Work Releases Reduce Unemployment for Aboriginal Offenders

Key Words

work release, Aboriginal offenders, community reintegration, dosage effects employment.

What it means

Among Aboriginal offenders, participating in a work release (WR) contributes positively to community reintegration. Participating in at least one WR was associated with reductions in unemployment post-release, and the more WRs the offender participated in, the greater the benefits.

What we found

Approximately 2.5% of Aboriginal offenders participated in a WR during their sentence. Overall, 81% of Aboriginal offenders who had a WR found employment in the community within 2 years of release, compared to 47% of Aboriginal offenders who did not have a WR. After controlling for risk-relevant differences between the two groups, participating in a WR significantly reduced the odds of unemployment by 64%. There was also a significant dosage effect: the more work releases the offender participated in, the greater the benefits in terms of post-release employment.

Why we did this study

WRs allow offenders to leave the institution for designated periods of time to obtain work experience in the community. The objective of this study was to determine if participating in a WR improved employment outcomes in the community. Given the cost associated with administering WRs and the potential risk to public safety inherent in allowing inmates to have conditional access to the community, it is necessary to determine whether WRs assist in the community reintegration process.

What we did

The purpose of the study was to examine the impact of participating in WRs on subsequent employment post-release. The final sample included 5,039 Aboriginal offenders released to the community between April 1, 2005 and March 31, 2011.

The outcome was whether the offender obtained any type of employment (full-time or part-time) within two years of their first release from their federal sentence during the study period.

Using variables that predicted WR participation among Aboriginal offenders, a propensity score was calculated using logistic regression to create one value summarizing differences in who received a WR and who did not (Table 1). For analyses of WR effectiveness controlling for this propensity score, the sample size was reduced to 4,458 because not all offenders had sufficient data to calculate the propensity score.

Analyses of the effectiveness of WR participation used logistic regression and controlled for the propensity score. Analyses of dosage effects examining the impact of the number of work releases the offender participated in, after controlling for the dichotomous effect of any WR participation, and propensity scores.

Variables used to create propensity score
  • Current violent offence
  • Reintegration Potential Rating
  • Security risk score on the Custody Rating Scale
  • Institutional adjustment score on the Custody Rating Scale
  • Static Factors risk assessment (moderate risk vs. other risk levels)
  • Previous segregation for disciplinary infractions
  • Criminal History subscale score on the Static Factors Assessment
  • Dynamic Factors risk assessment (moderate need vs. other need levels)
  • Previous youth offences
  • Sentence lengthFootnote 1
  • Prior sex offence
  • Motivation rating

For more information

Please e-mail the Research Branch or contact us by phone at (613) 995-3975.
You can also visit the Research Publications section for a full list of reports and one-page summaries.


Footnote 1

For sentence length, lifers were scored as having a sentence length of 31 years as the longest determinate sentence length was 30 years.

Return to footnote 1