Skip navigation

Massachusetts Sentencing Commission Recidivism Study 2002

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
Comprehensive Recidivism Study

A Report to the
House Committee on Ways and Means
Senate Committee on Ways and Means
Joint Committee on Criminal Justice
and the
Joint Committee on Public Safety

Massachusetts Sentencing Commission
Honorable Robert A. Mulligan, Chairman
June 1, 2002

Massachusetts Sentencing Commission
90 Devonshire Street, Room 2001
Boston, MA 02109
Voice: 617-788-6867
Fax: 617-788-6885

Comprehensive Recidivism Study

A Report to the
House Committee on Ways and Means
Senate Committee on Ways and Means
Joint Committee on Criminal Justice
and the
Joint Committee on Public Safety

Massachusetts Sentencing Commission
Honorable Robert A. Mulligan, Chairman
June 1, 2002

MEMBERSHIP OF THE
MASSACHUSETTS SENTENCING COMMISSION
Judges

Parole Board

Honorable Robert A. Mulligan, Chairman
Associate Justice, Superior Court

Michael J. Pomarole, Esq.
Chairman, Parole Board

Honorable Margaret R. Hinkle
Associate Justice, Superior Court

Public Safety
James P. Jajuga, Secretary
Executive Office of Public Safety

Honorable Mark H. Summerville
Associate Justice,
Boston Municipal Court

Victim Witness Board
Maria F. Rodriguez, Esq.
Chief of the Family Protection Unit
Hampden County District Attorney's Office

Prosecutors
Honorable S. Jane Haggerty
Associate Justice, Superior Court

Department of Correction
Pamela L. Hunt, Esq.
Office of the Attorney General

David Slade, Esq.
Office of the General Counsel
Department of Correction

Geline W. Williams, Esq.
Executive Director
MA District Attorneys’ Association

Probation
John J. O’Brien, Commissioner
Office of the Commissioner of Probation

Defense
Michael J. Traft, Esq.
Private Attorney, Boston

District Court Liaison
Honorable Timothy Gailey
Presiding Justice
Chelsea District Court

William W. Robinson, Esq.
Committee for Public Counsel Services
Thomas G. Murray, Esq.
Private Attorney, Boston (deceased)

Staff
Francis J. Carney, Jr., Executive Director
Linda K. Holt, Research Director
Lee M. Kavanagh, Research Analyst
Valerie C. Caldwell, Executive Assistant

Sheriffs Association
Mariellen H. Fidrych
Executive Director
MA Sheriffs’ Association

i

ii

Acknowledgments

This research would not have been possible without the cooperation of administrators and
the assistance of researchers and practitioners in several criminal justice agencies and
entities. The Commission would like to acknowledge that cooperation and assistance as
follows:
Franklin County House of Correction

Department of Correction

Sheriff Frederick B. Macdonald
Lt. John Zewski

Michael Maloney, Commissioner
Rhiana Kohl, Ph.D., Research Director
Hollie A. Matthews Hoover, Research Analyst

Hampden County House of Correction
Sheriff Michael J. Ashe, Jr.
J. John Ashe, Deputy Superintendent
Martha A. Lyman, Ed.D., Research Director

Massachusetts Parole Board
Michael J. Pomarole, Chairman
Timothy F. App, Executive Director
Sallyann V. Sweeney, Director of Support Services
Joseph C. Balbo, Systems Analyst

Norfolk County House of Correction
Sheriff Michael Bellotti
George Klier, Assistant Deputy Superintendent

Office of the Commissioner of Probation
Plymouth County House of Correction
John J. O’Brien, Commissioner
Sheriff Joseph F. McDonough
Lt. Brian C. Case

Office of Community Corrections
Stephen V. Price, Executive Director
Phyllis Buccio-Notaro, Statewide Program Supervisor
Vincent Lorenti, Program Specialist
John Quinn, Administrative Staff Attorney
Sam Bellistri, Regional Manager
Pamerson Ifill, Regional Manager
Kevin Kearney, Regional Manager
John Monahan, Regional Manager

Suffolk County House of Correction
Sheriff Richard J. Rouse
Patrick Bradley, Deputy Sheriff
Stefan Lobuglio, Deputy Superintendent
Worcester County House of Correction
Sheriff John M. Flynn

iii

Honorable Joseph I. Dever
First Justice, Lynn District Court

Community Corrections Centers
Maureen Richitelli, Manager
Barnstable County, Barnstable

Honorable Michael C. Lauranzano
Associate Justice, Lynn District Court

Rick Kassiotis, Jr., Manager
Essex County, Lawrence

Honorable Herbert H. Hodos
First Justice, Greenfield District Court

Jack Fitzgerald, Manager
Hampden County, Springfield

Honorable M. John Schubert, Jr.
Acting First Justice, Orange District Court

Tim Simons, Manager
Hampshire County, Northampton

Honorable Robert V. Greco
First Justice, Framingham District Court

Shawn Murray, Manager
Middlesex County, Cambridge

Honorable Sydney Hanlon
First Justice, Dorchester District Court

Richard McDonough, Manager
Norfolk County, Quincy

Honorable Milton L. Wright, Jr.
First Justice, Roxbury District Court

Patty Campatelli, Manager
Suffolk County, Boston

Honorable Robert P. Ziemian
First Justice, South Boston District Court

Jim Wright, Manager
Worcester County, Fitchburg
Worcester County, Webster

Criminal History Systems Board
Barry Lacroix, Executive Director

Drug Courts
Administrative Office of the Trial Court

Honorable Robert A. Cornetta
Chairman, Drug Court Working Group
Associate Justice, Salem District Court

William Marchant
Acting Director, Fiscal Affairs

Jerome S. Berg, Court Administrator
District Court Department
Student Interns
Paul Q. O'Donnell, Regional Coordinator
District Court Region II

Amanda E. Wall, College of the Holy Cross
Kara Fielder, Northeastern University

Honorable Kevin M. Herlihy
First Justice, Haverhill District Court

iv

Executive Summary
This report was prepared to meet the legislative requirement in Chapter 177 of the Acts of
2001 which directed the Massachusetts Sentencing Commission to “perform a
comprehensive study on the recidivism rate of offenders.”
The completed project is the culmination of a collaborative effort of a variety of criminal
justice agencies. A total of thirty different criminal justice agencies and programs
contributed data to this project or assisted in the compilation of the data and report.
The study reports on a one year follow-up of a sample of 3,751 offenders representing
drug courts, community corrections centers, houses of correction, Department of
Correction, and the Parole Board. The study defined a recidivist as:
For offenders released from incarceration:
•
•

an offender with a new arrest (arraignment) in the year following release from a
correctional facility; or,
an offender with a technical violation of probation or parole that resulted in
incarceration in the year following release from a correctional facility;

For offenders in community corrections centers or drug court programs:
•
•

an offender with a new arrest (arraignment) in the year following the date of entry
into a community corrections center or drug court program; or,
an offender with a technical violation of probation or parole that resulted in
incarceration in the year following the date of entry into a community corrections
center or drug court program.

Based on this definition, the one year recidivism rate was estimated to be 49.1% for all
samples. The report discusses the method used to measure recidivism and presents data
on recidivism rates for each of the criminal justice cohorts. The report provides further
information on the nature of the recidivism behavior and explores the relationship
between recidivism rates and program involvement, current offense, criminal history, and
demographic characteristics.
Future research is suggested that would build on the model presented here but address
some of the limitations of the present analysis (scope of the sample, length of the followup period, and limited program information).

v

vi

Table of Contents

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
RECIDIVISM RATES - MASSACHUSETTS AND NATIONAL RESEARCH . . . . . . 2
METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
SAMPLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Sampling Time Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Originating Criminal Justice Agencies . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
DATA COLLECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Recidivism Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Program Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Offender Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
DATA EXTRACTION AND DATA LINKAGES . . . . . . . . . . . . . . . . . . . . . . . 15
DATA ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
STUDY LIMITATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
RECIDIVISM RATES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Type of Recidivist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Time Until Recidivism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
New Arrest Offenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
PROGRAMS, SECURITY, AND COMMUNITY RELEASE . . . . . . . . . . . . . . 28
Correctional Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Community Corrections Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Drug Courts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
DEMOGRAPHIC CHARACTERISTICS OF OFFENDERS . . . . . . . . . . . . . . . 33
Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Race / Ethnicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
CURRENT OFFENSE AND CRIMINAL HISTORY . . . . . . . . . . . . . . . . . . . . 38
Current Offense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Type of Offense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Offense Seriousness Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Criminal History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Criminal History Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Juvenile Criminal History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Sentencing Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

vii

SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
APPENDIX - METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
SAMPLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
DATA COLLECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Recidivism Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
New Offenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Incarceration Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Parole Violation Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Probation Violation / Other Status . . . . . . . . . . . . . . . . . . . . . . . . 58
Program Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Offender Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
DATA EXTRACTION AND DATA LINKAGES . . . . . . . . . . . . . . . . . . . . . . . 60
DATA ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
APPENDIX - HOUSE OF CORRECTION SENTENCED OFFENDERS . . . . . . . . . . 66

List of Figures

Figure 1. Sentencing Guidelines Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

viii

List of Tables
Table 1. Originating Criminal Justice Agencies Included in Study Sample . . . . . . . . . . . 6
Table 2. Type of Sentence and Correctional Custody at Time of Release, Correctional
Releases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Table 3. Recidivism Classification by Criminal Justice Cohort . . . . . . . . . . . . . . . . . . . 21
Table 4. New Arrest Status, Parole Supervision Status, and Community Corrections
Center Participation Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Table 5. Time Until Recidivism Incident by Criminal Justice Cohort . . . . . . . . . . . . . . 24
Table 6. Offense Seriousness Level of Most Serious New Arrest Offense by Criminal
Justice Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Table 7. Type of Offense of Most Serious New Arrest Offense by Criminal Justice
Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Table 8. Recidivism Status by Security Level at Time of Release and Criminal Justice
Cohort, Correctional Releases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 9. Recidivism Status by Selected Program Characteristics, Community Corrections
Center Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Table 10. Recidivism Status by Selected Program Characteristics, Drug Court
Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table 11. Recidivism Status by Gender and Criminal Justice Cohort . . . . . . . . . . . . . . 35
Table 12. Recidivism Status by Age and Criminal Justice Cohort . . . . . . . . . . . . . . . . . 36
Table 13. Recidivism Status by Race / Ethnicity and Criminal Justice Cohort . . . . . . . 37
Table 14. Recidivism Status by Type of Offense (Current Offense) and Criminal Justice
Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Table 15. Recidivism Status by Offense Seriousness Level (Current Offense) and
Criminal Justice Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Table 16. Recidivism Status by Criminal History and Criminal Justice Cohort . . . . . . 44
Table 17. Recidivism Status by Juvenile Criminal History and Criminal Justice Cohort
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Table 18. Recidivism Status by Sentencing Grid Cell and Grid Assignment, All
Offenders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Table 19. Non-hierarchical Recidivism Classification by Criminal Justice Cohort . . . . 62
Table 20. Alternate Recidivism Measures by Criminal Justice Cohort . . . . . . . . . . . . . 63
Table 21. Conviction Status by Time Until Recidivism and Criminal Justice Cohort . . 65
Table 22. Selected Recidivism Characteristics by Type of Release and Custody at Time
of Release, House of Correction Sentenced Offenders . . . . . . . . . . . . . . . . . . . . 67
Table 23. Selected Offender Characteristics by Type of Release and Custody at Time of
Release, House of Correction Sentenced Offenders . . . . . . . . . . . . . . . . . . . . . . 69

ix

x

MASSACHUSETTS SENTENCING COMMISSION

COMPREHENSIVE RECIDIVISM STUDY

A REPORT TO THE
HOUSE COMMITTEE ON WAYS AND MEANS
SENATE COMMITTEE ON WAYS AND MEANS
JOINT COMMITTEE ON CRIMINAL JUSTICE
AND THE
JOINT COMMITTEE ON PUBLIC SAFETY

INTRODUCTION
Chapter 177 of the Acts of 2001 (Chapter 177), the FY2002 budget, directed the
Massachusetts Sentencing Commission (Commission) to “perform a comprehensive
study on the recidivism rate of offenders.” The legislature established a June 1, 2002
deadline for this study. This report was prepared to meet this legislative mandate. As a
measure of the subsequent involvement of offenders in the criminal justice system
following release from incarceration or following participation in a criminal justice
program, recidivism is a commonly used indicator of the outcome of various criminal
justice initiatives. This document reviews previous research on recidivism in
Massachusetts, presents the research method used in the present analysis, and discusses
the findings from this research initiative.

1

Massachusetts Sentencing Commission

RECIDIVISM RATES - MASSACHUSETTS AND NATIONAL RESEARCH
Studies of recidivism are both common in the criminal justice literature and diverse in
their definition and structure. For purposes of this report, numerous studies were
reviewed. The studies reported here concentrate on recent findings from Massachusetts
research. One national study is also reviewed.
The Department of Correction (DOC) has a long tradition of reporting recidivism rates
for offenders released from DOC facilities. The most recent report presented three-year
recidivism rates for offenders released in 1995. The DOC defined a recidivist as an
offender re-incarcerated for at least 30 days during the three-year follow-up period.1 This
definition included those re-incarcerated after conviction of a new offense, as well as
those re-incarcerated for a violation of the conditions of their parole or probation. After
one year the recidivism rate was 22%; after two years the recidivism rate was 37%; and
after three years the recidivism rate was 44%.2 One year recidivism rates were available
for prior years and had decreased from 29% for offenders released in 1990 to 22% for
offenders released in 1995.3 An analysis of the long term trends in DOC recidivism rates
indicated that age at incarceration was the variable most strongly associated with
recidivism risk for the DOC release population.4 As the proportion of offenders in the
high risk age group (those offenders who began their incarceration at age 25 or younger)
decreased, the overall recidivism rate decreased as well.
The Hampden County Sheriff’s Department initiated a research program to study the
recidivism of offenders released from the correctional facilities in Hampden County with
a goal to provide “objective, empirical data necessary for . . . planning, operation and
evaluation.”5 A study of the recidivism rates for offenders released in 1998 and 1999
defined a recidivist as an offender re-incarcerated during the follow-up period in a state or
1

Hollie A. M atthews Ho over, The Background Characteristics and Recidivism Rates of Releases from
Massachusetts Correctional Institutions During 1995, Massachusetts Department of Correction, May 2000, page 1.
2

Ibid., page 18.

3

Ibid., page 5.

4

Daniel P . LeClair, Recent C hange s in Rates o f Recidivism for Releas es from M assachu setts State
Correctional Institutions: A n Exploratory Evaluation o f Possible Causa l Factors, Stonehill College, Fall 1998,
page 16.
5

Martha A . Lyman, A Comprehensive Study of Recidivism at the Hampden County House of Correction,
Hampden County Sheriff’s Department, February 23, 2001, page 1.

2

Comprehensive Recidivism Study

county correctional facility in Massachusetts for either a new offense or violation of
release conditions.6 For those offenders released in 1998, the recidivism rate after one
year was 12.0% and the recidivism rate after two years was 28%; and, for those offenders
released in 1999, the recidivism rate after one year was 21.9%.7 The Hampden County
study also reported re-arraignment rates and re-conviction rates. The proportion of
offenders with a new arraignment within one year of release was 44.7% and 52.7% for the
1998 and 1999 cohorts, respectively. 8 The proportion of offenders with a new conviction
within one year of release was 23.1% and 31.7% for the 1998 and 1999 cohorts,
respectively. 9 The study attributed the change in recidivism rates between the 1998 and
1999 cohorts to a change in the definition of recidivism for parole and probation violators
and to changes in the policies and practices of other criminal justice agencies.
Massachusetts is similar to other jurisdictions in the recidivism rates that have been
observed by the DOC and Hampden County. A national study followed offenders
released from 11 large states for a period of three years. The study found that 62.5% had
been rearrested, 46.8% had been re-convicted, and 41.4% re-incarcerated by the end of
the three years.10 This study also found that the rate of re-arrest was highest during the
first year following release.
These studies indicate some of the important themes found in the recidivism literature:
recidivism rates vary by the type of behavior that is considered (arrest, conviction or
incarceration); recidivism rates vary by the length of the follow-up period; recidivism
rates vary by the composition of the offender population under study; and, recidivism
rates vary over time in response to changes in policy and practices in other criminal
justice agencies. While the recidivism rates derived in the current analysis will not be
directly comparable to these studies due to differences in the definition of recidivism,
differences in the population that was considered, and differences in the sampling time
period, they serve as a useful and important backdrop for the current study.

6

Ibid., page 3.

7

Ibid., pages 10 and 38.

8

Ibid., page 10.

9

Ibid.

10

Allen J. Be ck and B ernard E . Shipley, Recidivism of Prisoners Released in 1983, Bureau of Justice
Statistics, February 1997, page 1.

3

Massachusetts Sentencing Commission

METHOD
This research project is the result of a collaborative effort of a number of criminal justice
agencies. The approach selected for this research was based on the assumption that
reliable recidivism information could be collected in a semi-automated fashion by linking
a number of data-bases from cooperating criminal justice agencies in the commonwealth:
•

a sample of offenders released following a period of incarceration and a sample of
offenders participating in community based criminal justice programs was
provided by the relevant originating criminal justice agency - DOC, Parole
Board, sheriffs departments, drug courts, and Office of Community Corrections;

•

a measure of new arrests was obtained by linking the relevant offender record
from the originating criminal justice agency to the CARI (Court Activity Record
Information) data-base. New arrests were defined as arrests and arraignments
that occurred after the release from incarceration or entry date into the community
based program;

•

a measure of parole violations was obtained by linking the relevant offender
record from the originating criminal justice agency to the PATS (Parole’s
Automated Tracking System) data-base; and,

•

a measure of probation violations was obtained by linking the relevant offender
record from the originating criminal justice agency to the CARI data-base.

This section contains a brief discussion of the sample, the data collection process, the data
extraction process, and the data analysis. Additional details on the method are contained
in the Appendix.
SAMPLE
Chapter 177 directed the Commission to include “offenders who have been, or currently
are, incarcerated and/or sentenced to the facilities and/or programs of:
•
•
•

the Department of Correction,
the respective county houses of correction,
the Parole Board,

4

Comprehensive Recidivism Study

•

•

those community correction centers established pursuant to
chapter 211F that have been fully operational for a period of at
least one calendar year, and
any drug courts, so-called, operating within the district courts of
the commonwealth.”

Sampling Time Frame.
In order to allow for a reasonable follow-up period and to ensure that the project was
completed within the time-frame established by Chapter 177, the sample was defined as
offenders released / program participants during the period April 2000 through June 2000.
Some of the considerations in the development of the sample were:
•

The period following release / program participation was one year. A longer
follow-up period was not feasible due to the reporting deadline, the newness of
some of the programs, and the scope of the sample. While recidivism status was
computed at periods of less than one year (e.g. 6 months), the data collection and
analysis focused on an initial follow-up period of one year.

•

Data collection for recidivism was done no earlier than 15 months following
release / program participation; and,

•

A release / program participant cohort of one calendar quarter allowed for a
reasonable sample size of 3,751 offenders for the initial study effort.

Originating Criminal Justice Agencies.
The cooperation of the originating criminal justice agencies was essential to the
identification of the release / program participation samples. There were no known data
sources except the originating criminal justice agencies that could have provided this
information. A total of 3,751 offender records was included in the study sample. Table 1
shows the specific sites participating in the study.

5

Massachusetts Sentencing Commission

Table 1. Originating Criminal Justice Agencies Included in Study Sample
Release Populations
Department of Correction
All DOC Facilities
Houses of Correction
Franklin County House of Correction
Hampden County House of Correction
Norfolk County House of Correction
Plymouth County House of Correction
Suffolk County House of Correction
Worcester County House of Correction
Massachusetts Parole Board
All Parolees
Community B ased Pro grams
Community Corrections Centers
Barnstable County Community Corrections Cent er in Barnstable
Essex County Community Corrections Center in Lawrence
Hampden County Community Corrections Cent er in Springfield
Hampshire County Community Corrections Center in Northampton
Middlesex County Community Corrections Center in Ca mbridge
Norfolk County Community Corrections Center in Quincy
Suffolk County Community Corrections Center in Boston
Worcester County Community Corrections Center in Fitchburg
Worcester County Co mmunity Correct ions Center in Webster
Drug Courts
Dorchester District Court
Framingham District Court
Greenfield District Court
Haverhill District Court
Lynn District Court
Orange District Court
Roxbury District Court
South Boston District Court

Department of Correction. The DOC provided a sample of 705 offenders released during
the period April 2000 through June 2000. This represented 18.8% of the study sample
and included all offenders released via parole and discharge from a DOC facility. Table 2
shows additional details on the type of sentence and gender for the DOC offenders in the
study sample.
Houses of Correction. Six houses of correction provided samples of offenders released
during the period April 2000 through June 2000. The total sample from these six houses
of correction was 2,396 representing 63.9% of the study sample. An estimate of the
proportion of the house of correction population that was included in this study was
derived from a five-year sample of admissions to all houses of correction. The six
6

Comprehensive Recidivism Study

participating houses of correction represented 56.3% of all admissions during the five
year period and therefore it was estimated that the current study included 56.3% of all
releases from houses of correction during the sample period.11 Because the
Massachusetts Parole Board provided data on all parolees state-wide, it was assumed that
the study sample included all parolees from houses of correction and 56.3% of the
discharges. Table 2 shows additional details on the type of sentence and gender for the
house of correction offenders in the study sample.
Parole Board. The Parole Board provided a state-wide sample of 776 offenders released
to parole supervision during the period April 2000 through June 2000. This represented
20.7% of the study sample. This sample included all offenders released to parole
supervision from houses of correction and the DOC. This sample included offenders
released to parole supervision from the houses of correction participating in the study as
well as offenders released to parole supervision from all other houses of correction in
Massachusetts. Of the 776 offenders released to parole supervision, 222 were paroled
from DOC facilities, 303 were paroled from participating houses of correction, and 251
were paroled from the non-participating houses of correction. Table 2 shows additional
details on the type of sentence, gender, and facility at the time of release for the offenders
released to parole supervision in the study sample.
Community Corrections Centers. Eligible community corrections centers provided
samples of offenders who began a community corrections center placement during the
period April 2000 through June 2000. The Office of Community Corrections (OCC) was
established under G.L. c. 211F. The mission of OCC is “the establishment of
intermediate sanctions programs which offer a continuum of sanctions and services for
probation, sheriffs, parole and the Department of Correction.”12 The intermediate
sanctions are based at the community corrections centers in operation across the state.
They are described by OCC as follows:
Community Corrections Centers are community based, intensive
supervision sites, which deliver bundled sanctions and services, including
treatment and education, to high risk offenders via Intermediate Sanction

11

Robert J . Tenaglia, New Court Commitments to Massachusetts County Correctional Facilities During
2000, Massachusetts Department of Correction, September 2001, page 22.
12

Administrative Office of the T rial Court, O ffice of Com munity Corr ections, Overview of the
Department, http://www.state.m a.us/courts/ad min/occ/o ccovervie w.html.

7

Massachusetts Sentencing Commission

Levels. Among the sanctions delivered at community corrections centers
are:
•
•
•
•

electronic monitoring
community service
drug & alcohol testing
day reporting

Among the services provided at community corrections centers are:
•
•
•
•
•
•
•

substance abuse treatment
GED/ABE/ESL or comparable educational component
communicable disease prevention education
job readiness training and placement
referral to Department of Public Health or Department of Mental Health
service providers
women's services
bilingual services

Community corrections centers are designed to provide a criminal justice
solution for a specific group of offenders. Intermediate Sanction Level III,
IV is indicated for those offenders who possess a serious criminal history
and are chronic substance abusers . . .
Intermediate Sanction Level IV is the most intense level of community
based, criminal justice supervision. Sanctions and services required at this
level of supervision represent a twenty-four hour restriction upon the
liberty of the offender. Level IV participants are required to report to the
community corrections center for four to six hours per day, six days per
week. Additionally, offenders placed at Intermediate Sanction Level IV are
monitored twenty-four hours per day via electronic device, required to
submit to the highest category of random drug and alcohol testing, and
mandated to attend two four hour community service shifts per week.
Intermediate Sanction Level III is an intense level of community-based,
criminal justice supervision. Sanctions and services required at this level
of supervision represent a daily imposition upon the liberty of the
offender. Level III participants are required to report to the community
corrections center for one to four hours per day, three to five days per
8

Comprehensive Recidivism Study

week. Offenders placed at Intermediate Sanction Level III may be
monitored via electronic device. Level III also requires random drug and
alcohol testing, and attendance at one four hour community service shift
per week.13
Only those community corrections centers in operation by January 2000 were considered
eligible for participation in the study and all such centers contributed data. A total of 323
offenders began participation in an eligible community corrections center during the study
period. This represented 8.6% of the study sample.
Drug Courts. Eligible drug courts provided samples of offenders who began a drug
court placement during the period April 2000 through June 2000. The annual report of
the Trial Court describes a drug court as:
A drug court is not a separate court, but rather a special session of a
District Court. The drug court session usually addresses serious drug
offenders for whom a special court intervention is likely to be the last stop
before incarceration. Drug court typically involves a strenuous regimen of
treatment and accountability on the part of the offender, coupled with the
strong personal commitment by the offender to take control of his or her
life situation and eliminate drug use. Among the key elements to an
effective drug court program are intensive probationary and therapeutic
programming activities, frequent drug testing, and careful monitoring of
progress by the drug court judge.14
Only those drug courts in operation by January 2000 were considered eligible for
participation in the study and all such courts contributed data. A total of 76 offenders
began participation in these eight drug courts during the study period. This represented
2.0% of the study sample. Of the 76 drug court program participants, 20 or 26.3% began
the drug court program under the probation portion of a house of correction / split
sentence.

13

Ibid.

14

Massac husetts Supre me Judicia l Court, Annual Report on the State of the Massachusetts Court System Fiscal Yea r 2000, page 98.

9

Massachusetts Sentencing Commission

Criminal Justice Cohort
In each of the tables, the study sample was divided into the following criminal justice
cohorts:

Community Based Program Participants:
•
drug court program participants;
•
community corrections centers program participants;
Offenders released after incarceration:
•
house of correction sentenced / released via discharge;
•
house of correction sentenced / released via parole;
•
state prison sentenced / released via discharge; and,
•
state prison sentenced / released via parole.
The study sample of offenders released after incarceration in a house of correction or the
DOC included offenders who had served either a house of correction sentence, which
ranges from one day to 2 ½ years, or a state prison sentence, which ranges from one year
to life. For purposes of this analysis, recidivism rates were considered by the type of
sentence for which the offender was incarcerated (house of correction or state prison) and
by the type of release (parole or discharge).
As indicated in Table 2, of the 705 offenders released from DOC facilities an estimated
that 492 or 69.8% had served a state prison sentence and 213 or 30.2% had served a
house of correction sentence prior to release. Many female offenders serving a house of
correction sentence served that sentence at a DOC facility and were released from the
DOC. Some male offenders serving a house of correction sentence were transferred to
the DOC and released from a DOC facility. Of the 213 offenders released from the DOC
who had served a house of correction sentence, 165 or 77.5% were female and 48 or
22.5% were male. For purposes of this analysis, those offenders released from a DOC
facility after serving a house of correction sentence were included in the house of
correction sentence cohort. The statistical appendix contains tables that disaggregate
those offenders released after serving a house of correction sentence by the correctional
custody at the time of release (DOC or house of correction).
As indicated in Table 2, of the 2,396 offenders released from houses of correction an
estimated 2,393 or 99.9% had served a house of correction sentence and 3 or 0.1% had
served a state prison sentence prior to release. For purposes of this analysis, the three
10

Comprehensive Recidivism Study

offenders released from a house of correction after serving a state prison sentence were
included in the state prison sentence cohort. The statistical appendix contains tables that
disaggregate those offenders released after serving a house of correction sentence by the
correctional custody at the time of release (DOC or house of correction).
Finally, as indicated in Table 2, there were 251 offenders released by parole from the
houses of correction that did not participate in the study. Of these 248 were released by
parole after serving a house of correction sentence and 3 were released by parole after
serving a state prison sentence. The proportion of offenders released by parole varied by
type of sentence and correctional custody at the time of release:
•

32.3% of the offenders released from a DOC facility after serving a state prison
sentence were paroled and 67.7% were discharged;
29.6% of the offenders released from a DOC facility after serving a house of
correction sentence were paroled and 70.4% were discharged; and,
12.5% of the offenders released from houses of correction after serving a house of
correction sentence were paroled and 87.5% were discharged.

•
•

It should be noted that offenders with house of correction sentences of less than two
months were not eligible for parole. In FY 2000 an estimated 29.3% of all house of
correction sentences were for less than two months.15

15

Massac husetts Senten cing Com mission, Survey of Sentencing Practices, November 2001, page 35.

11

Massachusetts Sentencing Commission

Table 2. Type of Sentence and Correctional Custody at Time of Release, Correctional Releases
Custody a t Time of Relea se
and Gender
Released from DO C Facility
Female
Male
Sub-Total

House of Correction Sentence

State Prison Sentence

Discharge

Parole

Total

Discharge

Parole

Total

Total

127
23
150

38
25
63

165
48
213

14
319
333

9
150
159

23
469
492

188
517
705

132
1961
2,093

43
257
300

175
2,218
2,393

0
0
0

0
3
3

0
3
3

175
2,221
2,396

0
0
0

22
226
248

22
226
248

0
0
0

1
2
3

1
2
3

23
228
251

2,243

611

2,854

333

165

498

3,352

Released from Participating
House of Correction
Female
Male
Sub-Total
Released by Parole from
Other House of Correction
Female
Male
Sub-Total
Total

12

Comprehensive Recidivism Study

DATA COLLECTION
The data that was collected can be classified in three general areas:
•
•

•

recidivism - those measures of the criminal justice involvement by the offender in
the period following release / program participation;
program involvement - those measures (e.g. security level, intermediate sanction
level) that described the criminal justice program in which the offender was
involved prior to release or during program participation; and,
offender background - those demographic, offense, and criminal history
variables that describe the offender and that the research literature suggested were
related to the risk of re-offending.

Recidivism Variables
There are many different measures of recidivism. The length of the follow-up period may
vary. In the period following release / program participation, recidivism can include
measures of new criminal activity as well as technical violations of supervised release. In
terms of new criminal activity, recidivism can be defined from the presence of a new
arrest or new conviction, to a new sentence to incarceration. In general, recidivism rates
will vary according to the length of the follow-up period and the definition of recidivism
used. The study defined a recidivist as:
For offenders released from incarceration:
•
•

an offender with a new arrest (arraignment) in the year following release from a
correctional facility; or,
an offender with a technical violation of probation or parole that resulted in
incarceration in the year following release from a correctional facility;

For offenders in community corrections centers or drug court programs:
•
•

an offender with a new arrest (arraignment) in the year following the date of entry
into a community corrections center or drug court program; or,
an offender with a technical violation of probation or parole that resulted in
incarceration in the year following the date of entry into a community corrections
center or drug court program.

13

Massachusetts Sentencing Commission

For all offenders in the sample new arrest information was collected from the CARI database, probation violation information was collected from the CARI data-base, and parole
violation information was collected from the Parole Board file. Those offenders under
parole or probation supervision may be returned to custody as the result of a new arrest or
some other technical violation of the conditions of parole or probation. Technical
violations can range from curfew violations, failure to attend and complete required
programs, or failing a drug test. Recidivism rates for offenders under parole or probation
supervision include returns to custody for technical violations in addition to new arrests
that occurred during the follow-up period. While considered as part of the overall
recidivism rates, the ability to return an offender to custody on a technical violation
before an arrest for a new crime is an important tool to promote public safety.
Program Involvement
Chapter 177 directed that the study consider the relationship between criminal justice
programs and recidivism rates:
. . . said study shall include, but not be limited to, information regarding
recidivism rates, by program and facility, including an analysis detailing
the effect of pre-release, post-release, diversionary and intermediate
sanction rehabilitative/supervisory programs on said recidivism rates . . .
Many of the originating criminal justice agencies provided programs and facilities at a
variety of security levels. Where available, this information was included in the analysis.
Offender Background
It was important to consider offender background variables in relation to observed
recidivism rates. The study included demographic, current offense, and criminal history
information. The study did not develop a comparison sample or control group. It was not
assumed that these were comparable offender groups with respect to recidivism risk.
Throughout the analysis, comparisons were made among the populations from the various
originating criminal justice agencies that indicate the substantial differences in these
populations that were related to the observed differences in recidivism rates.
The demographic characteristics of the population that were collected included: gender,
race / ethnicity, and age. The source of offender demographic information was the CARI
data-base.

14

Comprehensive Recidivism Study

The current offense and criminal history associated with the offender’s placement in
the originating criminal justice agency were collected from the adult and juvenile case
records in the CARI data-base. The case records were identified through a link via the
docket number where available.

DATA EXTRACTION AND DATA LINKAGES.
Originating criminal justice agencies were asked to provide a minimum set of data
elements in order to minimize the disruption of the research to their on-going operations.
The data elements included the offender’s name and date of release or date of program
entry. The criminal justice agency also provided either the offender’s date of birth or the
offender’s PCF number. If the participating agency provided the offender’s date of birth,
the name and date of birth were used to extract the PCF number from the CARI database. The PCF number was then used to establish the link between the various data sets.
The minimum set of required data elements was sufficient to establish a link with the
CARI data-base and to establish a starting point for the follow-up period. In all cases, the
resulting case extract was validated in order to ensure that a unique name, date of birth,
and PCF number match was made.
Of all cases submitted by participating agencies, only five cases or 0.1% were excluded
from the sample because they were not identified in the CARI data-base. This method
resulted in a very complete sample for the recidivism study.

DATA ANALYSIS
Some discussion of how recidivism rates were calculated, how the sample total was
estimated, and how the sentencing guidelines grid was used as a framework for some
sections of this analysis, may be helpful to the interpretation of the statistical information.
Recidivism Rates. In most of the tables in the report, three columns are presented: the
total number of cases, the total number of recidivists, and the recidivism rate. The
recidivism rate is expressed as the percentage of offenders classified as recidivists of the
total population of offenders. These columns are repeated for each of the criminal justice
cohorts in the study.
Sample Totals. In each of the tables a sample total has been calculated. No weights
were used in calculating the total. This total is only representative of the sample used for
15

Massachusetts Sentencing Commission

this particular study. Because not every criminal justice program was included in the
analysis, the total is not representative of all offenders in the criminal justice system. In
particular, not all offenders discharged from houses of correction during the sample
period were included in the study sample. Further, only those drug courts and community
corrections centers established prior to January 2000 were considered eligible for the
study. A sample total was also calculated for the house of correction and state prison
sentenced cohorts. No adjustment was made to the house of correction portion of the
sample in estimating totals, even though this sample includes all offenders paroled from
the houses of correction and an estimated 56.3% of the offenders discharged from the
houses of correction.
Sentencing Guidelines Grid. The proposed sentencing guidelines are in the form of a
sentencing guidelines grid which has been used as a framework for some parts of the
analysis of the recidivism data. The vertical axis of the grid is comprised of nine levels of
offense seriousness, ranging from level 1, the lowest level of offense seriousness, to level
9, the highest level of offense seriousness. The horizontal axis of the grid is comprised of
five categories of defendant criminal history, ranging from criminal history category A,
“No/Minor Record” to criminal history category E, “Serious Violent Record”. The
resulting grid, comprising a total of forty-five cells, is also divided into three sentencing
zones: the “incarceration zone”; the “discretionary zone”; and, the “intermediate
sanction” zone. The sentencing guidelines grid also contains a brief description of Level
III and Level IV intermediate sanctions which are provided by the community corrections
centers. The sentencing guidelines grid is shown in Figure 1. A detailed description of
proposed sentencing guidelines can be found in the Report to the General Court.16 The
definition of the criminal history groups is contained in the Appendix.

STUDY LIMITATIONS
There are certain limitations in the design of this research that should be acknowledged
before presenting the findings. There was no possibility of utilizing an experimental
design whereby offenders could be randomly assigned to treatment and control groups

16

Massac husetts Senten cing Com mission, Report to the G eneral Court , April 1996.

16

Comprehensive Recidivism Study

Figure 1. Sentencing Guidelines Grid
Level

Illustrative Offense

9

Murder

8

Sentence Range
Life

Life

Life

Life

Life

Rape of Child with Force
Aggravated Rape
Armed Burglary

96 - 144 Mos.

108 - 162 Mos.

120 - 180 Mos.

144 - 216 Mos.

204 - 306 Mos.

7

Armed Robbery (Gun)
Rape
Mayhem

60 - 90 Mos.

68 - 102 Mos.

84 - 126 Mos.

108 - 162 Mos.

160 - 240 Mos.

6

Manslaughter (Invol)
Armed Robbery (No gun)
A&B DW (Sign. injury)

40 - 60 Mos.

45 - 67 Mos.

50 - 75 Mos.

60 - 90 Mos.

80 - 120 Mos.

5

Unarmed Rob bery
Stalking ( Viol. of Orde r)
Unarmed Burg lary
Larceny ($50,000+)

12 - 36 Mos.
IS-IV
IS-III
IS-II

24 - 36 Mos.
IS-IV
IS-III
IS-II

36 - 54 Mos.

48 - 72 Mos.

60 - 90 Mos.

4

Larceny From a Person
A&B DW (Mod. injury)
B&E (Dwelling)
Larceny ($10,000-$50,000)

0 - 24 Mos.
IS-IV
IS-III
IS-II

3 - 30 Mos.
IS-IV
IS-III
IS-II

6 - 30 Mos.
IS-IV
IS-III
IS-II

20 - 30 Mos.

24 - 36 Mos.

3

A&B DW (No/minor injury)
B&E (Not dwelling)
Larceny ($250 to $10,000)

0 - 12 Mos.
IS-IV
IS-III
IS-II
IS-I

0 - 15 Mos.
IS-IV
IS-III
IS-II
IS-I

0 - 18 Mos.
IS-IV
IS-III
IS-II
IS-I

0 - 24 Mos.
IS-IV
IS-III
IS-II

6 - 24 Mos.
IS-IV
IS-III
IS-II

2

Assault
Larceny Under $250

0 - 6 Mos.

0 - 6 Mos.

IS-III
IS-II
IS-I

IS-III
IS-II
IS-I

0 - 9 Mos.
IS-IV
IS-III
IS-II
IS-I

0 - 12 Mos.
IS-IV
IS-III
IS-II
IS-I

IS-III
IS-II
IS-I
C
Serious
Record

0 - 3 Mos.
IS-IV
IS-III
IS-II
IS-I
D
Violent or
Repetitive

0 - 6 Mos.
IS-IV
IS-III
IS-II
IS-I
E
Serious
Violent

IS-III
IS-II
IS-I
1

Operate After Suspension
Disorderly Conduct
Vandalism

Criminal History Scale

IS-II
IS-I
A
No/Minor
Record

IS-III
IS-II
IS-I
B
Moderate
Record

Sentencing Zone

Intermediate Sanction Level
Incarceration Zone
Discretionary Zone (Incarceration/Intermediate Sanctions)

IS-IV
IS-III
IS-II
IS-I

24-Hour Restriction
Daily Accountability
Standard Supervision
Financial Accountability

Intermediate Sanction Zone

The numbers in each cell represent the range from which the judge selects the m aximum sentenc e (Not More Than);
The minimum sentence (Not Less Than) is 2/3rds of the maximum sentence and constitutes the initial parole eligibility date.

17

Massachusetts Sentencing Commission

(e.g., for drug court or community corrections participants) or of matching offenders
participating in certain programs with similar non-participating offenders for comparison
purposes in the follow-up period. This is significant because the research did indicate
some important differences in the characteristics of offenders in the various criminal
justice cohorts - including differences on factors that are highly associated with
recidivism risk. Therefore, it would not be valid to simply compare overall recidivism
rates across the criminal justice cohorts as a measure of the relative effectiveness of these
agencies or initiatives. This research was not designed to be an evaluation of the
programs or agencies involved. Rather, the goal was to document outcomes on one
dimension - recidivism - of the many dimensions that could be studied regarding these
programs or agencies.
Time and resource constraints also led to other limitations. It was not possible to compile
detailed program participation data on offenders in the various cohorts so that the
relationship between the degree of program involvement and recidivism could be more
fully explored. For some cohorts - e.g., drug courts - the sample size was fairly small.
The one year follow-up period was relatively brief and the use of arrest as the primary
indicator of recidivism poses some problems (some are found not guilty or have their
charges dismissed).
While this research has its limitations, the fact that some thirty criminal justice agencies
and programs could work together to meet the mandate of the Legislature and produce in
a timely fashion a comprehensive recidivism study which adopts a uniform definition of
recidivism and systematically applies it to offenders in different jurisdictions is a
noteworthy accomplishment.
FINDINGS
The analysis of recidivism covers four major areas. First, the overall findings on
recidivism rates are presented. Second, recidivism rates are considered in the context of
program related variables. This is done separately for the major program groups in the
study: correctional facilities (DOC and the houses of correction, including paroles and
discharges), community corrections centers, and drug courts. Third, recidivism rates are
considered in relation to the demographic characteristics of the population. Some
comparisons of the demographic composition of the various criminal justice cohorts are
provided, especially as they are related to recidivism rates. Finally, recidivism rates are
considered in relation to the offender’s current offense, criminal history and placement on
the proposed sentencing guidelines grid.

18

Comprehensive Recidivism Study

RECIDIVISM RATES
The definition of a recidivist used for this analysis was:
For offenders released from incarceration:
•
•

an offender with a new arrest (arraignment) in the year following release from a
correctional facility; or,
an offender with a technical violation of probation or parole that resulted in
incarceration in the year following release from a correctional facility;

For offenders in community corrections centers or drug court programs:
•
•

an offender with a new arrest (arraignment) in the year following the date of entry
into a community corrections center or drug court program; or,
an offender with a technical violation of probation or parole that resulted in
incarceration in the year following the date of entry into a community corrections
center or drug court program.

In this section, the overall recidivism rates are discussed along with a discussion of the
timing of the recidivist behavior and a description of that behavior.
Type of Recidivist
The overall recidivism rate was estimated to be 49.1% within one year of release or
program participation. Of the 3,751 offenders in the sample, 1,841 or 49.1% met the
definition of a recidivist.
Table 3 shows the manner in which offenders were classified as recidivists. The most
common reason for classifying an offender as a recidivist was one or more new arrests /
arraignments during the one year follow-up period. Of the 3,751 offenders in the sample,
an estimated 1,668 or 44.5% had one or more new arrests during the follow-up period.
The proportion of offenders with one or more new arrests was lower for the community
based programs: 38.2% of the drug court participants had one or more new arrests and
39.3% of the community corrections center participants had one or more new arrests. Of
the incarcerated offenders, the proportion of offenders with one or more new arrests was
lower among those released via parole than those offenders who were discharged. Of the
offenders released after serving a house of correction sentence, 49.8% of those discharged
had one or more new arrests compared to 38.5% of those released by parole; and, of the
19

Massachusetts Sentencing Commission

offenders released after serving a state prison sentence, 36.3% of those discharged had
one or more new arrests compared to 23.6% of those released by parole.
Other offenders had no new arrests but were classified as a recidivist solely on the basis
of a technical violation of parole or probation. Of all offenders in the sample, 173 or
4.6% were classified as recidivists based on a technical violation. Because all offenders
released by parole and all community corrections center and drug court program
participants were under supervision for at least part of the follow-up period, they were
expected to be subject to a higher rate of violation activity.
As indicated in Table 3, the proportion of offenders classified as recidivists based on
technical violations was higher for the community based programs: 11.8% of the drug
court participants were classified as a recidivist based on a technical violation and 12.7%
of the community corrections center participants were classified as a recidivist based on a
technical violation. Of the incarcerated offenders, the proportion of offenders classified
as recidivists based on a technical violation was higher among those released via parole
than those offenders who were discharged. Of the offenders released after serving a
house of correction sentence, 1.7% of those discharged were technical violators compared
to 8.7% of those released by parole; and, of the offenders released after serving a state
prison sentence, 0.0% of those discharged were technical violators compared to 18.8% of
those released by parole.
The nature of the new charges for which the offender was arraigned during the follow-up
period was also considered. The first characteristic of the new offense behavior that was
considered was whether the new charge was a felony or misdemeanor. In Massachusetts,
a felony is any offense for which the offender can be sentenced to state prison and
misdemeanors are all other offenses.
Table 3 shows the felony / misdemeanor breakdown associated with the most serious new
offense from among all of the charges for which an offender was arrested / arraigned
during the one year follow-up period. Of the 3,751 offenders in the sample, for 784 or
20.9% the most serious new offense was a felony and for 884 or 23.6% the most serious
new offense was a misdemeanor. The proportion of offenders whose most serious new
arrest offense was classified as a felony varied by criminal justice cohort. For example,
of the community corrections center participants, 46 or 14.2% had a felony as the most
serious new offense.

20

Comprehensive Recidivism Study

Table 3. Recidivism Classification by Criminal Justice Cohort

Sample Total

Drug Courts

House of Correction Sentence

Community
Corrections
Centers

Discharge

Parole

State Prison Sentence

Sample Total

Discharge

Parole

Sample Total
%

Type of Recidivist

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

Recidivism Rate

1,841

49.1%

38

50.0%

168

52.0%

1,156

51.5%

288

47.1%

1,444

50.6%

121

36.3%

70

42.4%

191

38.4%

One or More Arrests

1,668

44.5%

29

38.2%

127

39.3%

1,117

49.8%

235

38.5%

1,352

47.4%

121

36.3%

39

23.6%

160

32.1%

felony

784

20.9%

15

19.7%

46

14.2%

530

23.6%

104

17.0%

634

22.2%

67

20.1%

22

13.3%

89

17.9%

misdemeanor

884

23.6%

14

18.4%

81

25.1%

587

26.2%

131

21.4%

718

25.2%

54

16.2%

17

10.3%

71

14.3%

173

4.6%

9

11.8%

41

12.7%

39

1.7%

53

8.7%

92

3.2%

0

0.0%

31

18.8%

31

6.2%

Non-Recidivist

1,910

50.9%

38

50.0%

155

48.0%

1,087

48.5%

323

52.9%

1,410

49.4%

212

63.7%

95

57.6%

307

61.6%

Total

3,751 100.0%

76

100.0%

323

100.0% 2,243 100.0%

611

100.0% 2,854 100.0%

333

100.0%

165

100.0%

498

100.0%

Most serious:

No Arrests / Technical

21

Massachusetts Sentencing Commission

Time Until Recidivism
The behavior that was associated with the offender being classified as a recidivist could
occur at any point during the one year follow-up period. Offenders were considered as
recidivists if there was any arrest in the one year follow-up period regardless of the parole
or program status at the time of the new arrest. For many offenders released to parole,
especially those released after serving a house of correction sentence, the period of parole
supervision terminated before the end of the one year follow-up period. Similarly, for
offenders in community corrections center programs, where program cycles typically run
for 12 weeks, program participation was expected to be completed before the end of the
one year follow-up period. For this reason, Table 4 shows the relationship between new
arrests, parole supervision status, and participation status for the parole and community
corrections center cohorts.
Of those offenders paroled with a house of correction sentence, 235 (38.5%) had one or
more new arrests. However, the new arrest occurred after the end of the parole
supervision for 160 offenders. Of those paroled with a state prison sentence, 39 (23.6%)
had one or more new arrests and five occurred after the end of the period of parole
supervision.
Of the 127 offenders participating in a community corrections center program with one or
more new arrests, 27 occurred prior to program completion / termination, 31 occurred
after program completion / termination, and the completion / termination date was not
available for 69 offenders.
Table 5 shows the time until the recidivism behavior by criminal justice cohort. The
recidivism rates that are shown in Table 5 can be interpreted as interim or cumulative
recidivism rates. This data demonstrates the relationship between the length of the
follow-up period and estimated recidivism rates; that is, longer follow-up periods will be
associated with higher recidivism rates. Comparisons between rates might also change
over time. For example, when comparing the offenders in the community corrections
center programs with the offenders released from houses of correction, the recidivism rate
for the community corrections center programs was higher initially but the difference
between the rates declined over the follow-up period.

22

Comprehensive Recidivism Study

Table 4. New Arrest Status, Parole Supervision Status, and Community Corrections Center Participation Status
Community
Corrections Centers
Time of New Arrest

House Parole
Sample

State Parole
Sample

N

%

N

%

N

%

No New Arrests

196

60.7%

376

61.5%

126

76.4%

New Arrest - before end of parole / program

27

8.4%

75

12.3%

34

20.6%

New Arrest - after end of parole / program

31

9.6%

160

26.2%

5

3.0%

New Arrest - completion date not provi ded

69

21.4%

0

0.0%

0

0.0%

Sub-Total

127

39.3%

235

38.5%

39

23.6%

Total

323

100.0%

611

100.0%

165

100.0%

23

Massachusetts Sentencing Commission

Table 5. Time Until Recidivism Incident by Criminal Justice Cohort

Non-Recidivists

Cumulative
Recidivism Rate

New Recidivists

Cumulative
Recidivism Rate

New Recidivists

Cumulative
Recidivism Rate

New Recidivists

Cumulative
Recidivism Rate

New Recidivists

Cumulative
Recidivism Rate

New Recidivists

Cumulative
Recidivism Rate

Sample Total
(n = 498 )

New Recidivists

Parole
(n = 165 )

Cumulative
Recidivism Rate

Discharge
(n = 333 )

New Recidivists

Sample Total
(n = 2854 )

Cumulative
Recidivism Rate

Parole
(n = 611 )

New Recidivists

Less than 1 month
1 to 2 months
2 to 3 months
3 to 4 months
4 to 5 months
5 to 6 months
6 to 7 months
7 to 8 months
8 to 9 months
9 to 10 months
10 to 11 months
11 to 12 months

Discharge
(n = 2243 )

Cumulative
Recidivism Rate

Time Until Recidivism
Incident

Drug Courts
(n =76)

State Prison Sentence

House of Correction Sentence

New Recidivists

Sample Total
(n =3751)

Community
Corrections
Centers
(n =323)

223
219
215
187
196
148
128
116
116
103
88
102

5.9%
11.8%
17.5%
22.5%
27.7%
31.7%
35.1%
38.2%
41.3%
44.0%
46.4%
49.1%

4
9
7
2
3
3
2
2
3
1
0
2

5.3%
17.1%
26.3%
28.9%
32.9%
36.8%
39.5%
42.1%
46.1%
47.4%
47.4%
50.0%

27
23
19
21
14
15
10
7
11
9
8
4

8.4%
15.5%
21.4%
27.9%
32.2%
36.8%
39.9%
42.1%
45.5%
48.3%
50.8%
52.0%

147
133
138
113
131
98
80
69
70
67
43
67

6.6%
12.5%
18.6%
23.7%
29.5%
33.9%
37.4%
40.5%
43.6%
46.6%
48.6%
51.5%

31
38
26
29
25
19
19
24
21
14
24
18

5.1%
11.3%
15.5%
20.3%
24.4%
27.5%
30.6%
34.5%
38.0%
40.3%
44.2%
47.1%

178
171
164
142
156
117
99
93
91
81
67
85

6.2%
12.2%
18.0%
23.0%
28.4%
32.5%
36.0%
39.2%
42.4%
45.3%
47.6%
50.6%

8
11
17
12
15
9
11
11
7
7
7
6

2.4%
5.7%
10.8%
14.4%
18.9%
21.6%
24.9%
28.2%
30.3%
32.4%
34.5%
36.3%

6
5
9
10
7
4
6
3
4
5
7
4

3.6%
6.7%
12.1%
18.2%
22.4%
24.8%
28.5%
30.3%
32.7%
35.8%
40.0%
42.4%

14
16
26
22
22
13
17
14
11
12
14
10

2.8%
6.0%
11.2%
15.7%
20.1%
22.7%
26.1%
28.9%
31.1%
33.5%
36.3%
38.4%

1,910

38

155

1,087

323

24

1,410

212

95

307

Comprehensive Recidivism Study

New Arrest Offenses
Further characteristics of the new charges that occurred during the follow-up period were
also considered. In this section the offense seriousness level and the type of offense are
studied.
The offense seriousness level of all new arrest offenses was considered according to the
ranking of offenses proposed by the Sentencing Commission. Table 6 shows the level of
the most seriousness new arrest offense by criminal justice cohort. For all offenders with
one or more new arrests, the most serious new arrests were classified at level 4 through 9
for 21.0% of the offenders and at level 1 through 3 for 78.8% of the offenders. The
seriousness of the new arrests varied by criminal justice cohort. For example, of the
offenders with one or more new arrests, only 10.3% of the drug court program
participants with one or more new arrests had a new arrest at offense seriousness level 4
or above. In contrast, of the offenders released after serving a state prison sentence with
one or more new arrests, 30.6% had a new arrest at offense seriousness level 4 or above.
For those offenders with one or more new arrests, the most serious offense was classified
by type of offense. As shown in Table 7, for all offenders with one or more new arrests,
462 or 27.7% were charged with an offense against the person or sex offense as the most
serious new arrest offense and 1,206 or 72.3% were charged with some other offense
(property, drug, motor vehicle, etc.) as the most serious new offense.

25

Massachusetts Sentencing Commission

Table 6. Offense Seriousness Level of Most Serious New Arrest Offense by Criminal Justice Cohort
Sample
Total

Drug Courts

House of Correction Sentence

Community
Corrections
Centers

Discharge

Parole

State Prison Sentence

Sample Total

Discharge

Parole

Sample Total

Level of Most
Serious Offense

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

9

1

0.1%

0

0.0%

0

0.0%

1

0.1%

0

0.0%

1

0.1%

0

0.0%

0

0.0%

0

0.0%

8

6

0.4%

0

0.0%

0

0.0%

4

0.4%

0

0.0%

4

0.3%

1

0.8%

1

2.6%

2

1.3%

7

7

0.4%

0

0.0%

0

0.0%

6

0.5%

0

0.0%

6

0.4%

0

0.0%

1

2.6%

1

0.6%

6

31

1.9%

0

0.0%

4

3.1%

23

2.1%

1

0.4%

24

1.8%

2

1.7%

1

2.6%

3

1.9%

%

5

73

4.4%

0

0.0%

4

3.1%

47

4.2%

13

5.5%

60

4.4%

8

6.6%

1

2.6%

9

5.6%

4

232

13.9%

3

10.3%

15

11.8%

163

14.6%

17

7.2%

180

13.3%

26

21.5%

8

20.5%

34

21.3%

3

652

39.1%

11

37.9%

47

37.0%

449

40.2%

90

38.3%

539

39.9%

41

33.9%

14

35.9%

55

34.4%

2

407

24.4%

10

34.5%

38

29.9%

257

23.0%

67

28.5%

324

24.0%

27

22.3%

8

20.5%

35

21.9%

1

256

15.3%

5

17.2%

19

15.0%

166

14.9%

45

19.1%

211

15.6%

16

13.2%

5

12.8%

21

13.1%

Not Assigned

3

0.2%

0

0.0%

0

0.0%

1

0.1%

2

0.9%

3

0.2%

0

0.0%

0

0.0%

0

0.0%

Total

1,668

100.0%

29

100.0%

127

100.0%

1,117

100.0%

235

100.0%

1,352

100.0%

121

100.0%

39

100.0%

160

100.0%

% Level 4+

21.0%

10.3%

18.1%

21.8%

13.2%

20.3%

30.6%

30.8%

30.6%

% Level 1-3

78.8%

89.7%

81.9%

78.1%

86.0%

79.4%

69.4%

69.2%

69.4%

26

Comprehensive Recidivism Study

Table 7. Type of Offense of Most Serious New Arrest Offense by Criminal Justice Cohort
Sample Total

Type of Offense

N

%

Person

439

26.3%

Sex

23

1.4%

Drug Courts

N

Community
Corrections
Centers

State Prison Sentence

House of Correction Sentence
Discharge

Parole

Sample Total

Discharge

Parole

Sample Total

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

2

6.9%

31

24.4%

326

29.2%

43

18.3%

369

27.3%

29

24.0%

8

20.5%

37

23.1%

0

0.0%

1

0.8%

17

1.5%

3

1.3%

20

1.5%

1

0.8%

1

2.6%

2

1.3%

Property

426

25.5%

15

51.7%

26

20.5%

267

23.9%

77

32.8%

344

25.4%

30

24.8%

11

28.2%

41

25.6%

Drug

313

18.8%

5

17.2%

26

20.5%

206

18.4%

40

17.0%

246

18.2%

27

22.3%

9

23.1%

36

22.5%

Motor Vehicle

294

17.6%

7

24.1%

29

22.8%

183

16.4%

44

18.7%

227

16.8%

23

19.0%

8

20.5%

31

19.4%

Other

173

10.4%

0

0.0%

14

11.0%

118

10.6%

28

11.9%

146

10.8%

11

9.1%

2

5.1%

13

8.1%

Total

1668

100.0%

29

100.0%

127

100.0%

1117

100.0%

235

100.0%

1352

100.0%

121

100.0%

39

100.0%

160

100.0%

27

Massachusetts Sentencing Commission

PROGRAMS, SECURITY, AND COMMUNITY RELEASE
Exposure to the criminal justice programs varied for the offenders in the study sample.
The community based program samples included offenders who dropped out of the
program after a brief stay as well as offenders who successfully completed the program
after many weeks or months of participation. The correctional release samples included
offenders incarcerated for varying lengths of time in a wide range of security settings.
The relationship between recidivism rates and the nature of the criminal justice placement
are considered in this section. For those offenders released from the DOC or houses of
correction, the security level at the time of release was considered. For community
corrections centers, the intermediate sanction level, program completion status, and
length of time in the program were considered. For drug courts, the completion status
and length of time in the program were considered.
Correctional Facilities
For many offenders released from correctional facilities, the security level of the offender
at the time of release was available. All security levels were defined by the contributing
criminal justice agency and may not be comparable across the various criminal justice
cohorts. As indicated in Table 8, the recidivism rates of those offenders released from
lower security facilities were less than the recidivism rates of those offenders released
from higher security facilities. For all offenders, it was estimated that the recidivism rate
for releases from lower security facilities was 39.9% and the recidivism rate for releases
from higher security facilities was 53.1%. This pattern held for each of the correctional
cohorts.

28

Comprehensive Recidivism Study

Table 8. Recidivism Status by Security Level at Time of Release and Criminal Justice Cohort, Correctional Releases

House of Correction Sentence

State Prison Sentence

Sample Total

Total

Recidivists

Recidivism Rate

Total

Recidivists

Recidivism Rate

129

Recidivism Rate

50.0%

Recidivists

1

Sample Total

Total

2

Parole

Recidivism Rate

48.0%

Discharge

Recidivists

Total

61

Recidivism Rate

127

Recidivists

45.5%

Sample Total

Total

Recidivists

80

Parole
Recidivism Rate

Total

176

Recidivism Rate

Maximum

Recidivists

Security Level

Total

Discharge

62

48.1%

46

18

39.1%

1

0

0.0%

47

18

38.3%

Medium

1046

569

54.4%

707

412

58.3%

97

55

56.7%

804

467

58.1%

180

69

38.3%

62

33

53.2%

242

102

42.1%

Medium/Minimum

82

44

53.7%

67

35

52.2%

15

9

60.0%

82

44

53.7%

0

0

N.A.

0

0

N.A.

0

0

N.A.

Sub-Total Secure

1304

693

53.1%

901

508

56.4%

114

65

57.0% 1015

573

56.5%

226

87

38.5%

63

33

52.4%

289

120

41.5%

Minimum

475

194

40.8%

297

125

42.1%

90

41

45.6%

166

42.9%

57

20

35.1%

31

8

25.8%

88

28

31.8%

387

Minimum/Pre-release

113

41

36.3%

32

11

34.4%

13

6

46.2%

45

17

37.8%

37

11

29.7%

31

13

41.9%

68

24

35.3%

Pre-release

176

70

39.8%

76

34

44.7%

53

21

39.6%

129

55

42.6%

13

3

23.1%

34

12

35.3%

47

15

31.9%

Sub-Total Lower

764

305

39.9%

405

170

42.0%

156

68

43.6%

561

238

42.4%

107

34

31.8%

96

33

34.4%

203

67

33.0%

Not Available

1284

637

49.6%

937

478

51.0%

341

155

45.5%

1278

633

49.5%

0

0

N.A.

6

4

66.7%

6

4

66.7%

3352 1635 48.8% 2243 1156 51.5%

611

288

47.1%

2854 1444 50.6%

333

121

36.3%

165

70

42.4%

498

191

38.4%

Total

29

Massachusetts Sentencing Commission

Community Corrections Centers
Table 9 shows recidivism status by selected program characteristics of community
corrections center participants. Because all of these data elements were not required, the
proportion of unknowns varies extensively. However, the data does point to some
interesting preliminary trends. For example, the recidivism rate for program completers
(39.5%) was lower than the recidivism rate for those who terminated the program
(68.3%) - some of whom were terminated due to a new arrest or other violation - and the
recidivism rate was lower for those who were in the program longer. Further research
should place more emphasis on collecting information on the program level and the
program completion / termination information for this population.

Drug Courts
Table 10 shows recidivism status by selected program characteristics of drug court
participants, completion status and length of time in the program. Many of the drug court
program participants were still in the program at the end of the one year follow-up period.
Offenders in the “other” category included inappropriate referrals, medical problems, and
probation transfers.
As was the case with the community corrections centers cohort, the recidivism rate was
much lower for those who completed the program (21.4%) as well as for those who were
still in the program at the end of the one year follow-up period (16.7%). In light of these
findings, further research should place more emphasis on the distinction between program
completers and those offenders who terminated the drug court program.

30

Comprehensive Recidivism Study

Table 9. Recidivism Status by Selected Program Characteristics, Community Corrections Center Participants
Community Corrections Centers
Program Characteristic

Total

Recidivists

Recidivism Rate

Beginning Level
Level IV
Level III
Not Available
Total

81
192
50
323

40
97
31
168

49.4%
50.5%
62.0%
52.0%

Referral Source
Probation
Sheriff
Parole
Not Available
Total

212
24
1
86
323

123
12
0
33
168

58.0%
50.0%
0.0%
38.4%
52.0%

Completion Type
Completion
Termination
Not Available
Total

43
41
239
323

17
28
123
168

39.5%
68.3%
51.5%
52.0%

Change in Level
Level IV / no change
Level IV / Level III
Level III / no change
Not Available
Total

11
6
33
273
323

4
1
17
146
168

36.4%
16.7%
51.5%
53.5%
52.0%

Time in Program
1 Month or Less
2 to 4 Months
5 Months or More
Not Available
Total

47
74
18
184
323

30
41
7
90
168

63.8%
55.4%
38.9%
48.9%
52.0%

31

Massachusetts Sentencing Commission

Table 10. Recidivism Status by Selected Program Characteristics, Drug Court Participants
Drug Courts

Program Characteristic

Total

Recidivists

Recidivism Rate

Completion Type
Completion

14

3

21.4%

Still in Program at End of Year

18

3

16.7%

Termination

20

16

80.0%

Other Type

4

3

75.0%

Not Available

20

13

65.0%

Total

76

38

50.0%

4

2

50.0%

Time in Program
1 Month or Less
2 to 4 Months

7

6

85.7%

5 Months or More

20

10

50.0%

Still in Program at End of Year

16

2

12.5%

Not Available

29

18

62.1%

Total

76

38

50.0%

32

Comprehensive Recidivism Study

DEMOGRAPHIC CHARACTERISTICS OF OFFENDERS
The demographic characteristics of the offenders available for the analysis included:
gender, age, and race / ethnicity.
Gender
Table 11 shows recidivism status by gender and criminal justice cohort. Overall, there
were 446 (11.9%) female offenders and 3,305 (88.1%) male offenders in the sample.
The proportion of female offenders varied across the various criminal justice cohorts:
•
•
•
•

30.3% of the drug court program participants were female;
12.7% of the offenders released after serving a house of correction sentence were
female;
11.5% of the community corrections center program participants were female;
and,
4.8% of the offenders released after serving a state prison sentence were female.

Overall, the recidivism rate for female offenders was slightly lower than that of male
offenders (47.3% and 49.3%, respectively). This pattern held for all of the criminal justice
cohorts. Further research could consider some of the specialized community corrections
centers for female offenders that have been developed by the Office of Community
Corrections but were not in operation early enough to be included in this study.

Age
Age was calculated at the time of release from a correctional facility or the time the
offender began the community based program. The age of offenders in the sample ranged
from 17 to 73 years. The mean age was 32.0 years and the median age was 31 years.
The age of offenders varied by criminal justice cohort. To the extent that the particular
criminal justice agency services those younger “high-risk” offenders, it was expected that
the corresponding recidivism rate would be higher. Those offenders in drug courts and
those offenders released from the DOC with state prison sentences were the oldest on
average, while those offenders participating in community corrections centers were the
youngest on average:

33

Massachusetts Sentencing Commission

•
•
•
•

57.6% of the community corrections center program participants were
29 years or younger;
45.1% of offenders released after serving a house of correction sentence
were 29 years or younger;
34.1% of offenders released after serving a state prison sentence were
29 years or younger; and,
23.7% of the drug court program participants were 29 years or younger.

Table 12 shows recidivism rates by age. Younger offenders were more likely to be
classified as recidivists than older offenders. The recidivism rate for offenders under 20
years of age was 61.5% while the recidivism rate for offenders age 50 to 59 was 25.5%.
For all offenders age 29 or younger, the recidivism rate was 55.2% and for all offenders
age 30 and older, the recidivism rate was 44.2%. This pattern held for all of the criminal
justice cohorts with the exception of the drug court program participants.
Race / Ethnicity
Table 13 shows recidivism rates by race / ethnicity. Overall, the sample was 53.3% white
and 46.7% racial / ethnic minorities. Race / ethnicity was not available for 1.1% of the
sample. The racial / ethnic composition of the population varied by criminal justice
cohort:
•
•
•
•

62.7% of the offenders released after serving a state prison sentence were racial /
ethnic minorities;
44.8% of the offenders released after serving a house of correction sentence were
racial / ethnic minorities;
41.0% of the community corrections center program participants were racial /
ethnic minorities; and,
35.6% of the drug court program participants were racial / ethnic minorities.

Recidivism rates were lower for whites than for racial / ethnic minorities across each of
the cohorts.

34

Comprehensive Recidivism Study

Table 11. Recidivism Status by Gender and Criminal Justice Cohort
House of Correction Sentence

Recidivists

Recidivism Rate

Total

Recidivists

Recidivism Rate

Total

Recidivists

Recidivism Rate

117

36.7%

155

66

42.6%

474

183

38.6%

47.1% 2854

1444 50.6%

333

121

36.3%

165

70

42.4%

498

191

38.4%

23

11

47.8%

37

15

40.5%

259

125

48.3%

103

52

50.5%

362

Ma le

3,305 1,630 49.3%

53

27

50.9%

286

153

53.5% 1,984 1031 52.0%

508

236

Total

3,751 1,841 49.1%

76

38

50.0%

323

168

52.0% 2,243 1156 51.5%

611

288

% Fe male

11.9%

Total

Total

319

Recidivism Rate

1267 50.8%

Recidivists

46.5% 2,492

Total

33.3%

Recidivism Rate

8

Recidivists

24

Total

40.0%

Recidivism Rate

4

Recidivists

10

Total

28.6%

Recidivism Rate

4

Recidivists

14

Total

48.9%

47.3%

Fema le

446

30.3%

11.5%

11.5%

16.9%

35

Sample Total

Paro le

177

211

Gender

Discharge

Total

Sample Total

Recidivism Rate

Paro le

Recidivism Rate

Discharge

Recidivists

Community
Corrections
Centers

Drug Co urts

Recidivists

Sample Total

State Prison Sentence

12.7%

4.2%

6.1%

4.8%

Massachusetts Sentencing Commission

Table 12. Recidivism Status by Age and Criminal Justice Cohort

68

40

58.8%

223

132

59.2%

3

2

66.7%

2

1

50.0%

5

3

60.0%

59.2%

132

67

50.8%

558

319

57.2%

35

18

51.4%

22

10

45.5%

57

28

49.1%

Total

R ecidivists

59.4%

252

Total

92

426

Total

R ecidivism R ate

155

51.8%

R ecidivists

70.2%

44

Total

40

85

R ecidivists

57

22.2%

Total

100.0%

R ecidivists

1
2

R ecidivists

1
9

Total

61.5%
55.4%

R ecidivists

176

R ecidivists

R ecidivism R ate

Sample Total

393

Total

R ecidivism R ate

Paro le

286

Total

R ecidivism R ate

Discharge

709

R ecidivists

R ecidivism R ate

State Prison Sentence
Sample Total

20 to 24

Total

R ecidivism R ate

Paro le

R ecidivism R ate

Discharge

R ecidivism R ate

House of Correction Sentence

Commu nity
Corrections C enters

R ecidivists

Drug Co urts

Under 20

Age at Release or
Program

R ecidivism R ate

Sample Total

25 to 29

665

348

52.3%

8

4

50.0%

44

28

63.6%

385

214

55.6%

120

58

48.3%

505

272

53.9%

71

27

38.0%

37

17

45.9%

108

44

40.7%

30 to 34

634

327

51.6% 23

13

56.5%

49

26

53.1%

370

200

54.1%

91

45

49.5%

461

245

53.1%

65

23

35.4%

36

20

55.6%

101

43

42.6%

35 to 39

640

308

48.1% 10

8

80.0%

36

17

47.2%

409

202

49.4%

97

50

51.5%

506

252

49.8%

64

23

35.9%

24

8

33.3%

88

31

35.2%

40 to 49

645

248

38.4% 22

9

40.9%

40

10

25.0%

399

169

42.4%

82

24

29.3%

481

193

40.1%

71

24

33.8%

31

12

38.7%

102

36

35.3%

50 to 59

145

37

25.5%

3

1

33.3%

10

3

30.0%

85

24

28.2%

16

3

18.8%

101

27

26.7%

20

4

20.0%

11

2

18.2%

31

6

19.4%

60 to 69

23

4

17.4%

0

0

N.A .

1

0

0.0%

13

3

23.1%

5

1

20.0%

18

4

22.2%

2

0

0.0%

2

0

0.0%

4

0

0.0%

70 to 79

4

0

0.0%

0

0

N.A .

1

0

0.0%

1

0

0.0%

0

0

N.A.

1

0

0.0%

2

0

0.0%

0

0

N.A.

2

0

0.0%

Total

3751

1841 49.1% 76

38

50.0%

323

168

52.0% 2243

1156 51.5%

611

288

47.1% 2854

1444 50.6%

333

121

36.3%

165

70

42.4%

498

191

38.4%

29 and Under

1660

917

55.2% 18

7

38.9%

186

112

60.2%

966

558

57.8%

320

165

51.6% 1286

723

56.2%

109

47

43.1%

61

28

45.9%

170

75

44.1%

30 and Older

2091

924

44.2% 58

31

53.4%

137

56

40.9% 1277

598

46.8%

291

123

42.3% 1568

721

46.0%

224

74

33.0%

104

42

40.4%

328

116

35.4%

`
% 29 and Younger
Mean Age
Median Age

44.3%
32.0 Ye ars
31 Yea rs

23.7%
35.0 Ye ars
33 Yea rs

57.6%
29.1 Ye ars
28 Yea rs

43.1%
32.2 Ye ars
32 Yea rs

52.4%
30.4 Ye ars
29 Yea rs

36

`
45.1%
31.8 Ye ars
31 Yea rs

32.7%
35.0 Ye ars
34 Yea rs

37.0%
34.0 Ye ars
32 Yea rs

34.1%
34.6 Ye ars
33 Yea rs

Comprehensive Recidivism Study

Table 13. Recidivism Status by Race / Ethnicity and Criminal Justice Cohort
House of Correction Sentence

Community Corrections
Centers

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

1,976

879

44.5%

47

22

46.8%

190

78

41.1%

1,139

532

46.7%

414

186

44.9% 1,553

718

46.2%

122

36

29.5%

64

25

39.1%

186

61

32.8%

Black

885

488

55.1%

17

11

64.7%

47

27

57.4%

565

330

58.4%

91

45

49.5%

656

375

57.2%

113

49

43.4%

52

26

50.0%

165

75

45.5%

Hispa nic

776

431

55.5%

8

4

50.0%

83

61

73.5%

455

264

58.0%

94

53

56.4%

549

317

57.7%

93

32

34.4%

43

17

39.5%

136

49

36.0%

Other

71

29

40.8%

1

0

0.0%

2

2

100.0%

45

17

37.8%

12

4

33.3%

57

21

36.8%

5

4

80.0%

6

2

33.3%

11

6

54.5%

U n k n ow n

43

14

32.6%

3

1

33.3%

1

0

0.0%

39

13

33.3%

0

0

N.A.

39

13

33.3%

0

0

N.A.

0

0

N.A.

0

0

N.A.

Total

3,751 1,841 49.1%

76

38

50.0%

323

168

52.0%

611

288

333

121

36.3%

165

70

42.4%

498

191

38.4%

% W hite

53.3%

64.4%

59.0%

51.7%

67.8%

55.2%

36.6%

38.8%

37.3%

% M inority

46.7%

35.6%

41.0%

48.3%

32.2%

44.8%

63.4%

61.2%

62.7%

% Missing

1.1%

3.9%

0.3%

1.7%

0.0%

1.4%

0.0%

0.0%

0.0%

2,243 1,156 51.5%

37

Total

R ecidivism R ate

White

Race /
Ethnicity

R ecidivism R ate

R ecidivists

Sample Total

Total

Paro le

R ecidivism R ate

Discharge

R ecidivists

Sample Total

Total

Paro le

R ecidivism R ate

Discharge

State Prison Sentence

R ecidivists

Drug Co urts

Total

Sample Total

47.1% 2854

1444 50.6%

Massachusetts Sentencing Commission

CURRENT OFFENSE AND CRIMINAL HISTORY
In this section the current offense and criminal history of the offender are discussed in
relation to observed recidivism rates. Both of these variables, current offense and
criminal history, were measured at the time of the current placement. For correctional
populations, current offense refers to that offense for which the offender had been
convicted and on which the offender was now being released. For the community-based
programs, current offense refers to that offense for which the offender was being
supervised at the time of the placement. For all populations, criminal history refers to
criminal activity that occurred prior to the current offense, excluding the current offense,
and excluding any new arrest behavior.
Current Offense
Type of Offense
Current offense was available for 2,675 (71.3%) of the study sample. The sample
included 732 (27.4%) person offenders, 77 (2.9%) sex offenders, 737 (27.6%) drug
offenders, 586 (21.9%) property offenders, 373 (13.9%) motor vehicle offenders, and 170
(6.4%) other (public order, prostitution, and weapons) offenders. The nature of current
offense varied by criminal justice cohort:
•
•
•
•
•

42.4% of the offenders released after serving a state prison sentence were person
offenders;
10.4% of the offenders released after serving a state prison sentence were sex
offenders;
26.8% of the community corrections center program participants were property
offenders;
52.6% of the drug court program participants were drug offenders; and,
17.5% of the offenders released after serving a house of correction sentence were
motor vehicle offenders.

Table 14 shows the current offense and recidivism status by criminal justice cohort. Of
the major offense categories, recidivism rates were lowest for sex offenders (20.8%) and
highest for property offenders (56.5%). These patterns were generally consistent across
criminal justice cohorts.

38

Comprehensive Recidivism Study

Offense Seriousness Level
Table 15 shows current offense classified according to the offense seriousness level
proposed by the commission, recidivism status and criminal justice cohort. There was an
inverse relationship between offense seriousness level and recidivism rates. In general,
the more serious the current offense, the lower the recidivism rate of the offender. The
recidivism rate for offenders with a current offense level of 4 or above was 40.2% and the
recidivism rate for offenders with a current offense level of 1 through 3 was 51.7%.

39

Massachusetts Sentencing Commission

Table 14. Recidivism Status by Type of Offense (Current Offense) and Criminal Justice Cohort

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

45.8%

Total

11

Sample Total

R ecidivism R ate

24

Paro le

R ecidivists

R ecidivism R ate

81.8%

Discharge

Total

R ecidivists

9

State Prison Sentence
Sample Total

R ecidivism R ate

Total

11

Paro le

R ecidivists

R ecidivism R ate

50.0%

Discharge

Total

R ecidivists

366

Total

732

House of Correction Sentence

Commu nity
Corrections C enters

Drug Co urts
R ecidivism R ate

Person

R ecidivists

Governing
Offense

Total

Sample Total

373

204

54.7%

114

58

50.9%

487

262

53.8%

139

54

38.8%

71

30

42.3%

210

84

40.0%

Sex

77

16

20.8%

0

0

N.A.

0

0

N.A.

22

4

18.2%

3

2

66.7%

25

6

24.0%

45

9

20.0%

7

1

14.3%

52

10

19.2%

Property

586

331

56.5%

15

7

46.7%

34

23

67.6%

287

160

55.7%

159

91

57.2%

446

251

56.3%

62

32

51.6%

29

18

62.1%

91

50

54.9%

Drug

737

344

46.7%

40

14

35.0%

50

27

54.0%

333

180

54.1%

196

90

45.9%

529

270

51.0%

66

17

25.8%

52

16

30.8%

118

33

28.0%

Mo tor V ehicle

373

137

36.7%

4

3

75.0%

14

7

50.0%

236

93

39.4%

110

30

27.3%

346

123

35.5%

8

3

37.5%

1

1

100.0%

9

4

44.4%

Other

170

99

58.2%

6

5

83.3%

5

4

80.0%

112

63

56.3%

29

17

58.6%

141

80

56.7%

13

6

46.2%

5

4

80.0%

18

10

55.6%

Not A vailable

1076

548

50.9%

0

0

N.A.

196

96

49.0%

880

452

51.4%

0

0

N.A.

880

452

51.4%

0

0

N.A.

0

0

N.A.

0

0

N.A.

Total

3751

1841 49.1%

76

38

50.0%

323

168

52.0% 2243

611

288

333

121

36.3%

165

70

42.4%

498

191

38.4%

% Person
% Sex

27.4%

14.5%

18.9%

1156 51.5%

27.4%

18.7%

47.1% 2854

1444 50.6%

24.7%

41.7%

43.0%

42.2%

2.9%

0.0%

0.0%

1.6%

0.5%

1.3%

13.5%

4.2%

10.4%

% Prop erty

21.9%

19.7%

26.8%

21.1%

26.0%

22.6%

18.6%

17.6%

18.3%

% Drug

27.6%

52.6%

39.4%

24.4%

32.1%

26.8%

19.8%

31.5%

23.7%

% M Vehic le

13.9%

5.3%

11.0%

17.3%

18.0%

17.5%

2.4%

0.6%

1.8%

6.4%

7.9%

3.9%

8.2%

4.7%

7.1%

3.9%

3.0%

3.6%

% Other

40

Comprehensive Recidivism Study

Table 15. Recidivism Status by Offense Seriousness Level (Current Offense) and Criminal Justice Cohort
House of Correction Sentence

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

26.3%

R ecidivism R ate

15.4%

10

R ecidivists

4

38

Total

26

R ecidivism R ate

0.0%

R ecidivists

0

Total

4

R ecidivism R ate

R ecidivism R ate

Sample Total

R ecidivists

Paro le

R ecidivists

Discharge

Total

Sample Total

R ecidivism R ate

Paro le

R ecidivists

Discharge

State Prison Sentence

Total

Commu nity
Corrections C enters

Drug Co urts

Total

Sample Total

4

0

0.0%

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

4

0

0.0%

26

4

15.4%

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

20

3

15.0%

6

1

16.7%

38

10

26.3%

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

0

0

N.A.

30

6

20.0%

8

4

50.0%

6

137

50

36.5%

0

0

N.A.

3

0

0.0%

0

0

N.A.

2

0

0.0%

2

0

0.0%

88

34

38.6%

44

16

36.4%

132

50

37.9%

5

116

49

42.2%

0

0

N.A.

4

3

75.0%

23

10

43.5%

3

2

66.7%

26

12

46.2%

61

22

36.1%

25

12

48.0%

86

34

39.5%

4

447

196

43.8%

10

3

30.0%

17

10

58.8%

187

91

48.7%

143

60

42.0%

330

151

45.8%

50

18

36.0%

40

14

35.0%

90

32

35.6%

3

1025

537

52.4%

28

17

60.7%

46

29

63.0%

579

303

52.3%

255

129

50.6%

834

432

51.8%

79

36

45.6%

38

23

60.5%

117

59

50.4%

2

644

316

49.1%

37

18

48.6%

46

23

50.0%

405

202

49.9%

155

72

46.5%

560

274

48.9%

1

1

100.0%

0

0

N.A.

1

1

100.0%

Governing
Offense Level

9
8
7

1

221

124

56.1%

1

0

0.0%

9

5

55.6%

161

95

59.0%

50

24

48.0%

211

119

56.4%

0

0

N.A.

0

0

N.A.

0

0

N.A.

Not Assigned

17

7

41.2%

0

0

N.A.

2

2

100.0%

8

3

37.5%

3

1

33.3%

11

4

36.4%

4

1

25.0%

0

0

N.A.

4

1

25.0%

Not A vailable

1076

548

50.9%

0

0

N.A.

196

96

49.0%

880

452

51.4%

0

0

N.A.

880

452

51.4%

0

0

N.A.

0

0

N.A.

0

0

N.A.

Total

3751

1841 49.1%

76

38

50.0%

323

168

52.0%

2243

1156

51.5%

611

288

47.1%

2854

1444 50.6%

333

121

36.3%

165

70

42.4%

498

191

38.4%

41

Massachusetts Sentencing Commission

Criminal History
Offenders were assigned to a criminal history group consistent with the definitions
proposed by the commission. The criminal history groups are defined in the Appendix.
The relationship between juvenile criminal history and recidivism rates was also
considered.
Criminal History Groups
Based on the data extraction and linkage procedures discussed in the methodology
section, offenders were assigned to a criminal history group as proposed by the
Commission. This assignment was made for 2,334 or 62.2% of the study sample.
Of the offenders in the sample, 55.7% were classified in criminal history groups A or B
and 44.3% were classified in criminal history groups C, D, or E. The proportion of
offenders in the more serious criminal history groups varied by criminal justice cohort:
•
•
•
•

51.7% of the offenders released after serving a state prison sentence were
classified in criminal history groups C, D, or E;
46.1% of the drug court program participants were classified in criminal
history groups C, D, or E;
43.7% of the offenders released after serving a house of correction
sentence were classified in criminal history groups C, D, or E; and
22.0% of the community corrections center program participants were
classified in criminal history groups C, D, or E.

Table 16 shows recidivism rates of offenders by criminal history group and criminal
justice cohort. Recidivism rates varied consistently by criminal history group. Those
offenders in the less serious criminal history groups had lower recidivism rates and those
offenders in the more serious criminal history groups had higher recidivism rates. For
example, offenders classified as criminal history group A, no / minor record, had a
recidivism rate of 36.5%. Offenders assigned to criminal history group D, violent or
repetitive record, had a recidivism rate of 60.8%. These patterns held for all of the
criminal justice cohorts.

42

Comprehensive Recidivism Study

Juvenile Criminal History
Juvenile criminal history was available for all offenders in the sample. For purposes of
this analysis all offenders were classified into two categories: (1) no prior juvenile
adjudications; or, (2) one or more prior juvenile adjudications. Overall, 23.6% of the
offenders had one or more prior juvenile adjudications. The proportion of offenders with
one or more prior juvenile adjudications ranged from 17.1% of the drug court program
participants to 26.1% of the offenders released after serving a state prison sentence.
Those offenders with one or more prior juvenile adjudications were younger than those
offenders with no prior juvenile adjudications, 25.7 years on average compared with 34.0
years.
Table 17 shows recidivism rates by this juvenile criminal history indicator. The
recidivism rate for offenders with one or more prior juvenile adjudications was 60.8%,
much higher than the rate for offenders with no prior juvenile adjudications, 45.5%. This
pattern held for all of the criminal justice cohorts, with the exception of the drug court
program participants.

43

Massachusetts Sentencing Commission

Table 16. Recidivism Status by Criminal History and Criminal Justice Cohort
House of Correction Sentence

Commu nity
Corrections C enters

State Prison Sentence

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

222 36.5%

16

5

31.3%

55

26

47.3%

337

131

38.9%

65

27

41.5%

402

158

39.3%

79

17

21.5%

57

16

28.1%

136

33

24.3%

Group B Mode rate Reco rd

691

336 48.6%

25

14

56.0%

44

28

63.6%

404

200

49.5%

114

59

51.8%

518

259

50.0%

72

23

31.9%

32

12

37.5%

104

35

33.7%

G r o up C Serious Re cord

526

271 51.5%

12

7

58.3%

20

14

70.0%

321

183

57.0%

67

30

44.8%

388

213

54.9%

75

22

29.3%

31

15

48.4%

106

37

34.9%

G r o up D Violent or Repetitive

505

307 60.8%

23

12

52.2%

8

4

50.0%

298

188

63.1%

27

18

66.7%

325

206

63.4%

105

58

55.2%

44

27

61.4%

149

85

57.0%

Group E Serious Violent

3

66.7%

0

0

N.A.

0

0

N.A.

1

1

100.0%

0

0

N.A.

1

1

100.0%

2

1

50.0%

0

0

N.A.

2

1

50.0%

Not Available or
Youthful Offender

1,417

703 49.6%

0

0

N.A.

196

96

49.0%

882

453

51.4%

338

154

45.6% 1,220

607

49.8%

0

0

N.A.

1

0

0.0%

1

0

0.0%

Total

3,751 1,841 49.1%

76

38

50.0%

323

168

52.0% 2,243 1,156 51.5%

611

288

47.1% 2854

1444

50.6%

333

121

36.3%

165

70

42.4%

498

191

38.4%

% C, D , or E

44.3%

R ecidivists

609

Total

Total

Sample Total

R ecidivism R ate

Paro le

R ecidivists

Discharge

Total

Sample Total

R ecidivism R ate

Paro le

R ecidivists

Discharge

Total

Drug Co urts

G roup A No/M inor Record

Crimina l
History
Group

2

R ecidivism R ate

Sample Total

46.1%

22.0%

45.6%

34.4%

44

43.7%

54.7%

45.7%

51.7%

Comprehensive Recidivism Study

Table 17. Recidivism Status by Juvenile Criminal History and Criminal Justice Cohort
House of Correction Sentence

State Prison Sentence

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

477

207

43.4% 2,195 1,032 47.0%

237

74

31.2%

131

48

36.6%

368

122

33.2%

60.8%

13

5

38.5%

83

52

62.7%

331

63.0%

134

81

60.4%

96

47

49.0%

34

22

64.7%

130

69

53.1%

Total

3,751 1,841 49.1%

76

38

50.0%

323

168

52.0% 2,243 1,156 51.5%

611

288

47.1% 2854

333

121

36.3%

165

70

42.4%

498

191

38.4%

% One or More

23.6%

885

538

17.1%

25.7%

525

23.4%

21.9%

45

659

412

R ecidivism R ate

48.0%

R ecidivists

R ecidivists

825

Total

Total

48.3% 1,718

One or M ore

R ecidivism R ate

R ecidivism R ate

116

2,866 1,303 45.5%

Total

240

R ecidivists

52.4%

Total

R ecidivists

Sample Total

R ecidivists

Paro le

Total

Discharge

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

33

None

R ecidivism R ate

Commu nity
Corrections C enters

Drug Co urts

63

Juven ile
Adjudications

R ecidivism R ate

Sample Total

62.5%

1444 50.6%

23.1%

28.8%

20.6%

26.1%

Massachusetts Sentencing Commission

Sentencing Grid
The current offense and criminal history were combined to place offenders on the
sentencing guidelines grid proposed by the commission. Those offenders convicted of
OUI offenses or mandatory firearms offenses were not assigned to the grid because the
proposed sentencing guidelines grid would not apply to these offenders. Table 18 shows
the number of offenders assigned to each grid cell, the number of recidivists in each grid
cell, and the resulting recidivism rate. The recidivism rate was higher at the lower
offense seriousness levels and the recidivism rate was higher for those offenders in the
more serious criminal history groups. For any particular offense level, the recidivism
rates are generally higher for those offenders in more serious criminal history groups and
for any particular criminal history category, the recidivism rates are generally higher for
those offenders in the less serious offense levels. The recidivism rate for OUI offenders
was 24.3%, lower than the overall rate of 49.1%, suggesting that further analysis of the
specialized programs for this population may be merited.

46

Comprehensive Recidivism Study

Table 18. Recidivism Status by Sentencing Grid Cell and Grid Assignment, All Offenders
ALL DEFENDANTS

RECIDIVISM RATE

RECIDIVISTS

A

B

C

D

E

N

A

B

C

D

E

N

9

4

0

0

0

0

4

9

0

0

0

0

0

0

8

8

6

10

2

0

26

8

2

0

1

1

0

7

17

7

7

7

0

38

7

6

0

3

1

6

48

27

30

29

0

134

6

11

11

10

5

29

29

21

32

0

111

5

5

11

4

110

97

83

54

0

344

4

34

3

174

231

181

239

1

826

3

2

116

167

114

74

0

471

1

35

69

55

37

0

N .A.

6

5

3

3

Total

547

638

504

477

Total

A

B

C

D

E

9

0.0%

N.A.

N.A.

N.A.

N.A.

4

8

25.0%

0.0%

10.0%

50.0%

N.A.

0

10

7

35.3%

0.0%

42.9%

14.3%

N.A.

18

0

50

6

22.9%

40.7%

33.3%

62.1%

N.A.

8

21

0

45

5

17.2%

37.9%

38.1%

65.6%

N.A.

46

41

32

0

153

4

30.9%

47.4%

49.4%

59.3%

N.A.

76

121

106

150

1

454

3

43.7%

52.4%

58.6%

62.8%

100.0%

2

57

94

62

46

0

259

2

49.1%

56.3%

54.4%

62.2%

N.A.

196

1

17

37

32

24

0

110

1

48.6%

53.6%

58.2%

64.9%

N.A.

0

17

N.A.

3

1

3

0

0

7

N.A.

50.0%

20.0%

100.0%

0.0%

N.A.

1

2,167

Total

211

321

266

293

1

1,092

Total

38.6%

50.3%

52.8%

61.4%

100.0%

Sentencing Zone
Assig ned to Sente ncing G rid
Incarceration Zone
Discretionary Zone
Intermediate Sanction Zone
Not Assigned
Sub-Total

N
309
1,566
275
17
2,167

Sentencing Zone
Assig ned to Sente ncing G rid
Incarceration Zone
Discretionary Zone
Intermediate Sanction Zone
Not Assigned
Sub-Total

Not A ssigne d to G rid
OUI Offenses
Mandatory Firearms Offenses
Not Available or Youthful Offender

148
19
1417

Not A ssigne d to G rid
OUI Offenses
Mandatory Firearms Offenses
Not Available or Youthful Offender

Total

3,751

Total

47

N
125
817
143
7
1,092

36
10
703
1,841

Sentencing Zone
Assig ned to Sente ncing G rid
Incarceration Zone
Discretionary Zone
Intermediate Sanction Zone
Not Assigned
Sub-Total

0.0%
15.4%
26.3%
37.3%
40.5%
44.5%
55.0%
55.0%
56.1%
41.2%
50.4%

Recidivism
40.5%
52.2%
52.0%
41.2%
50.4%

Not A ssigne d to G rid
OUI Offenses
Mandatory Firearms Offenses
Not Available or Youthful Offender

24.3%
52.6%
49.6%

Total

49.1%

Massachusetts Sentencing Commission

SUMMARY
A main purpose of this report has been to meet the legislative mandate set forth in
Chapter 177 by compiling recidivism statistics from a variety of criminal justice and
correctional entities using a consistent definition of recidivism. Among the highlights
are:
The project developed a reliable measure of the recidivism rates across a range of
criminal justice cohorts.
Based on the cooperation of the originating criminal justice agencies, a reliable
measure of the recidivism rates was achieved as directed by Chapter 177 in the
time frame provided by the legislature. Data was contributed from a variety of
different criminal justice agencies and programs and integrated in a manner that
yielded both reliable results and a basis for some comparisons of the populations
and associated recidivism rates across criminal justice cohorts.
This project demonstrated the utility and feasibility of conducting research across
traditional criminal justice agency lines.
This research project was the result of a collaborative effort among thirty different
criminal justice agencies and programs. The project succeeded without any
additional resources devoted to its completion. The ability of the various agencies
to collect data independently and to contribute unique agency data to the
development of a comprehensive research project was shown to be feasible. This
model is one that could be applied to other research topics or to the expanded
study of recidivism.
It was important to consider the nature of the recidivism behavior in addition to the
overall recidivism rates.
The analysis demonstrated that it is important to consider the nature of the
recidivism behavior in addition to comparing the overall recidivism rates. The
analysis provided an examination of the nature of the recidivism behavior (new
arrest or technical violation), the timing of the recidivism behavior, and the
seriousness of that behavior. As suggested by the data collected for this study,
offenders are subject to varying levels of supervision that can lead to technical
violations. Even if the overall recidivism rates were the same, an argument can be
made that there is a public safety benefit if the recidivism classification resulted
48

Comprehensive Recidivism Study

from a technical violation rather than a new arrest. That is, a swift and certain
response to a violation of the conditions of probation or parole may be viewed as a
preventive step that contributes to public safety by intervening in the life of a
supervised offender who may be in danger of slipping back into a pattern of
criminal behavior.
It was important to consider offender background information along with
recidivism information.
The study demonstrated both the strong relationship between certain offender
characteristics and recidivism rates and the large differences in the composition of
the offender population across criminal justice cohorts. In particular, the age of
the offender and the criminal history of the offender were shown to be strongly
related to the probability of recidivism across all of the criminal justice cohorts in
the study. It is important to recognize that differences in the age and criminal
history characteristics of the offender population in each of the criminal justice
cohorts are related to observed recidivism rates.
Additional data collection on program involvement and supervision variables
should be encouraged in any future studies of recidivism.
Future research should include program involvement data in order to shed more
light on the relative effectiveness of different programs and on what types of
programs are more (or less) effective with what types of offenders. Some of the
offenders discharged from correctional facilities might have been under probation
supervision during the follow-up period. Future research should also include the
probation supervision status of the discharged offenders.
Longer follow-up periods should be encouraged in any future studies of recidivism.
The time frames available for completion of this project allowed for a one year
follow-up. Longer follow-up periods may be interesting and useful for the
comparisons to be made. For example, at the end of the one year follow-up period
the recidivism rates for the community corrections centers and houses of
correction were very close. However, the interim rates were different, suggesting
that longer term recidivism rates may look different for these or other populations.

49

Massachusetts Sentencing Commission

Replication of the model developed for the study is feasible and should be
encouraged.
The methodology developed in the process of completing this study could be used
to provide a valuable basis to support on-going research and monitoring of
recidivism rates. The method provided a foundation for routine updates of this
initial effort. Although on-going studies are not required by the current legislative
mandate, continued comprehensive recidivism studies should be encouraged.
This model could be expanded to accommodate new criminal justice programs
such as additional community corrections centers, drug courts, or additional
houses of correction. Extensions to other probation supervised offender
populations could be considered as well. The model should allow for longer term
follow-up periods than were used in this study. Further, the model should be able
to sustain routine follow-up studies on succeeding time periods.
Much has been written in recent years about the importance of partnerships in criminal
justice. Community policing and community corrections are example of recent initiatives
that have emphasized cooperation and coordination among criminal justice agencies to
achieve common goals. This research endeavor has underscored the value of such
collaboration among criminal justice researchers to produce a comprehensive study that
goes beyond the purview of each of the participating entities. In demonstrating that such
a collaborative research process is feasible, this study has not only produced useful data
on recidivism across a spectrum of the criminal justice system, but has also established a
model that could be applied for future system-oriented research.

50

Comprehensive Recidivism Study

APPENDIX - METHODOLOGY
This appendix contains additional information of the method used to select the sample for
this analysis and the definition of the variables collected for the analysis.
SAMPLE
All samples were submitted by the originating criminal justice agencies and were
assumed to be complete. The data extraction and data linkage model provided for some
validation between samples, e.g. the parole board sample was linked to parolees from all
contributing correctional agencies. All validations indicated that the samples were
substantially complete and the links between the various data sets were valid.
Multiple Releases / Admissions. In all samples, the data were examined for cases with
two or more program admissions or two or more releases during the sample period. It
was possible that an offender could be released more than once during the three month
sampling period. Where the case was determined to be a valid duplicate admission or
release, both admissions or releases were included in the sample and subject to routine
data extraction and linkage routines. It was also possible that offenders would be
included as part of the sample of more than one criminal justice agency. A total of 22
cases were identified where the offender had two valid releases from the same facility or
program admissions during the sampling period. In addition, a total of 38 cases were
identified where the offender had two valid releases or program admissions from different
facilities over the sampling period.
Release Status. Offenders who were released by court or released to another correctional
placement (transfer, pre-trial placement, or forthwith sentence) were not included in the
analysis. Offenders released from a correctional facility from a “week-end” sentence
were included only once in the sample. The most recent release among multiple weekend
releases was selected. Those offenders released to parole from a correctional facility
following a period of incarceration as the result of a 15-day parole detainer were excluded
from the sample. For those offenders who were sentenced to a drug court or a community
corrections center a s part of the probation supervision on a split sentence, the one year
follow-up period began when the offender entered the program following release from the
incarcerated portion of the split sentence.
Conviction Status. It was not assumed that offenders participating in all of the
originating criminal justice agencies had been convicted at the time of program
participation. In particular, it was possible for a offender to be referred to a community
51

Massachusetts Sentencing Commission

corrections center or drug court program following a disposition of general continuance
or CWOF (Continued without a Finding). While a disposition of general continuance or
CWOF is not considered a conviction, offenders are generally under probation
supervision and may be required to attend a drug court or community corrections center
program as part of the conditions of that probation supervision. Offenders in drug court
or community corrections center samples were included in the analysis even if not
convicted of the current offense. The release samples from the DOC and houses of
correction included only those offenders released following a conviction and excluded
any releases from pre-trial detention status.
Length of Incarceration. Most recently, the DOC has defined a recidivist to be “any
release who is re-incarcerated . . . for at least 30 days within three years of their date of
release.”17 Because of the definition of a recidivist, the DOC includes “only inmates
released off of a commitment of at least 30 days”18 in the analysis. This definition was
not used as a sampling criterion, i.e. all inmates released during the sampling time frame
were included.
Juvenile Offenders / Youthful Offenders. The study sample included adult offenders.
Those offenders adjudicated delinquent who were participating in the specialized
community corrections centers or drug courts for juvenile offenders were not included in
the study.
Those offenders adjudicated youthful offenders, sentenced as adults, and released from an
adult correction facility were included in the study. Four youthful offenders were
identified and included in the study sample.
Out-of-State and Federal Offenders. Some of the offenders in the release / program
participation cohorts were originally sentenced in another state or federal court and
subsequently transferred to Massachusetts. Not all of those offenders were identified in
the CARI data-base. A total of 28 federal and out-of-state cases were identified and
excluded from the final sample.
Mortality Information. Any offenders who died within the follow-up period were
excluded from the analysis. For each offender in the study sample, the identification

17

Hoover, op. cit., page 1.

18

Ibid., page 32.

52

Comprehensive Recidivism Study

records and adult arraignment records in the CARI data-base were evaluated for
occurrences of the phrase “DECEAS.” A total of 31 offender records were identified with
this phrase. The date of death was taken from the Social Security Death Index (SSDI).
Of the 31 identified offenders: 19 were excluded from the analysis because the date of
death was verified and the death occurred within the one year follow-up period; 8 were
included in the analysis because the date of death was verified and occurred after the one
year follow-up period; and, 4 were included in the analysis because there was no
verifying information in the SSDI.
Whereabouts Unknown. An offender could be “whereabouts unknown” for some or all
of the follow-up period. that period of time for which the offender was “whereabouts
unknown” should be accounted for in the analysis. There were five parolees who were
classified as “whereabouts unknown” during the recidivism follow-up period and
remained in that classification until the end of that period. For the purposes of this
analysis, these offenders have been classified as recidivists because the status of these
offenders could not be determined for the entire follow-up period and they were counted
in the category, “no new arrest / technical violation.”
FY2002 Data Requirement. Chapter 177 contained a provision requiring that the study
contain data from FY2002. Because of the reporting deadline and the desire to have a
reasonable follow-up period, the study used data from time periods prior to FY2002.
While some data from FY2002 was collected during the research process, the focus of the
study was on data sets earlier than FY 2002.

DATA COLLECTION
Recidivism Variables
Data collection for new arrest information was from the CARI database. As such,
recidivism was measured as subsequent arrest / arraignment in Massachusetts and did not
systematically capture criminal activity that occurred in other states.
Data collection for parole and probation violations was from the PATS file and CARI
database. Of primary interest was identifying those technical violations of supervised
release that result in a period of incarceration. Those incarcerations resulting from 15-day
parole detainers were not considered as indications of recidivism. In a recent study of
recidivism by the DOC, of the 1,504 offenders defined as recidivists, 19.8% were re-

53

Massachusetts Sentencing Commission

incarcerated technical violators and 80.2% were re-incarcerated following a new arrest.19
For those cases in drug courts and community corrections centers where there was no
conviction of the current offense prior to the program admission, that is the offender was
under probation supervision with a disposition of a CWOF or a general continuance, a
violation that resulted in a guilty verdict being entered and a period of incarceration was
treated as a probation violation.
A number of offenders were arraigned for offenses prior to the correctional release or
program participation and convicted of those offenses during the follow-up period. Those
offenders who had an incarceration of less than 15 days resulting from a conviction of an
offense that was arraigned prior to the correctional release or program participation were
treated as non-recidivists and included in the sample. Those offenders who had an
incarceration of more than 15 days resulting from a conviction of an offense arraigned
prior to the correctional release or program participation and had no other new arrests or
technical violations of parole or probation were excluded from the sample because they
were not at liberty for the entire follow-up period. A total of 23 offenders were excluded
for this reason.
New Offenses. Six variables were included in the recidivism data base that describe the
offenses for which offenders were charged during the follow-up period.
The first three were selected from those charges that were arraigned during the first
arraignment event following release or program participation:
•
•
•

most serious charge;
most serious charge resulting in a conviction; and,
most serious charge resulting in a sentence of incarceration.

The second three variables were selected from all charges that were arraigned during the
one year period following release or program participation:
•
•
•

19

most serious charge;
most serious charge resulting in a conviction; and,
most serious charge resulting in a sentence of incarceration.

Ibid., page 2.

54

Comprehensive Recidivism Study

In all instances, the ranking that has been proposed by the Commission was used to
assign a seriousness level to an offense. For purposes of the statistical presentation only
two variables were selected: the offense level of the most serious charge among all new
charges and the offense level of the most serious charge resulting in a conviction. The
type of offense was assigned based on the most serious charge among all new charges.
Routine editing was performed on offenses that were arraigned in the district court so that
all offense levels were set consistent with district court jurisdiction. For example, an
arraignment charge of robbery in the district court was assumed to be an arraignment
charge of larceny person. If an offender was arraigned in district court during the followup period and indicted in the superior court after the follow-up period, the seriousness
level of the new offense may be under-stated.
The disposition and conviction status of all new offenses were determined based on the
standard methodology developed by the Commission for use in the Survey of Sentencing
Practices.20 If a defendant was arraigned on a new charge during the follow-up period
and the case resulted in a conviction after the follow-up period, for purposes of this
analysis, that charge was not counted as a conviction. Only new arraignments that
occurred during the one year follow-up period and case processing activity that occurred
during the one year follow-up period were included in the analysis of convictions.
Type of Offense. All new offenses were assigned to one of six general type of offense
categories consistent with the method used in the Survey of Sentencing Practices.21 The
six categories are:
•
•
•
•
•

person offenses - includes murder, robberies, assault & battery;
sex offenses - includes rapes, indecent assault & battery;
property offenses - includes larceny, burglary, breaking & entering;
drug offenses - includes possession and distribution of controlled substances;
motor vehicle offenses - includes driving with licence suspended and vehicular
homicide; and,
other offenses - includes weapons, public disorder, and other offenses such as
prostitution, disorderly person, dangerous weapons.

•

20

Massac husetts Senten cing Com mission, Survey of Sentencing Practices, FY 2000, November 2001.

21

Ibid.

55

Massachusetts Sentencing Commission

Incarceration Status. A number of variables was included in the recidivism data base
that indicate if the new offense resulted in a sentence to incarceration during the one year
follow-up period.
The first three were selected from those charges that were arraigned during the first
arraignment event following release or program participation:
•
•
•

number of charges resulting in a sentence of incarceration;
first date an incarceration sentence was imposed; and,
most serious charge resulting in a sentence of incarceration;

The next three were selected from all charges that were arraigned during the one year
period following release or program participation:
•
•
•

number of charges resulting in a sentence of incarceration;
most recent date an incarceration sentence was imposed; and,
most serious charge resulting in a sentence of incarceration;

In all instances, the ranking that was developed by the Commission was used to assign a
seriousness level to an offense. In most cases these variables were set based on an initial
sentence to incarceration. For those offenders who had no other sentence of incarceration
from any new offense behavior in the follow-up period but who did have one or more
sentences to probation followed by a violation of probation that resulted in an
incarceration for a new offense that occurred during the follow-up period, the recidivism
incarceration variables were set based on the incarcerated violation of probation
information. For purposes of the statistical presentation only one variable was selected:
the number of charges resulting in a sentence of incarceration.
Parole Violation Status. The Massachusetts Parole Board provided information on all
offenders released to parole supervision during the three-month time period under study.
For each of those offenders information on all administrative activity was available for
the year following release from the correctional facility. The administrative chronology
contained information on all detainers, warrants, and board decisions during the followup period.
If the Parole Officer believes that the parolee has lapsed or is about to lapse into criminal
ways or has violated the conditions of his parole and cannot remain in the community, the
Parole Officer with the consent of a parole supervisor will issue a warrant for temporary
custody, also known as a 15-day detainer. The 15-day detainer authorizes the parolee to
56

Comprehensive Recidivism Study

be detained for a maximum of 15 days during which time the Board will conduct a
preliminary revocation hearing. The parolee may postpone the preliminary revocation
hearing to obtain legal representation or the attendance of witnesses. A second detainer
may issue if the Board grants a postponement.
If the parolee is being held on the 15-day detainer, the preliminary parole revocation
hearing will be held by a Hearing Examiner during this time and the Parole Board, after
considering the recommendation of the Hearing Examiner, will vote whether to
provisionally revoke parole. If the Board votes to provisionally revoke parole, a parole
violation warrant, also known as a warrant for permanent custody will be issued. The
warrant for permanent custody is served when the parolee is physically detained under
the Board's warrant. Upon service of the warrant for permanent custody, a final
revocation hearing will be scheduled within 60 days.
•

If the Board provisionally revokes the parole, the parolee remains in custody until
the Board conducts a final revocation hearing and makes a final decision
regarding revocation.

•

If the Board does not provisionally revoke parole, the parolee is returned to the
community under parole supervision.

At the final revocation hearing, the Board members determine whether the parolee
violated the terms and conditions of parole and whether the parolee will be re-released on
parole supervision.
•

If the Board does not find by a preponderance of the evidence that the parolee violated
a condition of parole, supervision will resume within 24 hours unless more time is
necessary to notify victims, notify the District Attorney, approve the home plan, etc.
The Board may modify or add conditions to parole.

•

Where the Board finds by a preponderance of the evidence that the parolee did violate
the parole condition(s), it will 'affirm' the provisional parole revocation. The Board
may either set a reparole date pending the fulfillment of certain conditions or the
Board may vote to deny reparole, thus returning the parolee to an inmate status.

Many of the offenders in the study were on parole at the end of the one year follow-up
period. Other offenders successfully completed their parole during the one year followup period. An individual was considered to be a parole violator if a detainer was issued
before the end of the one year follow-up period and if parole was subsequently revoked.
57

Massachusetts Sentencing Commission

An individual was also considered to be a parole violator if a detainer was issued before
the end of the one year follow-up period and if the offender ended the parole supervision
while in custody and the period between the detainer and the end of the parole
supervision was more than 15 days. An individual was not considered a parole violator if
a detainer was issued but the parole was not revoked. Information on new arrests was
taken from the CARI database. It was assumed that offenders who were classified as
parole violators but who had no new arrests were technical parole violators.
Probation Violation / Other Status. An offender was considered a probation violator if
the adult criminal record included one or more charges that were arraigned prior to the
release from a correctional facility or placement in a community based program which
resulted in a violation of probation and a sentence to incarceration during the one year
follow-up period. For example, it is not unusual for a person to be placed on probation
subsequent to a period of incarceration as part of a split sentence or a from and after
probation. Such persons would be counted as recidivists if they violated probation and
were incarcerated during the one year follow-up period. Those violations of probation
that resulted in a non-incarceration disposition (e.g. probation extended or probation
conditions modified) were not used to classify the offender as a probation violator. If an
offender had multiple charges that resulted in an incarceration as a probation violator, the
earliest return as a probation violator was used.
Offenders placed in a residential facility (e.g. detox or in-patient substance abuse facility)
during the one year follow up period were not considered recidivists.
Program Involvement
For those correctional facilities that did not provide type of release, the parole board
information was used to classify the offender as a parolee or a dischargee. For parolees,
the security level of release was derived from the corresponding correctional record
where that was available. Those offenders for whom the Parole Board’s interest in the
case was closed on the same day as the release from the correctional institution were
included in the category of discharges. For these offenders there was no period of parole
supervision following the release.
Offender Background
Demographic Information. The principal source of offender demographic information
was the CARI data-base. Because the CARI data-base did not contain the variable race

58

Comprehensive Recidivism Study

for all offenders, supplemental data collection was done by the DOC and Parole Board to
improve the quality of information available for the analysis.
Current Offense and Criminal History. An optional data element that could be submitted
by the originating criminal justice agency was the docket number(s) associated with the
current placement. If the docket number(s) were submitted this allowed for the current
offense and criminal history to be extracted from the record.
An individual could have one or more docket numbers associated with the criminal
justice placement at the time of the program admission or correctional release. All docket
numbers submitted as part of the originating criminal justice agencies sample were
inspected and linked to the CARI data-base. The docket number associated with the
earliest arraignment date was used to set the parameters for the current offense. For
parolees, the current offense and criminal history were derived from the corresponding
correctional record where that was available.
For youthful offenders, the current offense was contained in the juvenile portion of the
CARI data-base and was assigned an offense seriousness level. Youthful offenders were
not assigned to one of the five criminal history groups and the juvenile criminal history
indicator excluded the adjudication for the current offense.
All current offenses were assigned to an offense seriousness level and type of offense
category in a manner consistent with the method established in the Survey of Sentencing
Practices. Each defendant was assigned to a criminal history group in accordance with
the five criminal history groups proposed by the Commission:
E

Serious Violent Record
Two or more prior convictions for offenses in level 7 through 9

D

Violent or Repetitive Record
One prior conviction for offenses in levels 7 through 9, or
Two or more prior convictions for offenses in levels 5 or 6, or
Six or more prior convictions in levels 3, 4, 5 or 6

C

Serious Record
One prior conviction for offenses at levels 5 or 6, or
Three to five prior convictions for offenses in levels 3 or 4

59

Massachusetts Sentencing Commission

B

Moderate Record
One or two prior convictions for offenses in 3 or 4, or
Six or more prior convictions for offenses in levels 1 or 2

A

No/Minor Record
One to five prior convictions for offenses in levels 1 or 2, or
No prior convictions of any kind

DATA EXTRACTION AND DATA LINKAGES
The Commission used the computing facilities of the Criminal History Systems Board for
data extraction and data linkage procedures relative to the CARI data-base.

DATA ANALYSIS
Missing Information. In all tables, cases with missing information have been reported.
In the narrative, cases with missing information were excluded prior to the calculation of
percentages that describe the characteristics of the population.
State Prison Sentence. The category of state prison sentence included offenders with a
life sentence, a state prison sentence, a state prison / split sentence, and a reformatory
sentence.
Non-Hierarchical Recidivism Classification. The definition of recidivism used for this
analysis was based on the presence of one or more new arrests or re-incarceration as a
parole or probation violator. Many offenders with parole or probation violations also had
one or more new arrests. Table 19 shows the number of offenders classified in each
category separately.
Alternate Definitions of Recidivism.. The definition of recidivism used for this analysis
relied on the presence of one or more new arrests or a re-incarceration due to a violation
of parole or probation during the one year follow-up period. An offender with a new
arrest was considered to be a recidivist even if that new charge had not reached
disposition at the end of the one year follow-up period or if the new charge resulted in a
non-conviction disposition (e.g., dismissed or CWOF) during the one year follow-up
period. In this section, alternate recidivism definitions are considered. Two alternate
recidivism rates were estimated:
60

Comprehensive Recidivism Study

•

•

First, alternate recidivism rates were estimated if only those with charges resulting
in a conviction during the one year follow-up period were considered to be
recidivists; and,
Second, alternate recidivism rates were estimated if only those with charges
resulting in a sentence to incarceration were considered to be recidivists.

In developing these alternate estimates the conviction decision or the sentence to
incarceration had to occur within the one year follow-up period for the offender to be
considered a recidivist.
Table 20 shows estimated recidivism rates using these alternate definitions. In all cases,
those offenders who were returned as the result of a parole or probation violation were
included as recidivists. Offenders who had an incarceration resulting from some other
offense were included in the conviction and incarceration category.
Overall, the recidivism rate, based on arrest / arraignment and including technical
violators, was estimated to be 49.1%. If only arrests that resulted in a conviction during
the one year follow-up period were considered, the estimated recidivism rate would be
30.3%. If only arrests that resulted in a conviction and a sentence of incarceration during
the one year follow-up period were considered, the estimated recidivism rate would be
20.7%.

61

Massachusetts Sentencing Commission

Table 19. Non-hierarchical Recidivism Classification by Criminal Justice Cohort
Sample Total

Drug Co urts

Type of Recidivist

Commu nity
Corrections C enters

House of Correction Sentence
Discharge

Paro le

State Prison Sentence

Sample Total

Discharge

Paro le

Sample Total

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

N

%

None

2,083

55.5%

47

61.8%

196

60.7%

1,126

50.2%

376

61.5%

1,502

52.6%

212

63.7%

126

76.4%

338

67.9%

One or mo re

1,668

44.5%

29

38.2%

127

39.3%

1,117

49.8%

235

38.5%

1,352

47.4%

121

36.3%

39

23.6%

160

32.1%

Total

3,751

100.0%

76

100.0%

323

100.0%

2,243

100.0%

611

100.0%

2,854

100.0%

333

100.0%

165

100.0%

498

100.0%

3,569

95.1%

76

100.0%

322

99.7%

2,243

100.0%

487

79.7%

2,730

95.7%

333

100.0%

108

65.5%

441

88.6%

New Arrest

Parole Violation R eturn
None
One or mo re

182

4.9%

0

0.0%

1

0.3%

0

0.0%

124

20.3%

124

4.3%

0

0.0%

57

34.5%

57

11.4%

Total

3,751

100.0%

76

100.0%

323

100.0%

2,243

100.0%

611

100.0%

2,854

100.0%

333

100.0%

165

100.0%

498

100.0%

3,438

91.7%

54

71.1%

240

74.3%

2,073

92.4%

584

95.6%

2,657

93.1%

327

98.2%

160

97.0%

487

97.8%

Probation V iolation Return
None
One or mo re

313

8.3%

22

28.9%

83

25.7%

170

7.6%

27

4.4%

197

6.9%

6

1.8%

5

3.0%

11

2.2%

Total

3,751

100.0%

76

100.0%

323

100.0%

2,243

100.0%

611

100.0%

2,854

100.0%

333

100.0%

165

100.0%

498

100.0%

62

Comprehensive Recidivism Study

Table 20. Alternate Recidivism Measures by Criminal Justice Cohort
House of Correction Sentence

R ecidivists

R ecidivism R ate

R ecidivists

R ecidivism R ate

R ecidivists

R ecidivism R ate

R ecidivists

R ecidivism R ate

R ecidivists

R ecidivism R ate

Sample Total
(n = 498)

R ecidivism R ate

Paro le
(n = 165)

R ecidivists

Discharge
(n = 333)

R ecidivism R ate

Sample Total
(n = 2854)

R ecidivists

Paro le
(n = 611)

R ecidivism R ate

Discharge
( n = 2243)

State Prison Sentence

R ecidivists

Commu nity
Corrections Centers
(n =323)

R ecidivism R ate

Drug Co urts
(n =76)

R ecidivists

Sample Total
(n =3751)

One or M ore Arrests

1841

49.1%

38

50.0%

168

52.0%

1156

51.5%

288

47.1%

1444

50.6%

121

36.3%

70

42.4%

191

38.4%

One or M ore Convictions

1136

30.3%

28

36.8%

116

35.9%

673

30.0%

206

33.7%

879

30.8%

52

15.6%

61

37.0%

113

22.7%

One or M ore Incarcerations

777

20.7%

23

30.3%

89

27.6%

419

18.7%

164

26.8%

583

20.4%

25

7.5%

57

34.5%

82

16.5%

Time Until Recidivism Incident

Note: Offenders incarcerated as a result of a violation of parole or probation or some other offense were included in all categories.

63

Massachusetts Sentencing Commission

The estimated recidivism rates (i.e. any new arrest, only convicted behavior, or only
incarcerated behavior) were sensitive to the time during the follow-up period that the
recidivism behavior occurred. Table 21 shows the relationships between these variables.
As expected, the probability of conviction within the one year follow-up period was
related to the time during the follow-up period when the recidivism behavior occurred.
For those offenders classified as a recidivist in the first month, 78.9% had one or more
convictions within the one year follow-up period. In contrast, for those offenders
classified as a recidivist in month eight, 44.8% had one or more convictions within the
one year follow-up period. In general, because a disposition decision in any case takes
some time, the use of a conviction definition biased recidivism rate estimates towards
those populations with recidivism behavior occurring more quickly. Because parole and
probation violators were included in all of the alternate recidivism measures, the relation
between the two measures was close for the parole and community based program
cohorts.

64

Comprehensive Recidivism Study

Table 21. Conviction Status by Time Until Recidivism and Criminal Justice Cohort
House of Correction Sentence

96.8%

178

140 78.7%

8

5

62.5%

%

30

All C onvictions

31

All Arrests

110 74.8%

%

147

All C onvictions

77.8%

Sample Total

All Arrests

All Arrests

21

%

%

27

Paro le

All C onvictions

All C onvictions

100.0%

Discharge

All Arrests

All Arrests

4

%

%

4

All Arrests

All C onvictions

78.9%

%

All Arrests

176

All C onvictions

%

223

Sample Total

All Arrests

All C onvictions

Less than 1 m onth

Paro le

%

Time U ntil
Recidivism

All Arrests

Discharge

State Prison Sentence

All C onvictions

Commu nity
Corrections C enters

Drug Co urts

All C onvictions

Sample Total

6

6

100.0%

14

11

78.6%

1 to 2 months

219

157

71.7%

9

7

77.8%

23

19

82.6%

133

83

62.4%

38

34

89.5%

171

117 68.4%

11

9

81.8%

5

5

100.0%

16

14

87.5%

2 to 3 months

215

163

75.8%

7

3

42.9%

19

18

94.7%

138

100 72.5%

26

22

84.6%

164

122 74.4%

17

12

70.6%

9

9

100.0%

26

21

80.8%

3 to 4 months

187

135

72.2%

2

2

100.0%

21

16

76.2%

113

79

69.9%

29

22

75.9%

142

101 71.1%

12

6

50.0%

10

10

100.0%

22

16

72.7%

4 to 5 months

196

128

65.3%

3

2

66.7%

14

10

71.4%

131

85

64.9%

25

20

80.0%

156

105 67.3%

15

4

26.7%

7

6

85.7%

22

10

45.5%

5 to 6 months

148

99

66.9%

3

3

100.0%

15

9

60.0%

98

63

64.3%

19

17

89.5%

117

80

9

3

33.3%

4

4

100.0%

13

7

53.8%

68.4%

6 to 7 months

128

77

60.2%

2

2

100.0%

10

8

80.0%

80

46

57.5%

19

11

57.9%

99

57

57.6%

11

5

45.5%

6

5

83.3%

17

10

58.8%

7 to 8 months

116

61

52.6%

2

1

50.0%

7

6

85.7%

69

33

47.8%

24

15

62.5%

93

48

51.6%

11

3

27.3%

3

3

100.0%

14

6

42.9%

8 to 9 months

116

52

44.8%

3

3

100.0%

11

4

36.4%

70

31

44.3%

21

9

42.9%

91

40

44.0%

7

3

42.9%

4

2

50.0%

11

5

45.5%

9 to 10 months

103

43

41.7%

1

1

100.0%

9

2

22.2%

67

24

35.8%

14

10

71.4%

81

34

42.0%

7

2

28.6%

5

4

80.0%

12

6

50.0%

10 to 11 months

88

32

36.4%

0

0

N. A .

8

3

37.5%

43

16

37.2%

24

9

37.5%

67

25

37.3%

7

0

0.0%

7

5

71.4%

14

5

35.7%

11 to 12 months

102

13

12.7%

2

0

0.0%

4

0

0.0%

67

3

4.5%

18

7

38.9%

85

10

11.8%

6

0

0.0%

4

2

50.0%

10

2

20.0%

1,841 1136 61.7%

38

28

73.7%

168

288

206

71.5%

1444

121

52

43.0%

70

61

87.1%

191

113

59.2%

Total

116 69.0% 1,156

673 58.2%

65

879 60.9%

Massachusetts Sentencing Commission

APPENDIX - HOUSE OF CORRECTION SENTENCED OFFENDERS
For reference purposes, Table 22 shows recidivism rates and associated offender
characteristics for house of correction sentenced offenders by type of release and the
correctional jurisdiction of release:
•
•
•
•

house of correction
house of correction
house of correction
house of correction

sentence / discharged / released from county facility;
sentence / paroled / released from county facility;
sentence / discharged / released from DOC facility;
sentence / paroled / released from DOC facility.

66

Comprehensive Recidivism Study

Table 22. Selected Recidivism Characteristics by Type of Release and Custody at Time of Release, House of Correction
Sentenced Offenders
House Sentence / Released from House
Recidivism
Char acter istic

Discharge

Paro le

H o u se S e nt e nc e / Re l ea s e d f r om D O C

Sample Total

Discharge

Paro le

Sample Total

N

%

N

%

N

%

N

%

N

%

N

%

Recidiv ism Ra te

1,082

51.7%

254

46.4%

1,336

50.6%

74

49.3%

34

54.0%

108

50.7%

One or M ore Arrests

1,049

50.1%

206

37.6%

1,255

47.5%

68

45.3%

29

46.0%

97

45.5%

Type of Recidivist

Most serious:
felony

507

24.2%

97

17.7%

604

22.9%

23

15.3%

7

11.1%

30

14.1%

misdemeanor

542

25.9%

109

19.9%

651

24.6%

45

30.0%

22

34.9%

67

31.5%

No Arrests / Technical

33

1.6%

48

8.8%

81

3.1%

6

4.0%

5

7.9%

11

5.2%

Non-Recidivist

1,011

48.3%

294

53.6%

1,305

49.4%

76

50.7%

29

46.0%

105

49.3%

Total

2,093

100.0%

548

100.0%

2,641

100.0%

150

100.0%

63

100.0%

213

100.0%

New
Cases
134

Cum
Rate
6.4%

New
Cases
29

Cum
Rate
5.3%

New
Cases
163

Cum
Rate
6.2%

New
Cases
13

Cum
Rate
8.7%

New
Cases
2

Cum
Rate
3.2%

New
Cases
15

Cum
Rate
7.0%

Time Until Recidivism Incident
Less than 1 m onth
1 to 2 months

125

12.4%

35

11.7%

160

12.2%

8

14.0%

3

7.9%

11

12.2%

2 to 3 months

134

18.8%

19

15.1%

153

18.0%

4

16.7%

7

19.0%

11

17.4%

3 to 4 months

102

23.7%

25

19.7%

127

22.8%

11

24.0%

4

25.4%

15

24.4%

4 to 5 months

122

29.5%

22

23.7%

144

28.3%

9

30.0%

3

30.2%

12

30.0%

5 to 6 months

94

34.0%

14

26.3%

108

32.4%

4

32.7%

5

38.1%

9

34.3%

6 to 7 months

75

37.6%

17

29.4%

92

35.9%

5

36.0%

2

41.3%

7

37.6%

7 to 8 months

65

40.7%

22

33.4%

87

39.2%

4

38.7%

2

44.4%

6

40.4%

8 to 9 months

67

43.9%

19

36.9%

86

42.4%

3

40.7%

2

47.6%

5

42.7%

9 to 10 months

59

46.7%

12

39.1%

71

45.1%

8

46.0%

2

50.8%

10

47.4%

10 to 11 months

42

48.7%

23

43.2%

65

47.6%

1

46.7%

1

52.4%

2

48.4%

11 to 12 months

63

51.7%

17

46.4%

80

50.6%

4

49.3%

1

54.0%

5

50.7%

67

Massachusetts Sentencing Commission
House Sentence / Released from House
Recidivism
Char acter istic

Discharge
N

%

Paro le
N

H o u se S e nt e nc e / Re l ea s e d f r om D O C

Sample Total
%

N

%

Discharge
N

%

Paro le
N

Sample Total
%

N

%

Offense Seriousness Level, Most
Serious New Arrest Offense
9

1

0.0%

0

0.0%

1

0.0%

0

0.0%

0

0.0%

0

0.0%

8

4

0.2%

0

0.0%

4

0.2%

0

0.0%

0

0.0%

0

0.0%

7

6

0.3%

0

0.0%

6

0.2%

0

0.0%

0

0.0%

0

0.0%

6

23

1.1%

1

0.2%

24

0.9%

0

0.0%

0

0.0%

0

0.0%

5

46

2.2%

13

2.4%

59

2.2%

1

0.7%

0

0.0%

1

0.5%

4

157

7.5%

16

2.9%

173

6.6%

6

4.0%

1

1.6%

7

3.3%

3

419

20.0%

84

15.3%

503

19.0%

30

20.0%

6

9.5%

36

16.9%

2

239

11.4%

52

9.5%

291

11.0%

18

12.0%

15

23.8%

33

15.5%

1

153

7.3%

38

6.9%

191

7.2%

13

8.7%

7

11.1%

20

9.4%

Not Assigned

1

0.0%

2

0.4%

3

0.1%

0

0.0%

0

0.0%

0

0.0%

Total

1,049

50.1%

206

37.6%

1,255

47.5%

68

45.3%

29

46.0%

97

45.5%

Type of Offense, M ost Serious
New Arrest Offense
Person

304

29.0%

41

19.9%

345

27.5%

22

32.4%

2

6.9%

24

24.7%

Sex

17

1.6%

3

1.5%

20

1.6%

0

0.0%

0

0.0%

0

0.0%

Property

249

23.7%

68

33.0%

317

25.3%

18

26.5%

9

31.0%

27

27.8%

Drug

198

18.9%

33

16.0%

231

18.4%

8

11.8%

7

24.1%

15

15.5%

Mo tor V ehicle

172

16.4%

39

18.9%

211

16.8%

11

16.2%

5

17.2%

16

16.5%

Other

109

10.4%

22

10.7%

131

10.4%

9

13.2%

6

20.7%

15

15.5%

Total

1,049

100.0%

206

100.0%

1,255

100.0%

68

100.0%

29

100.0%

97

100.0%

68

Comprehensive Recidivism Study

Table 23. Selected Offender Characteristics by Type of Release and Custody at Time of Release, House of Correction
Sentenced Offenders
House Sentence / Released from House

H o u se S e nt e nc e / Re l ea s e d f r om D O C

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

Maximum
Medium
Medium/M inimum
Sub-Total Se cure

127
597
67
791

61
352
35
448

48.0%
59.0%
52.2%
56.6%

2
67
15
84

1
38
9
48

50.0%
56.7%
60.0%
57.1%

129
664
82
875

62
390
44
496

48.1%
58.7%
53.7%
56.7%

0
110
0
110

0
60
0
60

N.A.
54.5%
N.A.
54.5%

0
30
0
30

0
17
0
17

N.A.
56.7%
N.A.
56.7%

0
140
0
140

0
77
0
77

N.A.
55.0%
N.A.
55.0%

Minimum
Minimum/Pre-release
Pre-release
Sub-Total Lower

293
0
72
365

125
0
31
156

42.7%
N.A.
43.1%
42.7%

72
0
51
123

31
0
20
51

43.1%
N.A.
39.2%
41.5%

365
0
123
488

156
0
51
207

42.7%
N.A.
41.5%
42.4%

4
32
4
40

0
11
3
14

0.0%
34.4%
75.0%
35.0%

18
13
2
33

10
6
1
17

55.6%
46.2%
50.0%
51.5%

22
45
6
73

10
17
4
31

45.5%
37.8%
66.7%
42.5%

Not Provided

937

478

51.0%

341

155

45.5%

1278

633

49.5%

0

0

N.A.

0

0

N.A.

0

0

N.A.

Total

2093

1082

51.7%

548

254

46.4%

2641

1336

50.6%

150

74

49.3%

63

34

54.0%

213

108

50.7%

Gender
Fema le
Ma le
Total

132
1961
2093

58
1024
1082

43.9%
52.2%
51.7%

65
483
548

32
222
254

49.2%
46.0%
46.4%

197
2444
2641

90
1246
1336

45.7%
51.0%
50.6%

127
23
150

67
7
74

52.8%
30.4%
49.3%

38
25
63

20
14
34

52.6%
56.0%
54.0%

165
48
213

87
21
108

52.7%
43.8%
50.7%

Offend er Ch aract eristic

Security Level at Release

69

Massachusetts Sentencing Commission
House Sentence / Released from House

H o u se S e nt e nc e / Re l ea s e d f r om D O C

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

Age at Release
Under 20
20 to 24
25 to 29
30 to 34
35 to 39
40 to 49
50 to 59
60 to 69
70 to 79
Total

149
413
358
336
375
374
74
13
1
2093

88
243
197
183
185
160
23
3
0
1082

59.1%
58.8%
55.0%
54.5%
49.3%
42.8%
31.1%
23.1%
0.0%
51.7%

65
119
108
81
83
72
16
4
0
548

37
60
51
41
39
22
3
1
0
254

56.9%
50.4%
47.2%
50.6%
47.0%
30.6%
18.8%
25.0%
N.A.
46.4%

214
532
466
417
458
446
90
17
1
2641

125
303
248
224
224
182
26
4
0
1336

58.4%
57.0%
53.2%
53.7%
48.9%
40.8%
28.9%
23.5%
0.0%
50.6%

6
13
27
34
34
25
11
0
0
150

4
9
17
17
17
9
1
0
0
74

66.7%
69.2%
63.0%
50.0%
50.0%
36.0%
9.1%
N.A.
N.A.
49.3%

3
13
12
10
14
10
1
0
0
63

3
7
7
4
11
2
0
0
0
34

100.0%
53.8%
58.3%
40.0%
78.6%
20.0%
0.0%
N.A.
N.A.
54.0%

9
26
39
44
48
35
12
0
0
213

7
16
24
21
28
11
1
0
0
108

77.8%
61.5%
61.5%
47.7%
58.3%
31.4%
8.3%
N.A.
N.A.
50.7%

Race / E thnicity
White
Black
Hispa nic
Other
U n k n ow n
Total

1029
544
438
43
39
2093

478
323
252
16
13
1082

46.5%
59.4%
57.5%
37.2%
33.3%
51.7%

361
87
90
10
0
548

159
41
51
3
0
254

44.0%
47.1%
56.7%
30.0%
N.A.
46.4%

1390
631
528
53
39
2641

637
364
303
19
13
1336

45.8%
57.7%
57.4%
35.8%
33.3%
50.6%

110
21
17
2
0
150

54
7
12
1
0
74

49.1%
33.3%
70.6%
50.0%
N.A.
49.3%

53
4
4
2
0
63

27
4
2
1
0
34

50.9%
100.0%
50.0%
50.0%
N.A.
54.0%

163
25
21
4
0
213

81
11
14
2
0
108

49.7%
44.0%
66.7%
50.0%
N.A.
50.7%

Type of Offense (Current Offense)
Person
Sex
Property
Drug
Mo tor V ehicle
Other
Not A vailable
Total

338
22
251
307
198
97
880
2093

184
4
143
165
80
54
452
1082

54.7%
18.2%
55.7%
54.1%
39.4%
56.3%
51.4%
51.7%

108
3
137
176
102
22
0
548

54
2
77
80
29
12
0
254

50.9%
66.7%
57.2%
45.9%
27.3%
58.6%
N.A.
46.4%

446
25
388
483
300
119
880
2641

238
6
220
245
109
66
452
1336

53.8%
24.0%
56.3%
51.0%
35.5%
56.7%
51.4%
50.6%

35
0
36
26
38
15
0
150

20
0
17
15
13
9
0
74

38.8%
20.0%
51.6%
25.8%
37.5%
46.2%
N.A.
49.3%

6
0
22
20
8
7
0
63

4
0
14
10
1
5
0
34

42.3%
14.3%
62.1%
30.8%
100.0%
80.0%
N.A.
54.0%

41
0
58
46
46
22
0
213

24
0
31
25
14
14
0
108

40.0%
19.2%
54.9%
28.0%
44.4%
55.6%
N.A.
50.7%

Offend er Ch aract eristic

70

Comprehensive Recidivism Study
House Sentence / Released from House

H o u se S e nt e nc e / Re l ea s e d f r om D O C

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Total

R ecidivists

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

R ecidivism R ate

Sample Total

R ecidivists

Paro le

Total

Discharge

Offense Seriousness Level (Current Offense)
9
8
7
6
5
4
3
2
1
Not Assigned
Not A vailable
Total

0
0
0
0
23
173
505
356
151
5
880
2093

0
0
0
0
10
84
268
178
88
2
452
1082

N.A.
N.A.
N.A.
N.A.
43.5%
48.6%
53.1%
50.0%
58.3%
40.0%
51.4%
51.7%

0
0
0
2
3
132
231
133
45
2
0
548

0
0
0
0
2
56
113
60
22
1
0
254

N.A.
N.A.
N.A.
0.0%
66.7%
42.4%
48.9%
45.1%
48.9%
50.0%
N.A.
46.4%

0
0
0
2
26
305
736
489
196
7
880
2641

0
0
0
0
12
140
381
238
110
3
452
1336

N.A.
N.A.
N.A.
0.0%
46.2%
45.9%
51.8%
48.7%
56.1%
42.9%
51.4%
50.6%

0
0
0
0
0
14
74
49
10
3
0
150

0
0
0
0
0
7
35
24
7
1
0
74

N.A.
N.A.
N.A.
N.A.
N.A.
50.0%
47.3%
49.0%
70.0%
33.3%
N.A.
49.3%

0
0
0
0
0
11
24
22
5
1
0
63

0
0
0
0
0
4
16
12
2
0
0
34

N.A.
N.A.
N.A.
N.A.
N.A.
36.4%
66.7%
54.5%
40.0%
0.0%
N.A.
54.0%

0
0
0
0
0
25
98
71
15
4
0
213

0
0
0
0
0
11
51
36
9
1
0
108

N.A.
N.A.
N.A.
N.A.
N.A.
44.0%
52.0%
50.7%
60.0%
25.0%
N.A.
50.7%

Criminal History Group
Group A - No/M inor Record
Group B - Mode rate Reco rd
Group C - Serious Re cord
Group D - V iolent or Repetitive
Group E - Serious V iolent
Not Assigned or Youthful Offender
Total

284
342
299
285
1
882
2093

110
167
170
181
1
453
1082

38.7%
48.8%
56.9%
63.5%
100.0%
51.4%
51.7%

50
87
52
21
0
338
548

19
43
25
13
0
154
254

38.0%
49.4%
48.1%
61.9%
N.A.
45.6%
46.4%

334
429
351
306
1
1220
2641

129
210
195
194
1
607
1336

38.6%
49.0%
55.6%
63.4%
100.0%
49.8%
50.6%

53
62
22
13
0
0
150

21
33
13
7
0
0
74

39.6%
53.2%
59.1%
53.8%
N.A.
N.A.
49.3%

15
27
15
6
0
0
63

8
16
5
5
0
0
34

53.3%
59.3%
33.3%
83.3%
N.A.
N.A.
54.0%

68
89
37
19
0
0
213

29
49
18
12
0
0
108

42.6%
55.1%
48.6%
63.2%
N.A.
N.A.
50.7%

Juvenile Criminal History
No Prior Adjudications
One or M ore Adjudications
Total

1579
514
2093

757
325
1082

47.9%
63.2%
51.7%

423
125
548

179
75
254

42.3%
60.0%
46.4%

2002
639
2641

936
400
1336

46.8%
62.6%
50.6%

139
11
150

68
6
74

48.9%
54.5%
49.3%

54
9
63

28
6
34

51.9%
66.7%
54.0%

193
20
213

96
12
108

49.7%
60.0%
50.7%

Offend er Ch aract eristic

71