Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

FORUM on Corrections Research

Warning This Web page has been archived on the Web.

Release outcomes of long-term offenders

by Sara L. Johnson and Brian A. Grant1
Research Branch, Correctional Service of Canada

While offenders serving long sentences have been studied to determine the impact of incarceration, there has only been limited research on their rate of recidivism. Offenders serving long sentences differ from the general inmate population in terms of the crimes they have committed, the age at which they are released from prison and the length of their criminal record. Given the seriousness of their crimes, offenders serving long sentences are often perceived by the general public to be at a high risk to re-offend; however, a few studies suggest that these offenders may be at a lower risk to re-offend than those serving shorter sentences. This article examines readmission and recidivism rates of three groups of long-term offenders released to the community from a Correctional Service of Canada penitentiary: those serving life and other indeterminate sentences, those serving long determinate sentences (10 years or more) and those serving shorter determinate sentences (less than 10 years).

Almost one-third of the approximately 13,000 offenders in Canadian federal penitentiaries are serving sentences of ten years or more. Given the size of this offender group, and the amount of time they will be incarcerated, a task force was created to study their special needs. The Report of the Task Force on Long-Term Offenders2 studied offenders serving long sentences to determine what interventions would be appropriate. The Task Force produced a report with 21 recommendations.

Numerous studies have examined the institutional behaviour and adjustment to incarceration among offenders serving long sentences.3 However, only a few studies have looked at how successful offenders serving long sentences are after they have been released to the community. These offenders differ from the general offender population on the basis of the type of offences they committed. Generally, they have committed the most serious violent offences, most frequently involving homicide (54%), but also including robbery, sexual assault and assault. Among Canadian federal offenders serving indeterminate sentences, over 90% are serving sentences for homicide or homicide-related offences.

The main purpose of this study is to determine if offenders serving long sentences differ from those serving shorter sentences in terms of their release outcomes and the patterns of recidivism observed. In addition, the results from this study make it possible to determine if offenders serving different types of sentences (determinate and indeterminate) differ in release outcome patterns. Comparisons like these will help to improve how these offenders should be risk managed after release to the community.

Comparison groups

In Canada, most sentences that include incarceration are for a fixed period of time and are referred to as determinate sentences. Determinate sentences do not require that the person remain in custody for the entire length of the sentence. Rather, the National Parole Board may grant them a discretionary release at various eligibility dates during their sentence. If they are still in custody at two-thirds of their sentence, they are normally released to serve the final one-third under supervision.

Offenders may also be sentenced to indeterminate sentences. The most common indeterminate sentence is a “life” sentence. Offenders sentenced to life remain under the authority of the Correctional Service of Canada until they die, but do not generally remain incarcerated for the entire period. Offenders convicted of first-degree murder are automatically required to serve 25 years in custody before they are eligible for parole. However, under section 745 of the Criminal Code (also known as “faint hope”) there is possibility of release by means of a judicial review at 15 years. For offenders sentenced to life for second-degree murder or for other offences, the judge sets the date they are eligible for parole with the minimum period being 10 years and the maximum 25 years. When offenders have completed the minimum custodial portion of the sentence, they are reviewed for parole and if granted, are released by the National Parole Board. Once released to the community, they are supervised by Correctional Service of Canada parole officers and may be returned to federal custody if they fail to meet the conditions of their release or if they commit a new offence.

Another form of indeterminate sentence, used with offenders who have shown a history of violent (including sexual) crimes, is the declaration by the court that the individual is a “dangerous offender.” Once declared a dangerous offender by the court, the person must remain in custody until the National Parole Board determines that the individual is no longer a danger to the community.

Each of the three groups of offenders used in the study are more completely defined below along with some of their demographic characteristics.

Indeterminate: Offenders serving life sentences and those declared dangerous offenders by the court.4 The group accounts for 2% of all releases and includes 280 offenders, of which 4% (11) are women. Most offenders in this group are Caucasian (84%), 11% are Aboriginal and 6% are from other racial groups. On average, these offenders were 34 years of age at admission and 44 years of age at time of release.

Long-term determinate: These offenders have determinate sentences of 10 years or more. The group accounts for 3% of all releases and includes 373 offenders, of which 2% (8) are women. Aboriginal offenders account for 8% of these offenders, while 78% are Caucasian and 14% are from other racial groups. On average, these offenders were 32 years of age at admission and 38 years of age at release.

Shorter-term determinate: Offenders with determinate sentences of less than 10 years. They account for the majority of the federal offender population released in any one year approximately 95%. Over the two years of releases studied, the group included 11,521 offenders, of which 3% (311) were women. Approximately 13% of these offenders are Aboriginal, similar to their representation in the general federal offender population, while 78% were Caucasian and 9% were from other racial groups. On average, these offenders were 32 years of age at admission and 33 years of age at time of release.


All offenders released from federal penitentiaries in 1993 and 1994 were followed from their release date to January 2000. This provided a minimum of four years for follow-up and a maximum of seven years. Survival analysis was used to evaluate the differences between groups on rates of being offence free (no recidivism) over time.

Survival analysis provides a number of unique benefits. First, it makes it possible to include all cases without regard to length of time they could be followed-up. Second, survival analysis provides a visual examination of the data, showing the rate of failure across time, so that groups that fail very quickly after release can be identified and compared to those that fail more slowly and over a longer period of time.

Two measures of recidivism are used new conviction and new violent conviction. New conviction includes any offence that occurred after release for which an offender is required to serve a period of incarceration in a federal penitentiary. New violent conviction is similar to the new conviction measure but includes violent offences only, such as all murder related offences, assault, sexual assault and robbery.

Current offence

The most serious offence that offenders had been convicted of up to the time of their release is presented in Table 1. The order of the offences represents the relative seriousness of the offences for the analysis. An offender with a murder conviction and a sexual assault would only be counted in the murder group for the purposes of this study.

Table 1

Offence History
Long- term determinated
Shorter-term determinated
4.3 (12)
85.0 (238)
Manslaughter / Attempt / Conspire Murder
3.9 (11)
25.3 (94)
4.3 (486)
Sexual offence
3.2 (9)
8.3 ( 31)
13.4 (1 549)
1.4 (4)
14.2 ( 53)
17.9 (2 058)
1.4 (4)
26.1 ( 97)
22.6 (2 604)
Other offences
0.7 (2)
25.6 (75)
41.8 (4 822)
Number of offences
1.includes Capital Murder;
2. includes Non- capital Murder;
3. includes international transfers with determinate sentences.

Most of the offenders (85%) in the indeterminate group (“lifers”) were convicted of second-degree murder. Offenders released with a first-degree murder charge were convicted prior to 1976 when a change in legislation required those convicted of first-degree murder to serve 25 years in custody before being eligible for parole. None of the offenders convicted after 1976 were eligible for parole in 1993 and 1994. For the long-term determinate group, about one quarter had a homicide-related offence, another quarter were convicted of robbery and another quarter were convicted of either sexual or other assaults. The remaining offenders in the group were convicted of a variety of other offences. By comparison, for federal offenders serving determinate sentences of less than 10 years, robbery was the most common offence (23%), followed by assaults (18%) and sexual assaults (13%).

Release Type

Offenders serving indeterminate and determinate sentences are not eligible for the same types of release and this could affect the follow-up results. Offenders serving indeterminate sentences are only eligible for day parole and full parole whereas offenders serving determinate sentences have four release options. Offenders serving determinate sentences are eligible for parole and day parole early in their sentence. If these offenders have not been released after serving two-thirds of their sentence they are automatically eligible for statutory release. In special cases, offenders may be detained in custody until the end of their sentence. In general, offenders released on parole remain under supervision longer than offenders released on their statutory date.

To determine the effect the different release options might have on the follow-up results, it is important to look at the types of release for each of the offender groups. These data are presented in Figure 1. For offenders with determinate sentences, as sentence length increases the percentage of offenders released on statutory release increases. Almost three-quarters (72%) of the offenders serving indeterminate sentences are first released on day parole with the balance released directly to full parole. Over half (56%) of the offenders serving long-term determinate sentences are first released on day parole and another one-quarter (26%) are released on full parole. Most of the balance of the these offenders are released after serving two-thirds of their sentence (statutory release) and a very small percentage are released to the community only at the end of their sentence. Offenders serving shorter-term determinate sentences are most likely to receive statutory release (41%) or day parole (37%). A slightly higher percentage of shorter-term offenders, than the long-term offenders, are released at the end of their sentence.

Figure 1
Type of Release by Offender Group

The data on release types suggests that all other things being equal, both readmission and new conviction with a federal sentence (two years or more) will be highest for offenders serving indeterminate sentences, followed by long-term determinate and lowest for the offenders in the shorter term determinate group. These differences would occur because offenders serving longer sentences are under supervision for longer periods of time and therefore offer a greater time at risk in the community for breaches of release conditions or new offences.

Release outcome: new offences

At the end of the seven year follow-up period, approximately 60% of offenders serving determinate sentences had not been convicted of a new offence, while 73% of offenders serving indeterminate sentences had not been convicted of a new offence. While the offence free rate by the end of seven years was similar for the two determinate sentence groups, the pattern over the release period was quite different, as shown in Figure 2.

Figure 2
Survival Rates of Long- Term Offender Releases
( Offence Free)

The shorter-term determinate group has the poorest survival curve with the fastest decline. Just over 20% were convicted of a new offence in the first 12 months after release and 30% were convicted of new offences in the first two years. By comparison, slightly more than 10% of the long-term determinate group and about 5% of the indeterminate group had a new conviction within the first year of release.

The survival curves for all three groups remain different throughout the follow-up period except that shortly after the five-year point, the curve for the long-term determinate group starts to match that of the shorter-term determinate group; however, the indeterminate group performs the best out of the three groups.

Release outcome: new violent offences

A survival analysis using new violent offence (see Figure 3) indicates that after seven years 89% of the offenders in the indeterminate group remained free of a new violent offence compared with approximately 80% of offenders in the shorter- and long-term determinate sentence groups. The pattern of re-offending with a new violent offence was very similar for the two determinate groups, which is different than that observed for the new convictions, in general.

Figure 3
Survival Rates of Long- Term Offender Releases
( Violent Offence Free)

Analyses were also conducted to look at the number and type of new convictions. In addition to being less likely to have a new conviction, offenders serving indeterminate sentences had fewer new convictions over the entire follow-up period. For example, while 1% of offenders in the indeterminate group had more than five new convictions, this increased to 5% for the long-term determinate group and 10% for the shorter-term determinate group. More specifically, the most serious offence committed was non-violent for over half of the indeterminate group who had a new offence. In addition, no offenders in the indeterminate group had a new conviction for a homicide-related offence.


Overall, the results indicate that offenders serving indeterminate sentences have a lower recidivism rate than offenders serving determinate sentences. Offenders serving long-term determinate sentences have better survival rates, particularly in the early years after release, than offenders serving shorter-term sentences. Differences between the groups, in terms of new convictions, are meaningful and significant during the first five years of follow-up and this is consistent with previous research. However, after five years the curves for the two determinate groups merge and the differences seem to disappear, but the indeterminate group remains different with a high survival rate, with almost three-quarters of offenders remaining free of new convictions.

Unfortunately, the reasons for the differences between the groups cannot be identified with from this research; however, data presented showed that the comparison groups differ in terms of their type of offence, the type of release and age. For example, offenders with indeterminate sentences are older, more likely to have received a conviction for homicide and more likely to be released on day parole than offenders in the other groups. Earlier research has shown that sentence length by itself does not have an effect on recidivism5 so other factors must be responsible. The recidivism data suggest that within the group of offenders serving indeterminate sentences, there is a proportion that is at low risk to reoffend. Additional research will be needed to develop tools to identify this group. The challenge for correctional staff is to work with this group to ensure they can safely be released at the earliest possible date.

1. 340 Laurier Avenue West, Ottawa, Ontario K1A 0P9.

2. Implementing the Life Line Concept: Report of the Task Force on Long-Term Offenders, Ottawa, ON, Correctional Service of Canada, February, 1998.

3. Zamble, E. “Behavior and Adaptation in Long-Term Prison Inmates: Descriptive Longitudinal Results,” Criminal Justice and Behavior, vol. 19, no 4, 1992, p. 409-425.

4. This group also includes a few offenders who, under previous legislation, had been declared dangerous sexual offenders and habitual offenders.

5. Motiuk, L.L., R. Belcourt and J. Bonta. “Managing High Risk Offenders: A Post-Detention Follow-up,” R-39, Correctional Service of Canada, 1995.