Correctional Service Canada
www.csc-scc.gc.ca
Common menu bar links
FORUM on Corrections Research
Recidivists Tend to Be...
In an attempt to come to grips with the concept of the "repeat offender," research on recidivism has
identified numerous characteristics associated with an offender's likelihood of committing further
offences. For example, marital status, age at first conviction and criminal history are among the
variables associated with recidivism. As well, the nature of an individual's criminal history seems
to be connected with varying rates of recidivism and with the inclination to commit certain
offences.
The Corrections Branch of the Solicitor General Secretariat, together with the Correctional Service
of Canada, created extensive data bases on more than 50,000 inmates who were admitted and/or released
from federal institutions between 1971 and 1985.(1)
This article summarizes a report on 3,348 male, female and aboriginal inmates released from federal
institutions in 1983-1984. The group was followed up for three years.
A recidivist was defined as any released federal offender who was convicted within the three-year
period following release of a new indictable offence that led to a custodial sentence.
A non-recidivist was defined as any released federal offender who was not subsequently convicted of,
and incarcerated for, an indictable offence.
It has been estimated that a three-year follow-up period would capture 90% of those offenders who
ultimately reoffend.
Recidivists included both federal inmates who returned to federal custody and federal inmates
who received provincial sentences of incarceration. Non-custodial sentences (such as probation) were
not included.
In 1983-1984, 3,556 offenders were released from federal institutions. Sufficient data were available
on 3,348 of them. Of the 3,348 released inmates, 2,985 were non-aboriginal men, 282 were aboriginal
men and 81 were women.
The small number of female and aboriginal offenders restricted detailed analysis, preventing direct
comparisons with the larger "male offender" or "non-aboriginal male offender" sample. Female and
aboriginal offenders have thus been dealt with separately.
Complete data were available on 81 female offenders released from federal custody in 1983-1984.
Approximately one third (36%) committed a further offence within three years of release. Of those
who reoffended, more than three quarters (79%) had more than one conviction for an indictable
offence, compared with slightly more than half (54%) of non-recidivists. This result is statistically
significant.
Of the 39 female offenders who had served a period of incarceration before their current term,
slightly more than half (55%) committed further offences and just less than half (44%) did not. This
result is not significant.
As well, female offenders who reoffended were, on average, younger at the time of their first adult
conviction (age 20) than were those who did not reoffend (age 24).
Most releases from incarceration were either on parole (54%) or on mandatory supervision (32%).
Female offenders who reoffended were significantly more likely to have been on mandatory supervision
(48%) than on full parole (31%) when they reoffended.
In total, 282 male aboriginal offenders were included in this study, and almost all of them (92%) had
a previous conviction for an indictable offence. Two thirds (66%) of the aboriginal offenders
committed a further offence within three years of release, and a significantly high proportion of
both recidivists (95%) and non-recidivists (87%) had a previous indictable offence.
More than three quarters (84%) of the aboriginal offenders had been incarcerated previously. The
number of offenders with previous incarcerations was significantly higher among recidivists (90%)
than among nonrecidivists (72%).
A comparison of types of conditional release produced results similar to those found for female
recidivists. While only one third of aboriginal offenders (33%) released on parole reoffended, three
quarters (75%) of those released on mandatory supervision did.
The recidivism rate for single offenders (71%) and for those in common-law relationships (68%) was
significantly higher than that for aboriginal offenders who were married (44%).
The male study group consisted of both aboriginal and non-aboriginal inmates released from federal institutions in 1983-1984. Of these 3,267 inmates, nearly half (49%) reoffended in the three-year follow-up period. For those inmates who reoffended, more than half (58%) did so within the first year, almost one tenth (9%) did so during the second year and one third (33%) reoffended in the third year following release.
The Age Factor
Recidivists were younger than nonrecidivists at the time of their first conviction for an indictable
offence. More than three quarters (85%) of the recidivists were less than 21 years old (average age
18), compared with less than two thirds (63%) of nonrecidivists (average age 22).
Other studies have also found that offenders who reoffend tend to be younger than those who do not.
In this study, recidivists were, on average, three years younger (26) than nonrecidivists
(29) at the time of the current conviction (that is, the sentence from which they were
released).
Married inmates reoffended less frequently than those who were single, divorced or in a common-law relationship. While single offenders and those in common-law relationships had comparable recidivism rates (53% and 54% respectively), recidivism rates of married offenders were substantially lower (29%).
The Criminal History Factor
As in other studies, previous criminal history was strongly related to recidivism. Inmates who
reoffended also had more-extensive criminal records than non-recidivists, having approximately twice
as many previous incarcerations and indictable convictions.
One quarter (25%) of nonrecidivists had no previous indictable convictions, compared with less than
one tenth (8%) of recidivists. Further, more than a third (40%) of recidivists had more than 10
indictable convictions compared with less than one fifth (18%) of non-recidivists.
The relationship between previous incarcerations and recidivism was similar to that between previous
indictable convictions and recidivism. More than a third (38%) of non-recidivists had no previous
incarcerations and only one tenth (10%) had more than four incarcerations, while less than a fifth
(15%) of recidivists had no previous incarcerations and almost a quarter (22%) had more than four
incarcerations.
Eight offence type categories (such as break and enter, robbery, weapons, drugs) were examined to
assess the relationship between the type of offence and recidivism.
Offenders whose current offence fell into more than one category were counted in each category. For
example, an offender with only robbery charges was recorded in only the robbery category, but if the
offender' current term included both robbery and weapons convictions, he was recorded in both
categories.
As shown in Figure 1, more than half the offenders whose current term
was for property-related offences (break and enter, other property and
robbery) reoffended (63%, 57% and 53% respectively). Offenders whose current
sentence was included in other offence types reoffended or had their release
revoked less frequently.
Figure 1
Further analysis showed that the offences committed by inmates after release more often differed from their current-term offence. Therefore, the current offence does not necessarily predict subsequent offence patterns. That is, recidivists did not tend to "specialize" in the types of offences they subsequently committed.
Even though more than a quarter (28%) of recidivists with current terms for break and enter reoffended with another break and enter - compared with less than one fifth (18%) of recidivists in other offence categories - almost three quarters (72%) reoffended with a different type of offence.
Conditional Release and Recidivism
As was found with the female recidivists and aboriginal recidivists, type of conditional release was
strongly associated with recidivism. The type of release for male offenders was split almost equally
between mandatory supervision (50%) and full parole (46%). As shown in Figure 2, of those who
reoffended, two thirds (67%) did so while on mandatory supervision, while only 30% did so while on
full parole. Thus, recidivists were twice as likely to have been released on mandatory supervision as
on full parole. This is most likely a reflection of the fact that offenders who are at lower risk are
more likely to be released, at the Parole Board's discretion, on full parole, while higher-risk
offenders generally must wait for release on mandatory supervision.
Despite the higher recidivism rate of offenders on mandatory supervision,
there was no difference in the types of offences subsequently committed
by offenders on parole and those on mandatory supervision.
Figure 2
Summary
Overall, about one half of male offenders released from federal institutions in 1983-1984 committed
further offences, as did two thirds of aboriginal male offenders and roughly one third of female
offenders.
Recidivists tended to be younger at the time of their first adult conviction,
to have more extensive criminal histories and, generally, to be unmarried.
Further, recidivists were significantly more likely to have been released
on mandatory supervision. These findings are generally consistent with
previous research.
J. Bonta, S. Lipinski and M. Martin, Characteristics of Federal Inmates Who Recidivate (Ottawa: Statistics Canada, 1992).
(1) R.G. Hann and W.G. Harman, Predicting General Release Risk for Penitentiary Inmates, User Report (Ottawa: Solicitor General Canada, 1992).
