The recidivism of federal offenders

The recidivism of federal offenders PDF Version (10KB)

Research summary
Vol. 8 No. 4
July 2003


After release from federal penitentiaries, how many offenders re-offend?


One of the most frequently asked questions about any prison system is the recidivism of released offenders. This question is often asked of the Canadian federal prison system, which is responsible for offenders serving sentences from two years to life. Knowledge of the recidivism rate of released inmates is important because it is one of many indicators of success of a prison system's attempt to reintegrate offenders safely into the community.

"Recidivism" can be measured in different ways for different purposes. There is no single measure of recidivism that does not have a disadvantage. The various measures that have been used (e.g., re-arrests, reincarceration) all have shortcomings but also certain advantages that justify their continued use. Bearing this in mind the need to communicate to the public an estimate of the re-offending rate of federal offenders that is easily understood, the measure of recidivism chosen for this study was any reconviction during a fixed follow-up period after release.


Reconviction, as a measure of recidivism, has a number of advantages over other measures. First, compared to re-arrest, reconviction requires a plea or finding of guilt in court. Thus, it minimizes the likelihood that someone will be counted as committing a new crime when in fact the person did not. Second, this definition of reconviction includes the full range of crimes from the least to the most serious (where as reincarceration would normally indicate only the more serious offences). The Royal Canadian Mounted Police's Finger Print Service, arguably the most comprehensive and reliable data source available, provided reconviction data.

Reconviction rates were calculated for all federal offenders released during the year from April 1, 1994 to March 31, 1995 and for two samples of releases in the following two fiscal years (1995/96 and 1996/97). The follow-up period was two years from the date of release. Two years is often selected as the follow-up period in studies of recidivism. One further advantage of using a two-year follow-up is that approximately one-half of the offenders would have completed their sentence and thus it provides an estimate of the recidivism of offenders beyond the end of sentence and any period of supervision.


The reconviction rate for all the releases in the first year was 44% with the reconviction rate for violence considerably lower (14%). The non-violent reconviction rate was 30% accounting for the majority of reconvictions. In each successive year, the reconviction rate decreased (43% and 41%). Significant differences in rates were found between men and women, with women being reconvicted at much lower rates then men. The reconviction rate for women fell during the three years studied from 30% to 23%. In addition, reconviction rates were higher for Aboriginal male offenders than Non-Aboriginal males (58% vs. 42%). As with women, the reconviction rate for Aboriginal offenders also decreased (57% to 53%).

Finally, slightly more than half of all reconvictions occurred after the offenders completed their sentences. Approximately 15% to 19% of reconvictions occurred while under supervision and 24% to 26% of reconvictions were after the completion of sentence.

Policy implications

  1. A reconviction rate in the range of 41% to 44% provides a possible measure that can be used for assessing the success of reintegrating offenders into the community. In addition, it can provide a basis to compare the relative success of programs over time and with each other.
  2. The reconviction rates found in this study compare favourably to other countries that have used a similar methodology. For example, England and Wales has reported a 50% reconviction rate for adult male inmates and 45% for adult female inmates released from their prisons in 1995. Considering the seriousness of many federal offenders, the findings suggest that the federal correctional system is working well.
  3. The higher reconviction rates found among Aboriginal offenders highlight the importance of focusing resources on this group while the significantly lower rates for women support the timely reintegration of female offenders into the community.


For further information

James Bonta, Ph.D.
Director, Corrections Research
Public Safety Canada
340 Laurier Ave. West
Ottawa, Ontario
K1A 0P8
Phone: 613-991-2831
Fax: 613-990-8295

Date modified: