Research Reports

Day Parole Outcome, Criminal History and Other Predictors of Successful Sentence Completion

by:
Brian A. Grant
and
Christa A. Gillis

Research Branch,
Corporate Development
Correctional Service of Canada

July, 1999

EXECUTIVE SUMMARY

The study was conducted to determine the factors associated with successful sentence completion for offenders who had received day parole and is a follows-up to the report, Day parole program review: Case management predictors of outcome (Grant et al., 1996). The results of the study indicate that successful completion of day parole is associated with lower readmission and new offence rate. Only 15% of the day parole sample that successfully completed day parole were readmitted after day parole completion, whereas 44% who were unsuccessful on day parole were readmitted. Furthermore, those who were unsuccessful in completing day parole reoffended at three times the rate of offenders who successfully completed day parole (30% compared to 10%, respectively). The study also demonstrates that day parole serves the needs of low risk offenders released early in their sentence. Additionally, day parole responds to the needs of higher risk offenders released shortly before their statutory release date, and contributes to community protection.

Day parole program review: Case management predictors of outcome (Grant et al., 1996). The results of the study indicate that successful completion of day parole is associated with lower readmission and new offence rate. Only 15% of the day parole sample that successfully completed day parole were readmitted after day parole completion, whereas 44% who were unsuccessful on day parole were readmitted. Furthermore, those who were unsuccessful in completing day parole reoffended at three times the rate of offenders who successfully completed day parole (30% compared to 10%, respectively). The study also demonstrates that day parole serves the needs of low risk offenders released early in their sentence. Additionally, day parole responds to the needs of higher risk offenders released shortly before their statutory release date, and contributes to community protection.

Offenders released on day parole are more likely than the general offender population to be granted full parole. In addition, their success rate on full parole and statutory release is higher compared to the general offender population.

Aboriginal offenders released on day parole are less likely (36%) than the general population of day parole offenders (45%) to be released on full parole, and are more likely to be convicted of a new offence after day parole. Failure during full release (full parole or statutory release) is higher for Aboriginal offenders than the general population of day parole offenders, even for those who successfully completed their day parole.

Female offenders were slightly more likely to be reincarcerated for a period immediately after their day parole than males (43% compared to 30%, respectively). Once released on full parole or statutory release, however, they were more likely to complete their sentence without being readmitted and were less likely to commit a new offence.

Regional differences in post-day parole outcome were minimal. However, the Pacific region had the lowest readmission (16%) and reoffence (15%) rates, while the Atlantic region had the highest rate of new offences (24%).

The study also looked at risk, needs, risk/needs, motivation, dynamic risk predictors and previous criminal history as factors influencing both day parole and full release outcome. Results confirmed the relationship between release outcome and risk and need, with higher risk and needs associated with higher levels of failure. High-risk offenders were readmitted at nearly three times the rate of low risk offenders (28% compared to 11%) and reoffend at twice the rate (18% versus 8%, respectively). Offenders classified at the high case need level were five times as likely to be readmitted than offenders identified at the low case needs level (35% compared to 7%). Moreover, new offence rates were more than three times greater for high (21%) versus (6%) case needs offenders. In addition, the results indicate that offender rated as low risk but with high levels of needs are at greater risk to fail on release (29%) than are some high risk offenders (14% for high risk/low needs group). For offenders eligible for day parole, needs may be a more important predictor of outcome than risk.

Motivation was shown to be a useful predictor of day parole and full release outcome, with motivated offender more likely to succeed than non-motivated offenders (21% failure rate compared to 30%, respectively). Dynamic risk predictors such as change in motivation and change in institutional security were also shown to be useful predictors of outcome, although security level at time of release was as good a predictor of full release outcome as change in security level.

Previous criminal history, consisting of federal and provincial offences, was related to both day parole and full release outcome. Offenders with less than four convictions had very high success rates on both day parole (85%) and full release (77%). However, offenders with more than 10 previous convictions had much lower success rates on day parole (68%) and full release (less than 50%).

The study demonstrates that day parole is an important predictor of full release outcome. It provides a form of early release for low risk offenders and is a method of preparing and assessing higher risk offenders for full release.

ACKNOWLEDGEMENTS

Linda Lefebvre assisted with the development of the main day parole data set. Moira Law and Chris Beal coded the pre and post-day parole offence data. Elli Caparelli provided English editorial and Cathy Delnef provided French editorial assistance. Laura Vandette was responsible for the preparation of the final document.

TABLE OF CONTENTS

• EXECUTIVE SUMMARY
• ACKNOWLEDGEMENTS
• LIST OF TABLES
• LIST OF FIGURES
• INTRODUCTION
• METHOD
• RESULTS
• DISCUSSION
• REFERENCES
• APPENDIX A: - CODING MANUAL

LIST OF TABLES

• Table 3-1: Type of release after day parole, day parole outcome and post-day parole outcome
• Table 3-2: Percentage of post-day parole failures by type of failure and day parole completion
• Table 3-3: Post-day parole outcome by time of day parole release
• Table 3-4: Potential number of days, actual number of days and proportion of potential days in the community by day parole completion
• Table 3-5: Potential number of days, actual number of days and proportion of potential days in the community by post-day parole outcome
• Table 3-6: Number of previous offences (prior to the current admission) by day parole completion
• Table 3-7: Number of previous federal offences (prior to the current admission) for day parole outcome
• Table 3-8: Number of previous offences (prior to the current admission) for day parole offenders by post-day parole outcome
• Table 3-9: Risk by post-day parole outcome
• Table 3-10: Need by post-day parole outcome
• Table 3-11: Risk/need by post-day parole outcome
• Table 3-12: Percentage failures by motivation and day
• Table 3-13: Change in motivation by day parole
• Table 3-14: Change in motivation by post-day parole
• Table 3-15: Change in institutional security level by day parole completion
• Table 3-16: Change in institutional security level by post-day parole outcome
• Table 3-17: Region by post-day parole outcome
• Table 3-18: Type of release after day parole release and post-day parole outcome (Aboriginal offenders)
• Table 3-19: Percentage of post-day parole failures by type of failure and day parole completion (Aboriginal offenders)
• Table 3-20: Post-day parole outcome by tine of day parole release (Aboriginal offenders)
• Table 3-21: Number of previous offences (prior to the current admission) by day parole completion¹ (Aboriginal offenders)
• Table 3-22: Number of previous offences by post day parole outcome (Aboriginal offenders)
• Table 3-23: Type of release after day parole release and post-day parole outcome (female offenders)
• Table 3-24: Post-day parole outcome by time of day parole release (female offenders)
• Table 3-25: Previous offences by post-day parole outcome (female offenders)

LIST OF FIGURES

• Figure 3-1: Post-day parole outcome by time of day parole release

INTRODUCTION

It is important to determine the effectiveness of correctional programs to ensure that those that reduce the likelihood of reoffending are maintained and encouraged and unsuccessful ones discontinued. Most outcome research evaluates the effectiveness of specialized treatment programs such as substance abuse and skills training. However, other correctional activities such as parole and statutory release, which have become integrated into operations, are frequently not evaluated to determine their effect on offenders. Day parole, which permits offender to be in the community during the day but requires them to return each evening to a residential facility, is one of these programs. The current study examines how participation in the day parole program is related to recidivism after and offender is released to the community on full parole or statutory release.

Overview of Day Parole

The purpose of conditional release programs (day parole, full parole, and statutory release) is to allow offenders to serve a portion of their sentence in the community. This provides for a gradual change from a high controlled prison environment to a less controlled life in the community.

Day parole is the earliest conditional release possible¹ and is usually accompanied by the most stringent conditions. In particular, most day paroles require that offenders live in an approved residential setting such as a Community Correctional Centre (CCC) operated by the Correctional Service of Canada (CSC), a Community Residential Centre (CRC) operated on a fee-for-service basis by a contractor, or a correctional institution. From these settings, offenders may attend school or treatment programs and/or maintain a job, but they are required to return each evening before their curfew.

Day parole also differs from full parole and statutory release in that it is granted for a relatively short period of time (usually six months) and its completion does not coincide with expiry of the offender’s sentence. Upon completion of a day parole, a new day parole, or other type of release, must be granted if the offender is to remain in the community, otherwise he or she must return to prison. On full parole and statutory release, the offender remains in the community until the end of the sentence. In all types of conditional release, offenders are returned to prison if hey fail to meet their release condition.

During the conditional release period, legally binding conditions on residency and behaviour are imposed. These conditions permit the correctional system to monitor and respond quickly to offenders who deviate from them. If an offender is suspected of illegal activity, or fails to meet the provisions of the release, the offender may be returned to a correctional institution until a proper hearing and review of the case can be completed. In this way, conditional release provides a mechanism for responding to problems early, before they become serious.

Day Parole Follow-Up

Day parole should contribute to the overall successful reintegration of the offender into the community. Therefore, an important factor in evaluating day parole is its association with performance on full parole or statutory release after the day parole is completed. While the report, Day Parole Program Review: Case Management Predictors of Outcome (Grant, et al., 1996) provides an analysis of the relationship between many variables and outcome of the day parole release, it does not address the relationship between day parole outcome and subsequent success in the community.

To evaluate the relationship between day parole outcome and recidivism, data were collected on the type of release after day parole, the time of day parole release, readmission, post-release offences and time in the community. Analyses are also presented on the association with other factors, including risk/needs, motivation, releasing region and outcome following day parole.

In addition, the relationship between adult criminal history and day parole outcome are addressed using data on criminal convictions which were not collected as part of the Grant et al. (1996) study. The study also presents outcome results for female and Aboriginal offenders. These results help to identify differences in the responses of these subgroups of the offender population to day parole.

Impact of Changes to Legislation Affecting Day Parole

In November 1992, Corrections and Conditional Release Act (CCRA) came into force, replacing the Parole Act and the Penitentiaries Act. The Act revised the purpose of day parole and the date at which offenders were eligible. Under the CCRA, day parole eligibility is set at six months prior to full parole eligibility for most offenders² rather than at one sixth of the sentence. The Act also states that the purpose of day parole is to prepare offenders for release on either full parole or statutory release. Additionally, the Act eliminated automatic review by the National Parole Board (NPB), making it necessary for offenders to apply for day parole. A more complete history of day parole is presented in the Grant et al. (1996) study.

The changes implemented by the CCRA may have reduced the number of offenders granted day parole, but the relationships identified in this study are unlikely to have been affected. The change in eligibility date only affected about 8% of the day parole cases, since 92% of releases already occurred after six months prior to full parole eligibility. Almost all releases extend into the parole eligibility period, suggesting that day parole was used largely as preparation for full parole or statutory release as required by changes in the CCRA.

Eliminating automatic review for day parole is the change most likely to have impacted upon the results of this study. However, this change would result in a reduction in the number of offenders who received day parole, rather than affecting the relationship between pre-release factors and post-release outcome. Therefore, the results of the study should be equally valid today, even with the changes introduced by the CCRA.

METHOD

Data Sources

The data used in the study are derived from two sources: the file review data from the Day Parole Review project (Grant et al., 1996) described briefly below and data created specifically for this study, which includes offence history and recidivism data.

File Review

The report, Day Parole Review: Case Management Predictors of Outcome (Grant et al., 1996), employed both administrative data systems and offender files to examine background and offence variables related to day parole outcome. The file review component of the day parole review was conducted by nine Masters-level students and was divided into three sections: background (e.g., demographics, education and employment history, marital and family history, young offender history, etc.), NPB history (e.g., type and conditions of release, previous NPB decisions, status changes, etc.) and CSC case management history (institutional rating, program activity, risk/needs level, suspension information, etc.). Coders were trained in how to use the coding manual and how to read both NPB decision files and CSC offender files. Practice coding was included in the training, which took place over the course of two days. Several revisions were made to the manual as problems with its use were identified and corrected.

As Grant et al. (1996) has described the sample used in the follow-up study in detail, only a brief description of the subjects is presented. The study used a national sample selected from all day parole releases in 1990-91. Specifically, a one third sample (n = 1100) was randomly selected from about 3,800 day parole completions that were taken by approximately 3300 offenders. Following a series of analyses, Grant et al. (1996) concluded that the day parole sample was represented of the offender population that completed day parole in 1990-91.

Offence History

The present study extended the day parole follow-up period in order to ascertain if performance on day parole predicts outcome of full release. The follow-up study also examined collateral information relating to previous provincial and federal offences to explore whether previous offence patterns contributed to outcome after day parole completion. A review of selected administrative data systems was undertaken to examine the period extending from day parole completion to March 31, 1994.

A coding manual (see Appendix) was developed to code information from the Offender Management System (OMS), Canadian Police Information Centre (CPIC) files and Correctional Research and Development (CRD) admission and release databases. Federal offences were derived from the OMS and provincial offences obtained from CPIC files; the admissions and release databases were used to confirm information obtained from the OMS and CPIC files.

The following information was extracted using the coding manual:

• number and type of previous provincial and federal offences (prior to day parole sentence);
• number and type of current federal offences;
• number and type of readmission(s) after day parole completion (terminations, revocations, technical and new offences);
• number and type(s) of offence(s) committed (technical or new offence);
• number of type of subsequent releases (subsequent day parole releases, full parole, mandatory supervision); and
• number and type of federal offences post-day parole completion.

Coding Procedure

Three university students trained by the authors coded the follow-up data. They were provided with written instructions on how to read and record the data and practice trials were used to ensure that a consistent approach to coding was adopted. During the coding process, any problems with data interpretation were discussed among the coders to ensure minimal inconsistency among the techniques employed. Upon completion of the coding, the data were entered in a database using the Statistical Analysis System (SAS; 1990). After the data were checked and cleaned, preliminary analyses were conducted to ensure the integrity of the data, and the follow-up database was merged with the initial day parole database.

The follow-up period commenced with each participant’s index day parole completion (the day parole which resulted in inclusion in the study) and terminated on March 31, 1994, in an attempt to follow most offenders to the end of their sentence. The first step in coding the data was to identify the commencement of the current sentence. The number and type of federal and provincial offences prior to the day parole sentence were recorded. Coders then examined the type (no release, full parole, statutory release) and immediacy of release after day parole. The remainder of the coding manual accounted for subsequent release pre and post-warrant expiry date, and all federal admissions, releases and convictions to March 31, 1994.

Measures

Day parole outcome

Day parole outcome was divided into the following categories: successful (no new admissions/no new offences) and unsuccessful (day parole revocation for either a technical violation or new offence).

Recidivism

Four primary measures of recidivism were recorded to provide an indication of post-day parole release performance: readmission, technical violation, offence and violent offence. First, all readmissions after day parole completion were recorded; readmission refers to re-entry to an institution resulting from either new convictions or technical violations. Technical violations refer an admission without new offences, including termination of the day parole. This is the least reliable measure since a technical violation may be used to end the day parole while police investigate a case. Given that new criminal charges relating to the violation may come some time later, technical violations may include some cases where an offence was committed. Offence refers to a new criminal offence(s) conviction. Violent offence indicates that the offender committed an offence classified as violent.³

Overall measures of offending were computed for number and type (violent, non-violent) of offences. The overall number of offences was calculated by totalling the number of new offences from day parole completion to warrant expiry day (WED).

In addition, time to readmission, which refers to the time an offender spent in the community from day parole completion to the first new admission date, was calculated by subtracting the day parole completion date from the first institutional admission after day parole completion. Admission was defined as 15 days or more of incarceration prior to subsequent release.

Finally, as offenders spent different amounts of time in the community, a measure of time at risk was calculated for comparison purposes. First, by subtracting the offender’s day parole completion date from WED, a measure of their potential community time was calculated. Subsequently, actual community time was computed by subtracting all time spent in the institution [release date(s) – admission date(s) from the potential time an offender could spend in the community.

Risk/Needs

An offender’s readiness for release is largely contingent on the level of risk for recidivism presented and the offender’s various needs related to criminal reoffending. Case management officers consider a variety of factors in their assessment of offender’s risk levels, including previous criminal history, current offence, institutional behaviour and program participation, and evidence of adjustment that would indicate a more pro-social orientation.

The Community Risk/Needs Management scale (CRNMS) is a semi-structured risk/needs assessment instrument completed shortly after offenders are released to the community. The CRNMS evaluates 12 areas that have been identified as potentially placing offenders at risk for future criminal behaviour:

1. academic and vocational skills,
2. employment pattern,
3. financial management,
4. marital and family arrangements,
5. companions and significant others,
6. living arrangements,
7. behavioural and emotional stability,
8. alcohol usage,
9. health,
10. mental ability,
11. health,
12. and attitude.

Following a review of the offenders’ needs and criminal history, parole officers rate the offenders on a three-point (low, medium, high) needs scale and a two-point (low, high) risk scale. These ratings can then be combined to produce a risk/needs rating.

Analyses

Offenders who did not reach their warrant expiry date by March 31, 1994 (112) or who were not released after day parole completion (18) were excluded from the analyses. As noted earlier, only ordinary day parole cases were studied as the number of other types of releases was too small to allow for meaningful analyses. The final follow-up sample consisted of 588 male offenders, 26 female offenders and 59 Aboriginal offenders.

A series of chi-square analyses were performed to examine the relationships between the various background variables and follow-up measures. More specifically, day parole outcome, timeliness of release, and offenders’ levels or risk and need were examined in relation to number of offences, type of offence (violent versus non-violent), technical violations and institutional admissions in the period following day parole completion.

RESULTS

Results are presented in three main sections. First is a presentation of the analyses using the sample of male offenders who received ordinary day parole, which accounts for 85% of all day parole releases. Analyses in this section are further divided into those relating day parole to post-day parole outcome and other factors related to recidivism. This is followed by the analyses for the Aboriginal sample, and finally, results are presented for female offenders. The analyses for Aboriginal and female offenders are limited by the small sample sizes.

Male Offenders

Relationship Between Day Parole and Post-Day Parole Outcome

Post-Day Parole Release

If day parole is effective, it should lead to release on full parole. Alternatively, an offender released very early in his or her sentence might have the day parole extended to reach the full parole eligibility date,⁴ or in the case of higher risk offenders who are not considered good candidates for full parole, the day parole might be extended until the statutory release date. Offenders who do not perform well on day parole are usually returned to prison to serve additional time in custody.

An examination of the type of release following day parole completion (Table 3-1) indicates that nearly half of the offenders (45%) were released on full parole after they completed the day parole, and another 19% reached their statutory release date during or immediately after day parole completion. Six percent of the sample were given at least one additional day parole prior to full parole or statutory release.

Almost one third of the offenders (30%) had to serve some additional time in prison before a full release, and the majority of these offenders (73%) were released at their statutory release date. Most offenders (82%) who returned to prison after their day parole remained incarcerated for more than two months.

Table 3-1: Type of release after day parole, day parole outcome and post-day parole outcome

Type of Release	Number of Cases	Percentage of Cases	Percentage successfully completed day parole	Percentage successfully completed sentence
Full Parole	250	45.0	93.20	85.6
Statutory release	105	18.9	97.14	78.1
Full parole after subsequent day parole(s)	19	3.4	94.74	89.5
Statutory release after subsequent day parole(s)	12	2.2	91.67	91.7
Full parole after incarceration	40	7.2	60.00	67.5
Statutory release after incarceration	129	23.2	19.35	58.9

 

The results presented in Table 3-1 also indicate the percentage of offenders who successfully completed their sentence (i.e., no new admission) for each type of release following day parole. The most successful groups were released on multiple day paroles, leading either to full parole or statutory release. Approximately 90% of these cases successfully completed their sentence, although the size of these groups is quite small, limiting the generalizability of the results. Of offenders granted full parole after their day parole, 86% completed their sentence successfully. The percentage of successful completion drops 78% for those released at their statutory release date. However, when the full parole was followed by a period of incarceration (indicating that the day parole performance was unsatisfactory), only 67% of offenders released on full parole were successful, and only 59% released on their statutory release date successfully completed their sentence without readmission.

Day Parole Outcome and Post-Day Parole Recidivism

To determine the relationship between day parole performance and full release outcome, new admissions and offences were monitored until the end of the offender’s sentence. A number of cases were not included in the follow-up because they were never released after their day parole (2%), or they did not reach the end of their sentence (17%). The following results apply to the remaining 550 cases.

The average follow-up period was 21 months (median of 19 months) and the range for follow-up was 4 to 47 months. More than three quarters of the cases had follow-up periods greater than 12 months.

Overall, 77% of offenders who had been released on day parole had no new admissions to a federal prison prior to the end of their sentence. For offenders who ended their day parole successfully, 85% completed their sentence without a new admission. However, only 56% of offenders who did not successfully complete their day parole completed their sentence without a new admission. These results are presented in Table 3-2.

An offender may be readmitted to federal prison for a technical violation⁵ of the conditional release (parole or statutory release) or for committing a new criminal offence.

Of offenders who were successfully on day parole, about 8% were readmitted for a technical violation and 10% were readmitted with a new offence. Of the offenders who were unsuccessful on day parole, 20% had their subsequent full release revoked for a technical reason, and 30% committed a new offence. These rates indicate that offenders who are unsuccessful on day parole are twice as likely to fail on full release for a technical violation and three times more likely to commit a new offence than offenders who successfully complete their day parole release.

Table 3-2: Percentage of post-day parole failures by type of failure and day parole completion

Day Parole Completion	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Successful	15.53	8.50	9.98	3.40	412
Unsuccessful	44.20	20.29	29.71	7.97	138
Percentage of all day parole sample cases	22.73	11.45	14.94	4.54	550
Chi-squared values	48.38	14.18	31.67	6.06
(df = 1, N = 550)	p<.001	p<.001	p<.001	p<.05

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group.

 

The results also indicate that only 3% of offenders who successfully completed their day parole committed a violent offence before the end of their sentence. Approximately twice as many offenders who were unsuccessful during their day parole release (8%) committed a violent offence later in their sentence.

Time of day parole release

In an earlier report (Grant et al., 1996), it was shown that time of the day parole release was associated with the success or failure of the day parole. More specifically, offenders released on day parole after their full parole eligibility date were less likely to complete their day parole. Successful completion of the sentence is also associated with the time of day parole release as shown in Table 3-3 and Figure 3-1. The patterns, however, are not as clear-cut as they were with day parole outcome. The Chi-square analyses also indicate that only the relationships between time of release and readmissions and technical violations are statistically reliable. However, there are interesting patterns in the data that are worth exploring.

Table 3-3: Post-day parole outcome by time of day parole release

Time of day parole release2	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Pre-full parole eligibility	17.27	7.63	12.05	3.21	249
before 25% of time to statutory release date	23.64	7.27	17.43	6.36	109
before 50% of time to statutory release date	32.89	23.68	19.74	6.53	76
before 75% of time to statutory release date	24.19	12.90	12.90	3.23	62
75% of time to statutory release date	28.95	23.68	13.16	5.26	38
Chi-squared values	9.73	22.19	ns	ns
(df = 1, N = 534)	p<.05	p<.001

sup>1The failure group are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

²The period from the full parole eligibility date and the statutory release date was divided into four equal parts.

 

The lowest rate of readmission before completion of the sentence (17%) was for offenders released before their parole eligibility date. Moreover, the early day parole release group had the lowest rates of recidivism for three of the four outcome measures. The highest readmission rate (33%) was for offenders released between one quarter and one half of the time between their parole eligibility date and their statutory release date. These offenders also had elevated recidivism rates (highest or second highest level) for technical violations and new offences.

The group of offenders released late in the sentence was expected to have the highest recidivism rates since they were likely kept in prison because they were viewed as high-risk offenders. While this group did have the second highest rate of readmission (29%), this was largely the result of technical violations. These offenders also had one of the lowest rates of new offences (13%), and a recidivism rate for new violent offences (5%) that was lower than that of offenders released between 25% and 50% of the time between full release and statutory release.

Figure 3-1: Post-day parole outcome by time of day parole release

 

Proportion of time in the community

In addition to measuring the outcome of day parole in terms of recidivism, it is also possible to measure outcome in relation to the amount of time the offender spent in the community under supervision after completion of the day parole. The greater the proportion of time spent under supervision, the less time that was spent in an institution. Since institutional time is much more expensive, these analyses provide an indication of the costs involved in failures.

Table 3-4 presents the potential community time (time from full release until the end of the sentence or warrant expiry date), the actual amount of time spent in the community and the proportion of time spent in the community. If an offender did not return to an institution, then the potential and actual community time are the same, and the proportion of time in the community is one. However, offenders may return to prison for either a technical violation or commission of a new offence.

 

Table 3-4: Potential number of days, actual number of days and proportion of potential days in the community by day parole completion

Day parole completion	Potential community days from day a parole completion to warrant expiry date	Actual community days from day parole completion to warrant expiry date	Proportion of community days	Number of cases
Successful	589	5571	.95	411
Unsuccessful	424	3371	.84	138

¹ F (1,546) = 53.8, p < .0001

 

The results indicate that offenders who completed their day parole successfully spent 95% of their eligible time in the community, whereas those were unsuccessful during the day parole spent only 84% of their eligible time in the community (F (1,546) = 53.8,p < .0001). In addition, offenders who were unsuccessful on day parole had a smaller number of days of potential time in the community because their day parole release occurred later in the sentence. The data indicate that on average, offenders who were successful on day parole spent an additional one month in prison after their full release while offenders who were unsuccessful on day parole spent an average of three months in prison after their full release.

Using the same measure of outcome it is possible to determine the effect of readmission or failure (after the day parole release) on proportion of time in the community, as presented in Table 3-5. The results indicate that offender who were readmitted prior to the end of their sentence would have been in the community longer (almost 600 days), but were in fact only in the community for about 400 days.

Specifically, they only spent 67% of the potential time in the community. Offenders who were not readmitted spent, on average, 530 days in the community.

 

Table 3-5: Potential number of days, actual number of days and proportion of potential days in the community by post-day parole outcome

Day parole completion	Potential community days from day a parole completion to warrant expiry date	Actual community days from day parole completion to warrant expiry date	Proportion of community days	Number of cases
Yes	596	3971	67%	128
No	530	5301	100%	426

¹ F (1,551) = 992.1, p < .0001

 

Other Factors Influencing Recidivism

Number of previous offences

To determine the relationship between day parole outcome, recidivism and previous offences, data were collected on all offences prior to the day parole sentence. These data were taken from the Canadian Police Information Centre (CPIC) records and include offences for which there were federal and provincial sentences.

The number of previous criminal offences shows a clear relationship to day parole outcome (Table 3-6). While 93% of offenders with no previous convictions were successful on day parole, the success rate dropped to approximately 80% for offenders with four to 10 convictions and to only 50% for offender with more than 20 convictions.

 

Table 3-6: Number of previous offences (prior to the current admission) by day parole completion

Number of previous offences	Successful	Unsuccessful	Number of Cases
0	92.86	7.14	84
1-3	84.55	15.45	110
4-10	77.27	22.73	132
11-20	68.10	31.90	116
21+	52.75	47.25	91
Chi-squared values	47.03
(df = 4, N = 533)	p<.001

Table 3-7 presents the relationships between number of federal offences prior to the day parole sentence and post-day parole outcome. Re-admission and technical violation rates tripled for offenders convicted of no offences to those convicted of 11 to 20 previous federal offences (from 19% to 67% and 9% to 33%, respectively). The differences in rates of new offending were also statistically significant, although less notable than the differences obtained for re-admission and technical violations. Interestingly, offenders convicted to one to three and four to ten offences evidenced higher rates of new offending than offenders convicted of 11 to 20 previous federal offences. The differences between violent offending rates were not statistically reliable.

Table 3-7: Number of previous federal offences (prior to the current admission) for day parole outcome

Number of previous offences	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
0	18.71	9.47	12.73	4.16	433
1-3	32.91	16.46	22.78	7.60	79
4-10	43.33	20.00	26.67	3.33	30
11-20	67.67	33.33	16.67	0.00	6
Chi-squared values	22.57	8.71	8.72	ns
(df = 3, N = 548)	p<.001	p<.05	p<.05

¹The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

 

Further analyses were conducted to explore the relationship between previous federal and provincial offences, and outcome variables. There is also a clear pattern of increasing recidivism with increases in the volume of criminal activity (number of previous convictions) for day parole offenders, as shown in Table 3-8. More specifically, whereas only 4% of offenders with no previous offences were readmitted, almost 40% of offenders with more than 10 previous convictions were readmitted.

There is little difference in readmission rates for offenders convicted of 11 to 20 offences and those convicted of 21 or more offences. Similarly, the rate of technical violations is comparable for offenders convicted of 11 to 20 offences, and 21 offences or more. Conversely, differences were noted in new offence rates (21% vs. 27%) for offenders convicted of 11 to 20 previous offences and 21 or more offences. Results in the table also indicate that violent recidivism is more likely for those with a higher number of previous convictions, although the numbers are not statistically reliable. Statistical tests indicate that the relationships between number of previous convictions and the other measures of recidivism are reliable.

Table 3-8: Number of previous offences (prior to the current admission) for day parole offenders by post-day parole outcome

Number of previous offences	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
0	3.57	0.00	3.57	1.19	84
1-3	7.27	1.82	8.18	2.73	110
4-10	23.31	12.03	15.04	3.76	133
11-20	36.97	21.01	21.19	9.24	119
21+	40.66	19.78	27.47	5.49	91
Chi-squared values	62.60	38.18	28.95	ns
(df = 4, N = 537)	p<.001	p<.001	p<.001

¹The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

 

Risk, Needs and Risk/Needs

The Community Risk/Needs Management Scale (CRNMS) was completed for offenders released into the community at the time of the study. The scale requires the parole officer to assess the offenders’ criminal history, risk and need level. Risk level (high or low) is determined on the basis of previous criminal history and other static risk factors. The level of need is rated as low, medium or high based on an evaluation of a set of twelve need areas for which the offender may require assistance. The full set of needs are listed in Chapter 2, and most have been shown to be criminogenic factors (see Andrews and Bonta, 1994). For a more complete discussion of the CRNMS see Grant et al. (1996) and Motiuk and Porporino (1989).

The CRNMS was not completed for all offenders, and other analyses indicate that those who did not have it completed were more likely to be higher risk offenders. As a result, differences between risk and needs groups are reduced, and thus underestimate the strength of the relationship (for details see Grant et al., 1996).

The relationship between risk and post-day parole outcome is shown in Table 3-9. Results in the table indicate that offenders rated ad high risk are three times more likely to be readmitted or to have a technical violation of their release. They are also twice as likely to have committed a new offence. Statistical tests indicate that these relationships are reliable. There were no differences between high and low risk offenders in violent re-offending.

Table 3-9: Risk by post-day parole outcome

Criminal History Risk Level	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases2
Low	10.55	3.52	8.08	3.02	199
High	28.37	14.18	18.44	4.26	141
Chi-squared values	17.79	12.84	8.14	ns
(df = 1, N = 340)	p<.001	p<.001	p<.01

¹The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

²The number of cases for these tables is lower than for the other tables because not all cases had the Community Risk/Needs Management Scale completed.

For evaluation purposes, three need levels are defined based on the number of needs identified. Low need offenders are those with zero to two needs, medium needs offenders have three needs identified, and high need offenders have more than three needs. Results presented in Table 3-10 show a clear relationship indicating that the need level increases, the probability of failure after day parole also increases. More specifically, only 7% of offenders with low needs were readmitted, while 35% of offenders with high needs were readmitted, a rate five times greater. For failure with a technical violation, the rate of high risk offenders doubles that of low risk offenders, and for failure with a new offence, the low need offenders have a failure rate of 6%, which triples to 21% for high need offenders. These relationships are statistically reliable. However, there is no difference across the need levels for failure with a violent offence.

Table 3-10: Need by post-day parole outcome

Case Need Level	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Low	6.80	6.80	5.83	3.88	103
Medium	18.39	5.17	13.29	2.29	174
High	34.92	17.46	20.63	4.76	64
Chi-squared values	21.05	9.81	8.12	ns
(df = 2, N = 341)	p<.001	p<.01	p<.05

¹The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

²The number of cases for these tables is lower than for the other tables because not all cases had the Community Risk/Needs Management Scale completed.

 

Risk level and need levels can be combined to produce a six-point scale of risk/needs as shown in Table 3-11. These results show a pattern very similar to that observed for needs. An increase in the level of risk/needs is clearly associated with an increase in the failure rate, with one notable exception. Offenders who are low risk, but high need have a higher readmission rate of reoffence rate than offenders who are high risk, but low needs. In addition, the high need/low risk offenders have the highest failure rate with a violent offence. This suggests that need levels are likely more important in predicting post-day parole outcome than risk level. The highest risk/need group is four to seven times more likely to recidivate (depending on the measure used) than the lowest risk/needs group. There is no statistically reliable difference between the risk/needs groups when recidivism is measured using violent reoffending; for most groups the rate is approximately 4%, with the exception of the high need/low risk group which had a failure rate of 7%.

Table 3-11: Risk/need by post-day parole outcome

>Risk/Need	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Low-Low	4.88	4.88	3.66	3.66	82
Low-Medium	11.76	1.96	8.91	.98	102
Low-High	38.57	7.14	21.43	7.14	14
High-Low	14.29	14.29	14.29	4.73	21
High-Medium	26.76	9.86	18.31	4.23	71
High-High	36.73	20.41	20.41	4.08	49
Chi-squared values	29.22	17.94	13.43	ns
(df = 5, N = 339)	p<.001	p<.01	p<.05

¹The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

²The number of cases for these tables is lower than for the other tables because not all cases had the Community Risk/Needs Management Scale completed.

 

Motivation

Motivation can be important mediating factors in trying to deal with problems associated with a criminal past. As part of the day parole study, an assessment was made to determine if the offender was motivated to participate in programs at the residential centre. Based on this assessment, an analysis was performed to determine the extent to which motivation contributed to the successful outcome of day parole and successful completion of the sentence.

Results indicate that level of motivation at the time of release on day parole was highly related to success on day parole (see Table 3-12). Only 16% of offenders rated as motivated were unsuccessful on day parole, whereas 48% of those classified as not motivated were unsuccessful. The strength of the relationship is reduced for post-day parole recidivism, but is still evident. That is, among offenders who were motivated to participate in day parole programs, only 21% were readmitted after full release, but 30% of offenders who were unmotivated were readmitted.

Table 3-12: Percentage failures by motivation and day parole outcome

Outcome	Motivated		Chi-squared values	Number of cases
	Yes	No
Day parole failure	16.33	48.00	c2 (1,N=492)=45.46 p<.001	492
Post-day parole failure(any new admission)	20.56	30.10	c2 (1,N=497)=4.23 p<.05	497

 

Dynamic risk predictors

Most predictors of offenders’ risk to offend are based on static factors, which cannot reflect positive changes over time. That is, these measures are based on factors such as offence pattern, previous offences and age at first offence. Bonta, Andrews and Motiuk (1993) have suggested that dynamic risk factors, which can change over time, may be more important in predicting successful release than static factors. As part of these analyses, two dynamic risk factors were tested to determine their relationship to day parole and post-day parole outcome. The two measured used were charged in institutional security and change in offender motivation.

Motivation

Change in motivation was determined by comparing the case manager’s rating of the offender’s expected response to supervision with the motivation level of the offender at the time of the day parole. Response to supervision is rated as either problematic or non-problematic when the offender is admitted to prison as part of the Forcefield Analysis of Needs (see Grant et al., 1996 for details). The coders reviewing reports in the offenders’ files rated motivation during day parole. Offenders were rated as either motivated or not motivated. Offenders who were rated as not having a problem in their response to supervision and who were motivated during day parole were classified as Motivated. Those initially classified as presenting a problem with supervision and who were later rated as motivated during day parole were classified as Changed. Finally, offenders who presented with a supervision problem prior to day parole, and who were not motivated on release were classified as Unmotivated. There were very few cases that showed a negative change from no problem with supervision to unmotivated. This group was therefore not included in the analyses.

Offenders in the Motivated group had a success rate on day parole of 83%, while only about half of the offenders in the Unmotivated group (47%) successfully completed their day parole. The Changed group had a day parole success rate of 78%. Offenders in the Unmotivated and Changed groups were both rated as having problems with supervision when they were admitted to prison. However, the Changed group adjusted and was motivated to succeed while on day parole, resulting in a 64% higher success rate than the Unmotivated group.

Table 3-13: Change in motivation by day parole completion

Change in Motivation	Day Parole Completion		Number of cases
	Successful	Unsuccessful
Motivated	82.51	17.49	223
Changed (to motivated)	77.48	22.52	151
Unmotivated	47.06	52.94	51
Chi-squared values (df = 2, N = 425)	29.13 p < .001

 

Results showing recidivism rates by change in motivation are presented in Table 3-14. The Motivated group had the lowest recidivism rates with only 16% readmitted and 12% committing an offence. Offenders who had exhibited a positive change in motivation, the Changed group, had a moderate rate of recidivism with 29% readmitted after release and 17% committing a new offence. Offenders in the Unmotivated group had the highest rates of recidivism with 42% readmitted and 26% committing a new offence. In addition, this group had the highest rate of new violent offences at 13%, between three and four times higher than the Motivated and Changed groups.

The relationship between change in motivation and both day parole outcome and recidivism is consistent. Offenders who were motivated had the highest success rates while those who were unmotivated had the lowest. However, the most interesting group in these analyses is the offenders for whom a change was identified. Had no change occurred, these offenders would have had the same result as the Unmotivated group. The positive change contributed to a 30% reduction in readmissions and a 35% reduction in new offences.

Table 3-14: Change in motivation by post-day parole outcome

Change in Motivation	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Motivated	15.63	8.48	11.61	4.02	224
Changed (to motivated)	28.95	18.42	17.22	2.63	152
Unmotivated	41.51	18.87	26.42	13.21	53
Chi-squared values	19.77	9.40	7.78	12.10
(df = 2, N = 429)	p<.001	p<.01	p<.05	p<.05

¹The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

 

Institutional Security

The second dynamic measure of risk studied was a shift in institutional security. Ideally, offenders move to increasingly lower levels of security during their sentence.

Alternatively, an offender’s institutional behaviour can result in a move to a higher level of security. It is also possible that a lower risk offender will be placed initially in a low security institution and therefore will have no opportunity for a shift in security level.

For these analyses, five categories of institutional security change were defined. Three categories include offenders who experienced no change in security level. Since the start point for each of these groups would reflect their risk level, a separate no change category was created for each institutional security level, thereby creating three categories: No change, minimum; No change, medium; No change, maximum. In addition to these categories, there were two change categories: Downward change (to a lower security level) and Upward change (to a higher security level).

The highest success rates for offenders while on day parole were for those offenders who started in a minimum security institution (84%) and those who had their prison security lowered (86%) (see Table 3-15). The offenders who experienced no change in security from medium to maximum institutions, and those who experienced an upward shift in security had day parole success rates ranging from 66% to 72%. A downward shift in security appears to be a good indicator of day parole success.

Table 3-15: Change in institutional security level by day parole completion

Change in Institutional Security Level	Day Parole Completion		Number of cases
	Successful	Unsuccessful
No change: minimum	84.11	15.89	107
No change: medium	68.61	31.39	223
Downward change	85.71	14.29	98
No change: maximum	72.22	27.78	18
Upward change	65.63	34.38	64
Chi-squared values (df = 4, N = 510)	18.62 p < .001

 

The relationship between change in security level and recidivism for the day parole offenders is presented in Table 3-16. The No change, minimum group have the lowest recidivism rates, with only 11% readmitted and 7% committing a new offence.

Offenders in the No change, medium and the Downward change groups had relatively similar recidivism rates at around 25% and new offence rates between 16% and 19%. The highest recidivism rates were for the No change, maximum and the Upward change groups which had readmission rates around 32% and reoffence rates around 22%.

Table 3-16: Change in institutional security level by post-day parole outcome

Change in Institutional Security Level	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
No change: minimum	11.21	5.61	6.54	1.87	107
No change: medium	24.34	11.50	15.56	2.65	226
Downward change	26.26	13.13	19.19	9.09	99
No change: maximum	33.33	16.67	22.22	5.56	18
Upward change	32.31	16.92	23.08	10.77	65
Chi-squared values	13.33	ns	11.02	16.31
(df = 4, N = 515)	p<.01		p<.05	p<.05

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

 

While a downward shift in institutional security level is associated with a relatively high success rate on day parole, it is also associated with a moderate level of recidivism after release. In both the day parole outcome and the recidivism analyses, an upward shift in security level was associated with greater risk of failure. A surprising finding in Table 3-16 is the relatively high rate of new violent offending for offenders who had a downward change in their security level. Approximately 9% of the downward change group committed a new violent offence, which is very close to the 11% for offenders who had an upward shift. It is unclear whether this is an anomaly in the data, but it should be examined in future studies.

Regional differences

Recidivism rates for each group are presented in Table 3-17. The rates vary across the five regions, but these variations are not statistically reliable as indicated by the Chi-squared tests. However, it is interesting to look at the trends. The Pacific region has the lowest readmission rate and the lowest failure rate for both new offences and violence offences. Three regions have readmission rates around 25%, but the Atlantic region is unique, with the highest rate of readmission with a new offence (24%) and one of the lowest rates for technical violations (8%).

Table 3-17: Region by post-day parole outcome

Region	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Atlantic	27.42	8.06	24.19	6.45	62
Quebec	25.42	11.02	16.10	5.93	118
Ontario	20.28	8.39	14.08	5.60	143
Prairies	26.24	14.18	14.89	4.26	141
Pacific	16.48	15.38	9.89	1.10	91
Chi-squared values	ns	ns	ns	ns

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

 

Aboriginal Offenders

The sample of Aboriginal offenders includes 74 males who completed day parole in 1990-91. Of these, eight men were not included in the follow-up analyses because they were not released on ordinary day parole, and seven did not reach their warrant expiry date by the end of the study period.

Post-Day Parole Release

The type of release granted after completion of the day parole is presented in Table 3-18. The results in the table indicate that slightly more than one third of Aboriginal offenders were released on full parole after they completed their day parole, compared to about 50% of male day parole offenders. An additional 24% of Aboriginal offenders were granted statutory releases at the end of their day parole, compared to about 21% of male day parole offenders. Finally, approximately 41% of the Aboriginal offenders were required to serve some additional time in custody prior to being granted either full parole or statutory release. This compares to about 31% of male day parole offenders. Overall, these results indicate that Aboriginal offenders are more likely to be reincarcerated before release after a day parole, and are less likely to be granted full parole after a day parole release.

Aboriginal offenders granted full parole at the end of their day parole were more likely to fail (38%) than those granted statutory release (29%). However, consistent with earlier results, reincarceration after day parole was associated with an increased likelihood of failure for Aboriginal offenders, while the failure rate reaching 67%. Success following the day parole period was poorer for Aboriginal offenders compared to male day parole offenders. While 80% or more of the male offenders were successful after release on full parole or statutory release, the success rate was less than 70% for Aboriginal offenders.

Aboriginal offender who were successful on day parole were less likely to be readmitted prior to the end of their sentence, or to have a technical violation of their conditional release than Aboriginal offenders who did not complete their day parole successfully. These results are shown in Table 3-19. However, the likelihood of committing a new offence, or of committing a new violent offence was not associated with day parole outcome. Approximately 27% of Aboriginal offenders who received day parole committed a new offence prior to the warrant expiry. New violent offences were committed by 10% of Aboriginal offenders who had been successful on day parole and by 15% of Aboriginal offenders who had been unsuccessful on day parole.

Table 3-18: Type of release after day parole release and post-day parole outcome (Aboriginal offenders)

Type of Release	Number of cases	Percentage of cases	Percentage who successfully completed their sentence
Full parole	21	35.6	61.9
Statutory release (including subsequent day parole)	14	23.7	71.4
Full parole or statutory release after incarceration	24	40.7	33.3

 

Table 3-19: Percentage of post-day parole failures by type of failure and day parole completion (Aboriginal offenders)

Day Parole Completion	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Successful	38.00	14.29	28.00	10.00	50
Unsuccessful	64.71	44.12	26.47	14.71	34
Percentage of all day parole sample cases	48.81	26.51	27.38	11.90	84
Chi-squared values (df = 1, N = 84)	5.78	9.17	ns	ns

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group.

 

Time of day parole release

Analyses were conducted to examine the relationships between time of day parole release and the outcome measures, as shown in Table 3-20. Although no relationships between time of release and outcome were statistically significant, several trends were noted.

The overall pattern suggests higher rates of readmission and technical violations for Aboriginal offenders released the earliest and latest in their sentence (i.e., a U-shaped curve). However, based on the risk principle, one would anticipate higher rates of readmission for offenders released later in their sentence (i.e., they are typically higher risk offenders and therefore present a higher risk to reoffend). Such findings seem counter-intuitive for offenders released early in their sentence, suggesting that they were not prepared for conditional release. These findings may be contrasted to those obtained for the male day parole sample, which suggested that offenders released earlier in their sentence evidenced lower rates of recidivism.

Overall, Aboriginal offenders reoffended with a new offence at higher rates than the full male sample (18% to 40% compared to 12% to 20%, respectively). Similarly, Aboriginal offenders released early in their sentence were three times more likely to be convicted of a violent offence than offenders in the male day parole sample (3% to 7% compared to 6% to 20%, respectively).

Table 3-20: Post-day parole outcome by time of day parole release (Aboriginal offenders)

Time of day parole release2	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Pre-full parole eligibility	50.00	31.25	25.00	6.25	16
before 25% of time to statutory release date	46.67	13.33	40.00	20.00	15
before 50% of time to statutory release date	36.36	18.18	18.18	18.18	11
before 75% of time to statutory release date	40.00	20.00	20.00	0.00	5
75% of time to statutory release date	54.55	36.36	18.18	0.00	11
Chi-squared values	ns	ns	ns	ns

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

² The period from the full parole eligibility date and the statutory release date was divided into four equal parts.

 

Number of previous offences

The relationship between number of previous offences (prior to the day parole sentence), day parole outcome and post release recidivism was also examined for Aboriginal offenders.

Aboriginal offenders with 10 or fewer previous offences are generally successful on day parole, with over 90% completing their day parole release without problems. However, Aboriginal offenders with more than 10 previous convictions have success rates of approximately 50%, dropping to less than 20% for offenders with more than 20 previous convictions. It should be noted that the size of these groups is relatively small, making the results less reliable. However, the Chi-squared test indicates that the relationship is statistically reliable.

Table 3-21: Number of previous offences (prior to the current admission) by day parole completion¹ (Aboriginal offenders)

Number of previous offences	Successful	Unsuccessful	Number of cases
1-3	90.91	9.09	11
4-10	90.91	9.09	11
11-20	52.63	47.37	19
21+	16.67	83.33	12
Chi-squared values	19.95
(df = 3, N = 53)	p<.001

 

The relationships between the number of previous convictions and the four recidivism measures are not statistically reliable as shown in Table 3-22. However, generally (with the exception of violent offending) they reflect a consistent pattern of increased recidivism with increased volume of criminal activity (number of previous convictions).

Most notably, offenders who were convicted of 11 or more offences were readmitted at more than twice the rate (50%) of offenders with less than four offences (18%).

The analyses also indicate that Aboriginal offenders with a greater number of previous convictions received more technical violations while on full release. Offenders with only one to three previous offences received no technical violations, whereas 27% of those in the mid-range (4-10) were convicted for technical offences. Offenders with 11 to 20 previous offences committed approximately the same percentage of technical violations (25%) as offenders in the previous category. More than 40% of Aboriginal offenders with 21 or more previous offences received technical violations during full release.

There is less variability with respect to previous convictions and new offences of full release. Again, offenders convicted of 11 or more previous offences showed higher rates of new offending (25%) than offenders convicted of 10 or less (18% or less). Conversely, the rate of violent reoffending remained relatively consistent across number of previous convictions.

These results demonstrate the value of considering the level of criminal activity in determining the likelihood of a positive outcome for both day parole and full release.

Table 3-22: Number of previous offences by post day parole outcome (Aboriginal offenders)

Number of previous offences2	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
1-3	18.18	0.00	18.18	9.09	11
4-10	36.36	27.27	18.18	18.18	11
11-20	50.00	25.00	25.00	10.00	20
21+	66.67	41.67	33.33	8.33	12
Chi-squared values	ns	ns	ns	ns

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

² One offender had no previous convictions.

 

Comparing these results to those from the male sample leads to the conclusion that while the pattern of results are similar, the recidivism rates are approximately 50% higher for Aboriginal offenders.

Female Offenders

The entire population of female offender’s (35) who completed day parole in 1990-91 was examined in the day parole report. The follow-up sub-sample consisted of only 26 women; three were excluded because they did not reach their warrant expiry date by the conclusion of the follow-up period and six did not receive ordinary day parole. Given the very small number of women in the sample, the following results should be interpreted cautiously.

Post-Day Parole Release

The majority of women (58%) were released on full parole or statutory release immediately following day parole or after completing a subsequent day parole. None of these women were readmitted to an institution prior to the expiry of their sentence. The success rate for women on full release was substantially poorer (73%) when full parole or statutory release was preceeded by a period of incarceration.

While 42% of females were reincarcerated after their day parole, only 30% of males were reincarcerated after their day parole. After full release, however, a greater percentage of female offenders successfully completed their sentence without readmission to a federal institution. An analysis of day parole outcome and full release outcome for female offenders revealed that those readmitted, and those who committed new offences, had been unsuccessful during their day parole.

Table 3-23: Type of release after day parole release and post-day parole outcome (female offenders)

Type of release	Number of cases	Percentage of cases	Percentage who successfully completed their sentence
Full parole and statutory release (includes SR after subsequent day parole)	15	57.7	100
Full parole and statutory release after incarceration	11	42.3	72.7

 

Table 3-24 presents time of day parole release in relation to full release outcome. As the numbers are small, the results are presented for the pre-full parole period, the period between parole eligibility and one half of the time to statutory release, and after one half of the time to statutory release. Of women released prior to their full parole eligibility date, 8% were readmitted to an institution. A total of 20% of women released before half of the time to statutory release date were readmitted, and no women who were released after this time were readmitted to an institution prior to warrant expiry. Although women released prior to their full parole eligibility committed no technical violations, over 20% of women released after their parole eligibility date were convicted of a technical violation. Conversely, whereas women released after parole eligibility committed no new offences, 8% of women released prior to full parole eligibility were convicted of a new offence. No women were convicted of violent offences after release on either full parole or statutory release.

Table 3-24: Post-day parole outcome by time of day parole release (female offenders)

Time of day parole release2	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
Pre-full parole eligibility	8.33	0.00	8.33	0.00	12
From parole eligibility day to 50% of sentence	20.00	20.00	0.00	0.00	10
From 50% of sentence to statutory release date	0.00	25.00	0.00	0.00	4
Chi-squared values	ns	ns	ns	ns

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

² The period from the full parole eligibility date and the statutory release date was divided into four equal parts.

 

Number of previous offences

The number of previous offences is based on a count of all offences prior to the index sentence. The number of offences was then used to create three offence groups: no previous offences, one to 10 offences and 11 or more offences. First time offenders were not convicted of any technical violations or new offences, and thus received no admissions after their day parole completion. There were few differences in the readmission rates and technical violations for offenders convicted of any previous offences; 18% of the sample with one to 10 previous offences were readmitted (for a technical violation) and 14% with 11 or more previous offences were readmitted (for a new offence and technical violation). No violent offences were committed by the female sample on full release.

Table 3-25: Previous offences by post-day parole outcome (female offenders)

Number of previous offences	Type of Post-Day Parole Failure1
	Re-admission	Technical violations	Offence	Violent Offence	Number of Cases
0	0.00	0.00	0.00	0.00	8
1-10	18.18	18.18	0.00	0.00	11
11+	14.29	14.29	14.29	0.00	7
Chi-squared values	ns	ns	ns	ns

¹ The failure groups are not mutually exclusive; an offender can be represented in more than one group. Percentages are the percent who failed from the total number of cases shown in the last column.

 

DISCUSSION

Day parole assists offenders in the transition from institution to community. For low risk offenders, it is a form of conditional release providing an opportunity to become contributing citizens before eligibility for full parole. For higher risk offenders, it is an opportunity to adjust to life outside of prison and to prepare for late release on full parole or statutory release. For the Correctional Service of Canada, it is a means of controlling the risk that offenders pose to the community by ensuring a stable and controlled living environment during the most critical period after release from prison. It also provides a method of determining if an offender will be able to remain in the community on conditional release.

A major finding of this study is the significant relationship between day parole success and outcome during a subsequent full release. Specifically, 85% of offenders who successfully completed their day parole also reached the end of their sentence without further intervention. In addition, 10% of offenders who were successfully on day parole committed a new crime while in the community before the end of their sentence. However, of offenders who were unsuccessful in their day parole, 56% completed their sentence without a readmission to prison, and 30% committed a new offence. Day parole is therefore an important indicator of success on full release.

Outcome measures from those released on day parole can be compared to those in the general offender population taken from a study by Grant (1996) which provides recidivism rates for a two-year follow-up of full parole and statutory release cases. Of the offenders released on full parole following day parole, only 13% were readmitted prior to the end of their sentence, whereas for the general population of offenders, 23% of full parole releases were readmitted. Likewise, only 22% of offenders granted statutory release after day parole were readmitted, while only 50% of offenders granted statutory release in the general population were not readmitted within two years.

Day parole also increased the likelihood of release on full parole. While 57% of offenders granted day parole received full parole, only about 38% of offenders in the general population received full parole.

When post-day parole outcome was investigated using a measure of time in the community, it was shown that offenders who were successful while on day parole served a much greater percentage of their sentence, after the day parole, in the community. This finding provides support for the argument that day parole may provide important cost benefits over and above any reduction in community risk which might be associated with a day parole release.

These results, taken together, suggest that day parole may increase the likelihood of success on full release, either on full parole of following statutory release. While some of this effect may be due to selection factors (i.e., the lower risk offenders are selected for day parole), it is also possible that some of the effect is the result of the day parole program. Additional analyses are required to isolate the effects of selection and day parole.

Multiple day paroles appear to be useful in some cases. While the number were small (only 6% of the sample), offenders who received multiple day paroles had the highest day parole success rate (over 90%). In addition, 90% of these offenders successfully completed their sentence.

Offenders who were reincarcerated prior to full release had high post-release failure rates, ranging from 30% to 40%. The inability to function effectively within the halfway house environment may be an important indicator of problems the offender will have after full release, whether on full parole or statutory release. These offenders should be targeted with additional resources at the time of the release to increase the likelihood of completing the sentence without a return to an institution.

Offenders released prior to their parole eligibility date were low risks. Nearly half of the sample was released prior to their parole eligibility date and 83% of these offenders successfully completed their sentence. Only 12% of the early release offenders committed a new offence before the end of their sentence. Offenders released on day parole after their parole eligibility date had higher recidivism rates, with the highest rate of readmissions (33%) for offenders released shortly before the half waypoint in their sentence.

While offenders released later on day parole have elevated failure rates, these rates do not vary a great deal for those released after one half of their sentence. That is, offenders released in the latter part of their sentence, shortly before their statutory release date did not increase the risk to society. It would appear that using day parole as preparation for statutory release is effective and does not increase the risk to society.

Given the clear relationship between success on day parole and post-day parole outcome, offenders who have had an opportunity to address their criminogenic needs at the institution should be considered as candidates for day parole, even if their level of risk is elevated. The day parole release will indicate if the offender’s programming achieved its goal, while still keeping the offender closely supervised. In this way, risk to society can be monitored and reduced.

As has been shown in previous research (Andrews and Bonta, 1994; Grant, et al., 1996; Motiuk & Porporino, 1989), this study demonstrates that risk, needs and risk/needs are good predictors of recidivism. In addition, the results indicate that among offenders granted day parole, criminogenic needs may be a more important indicator of serious problems than risk level. The high needs offenders, whether or not they were high risk, had the highest recidivism rates. Prior to, and during day parole, these offenders may require additional attention to ensure that their criminogenic needs are addressed, and that these need areas do not interfere with their ability to function in the community.

Results of the study show that motivation is an important factor in determining outcome, both on day parole, and during the completion of the sentence in the community. Offenders who were motivated, based on ratings by coders, were much more likely to succeed. This suggests that there is a need to explore the use of motivation as a factor in predicting release outcome.

It has been suggested that dynamic risk factors are important indicators of an offender’s success after release. Two dynamic risk factors, change in motivation and change in institutional security level, were studied to determine their relationship to post-day parole outcome. A change from unmotivated to motivated resulted in a 65% increase in the success rate of day parole, and a 21% increase in the success rate during full release. Likewise, a change in institutional security to a lower level was associated with a success rate on day parole equivalent to that of offenders released from minimum security institutions.

The number of previous convictions, which was not examined in the previous day parole study, shows a clear relationship with day parole outcome. As the number of previous criminal offences increases, the likelihood of failure on day parole increases. For example, offenders with less than four previous convictions had a day parole success rate of 85%, while those with more than 10 convictions had a success rate of only 68%.

The number of previous convictions was also associated with outcome after the day parole period. While offenders with less than four previous convictions had a readmission rate of 7%, those with more than 10 previous convictions had a readmission rate of almost 40%. Similarly, for new offences, offenders with more than 21 previous offences were three times more likely to have committed a new offence after full release (27%) than offenders with less than four previous convictions (8%).

Within the group of day parole cases studies here, the Pacific region had the lowest readmission rate of new offence rate compared to the other regions. While the Atlantic region had readmission rates similar to other regions, they had the highest percentage of readmissions with a new offence (24%).

Aboriginal offenders were less likely to have had their day parole followed by full parole and were more likely to be reincarcerated after their day parole than the day parole offenders in general. In addition, even among the lower risk Aboriginal offenders (those released early in their sentence), Aboriginal offenders had a higher failure rate on day parole. The recidivism rates for Aboriginal offenders were about 50% higher than for the other offenders. These results suggest that day parole may not be meeting the needs of Aboriginal offenders. Additional work may be needed to ensure that these offenders benefit from the day parole experience and receive effective programming and support in the community to reduce the likelihood of returning to prison.

Female offenders were slightly more likely to be reincarcerated after their day parole, but prior to full release, than their male counterparts. However, they were also more likely to complete their sentence, after release on full parole or statutory release, without readmission to an institution, and were less likely to commit a new offence.

Overall, the results of this study indicate that day parole does not substantially increase community risk, even when offenders are released close to their statutory release date. The post-day parole success of offenders suggests that the selection of cases for day parole is effective for ensuring community safety. Given that that goal of corrections is to safely reintegrate the offenders into the community, having a portion of the sentence served in the community is obviously useful.

An additional benefit of day parole is the potential cost savings that can occur. It is far less expensive to maintain an offender in a Community Correctional Centre or a Community Residential Centre than to maintain them in prison. Finally, there is also evidence that delivering treatment programs in the community can be more effective than in prison (Andrews and Bonta, 1994).

Day parole is an important component in the gradual release process that permits the effective management of offender risk and needs while ensuring community safety. For low risk offenders, day parole provides a form of early release and it is a method of preparing and assessing higher risk offenders for full release.

REFERENCES

Andrews, D.A. and Bonta, J. (1994). The psychology of criminal conduct. Cincinnati, Ohio: Anderson Publishing.

Grant, B.A., Motiuk, L., Brunet, L., Courturier, P. and Lefebvre, L. (1996). Day parole program review: Case management predictors of outcome. Ottawa: Research Branch, Correctional Service of Canada.

Grant, B.A. (1996). Inmates referred for detention (1989-90 to 1993-94): A comparative analysis. (Report #R-45). Ottawa: Research Branch, Correctional Service of Canada.

Corrections and Conditional Release Act, (R SC, C-20, 1992).

Parole Act, RSC, C-2 (1985).

SAS Institute Inc. (1990). SAS Language: Reference, Version 6, First Edition. Cary, NC: SAS Institute Inc.

SAS Institute Inc. (1990). SAS Version 6.11. Cary, NC: SAS Institute Inc.

SAS Institute Inc. (1990). SAS Sharing Version 6.10. Cary, NC: SAS Institute Inc.

APPENDIX A Coding Manual

FPS:_________________

I. Previous criminal history

1. Number of sentences: ______

2. Offences - first sentence:

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______

3. Offences - second sentence:

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______

4. Offences - all additional sentences:

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______

II. Current DP sentence

5. Current offences (sentence for which offender was released on DP):

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______

III. DP Period

6. Estimated DP offences (sentenced up to approximately 6-12 months after DP completion and they were incarcerated):

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______
Terminated* ______

* terminated and readmitted for reason other than technical violation (but with no new offences)

7. Estimated offences that occurred during subsequent DP releases, and prior to full releases (i.e., FP and/or SR releases):

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______
Terminated * ______

* terminated and readmitted for reason other than technical violation (but with no new offences)

8. Reached WED from start of study-DP with no new admission and/or no new offence (include DP period and post-DP completion):

Yes ______
No ______
Lifer ______
Did not reach WED by March 31, 1994 ______
N/A - no release after DP ______

IV. Post study-DP (post DP completion)

9. First full release after DP completion:

FP immediate ______
SR immediate ______
FP after subsequent DP(s) ______
SR after subsequent DP(s) ______
FP after incarceration* ______
SR after incarceration* ______
No release after DP ______

* incarceration of more than 14 days

Note: If study-DP is followed by subsequent DPs, then by incarceration and FP or SR: (1) if DP and/or subsequent DPs were successfully completed, code FP/SR after subsequent DP(s); (2) if DP and/or subsequent DPs were unsuccessfully completed (e.g., revoked), then code full release as FP/SR after incarceration

10. Immediacy of release after study-DP and/or series of DPs following immediately from study-DP:

Immediate ______
Within 2 months ______
More than 2 months ______
No release after DP ______

11. First readmission after FP or SR release and before WED:

N/A - no release after DP completion ______

No readmission ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______
Termination* ______

* terminated and readmitted for reason other than technical violation (but with no new offences)

12. All subsequent offences after FP or SR release and before WED:

N/A - no release after DP completion: ______

No offences ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______
Termination* ______

* terminated and readmitted for reason other than technical violation (but with no new offences)

13. Potential community time from study-DP completion to original WED:

Study-DP completion: __________________
WED: __________________

Did not reach WED by March 31, 1994: ______

14. Actual community time from study-DP completion to original WED:

Study-DP completion: __________________
WED: __________________

Did not reach WED by March 31, 1994: ______

Admissions post-DP completion but prior to FP or SR (accounts for subsequent DP releases and admissions) up to original WED:

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Study-DP completion: __________________
Post-DP release (FP or SR): __________________

N/A - no release after DP completion: ______

Admissions and releases post-FP or SR release up to original WED:

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Additional incarceration (days): __________

15. Actual community time from original WED to March 31, 1994:

Did not reach WED by March 31, 1994: ______

WED: __________________
March 31, 1994: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Additional incarceration (days): __________

16. Other post-DP release(s) (number) up to original WED (after first full release):

None ______
DP ______
FP ______
SR ______

V. Post original WED

Did not reach WED by March 31, 1994: ______

17. All time served after original WED (i.e., new sentences after completion of study- DP sentence - new SENTID):

No admissions: ______

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Admission date: __________________
Release date: __________________

Additional incarceration (days): __________

18. First readmission post-original WED:

No admissions ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______
Termination* ______

* terminated and readmitted for reason other than technical violation (but with no new offences)

19. All subsequent readmissions post-original WED:

No admissions ______
Homicide ______
Sex offence ______
Violence against person ______
Violence - other ______
Drug offences ______
Property ______
Non violent - other ______
Technical violation ______
Termination* ______

* terminated and readmitted for reason other than technical violation (but with no new offences)

20. Other post-WED releases (number) (any releases after original WED):

None ______
DP ______
FP ______
SR ______

FPS Coding

FPS: __________________

Date of first adult conviction: __________________

Pre-admission to federal system (should be prior to adm date):

Number of offences:

    Number of non-violent criminal code offences: ______

    Number of violent criminal code offences: ______

Type of sentence:

    Number of fine/conditional discharges/etc. ______

    Number of probation sentences: ______

    Number of jail sentences: ______

        Less than 30 days: ______

        Greater than 30 days: ______

Post completion of day parole sentence:

Offender reached WED: ______ Y ______ N
(if no, then finished with this case)

Number of offences:

    Number of non-violent criminal code offences: ______

    Number of violent criminal code offences: ______

Type of sentence:

    Number of fine/conditional discharges/etc. ______

    Number of probation sentences: ______

    Number of jail sentences: ______

    Less than 30 days: ______

    Greater than 30 days: ______