent of all male releases in 2002. MCI-Concord, the
male reception site for the DOC, had the highest number of inmates released and returned to
prison within three years. The recidivism rate of 46 percent among inmates released from MCIConcord was most likely related to MCI-Concord having released the most parolees/parole
violators in the department in 2002. As seen in table 15, as security levels increase, recidivism
rates increase; this finding was statistically significant.
Table 15. Recidivism Rate for Security Level of Releasing Institution

2002 Male Release Cohort Recidivism Data
Number of
releases

Number of
recidivists

Recidivism
rate (%)

Maximum

193

96

50%

Medium

985

392

40%

Minimum/prerelease

443

153

35%

Treatment and support facilities*

165

59

36%

Total

1786

700

39%

Security level of releasing institution

*Treatment facilities as of 2002 included Massachusetts Alcohol and Substance Abuse Center (MASAC), Bridgewater State Hospital
(BSH), the Massachusetts Treatment Center (MTC, a facility for sex offenders), Shattuck Hospital Correctional Unit, and the
23
Massachusetts Boot Camp.

23

The recidivism rate among the 63 inmates released from the Boot Camp was 49 percent, which was closed in 2002. The
recidivism rate at MTC was 21 percent. Many of the inmates housed at MTC and most of the inmates at BSH and MASAC were
civil commitments.

25

SECTION 6.

IN-PRISON REENTRY PREPARATION
Participation in Transition Planning Workshop
During 2002, a five-day workshop was available at most facilities24 for inmates with a parole
eligibility date or projected release date within a year.25 Approximately 900 of the release cohort
participated in this Transition Workshop prior to their 2002 release.26 Participants recidivated at a
higher rate (43 percent) than those who did not participate (35 percent). The higher recidivism
rate among participants may be more directly associated with the fact that 78 percent of them
released from higher-security facilities (12 percent from maximum and 66 percent from medium
security facilities). In comparison, 61 percent of those who did not participate in the workshop
were released from a maximum (10 percent) or medium (51 percent) security facility.27
Table 16. Transition Workshop Participation by Release Type

Release Type
Parole to Street
Transition
Workshop

Expiration of Sentence

Number of
releases

Number of
recidivists

Recidivism
rate (%)

Number of
releases

Number of
recidivists

Recidivism
rate (%)

No Participation

341

128

38%

546

185

34%

Participation

282

154

55%

617

233

38%

Total

623

282

45%

1,163

418

36%

Note: Workshop participation data reflect only 629 of the 899 participants who completed the workshop. 28

24

This transition workshop continues to be offered, on a voluntary basis by inmates, at the Massachusetts DOC in an increased
number of facilities compared to the availability in 2002 and has been expanded to cover ten days rather than five. In 2002, most
DOC facilities offered the transition workshop, except for the treatment-type facilities and female prisons. Thus, if an inmate was
housed in one of the treatment facilities at the time of eligibility, he would not have been able to participate.
25
Since inmates become eligible for the workshop one year prior to their parole eligibility date and participation is voluntary, the
time between when an inmate participates in the workshop and is released runs the gamut from weeks to years. Inmates may take
the workshop more than once within a sentence term but are only eligible for earned good time credits for the first time they
complete the program.
26

Due to limitations in the data tracking for this program, it was difficult to obtain an accurate number of participants or
determine whether they took the workshop.
27

Approximately 45 percent of those paroled in 2002 participated in the transition workshop. When comparing workshop
participants with those who did not participate by whether they were released via expiration of sentence or paroled, those paroled
who did participate recidivated at the highest rate (55 percent).

28

Reasons for not completing the Transition Workshop include "dropping out" as well as being released or transferred from the
facility offering the program.

27

Offender-Reported Postrelease Statistics
Many DOC postrelease survey29 questions and topics were captured in the Massachusetts recidivism
survey30 and are reported here. A major difference between the two surveys is that the postrelease survey
was based on inmates who had not returned to prison since being released, whereas the recidivism
survey was based on those who did return to prison and was conducted at the point of that
reincarceration. Furthermore, questions posed by telephone for the postrelease survey were general in
nature, whereas those asked in person delved more deeply into the topics. The following discussion
explores data for housing, family relations, employment, substance abuse, and health care.

Housing
DOC postrelease survey respondents were more likely to live in permanent housing, with 55 percent
living in permanent housing at 30 days and 66 percent at six months after release. One reason for this
was the nature of the survey sample, which consisted of inmates released to the street who could be
reached via telephone.31 The respondents interviewed for the Massachusetts Recidivism Study seemed
less likely to live in stable housing. Fifty-seven percent moved once and 23 percent moved multiple times.
Three-quarters of the postrelease survey respondents who reported living in a private residence at 30
days after release stated that they were living with a family member, including a spouse. Only 9 percent
reported living alone. By six months, only half reported living with family, and those living alone increased
to 21 percent. During most of their time in the community before reincarceration, 37 percent of the
Massachusetts recidivism survey respondents lived in their own house or apartment, and 22 percent lived
in the home of a family member. Eighty-five percent did not live alone, living with others in their own
home or living in the home of a family member or friend.
Among the 126 postrelease survey respondents at 30 days who were living in temporary housing, 42
percent reported that they were having trouble finding permanent housing. Sixteen percent of the inmates
interviewed for the Massachusetts Recidivism Study said they had trouble finding a place to live, in most
cases because they lacked money for rent or deposit. Comparisons between the two respondent groups
were skewed by the fact that only those in temporary housing were asked about challenges of finding
permanent housing.

Marital and Family Support
Most respondents for both surveys reported that their families were highly supportive during their
transition out of prison: for postrelease survey respondents, 92 percent at 30 days and 99 percent at six
months, and for the recidivism study, 86 percent of inmates.
Most male respondents in the postrelease survey did not report having children under 18 years old. Of
those who did have minors, most reported not living with those children. Similar findings were reported by
inmates participating in the recidivism study; 69 percent of those with minor children reported not living
with their children at the time of their arrest.

29

Data included from the Massachusetts DOC postrelease survey reflect 502 telephone interviews with males conducted 30 days
and/or six months after release between May 2004 and December 2006: 382 at 30 days and 120 at six months after release. DOC
researchers made efforts to contact all eligible released inmates at each postrelease time interval unless it was determined that
they were back in prison (or deceased). Efforts to reach all released offenders at each interval were made regardless of whether
they had been reached previously. For example, a released offender may have participated in completing a survey six months
after release even if he or she had not participated in the 30 day postrelease survey, so all the same respondents were not reached
at both one and six months after release. Not all respondents answered all questions.
30

Complete findings from the Massachusetts recidivism survey are in the companion report for this study, “Reincarcerated: The
Experiences of Men Returning to Massachusetts Prisons.”

31

Released individuals living in temporary housing, such as shelters or residential programs, as well as those who were
homeless, were less likely to be reached due to the lack of a telephone or confidentiality issues at residential programs.

28

The postrelease survey respondents who did live with their children reported the children’s adjustment as
being above 7 on a scale of 1 to 10 with 10 being “excellent.” Seventy percent of recidivism survey
respondents reported it was easy for their children to adjust to their return from prison. Generally
speaking, findings from both surveys revealed a perspective that families adjusted well to an inmate’s
transition out of prison and were often viewed as being supportive in the process.

Employment
The postrelease survey respondents’ employment status appears to improve from 30 days to six months
out of prison. Although 44 percent of respondents were working at 30 days after release; by six months,
65 percent were working in some capacity, and only 35 percent reported not working at all. Seventyseven percent of the respondents from the Massachusetts recidivism survey worked at some point
between their release and rearrest, while only 56 percent were employed at the time of their arrest. Just
under a third of the unemployed postrelease survey respondents saw their criminal record as a handicap
to getting a job. In the Massachusetts recidivism survey, 68 percent of respondents, even those who did
find employment, felt their criminal records had affected their ability to find work.
As reported in the recidivism survey, structural work (construction) was the most common type of
employment (32 percent) among postrelease survey respondents. Other employment types at 30 days
included factory and warehouse work (11 percent) closely followed by food services (10 percent) with
similar findings at six months out. Forty-one percent of Massachusetts recidivism survey respondents had
been employed in construction with no other occupation accounting for more than 10 percent of reported
jobs.
At 30 days, 30 percent of postrelease survey respondents reported finding their job through a family
member, 24 percent from a job listing, and 12 percent through a nonfamily contact. At six months, the
percentages were slightly higher.32 This is consistent with more respondents reporting employment.
Altogether, approximately 10 percent of the postrelease survey respondents reported having had a job
lined up prior to having been released from incarceration. At 30 days after release, 2 percent reported
having jobs arranged by the Department of Correction, and 7 percent had participated in prerelease work
while in prison; by six months, 1 percent of those employed reported having had participated in
prerelease work. Seventeen percent of inmates interviewed for the Massachusetts Recidivism Study
participated in prerelease work or vocational training, which might explain the higher proportion reporting
being employed.
Despite the fact that a small percentage (9 percent at 30 days and 13 percent at six months) of the
postrelease survey respondents who were working reported earning over $501, over 60 percent of the
173 ex-offenders responding at 30 days and the 79 ex-offenders responding at six months felt they
earned enough to support basic expenses. The median income for the Massachusetts recidivism survey
respondents was $2,200 per month (about $550/week), and the underlying theme was that many, if not
most, were not able to meet their financial obligations. Those in the postrelease survey averaged less
income but reported being able to meet their expenses. This apparent discrepancy may be related to their
not having as many additional costs, such as supervision fees, debt, and court fines.

32

Thirty-five percent found their job through a family member, 22 percent via a nonfamily contact, and 17 percent from a job
listing.

29

Substance Abuse
An average of 95 percent of postrelease survey respondents reported reducing their drug and alcohol
intake within 30 days of being interviewed for the survey whether at the 30 day or six month interval. The
majority reported having participated in self help/peer counseling groups, such as Alcohol Anonymous
(AA), Narcotics Anonymous (NA), Cocaine Anonymous, etc. (85 percent at 30 days and 94 percent at six
months). Eleven percent spent time in a special facility, and 3 percent spent time in a detoxification unit,
at the 30 day survey administration. Over a third of the Massachusetts recidivism survey respondents
reported that they had attempted to stop using drugs or alcohol in the months before their return to prison,
but many acknowledged abusing alcohol and/or illegal drugs. Among frequent users, 62 percent reported
having tried to stop using drugs or alcohol. Nearly half of the respondents reportedly attended some type
of substance abuse treatment, including AA and NA.

Health Care
In terms of having any medical insurance, 63 percent of those respondents to the postrelease survey
after one month from release reported having insurance and 76 percent of those after six months after
release. Just over half of the postrelease survey respondents conveyed they had Mass Health insurance.
Other means of coverage were through family or a job. Thirty-one percent at 30 days and 18 percent at
six months reported they did not have any health care coverage. Three-quarters of the Massachusetts
recidivism survey respondents had health care coverage, 57 percent were covered by Mass Health (on
par with the postrelease respondents), 11 percent through a job, 5 percent through Medicare, and 6
percent were covered by a family member.
Of the 67 postrelease survey respondents who responded to the question on medical and mental health
care at 30 days after release, 3 percent reported receiving inpatient medical care, and 7 percent inpatient
mental health care. In turn, 14 percent of the Massachusetts recidivism survey respondents had been
hospitalized for a physical health problem.
Due to the self-selected nature of the sample for the postrelease survey, released individuals who were
surveyed were arguably in relatively stable situations where they could be reached by telephone,
especially after six months. Most individuals in the release cohort were not reachable and many were not
included because they had already returned to prison. A broad assessment of the surveys combined
would speculate that, except for a few offenders who get into trouble immediately after being released,
most spend the first few months trying to get on their feet, and that six-month time frame seems to be
pivotal to their success.

30

SUMMARY OF FINDINGS AND DISCUSSION
In 2002, the Massachusetts Department of Correction (DOC) released 1,786 male inmates to the
community via expiration of sentence or parole. In total, 39 percent recidivated by being returned
to prison for a new sentence or on a technical parole violation within three years of release. On
average, the 700 men in the 2002 release cohort who returned to prison were young, single, and
more likely to commit nonviolent (i.e., property) crimes. These characteristics were consistent
with other findings in the literature. Likewise, a disproportionate number of black inmates
recidivated at a statistically higher rate of 44 percent compared with other races. Whether these
demographic characteristics are a function of behavior, environment, the criminal justice system,
or other factors needs to be further explored.
In general, inmates released from the Massachusetts DOC had dense and lengthy criminal
histories. When a comparison was made between recidivists and nonrecidivists in the cohort,
recidivists, on average, were younger when they became involved in the criminal justice system
and accumulated more arraignments, convictions, and incarcerations. The criminal histories of
the recidivist cohort among the 2002 releases and the criminal histories of the survey respondents
who participated in the recidivism survey were very similar. Within the recidivists, on average,
parolees returned for a technical violation had a less dense criminal history, perhaps because
their parole officers returned them to prison before they could commit additional offenses.
Most noteworthy was that at least 72 percent of the men in the release cohort had been
incarcerated at least once prior to entering prison for their current sentence. This was most
pronounced for the 700 recidivists in the release cohort, as well as those who had been released
via expiration of sentence. A critical component of effective reentry planning relies on risk and
need assessments at the onset of incarceration. Such assessments should expand beyond
identifying key static and dynamic factors identified from the research and address individual
experiences in prior incarcerations as well as failures in the community that resulted in being
reincarcerated.
Often the public perceives those who have committed serious violent crimes as having the
highest “risk to recidivate.” Research has demonstrated that committing a violent crime is not
necessarily a predictor of having a higher risk to reoffend. Among the 2002 cohort of male
releases, nonviolent offenders recidivated at a rate of 43 percent, which was significantly higher
than the 36 percent among violent offenders. These figures were driven largely by the extremely
high recidivism rate of 57 percent among property offenders, which is consistent with
Massachusetts and national recidivism studies. The misconception that serious violent offenders,
including sex offenders, will reoffend may be due in part to misrepresentations in the media and
a general fear of these particular offenders.
During the “war on drugs,” Massachusetts implemented mandatory minimum sentence statutes
on specific types of drug offenses, which not only require the minimum sentence be at least that
promulgated by statute, but also restrict inmates from participating in work release or any other
community-based program during their mandatory term and often their entire incarceration.
Currently, this issue is being reassessed politically as the restrictions impact reentry preparation.
Our analysis compared inmates serving mandatory minimum sentences with those without

31

among all 426 drug offenders in the cohort. The results showed drug offenders with mandatory
minimum sentences recidivated at a statistically significant lower rate (29 percent) compared
with the “nonmandatory” drug offenders (46 percent).
Further exploration into this discrepancy should factor in differences in time served, release type
(i.e., parole), and the nature of the drug offense. By having mandatory minimum terms, do
inmates serve longer sentences than those who do not? Data on overall time served supported
other findings concluding that longer prison terms correlated with being less likely to recidivate.
This finding is often associated with the concept of inmates “aging out” in relation to older
inmates being less likely to recidivate.
In addition, a closer look at the nature of drug offenses that have mandatory minimum statutes
may reveal what appears to be a slight difference between drug crime types. Most drug offenders
with mandatory terms were serving drug trafficking offenses, compared with 86 percent of the
nonmandatory drug offense types, which were more likely first-time offenses for “distribution or
possession with intent to distribute.” This raises the question of whether a substance abuser who
sells drugs to maintain his or her habit should be treated differently than someone who sells
drugs for business.
Many inmates sentenced with mandatory minimums are given a one-day difference between
their minimum and maximum sentence terms, essentially eliminating the possibility of being
paroled, which may also impact recidivism rates. Whether inmates are paroled or released via
expiration of sentence has many implications on their return to the community and the nature of
their recidivism rates. Over the past 10 years or more, including the 2002 release cohort,
approximately a third of inmates were paroled from the DOC to the community. The recidivism
rate among inmates paroled in the cohort was 45 percent, compared with 36 percent for those
released via expiration of sentence. When technical violators were excluded from the calculation,
recidivism rates among parolees dropped to 36 percent. When returns to prison for a parole
technical violation are excluded, the recidivism rate for the overall release cohort dropped from
39 to 35 percent.
Fluctuations in parole rates and the corresponding number of release cohorts can affect
recidivism rates in the context of parole violators. Recidivism rates of inmates released on parole
are consistently higher, in general, than of inmates released via expiration of sentence. Thus, the
more inmates paroled, the more likely recidivism rates will be higher. In addition, if the percent
of parolees returned to prison remains constant or increases, the DOC admissions rate will
increase as parole rates and corresponding numbers of parole revocations increase. Though this
may appear counterintuitive, the revocation of technical violators by parole is a significant factor
in recidivism rate calculations and is understandable in the context of parolees being under
supervision and closely monitored.
Both technical violation rates and the number of recidivists, in general, were highest during the
first year after release. Three-quarters of all parole revocations occurred during the first year. The
majority (67 percent) of those who recidivated did so by 18 months after release. Since
recidivism rates are based on each release in a three-year timeframe, some parolees are revoked
and rereleased from the same sentence. Recidivism rates are higher among inmates rereleased

32

(44 percent) compared with those from their first release. This finding is not only consistent with
national trends, but raises further questions about the revocation and rerelease process.
Among the 700 recidivists in the cohort, 25 percent were returned on any type of parole technical
violation. Just over half (53 percent) were returned to prison for a county sentence, and 19
percent for a state sentence. Almost a quarter (23 percent) of those inmates returning to prison on
a new (state or county) sentence had been released on parole but were not necessarily under
parole supervision when they committed a new crime or were convicted and sentenced for the
new offense.33 Forty percent of recidivists in the release cohort had been paroled, compared with
76 percent of those interviewed for the recidivism survey. Considering that almost all inmates
admitted to the DOC while the recidivism survey was conducted who qualified as recidivists
(previously released from the DOC within three years) participated, that sample reflects a
window of male recidivists for July 2006 through June 2007. Thus, further inquiry as to whether
the survey sample reflects an increase among inmates paroled, recidivating, or other related
factors is needed to better understand the relationship between parole and recidivism.
Another factor associated with recidivism rates is the security level of the prison facility from
which the inmate was released. The majority (58 percent) of inmates released from the DOC
were in medium custody. Not surprisingly, the recidivism rate among those released from
medium security facilities (40 percent) was consistent with the overall recidivism rate of 39
percent. It was no less surprising and consistent with other findings that the 193 inmates released
from a maximum security level recidivated at this higher rate (50 percent) compared with lower
security levels and treatment facilities.
The challenge in classifying inmates to minimum and prerelease facilities prior to release is, to a
degree, limited by statutory restrictions (i.e., mandatory minimum sentencing) as well as the
balance between risk and public safety. This poses both a practical and philosophical challenge
because a more gradual process of transitioning an inmate provides better reintegration into the
community. If an inmate is going to be released into a community without being paroled (as
current trends indicate), policymakers should devise a method of reducing, if not eliminating, the
number of inmates who live in a medium or maximum custody prison one day and in the
community the next. This could be achieved in a variety of ways, such as using transition houses
or a graduated step-down process with more intensive community supervision.
Until these challenges can be better addressed, correctional facilities must continue to provide
adequate reentry programming and transitional assistance for inmates. Though the DOC provides
a multitude of evidence-based inmate programming designed to reduce recidivism, this study
was only able to collect data on the transition workshop designed primarily to enhance an
inmate’s awareness and realization of what he needs to do to prepare for his transition back into
the community. The recidivism rate among the 899 workshop participants within the release
cohort was 43 percent, compared with 35 percent among nonparticipants. These results are based
on participation occurring prior to or in 2002, and since that time, the workshop has been

33

Parole discharge dates were not available.

33

expanded in length and content. Other factors that may have contributed to the discrepancy in
recidivism rates may include the release of a disproportionate number of participants from higher
(maximum and medium) security facilities and the unknown time between when an inmate
participated and was released. There was no way to measure whether an inmate was an active
participant in the workshop or followed through after the workshop and stayed actively involved
in transition planning or refused assistance, as many inmates do. These results also beg the
question if an inmate who chooses to participate in such a workshop is a different type of
offender.
While much is accomplished by devising an initial reentry plan as an outcome of the workshop,
the key “reality checks” are addressed during the DOC triage reentry process that follows. This
individualized case management process focuses on inmates’ inappropriate or unrealistic plans,
and efforts are made to work with them to develop reentry plans that realistically address needs
in housing, employment, medical/mental health, treatment, education, social well-being, and
more. Again, the success of these plans can largely depend on the active role an inmate takes in
the planning process and varies from case to case. Among inmates who are paroled, packets of
information, including reentry plans and referrals, are shared with the corresponding parole
regional reentry centers, where those inmates report upon being released from the DOC.
What an inmate encounters during the transition back into the community and how well he is
equipped and supported in handling that process is key to a successful transition. The
comparison between postrelease surveys, conducted by telephone with inmates one and/or six
months after release, to results from the in-person recidivism surveys demonstrate the many
challenges released inmates experience. Though most inmates reported “stable” or “permanent”
housing, more individuals who were still in the community when contacted had stable housing
compared with recidivists reflecting back on their time in between incarcerations. This
seemingly stable housing trend corresponded with the overwhelming perception that inmates’
families were very supportive. Offenders reported challenges regarding employment and more
recidivists reported problems associated with substance abuse. Participation in substance abuse
treatment varied but seemed higher immediately after being released for both groups.
The analysis of administrative data from the DOC of the 2002 release cohort was consistent with
and expanded on static factors associated with the risk to recidivate. Armed with these and new
analyses, the DOC is better positioned to focus transitional resources by targeting the subgroups
who recidivate the most—“short-termers,” those being released from higher security levels,
younger inmates with property offenses, those with long criminal histories who started their
criminal careers early, and those with a history of certain offense types, largely associated with
drug use. The new risk and needs assessment tool, COMPAS34 soon to be implemented by the
DOC and parole, captures most of these static variables and will be instrumental in assisting both
agencies in focusing resources. Additional partnerships between not only the DOC and parole,
but among sheriffs, Department of Youth Services, probation, the courts, community service
providers, and other related criminal justice and health and human service agencies are

34

COMPAS stands for Correctional Offender Management Profiling for Alternative Sanctions.

34

imperative to improve the sharing of information that may not be readily captured and available,
such as program participation, medical and mental health histories, family involvement and
support, and behavior patterns while incarcerated or under community supervision. Such
information would greatly enhance efforts for each phase and responsible agency in planning and
providing services for individuals as they move through the criminal justice system. What is
needed is not only the sharing of information and relevant background information, but a
continuation of care and case management.
Ultimately, the transition from prison to the community relies on more than preparing inmates
for a successful transition back to the community. There must be partnerships with appropriate
community service providers, and support systems must be available in collaboration with all
supervisory agencies (parole and probation) when involved. Results from this analysis illustrate
that that transition, especially the first year after release, is fraught with challenges. By knowing
more about what these challenges are, we can better focus our resources and partnerships to
facilitate a more successful process that will result in less recidivism and increased quality of life
for everyone.

35