Using a Multistate Perspective to Examine Recidivism Outcomes

By Matthew Durose & Joshua Markman, Bureau of Justice Statistics

The authors would like to acknowledge April Trotter, Writer/Editor, Lockheed Martin and Anastasios (Tom) Tsoutis, Chief of the Recidivism Statistics Unit, Bureau of Justice Statistics for helping to edit the article.

Recidivism outcomes are a key criminal justice statistic. Historically, collecting and processing multistate criminal history records into a common coding structure suitable for statistical analysis has been time consuming and labor intensive. In recent years, the Bureau of Justice Statistics (BJS) has worked with the International Justice and Public Safety Network (Nlets) and NORC at the University of Chicago to build a data-processing infrastructure to improve the methods for converting large samples of multistate criminal history records into nationally standardized research databases.1

BJS partnered with Nlets to develop a new, automated data collection process that reduces the time needed to collect and consolidate individual state and federal criminal history records into a uniform file layout. In addition, BJS worked with NORC to develop software that converts the state- and federal-specific fields from the criminal history records (e.g., offense statutes and disposition codes) into a common coding structure that supports national-level criminal history research

BJS's efforts to automate and standardize the criminal history data collection process enables the use of larger samples of criminal history records to study recidivism patterns in greater depth. In addition to disseminating information that provides key insight on recidivism, BJS's data analyses also support efforts to improve the accuracy and comprehensiveness of criminal history records. These analyses can inform state repositories and other criminal justice agencies about data quality issues in their criminal history records.

Recidivism rates of former prisoners

BJS's recent data collection of the criminal histories of released prisoners was the first to utilize the FBI's Interstate Identification Index (III) to directly access the criminal history record systems of all 50 states.2 BJS received approval from the FBI to conduct this record collection. The data collection captured the criminal histories of prisoners released in 2005 in 30 states, which was twice the number of states included in the previous BJS recidivism study on prisoners released in 1994. These 30 states were responsible for about three-quarters of all state prisoners released in 2005 nationwide. Such expanded analyses are important given the complexities and nuances of examining recidivism.

A total of 68,597 released prisoners were randomly selected to represent the 404,638 inmates released in 30 states in 2005. States were included in this study due to their ability to provide the necessary prisoner records through the BJS National Corrections Reporting Program. While prior BJS recidivism studies tracked inmates for three years following release, this study used a five-year follow-up period to provide a more complete assessment of prisoners' post-release recidivism outcomes. To measure the percentage of released prisoners arrested, convicted, or returned to prison within five years, the collection of multistate criminal history records through III via Nlets was conducted in 2011. The study used several outcome measures to allow researchers to consider recidivism from different perspectives. This data collection will support a series of reports addressing a wide range of issues on recidivism, criminal careers, and the cross-jurisdictional mobility of criminal offenders.

The first publication from this data collection, Recidivism of Prisoners Released in 30 States in 2005: Patterns from 2005 to 2010, was published in 2014.3 It summarized the offending patterns of the released prisoners by offender characteristics, prior criminal history, and commitment offense. Overall, about two-thirds (68%) of the prisoners released in 30 states in 2005 were arrested for a new crime within three years, and three-quarters (77%) were arrested within five years. Among prisoners released in 2005 in 23 states (with available data on inmates returned to prison), about half (50%) had either a parole or probation violation or an arrest for a new crime within three years that led to imprisonment. Within five years of release, more than half (55%) returned to prison for a parole or probation violation or an arrest for a new crime.

The recidivism rates of prisoners released in 2005 varied with offender attributes, commitment offense, and criminal history. At the end of the five-year follow-up period, 61% of former inmates with four or fewer arrests in their prior criminal history were arrested, compared to 86% of those who had 10 or more prior arrests. Prisoners who served time for a property offense (82%) were more likely to be arrested within five years of release than drug offenders (77%), public order offenders (74%), and violent offenders (71%). Male prisoners (78%) were more likely than female prisoners (68%) to be arrested for a new crime. Additionally, 84% of prisoners who were age 24 or younger at release were arrested within five years of release, compared to 79% of inmates ages 25 to 39 and 69% of those ages 40 or older.

Multistate criminal history patterns

A second publication from this data collection, Multistate Criminal History Patterns of Prisoners Released in 30 States, was released in September 2015.4 It examined the extent to which the former prisoners were arrested in states other than the one where they had served time. Among the prisoners released in 30 states in 2005, an estimated 25% had at least one arrest in another state prior to their release in 2005. This rate ranged from 17% to 57% across the study's 30 states. Among the released prisoners who had prior arrests in more than one state, 72% had arrests in two states, while 18% had arrests in three states, and 9% had arrests in four or more states.

Within five years of their release, an estimated 11% of the prisoners released in 30 states during 2005 were arrested at least once in another state. The rate at which these prisoners were arrested in another state within five years varied by the state of release from 6% to 26% (see figure 1).

Figure 1. Percent of prisoners released in 30 states in 2005 who were arrested in another state within five years, by state of release

This report also examined the differences in prisoner recidivism rates based on multistate (i.e., national) criminal history records compared to those rates based only on criminal history records from within the state that released the prisoner. The findings allow those who track recidivism patterns solely within a particular state, to gauge the extent to which their prison populations commit crimes in other states and to which their recidivism statistics underrepresent the actual recidivism rates.

For instance, 77% of the released prisoners were arrested either within or outside the state of release after five years. However, when examining arrests during the five-year follow-up period within the state of release only, an estimated 72% of prisoners were arrested. During the follow-up period, 5% of the prisoners had only out-of-state arrests and were not arrested within the state that released them. The difference of in-state recidivism rates when out-of-state arrests were included shows the limitations of recidivism studies that have access only to in-state criminal history information. This illustrates the value to recidivism research of access to out-of-state criminal history information.

Disclaimer: The analysis and conclusions presented here are those of the authors and do not necessarily represent the views and opinions of the Bureau of Justice Statistics or the U.S. Department of Justice.

1Durose, M. Standardized Approach to Collecting and Processing Multistate Criminal History Information for Statistical Analysis, JRSA Forum, Vol. 28, No. 2 (June 2010).

2The FBI's III is an automated pointer system that allows authorized agencies to determine whether any state repository has criminal history records on an individual. Nlets is a computer-based network that is responsible for the interstate transmissions of federal and state criminal history records. It allows users to query III and send requests to states holding criminal history records on an individual.

3Durose, M, Cooper, A, and Snyder, H. Recidivism of Prisoners Released in 30 States in 2005: Patterns from 2005 to 2010, Bureau of Justice Statistics (April 2014).

4Durose, M, Snyder, H, and Cooper, A. Multistate Criminal History Patterns of Prisoners Released in 30 States, Bureau of Justice Statistics (September 2015).