Introduction
Recidivism is a key measure of the effectiveness of the criminal justice system in meeting the goals of protecting public safety, evaluating the effectiveness of rehabilitative programs, and maximizing the efficient use of public resources. The 2018 Criminal Justice Reform Bill (CJR) requires the Secretary of the Executive Office of Public Safety and Security (EOPSS) to report out on recidivism rates across criminal justice agencies in the Commonwealth tracked over one-, two-, and three- year periods. Specifically, EOPSS is to report on reconviction, reincarceration, and rearraignment and to allow the user to categorize the data by race/ethnicity, gender, and age.
The linked dashboards create for the first time a comprehensive view of recidivism across all custodial agencies displaying data from 2017 through the present. Additionally, the dashboards display the three distinct measures of recidivism: reconviction; reincarceration; and rearraignment (collectively referred to as "recidivistic event(s)") that occurs during the first three years of the releasee's return to the community. Several criminal justice agencies collect and report recidivism rates outside of this initiative using their own definitions and metrics unique to their agencies' needs. Through Cross-Tracking, for the first time, all criminal justice agencies across the Commonwealth are utilizing standardized definitions and follow-up periods to ensure consistency in recidivism reporting.
About the Information
The dashboards display all Massachusetts criminally sentenced individuals released to the community from a custodial agency, including those released by court, those released at the end of a sentence, and those released with or without supervision. Those released from a custodial agency by virtue of escape, death, or transfer to another custodial agency are not included as they are not released to the community. All data for individuals released to the community is tracked for one-, two-, and three-year periods to determine if a recidivistic event occurred. The dashboards capture the first recidivistic event following an individual’s release.
The recidivism data displayed is created using several systems across the contributing partners, including, but not limited to:
-
The Department of Correction (DOC) Inmate Management System (IMS)
-
The County Sheriffs’ 12 independent Offender Management Systems (OMS) and Hampden County’s Jail Management System (JMS)
-
The Trial Court and Probation Department’s MassCourts System (including the BOP)
-
The Parole Board’s Record Management System (SPIRIT II)
-
State Police State Identification Section (OneIII)
Given that data submitted by the participating criminal justice agencies predates the CJR data reporting standards outlined in the new governing regulations, EOPSS has taken great care to review and match the records from the various agencies. A conservative method of matching records was used in developing the dashboard to ensure that recidivistic events were not missed. All releases in the dashboard had an exact match between the last name, first name, and date of birth as recorded in the custodial agency and the court record. Release records that did not meet these matching criteria were excluded from the dashboard.
As a user, you can hover over data points to see more detailed information. You can also download the data behind any visualization through a link at the bottom of each page. For all dashboards, users can filter by agency (individual Department of Correction institutions or individual Sheriff's Offices) or groups of agencies (Department of Correction or Sheriff's Offices by county). Filters for sex, age, and race/ethnicity are also available.
For more information about the dashboard, you can access FAQs here.
For any other questions, please write to CJ360help@mass.gov
Models Explained
Through discussions with stakeholders and development of the dashboard, it was determined that two recidivism models were necessary to display the data effectively: the CMR Model and the Query Model. The differences are discussed below but the main difference is that the CMR Model begins tracking recidivism at the time Parole supervision (if any) is complete, while the Query Model begins tracking the moment an individual is released from custody whether or not they are subject to Parole supervision.
CMR Model
The CMR model is based off of 501 CMR 18, Data Collection and Reporting Standards, promulgated in December of 2021.
As defined by 501 CMR 18, the three measures of recidivism are reconviction, reincarceration, and rearraignment:
-
"Reconviction" is defined as the date when an Offender is convicted of an offense within one, two or three years of their Release Date.
-
"Reincarceration" is defined as the date when an Offender is recommitted to custody within one, two or three years of their Release Date.
-
"Rearraignment" is defined as the date when an Offender is arraigned for an offense within one, two or three years of their Release Date.
“Release Date” is defined in 501 CMR 18 as the calculated date on which an Offender will be released from the Offender’s sentence, based on calculations of a sentence’s terms without accounting for sentence deductions that an Offender may earn while serving the Offender’s sentence. Essentially, this is the date on which an Offender will be released that is calculated at the time the individual is committed to a custodial agency. It does not take into consideration any earned credits or deductions and is not necessarily, in fact oftentimes it is not, the date the Offender is actually released from custody. 501 CMR 18 accounted for this by defining “Actual Release from Sentence Date” as:
The date on which an Offender is released from the custody of the Department of Correction or Sheriff's Office,
(a) having completed the Offender's entire sentence, including any parole supervision, based upon an initial calculation of the Offender's Release Date minus any sentence deductions such as Earned Time; or
(b) having been released from the Offender's sentence by post-sentencing action of a court (e.g., revise and revoke, sentence or conviction vacated, conviction overturned by appellate court with no retrial, reduction in sentence terms by Appellate Division or new trial).
Actual Release from Sentence Date shall not apply to pre-trial detainees.
At the time of developing these dashboards, EOPSS, along with its stakeholders, decided to apply the recidivism terms (reconviction, reincarceration, and rearraignment) to its definition of Actual Release from Sentence Date. As such, for purposes of the CMR Model, the beginning point of tracking an individual for any recidivistic event (the “base date”) begins at the date an Offender is released from custodial supervision, including parole supervision, and is followed for three years.
Areas to note about this model:
-
Individuals who leave the custody of the Department of Correction or Sheriff's Office to the community under parole supervision are not included until they are discharged from parole supervision. Any arraignment, conviction, or incarceration that occurs while under parole supervision is not counted as a recidivistic event. In contrast, individuals who leave the custody of the Department of Correction or Sheriff's Office with probation supervision or with no post-release supervision are included in the dashboard and any subsequent arraignment, conviction, or incarceration is counted as a recidivistic event;
-
Instances where an individual is subsequently admitted to a custodial agency on a pre-trial status that is later deemed “time served” because they are subsequently convicted and sentenced to incarceration are included in the definition of reincarceration; and
-
Age of Offender is calculated at the date the Offender is released from custody, including completion of any parole supervision.
Query Model
The Query Model begins to track an individual for a recidivistic event (the “base date”) beginning on the date that the Offender is released from custody of the Department of Correction or Sheriff's Office to the community, whether or not the Offender was under parole supervision. The base date for the Query Model is referred to below as “Date Exiting Custody.” For those not released to parole supervision, the results of the CMR Model and the Query Model will be identical; however, EOPSS and its stakeholders determined that the Query Model may display more recidivistic events and was consistent with the manner in which some criminal justice agencies tracked recidivism in the Commonwealth prior to CJR. As such, this model displays any recidivistic event that occurs when an individual physically leaves the custody of the DOC or Sheriff's Office whether or not they are subject to parole supervision.
Areas to note about this model:
-
Individuals who are released from a custodial agency to Parole supervision who subsequently have their parole revoked and are re-committed to custody will be included in this dashboard;
-
Age of Offender is calculated at the date they are released from custody, whether or not they are subject to Parole supervision; and
-
In comparison to the CMR Model, the number of recidivistic events will change due to the change in the date from which recidivism is tracked.
Additional notes on the methodology used in developing recidivism estimates can be found in this document [FAQ]
Recidivism Model Results
Downloads
Revision History
| Details | Date |
|---|---|
| Updated On | 12/01/2023 |
| Published On | 06/30/2023 |