Skip to content | Skip to institutional links

Common menu bar links

Institutional links

Research Reports

Warning This Web page has been archived on the Web.

Work Release Program: How it is used and for what purposes

Brian A. Grant and Chris A. Beal
Research Branch
Correctional Service of Canada

February, 1998

This report is part of a series of 24 research/evaluation reports (listed below) that were prepared as background to the Consolidated Report of the Working Group studying the provisions and operations of the Corrections and Conditional Release Act and related Consultation Paper.

The Working Group is composed of representatives from the following agencies:

Correctional Service Canada
National Parole Board
Correctional Investigator
Justice
Department of the Solicitor General
Research/Evaluation Reports:
Information about Offenders
Security Classification of Inmates
Judicial Determination

The Temporary Absence Program: A Descriptive Analysis

Personal Development Temporary Absences
Work Release Program: How it is used and for what purposes
Day Parole: effects of the CCRA (1992)
Case Management: Preparation for Release and Day Parole Outcome
Accelerated Parole Review
Statutory Release and Detention Provisions
Community Supervision Provisions

Provisions Relating to Victims

Observers at National Parole Board Hearings
The National Parole Board Registry of Decisions
CSC Human Resources
Administrative Segregation
Search, Seizure and Inmate Discipline
Offender Grievance System
Urinalysis Testing Program
Inmate's Input in Decision-making
Information to Offenders
Aboriginal Offenders
Health Services
Women Offenders

Executive Summary

The report examines the use of work release in the Correctional Service of Canada as part of the five year review of the Corrections and Conditional Release Act (CCRA).

Work release provides opportunities for offenders to work away from the institution, but generally requires a return to custody or a halfway house each day. Work releases are granted by the Correctional Service of Canada and may be used for projects that directly meet the needs of the offender or for projects that provide services to the community. Unlike other jurisdictions, where work release is an early form of parole or statutory release, within the Correctional Service of Canada it may be granted at any time in the sentence and the offender generally returns to full custody when the work period is completed.

From the implementation of the CCRA in 1992 until September 30^th, 1996 there were a total of 4,659 work release granted to 1,167 offenders. Results for the last two years suggest that there are about 800 work releases per year granted to approximately 300 offenders per year.

Regional differences in the number of work releases granted were evident. While the Pacific and Ontario regions had the highest average number of work release, the Atlantic and Pacific regions granted work releases to the largest percentage of their on-register offenders (3%).

Female offenders accounted for about 1% of work releases, slightly less than their representation of the inmate population (2% to 3%). Aboriginal offenders accounted for 8% of the work release population, also less than their representation in the inmate population (11% to 12%).

Offenders serving determinate sentences spent an average of 3 years incarcerated before receiving a work release while offenders serving life sentences served an average of 12 years incarcerated before their work release.

Work release appears to be associated with an increased chance of being granted day parole. Only 24% of the work release population who applied for a day parole prior to their first work release were granted a day parole, whereas 73% of those who applied for day parole after their work release were granted day parole. Slightly more than half of the work release offenders were released on their statutory release date with the balance receiving full parole. After their full release (full parole or statutory release), 35% of offenders granted a work release were readmitted and one-fifth were readmitted for a new offence. The recidivism rates for work release cases was slightly higher than for offenders released on day parole.

A detailed file review was also conducted on a sample of the total work release population. These analyses suggest that documentation for the work release is not always complete, but critical issues such as risk to the community are usually evaluated. For two-thirds of the work releases the goal was preparation for release.

File review data also indicated that most work releases are for manual labour tasks with more than one-quarter providing community assistance for non-profit groups. Approximately one quarter of work release are supervised by CSC staff, one-third are supervised by organizations outside of CSC and one-fifth are supervised by private citizens.

Work releases provide an opportunity for offenders to be engaged in productive activities outside the prison. The results suggest that work release meets the general goals of the CCRA, including providing preparation to offenders for their eventual release. In addition, experience on a work release is associated with increased likelihood of a day parole release.

Acknowledgements

A number of people were involved in this project. Ray Belcourt extracted the data on work releases from OMS and provided other data that was used for the follow-up. Marie-France Lapierre assisted with the development of the coding manual for the file review and coded most of the cases. David Joubert also assisted with the coding. Laura Vandette assisted with the final formatting of the report. The National Parole Board provided assistance with identifying day parole cases pre and post-work release.

Executive Summary
Acknowledgements
List of Tables
List of Figures
CHAPTER 1: INTRODUCTION
CHAPTER 2: METHODOLOGY
CHAPTER 3: RESULTS
CHAPTER 4: REVIEW OF FILE INFORMATION
CHAPTER 5: DISCUSSION
References
Appendix A Coding Manual for Work Release Study

List of Tables

Table 1: Number of work releases, offenders granted work release and on-register offenders by fiscal year
Table 2: Number of work releases, offenders granted work release and on-register offenders by region and fiscal year
Table 3: Percentage of offenders by number of work releases days and region
Table 4: Percentage of offenders having at least one federal conviction for each type of offence prior to first work release
Table 5: Most serious offence as a percentage of offenders for current admission
Table 6: Number of day parole applications before and after first work release
Table 7: Percentage of offenders released at or before specific points in their sentence (offenders with determinate sentences)
Table 8: Percentage and number of reintegration ETAs per offender before first work release
Table 9: Percentage and number of reintegration UTAs per offender before first work release
Table 10: Completion of Work Release Departures
Table 11: Release type after work release
Table 12: Admission type after completion of work release
Table 13: Percentage of offenders with criteria mentioned in progress summary report
Table 14: Specific Criteria mentioned in Progress Summary Report
Table 15: Percentage of work releases designed to achieve goals in correctional plan
Table 16: Correctional goals outlined for each offender
Table 17: Type of work placement
Table 18: Areas addressed in work release evaluation reports
Table 19: Additional conditions imposed on work release

List of Figures

Figure 1: Median number of work releases days per offender by region and fiscal year
Figure 2: Sponsoring agencies for work release
Figure 3: Supervising agencies for work releases

Chapter 1: Introduction

The Corrections and Conditional Release Act (CCRA), which was implemented in November 1992, made a number of changes to the law governing the release of inmates from federal prisons in Canada. These changes affected temporary absences, day and full parole, and statutory release.

One important change in the CCRA was the introduction of work release. Work release provides opportunities for offenders to work away from the institution, but generally requires a return to custody, or a halfway house each day. Granting authority for work releases rests with the Correctional Service of Canada and these releases do not require approval from the National Parole Board. The supervision and other criteria for work releases make them similar to temporary absences. However, the length of work releases, 60 days with opportunity for renewal, makes them similar to day parole.

In many jurisdictions (Katz & Decker, 1982; Marlette, 1990; Waldo & Chircos, 1974a; Waldo & Chircos, 1974b; Waldo & Chircos, 1977) work release is used to provide full time employment opportunities prior to full parole or sentence completion, but its function is different in Canada.

In the Canadian federal prison system, the purpose of work release is to provide offenders with meaningful work opportunities at any point in the custody portion of their sentence, not just close to their release date. Work releases provide offenders with opportunities to work outside the prison as part of community projects, in support of non-profit organizations and for paid employment such as work on forest fire crews and crop harvesting.

While the work project must be clearly specified along with the conditions of the work, including supervision, it is not necessary that the work be directly related to the offenders correctional plan. In this way, work release is a very flexible program that allows correctional managers to respond to community projects and local needs for labour. While the releases do not need to meet correctional goals, it is likely that they provide useful opportunities for community restitution, development of work habits and in some cases, skills that can be used after release.

Prior to the introduction of the CCRA (1992), the Correctional Service of Canada, with the approval of the National Parole Board, could establish a day parole special project which would allow offenders to work on community projects. The CCRA changed the criteria for day parole such that it had to be in preparation for full release and it could only be granted six months before the parole eligibility date. As a result of these changes, day parole special projects could no longer meet the requirements for participation in community work projects and seasonal employment opportunities.

Relative to other short term release programs, work release is used with a relatively small number of offenders . In 1995-96 approximately 315 offenders were released on a work release while 2,000 offenders were released on escorted temporary absences, 800 were released on unescorted temporary absences, and 2,600 offenders were released on day parole.

The CCRA required that a review of its provisions be made five years after it came into force. The review is to be completed at the end of 1997 and this report is one of a number which address various components of the CCRA. Other reports related to work release include two on the temporary absence program (Grant & Millson, 1998; Grant & Johnson, 1998) and two on day parole (Grant & Gal, 1998; Grant, 1998).

The study was designed to provide answers to the following questions:

How many work releases have there been since the CCRA, how many offenders have been affected, and has the frequency of use changed since work releases became available?
What are the characteristics of offenders granted work releases?
What types of full releases do offenders receive after they have been granted a work release?
What is the success rate of work releases?
Are detailed plans prepared for the work releases?
What are the supervision requirements while on work release?
How frequently are work releases used for different purposes?

The next chapter of the report presents the methodology used to obtain the information used in the study. This is followed by a description of the results for all work releases and for a sample of work releases for which case file information was obtained. The final chapter addresses the questions presented above and provides a discussion of work release in the context of other release programs.

Chapter 2: Methodology

Data Development

Data for this study was obtained from the Offender Management System (OMS) of the Correctional Service of Canada. OMS is an automated administrative records system that includes records for all offenders under the jurisdiction of the Correctional Service. As an administrative data system it is designed to meet the day to day requirements of managing approximately 25,000 offenders.

Due to the constant updates of new records within OMS, analysis becomes difficult since the number of cases and case descriptions constantly change. As a result, specialized static data sets are created from the administrative records to ensure a stable database from which to conduct research. A consequence of using a specialized database is that the numbers presented in the report may not match official figures presented elsewhere, although the differences should be minimal and have no significant effect on the conclusions reached.

Databases that have been created from OMS contain problems that are found in most administrative systems. In certain instances there are duplicate records of events or changes in personal identifiers which make it difficult to track offenders though time, and individual data fields are not always up to date. These problems can generally be corrected by obtaining information from different components of OMS. In the worst case scenario, cases may be excluded because they lack critical information, but the number of cases lost this way is a very small percentage of the total.

A specialized data set of work releases was created from the national OMS database. It included all work releases from November 1992, when work releases were permitted by the CCRA, to September 30^th, 1996. During the study period there were a total of 4,569 work releases granted to 1,167¹ offenders. When presenting annual trends the years 1992-93 and 1996-97 are excluded because they include data for less than six months. The number of work releases is based on the number of departures from institutions and is not a count of the number of work release programs. The number of offenders granted work releases may be a better indicator of the number of work release programs, but this underestimates the number, since many offenders are granted more than one work release program.

The study also makes use of release, admission and offence databases created for the study. These databases make it possible to conduct follow-up analyses and to collect information on previous criminal activity.

A second data set was created to obtain more detailed information on work releases. This data set includes cases from the fiscal year 1994-95 and was designed to provide information contained in text-based records. These text-based records describe work release in more detail than is available in the other data sets. A coding manual was developed to obtain information on the type of activities involved in work release, the objectives of work releases, and the periods of involvement for individual offenders on work releases etc. Records for 223 offenders were coded by two research assistants. The coding manual is presented in Appendix A and was designed to provide forced choice responses to items. In cases in which the file information was written in French, a francophone read and coded the data, while files that were written in English were coded by an anglophone. All file information was obtained from OMS.

Analyses

Analyses conducted on the full work release data set use the full population of cases. This eliminates the need to use statistical techniques that estimate population parameters. Any observed differences are actual differences and do not need inferential statistics to determine if they are statistically reliable. The only criteria to apply when comparing groups in the study is whether the observed differences are meaningful from a correctional point of view.

The analyses that were conducted on the file data are primarily descriptive and are designed to show how work releases are used.

Chapter 3: Results

All Work Releases

Work release data were obtained for five fiscal years, from 1992-93 to 1996-97. Some analyses are based on all cases (4,659 work releases and 1,167 offenders), but when trends over years are presented only the three full fiscal years (1993-94 to 1995-96) are included. The other two years contain only partial data and are therefore not appropriate for trend analysis. Prior to November 1992, work release was not available and thus there are no pre-CCRA comparisons².

The number of work releases, offenders granted work releases and the on-register population for each fiscal year from 1993-94 to 1995-96 are presented in Table 1. The number of work releases granted over the three-year period was 3,751, with almost 60% of those occurring in the year 1993-94. Upon further investigation, it was found that the majority (1,608) of work releases that were granted in 1993-94 were from the Pacific region, and almost all of them were for one day (1,553). In the next year, the number of work releases granted decreased significantly, suggesting a shift from one day work releases to multiple day work releases. Results for the last two years suggest there were about 800 work releases per year. Although the total number of work releases varied over the years, the number of offenders granted a work release remained consistently around 300 per year indicating that the use of this release option has not increased as case managers became more familiar with its use.

Table 1: Number of work releases, offenders granted work release and on-register offenders by fiscal year
Fiscal Year	Work Releases	Offenders	On-register Offenders
1993-94	2,165	300	13,863
1994-95	742	286	14,539
1995-96	844	315	14,459
Total	3,751	901	42,861
Three Year Avg.	1,250	300	14,287

Regional Comparisons

The number of work releases, offenders granted work releases and on-register offenders for each region and fiscal year are presented in Table 2. The Pacific region made the most use of work releases in 1993-94 with 1,608 departures for 147 offenders, the highest number for any region and year included in the study. No other region granted work releases to as many as 147 offenders in any year after their introduction. Analyses indicated that most of the work releases in 1993-94 (97%) were single day work releases. In subsequent years, the Pacific region work release numbers are more consistent with other regions. Therefore, 1993-94 may be considered an anomaly and numbers for this year are not included in the averages presented below.

The Pacific region had the highest average number of work releases per year (244), based on the years 1994-95 and 1995-96, with Ontario having the next highest average (200), while the Atlantic region had the lowest (75). However, both the Atlantic and Pacific regions granted work releases to more than 3% of their on-register offenders while the other regions granted work releases to 2% or less of their on-register offenders.

Table 2: Number of work releases, offenders granted work release and on-register offenders by region and fiscal year
Fiscal Year	Atlantic			Quebec			Ontario			Prairie			Pacific
Fiscal Year	WRs¹	Offs²	On-reg³	WRs	Offs	On-reg	WRs	Offs	On-reg	WRs	Offs	On-reg	WRs.	Offs	On-reg
1993-94	72	46	1,341	109	48	3,747	324	73	3,865	52	37	3,087	1,608	147	1,823
1994-95	92	52	1,464	199	94	3,891	162	90	3,875	68	44	3,325	221	48	1,984
1995-96	60	41	1,450	180	107	3,804	115	68	3,852	223	74	3,360	266	96	1,993
Total	224	139	4,255	488	249	11,442	601	231	11,592	343	155	9,772	487	144	5,800
Three year avg.(%)⁴	75 (9%)	46 (13%)	1,418 (10%)	163 (20%)	83 (25%)	3,814 (27%)	200 (25%)	77 (23%)	3,864 (27%)	114 (14%)	52 (16%)	3,257 (23%)	244⁵ (31%)	72⁵ (22%)	1,933 (14%)

¹ Number of work releases.

² Number of offenders granted work releases.

³ Number of on-register offenders.

⁴ The percentage is the percentage across all regions. For example, the 9% of work releases in the first column means that 9% of all work releases were in the Atlantic region and 13% of all offenders granted work releases were in the Atlantic region.

⁵ Average is based on years 1994-95 and 1995-96.

The distribution of offenders granted work releases was not generally different from the distribution of on-register offenders in each region as shown in Table 2. The Atlantic, Quebec, and Ontario regions had small differences (2% to 4%) between the percentage of offenders granted work releases and the percentage of on-register offenders in their regions. However, the results indicate that the Pacific region, with 14% of the inmate population accounted for 22% of the offenders granted work release. The Prairie region, which accounts for 23% of the inmate population, accounted for only 16% of the work releases.

Gender Differences

Female offenders accounted for only about 1% of work releases, which is less than the 2% to 3% of the inmate population that they represent. In terms of numbers, 45 work releases were granted to 13 female offenders. Work releases may be underutilized for female offenders.

Aboriginal Offenders

Aboriginal offenders accounted for 8% (395) of the work release population, slightly lower than their representation in the inmate population (11% to 12%).

Age

The average age at which the work release population received their first federal sentence was approximately 29 years of age, with the median age being 27 years. The average age at current admission³ was 32 years. The average age at which each offender received their first work release was approximately 37 years, considerably later than their admission age.

Work Release Days

The number of days offenders were on work release throughout their sentence was calculated because determining the length of individual work releases was impossible. In some instances the data showed that an offender received a subsequent work release, but started the work release before his/her previous work release had ended. When this occurred the overlap between the two work releases was eliminated from the calculation allowing for a more accurate representation of work release days. For example, if an offender received two consecutive sixty-day work releases, the number of work release days would be 120 days. If the second work release started after the first one, the number of work release days would decrease to 70 days. It should be noted that if an offender left and returned to an institution on the same day this was counted as a one day work release.

As specified in the CCRA, the length of a work release is 'normally not to exceed 60 days' (CCRA, sec 18(5), 1992)), although there is no restriction on the total number of days an offender may be on work release. The median number of work release days from 1993-94 to 1995-96 was 59. Figure 1 presents the median number of work release days per offender for the years 1993-94 to 1995-96. The median is the middle point in a series of numbers with 50% of the cases above the median and 50% of cases below the median. When a set of numbers has some extreme scores the median is a better indicator of the average than the mean.

Figure 1: Median number of work releases days per offender by region and fiscal year

In Figure 1 it can be seen that 60 days was the median number of work release days in the Atlantic, Quebec, Ontario and Prairie regions during 1995-96. This represents the total number of days that an offender was on a work release and may reflect multiple work releases. It should be noted that, of the offenders that had a work release, 71% had at least one work release that was approximately 60 days. Increases in work release days in the Ontario, Prairie and Pacific regions indicate that the regions were adapting to the new program and extending the number of work release days. Both Quebec and Pacific regions have one year in which the median number of work release days exceeded 100 days.

Table 3 presents the percentage of offenders by number of work release days for each region for the years 1993-94 to 1995-96. Except for the Prairie and Pacific regions, the percentage of offenders that were on a work release for a total of one day was less than 4% in each region. Ontario and Pacific regions appear to use shorter work releases with around 50% and 75% of offenders having less than 50 days on work release, respectively. In Atlantic, Quebec, and Prairie regions, 20%, or more, of their offenders received more than 70 work release days indicating that offenders were being granted multiple work releases. Almost all of the regions had a high percentage of offenders with work release days between 51 and 70 days.

Table 3: Percentage of offenders by number of work releases days and region
Number of Days	Atlantic	Quebec	Ontario	Prairie	Pacific
1	3.6%	1.0%	3.7%	11.1%	4.9%
2-50	10.3%	12.5%	42.8%	27.4%	75.5%
51-70	48.2%	42.2%	30.8%	32.1%	7.0%
71-129	22.8%	23.2%	11.3%	20.4%	6.9%
130 or more	15.2%	21.1%	11.5%	9.0%	5.7%

¹ Percentage represents work release days per offender in a specific region. For example, 3.6% of the offenders granted a work release in the Atlantic region spent a total of 1 day on work release.

Offence Classification and Sentence Length

As with any type of release there is a concern with the types of offenders that are being released into the community. Table 4 presents the offence classification and the percentage of offenders who had at least one offence of each type for all work release offenders and includes all previous federal offences prior to the first work release. The eight offence categories are not mutually exclusive, meaning that an offender who has been convicted of a sex offence might also be convicted of another offence. Almost 50% of the offenders released on work release had committed at least one non-sexual violent offence, 14% had committed a sexual offence and 11% had been convicted of murder.

The average length of the current sentence for offenders serving determinate sentences was just over 7 years.

Table 4: Percentage of offenders having at least one federal conviction for each type of offence prior to first work release
Offence	Percentage¹	Offenders with at least one conviction
Murder	11.4%	133
Manslaughter	7.1%	83
Sexual Offence	13.6%	159
Non-sexual violent	48.2%	562
Drug	26.0%	304
Property	30.3%	354
Break and Enter	30.2%	352
Other Non-Violent	69.0%	805

1 The percentage represents the percentage of offenders that had at least one conviction of a certain offence. For example, approximately 11% of the work release population (n = 1,167) was convicted of a murder offence.

Table 5 provides another view of the offences committed by offenders granted a work release. Unlike Table 4, which presents all offences, Table 5 presents the most serious offence, so each offender is only counted once in the percentages. The order of seriousness of the offences is represented by the order in the table. Two thirds of the work release offenders had been convicted of a violent offence, which is lower than the 76% of the on-register population (Basic Facts, 1994). Sexual offences were the most serious for 12% of work release offenders and 36% committed a non-sexual violent offence. Offenders convicted of drug offences accounted for 14% of work release offenders.

Table 5: Most serious offence as a percentage of offenders for current admission
	Percentage¹	Number of Offenders with Conviction
Murder	11.6%	133
Manslaughter	6.0%	69
Sexual Offence	12.5%	143
Non-sexual violent	36.5%	418
Drug	13.5%	155
Property	8.6%	98
Break and Enter	4.2%	48
Other Non-Violent	4.4%	80

1 Percentage is based on number of offender with non-missing data (n = 1,144).

Timing of Work Release

Unlike other forms of release outlined in the CCRA, there is no specification as to the minimum amount of time that an offender is required to spend incarcerated before being permitted a work release. Four issues are raised with the timing of the work release program; 1) how many offenders applied for day parole before and after the completion of their first work release?; 2) at what point in the offenders' sentences are they being granted work releases in relation to parole eligibility dates?; 3) what is the relationship between work release and other forms of temporary absences?; and 4) how much time is spent incarcerated before being granted a work release?

The number of successful day parole applicants prior to first work release helps provide an indication of potential risk that is associated with each offender. Table 6 presents the number of applications for day parole within one year prior to first work release. Thirty-seven percent (430) of the work release population made an application for day parole prior to their first work release, and only 24% of the applications (107) were granted.

Table 6: Number of day parole applications before and after first work release
	Pre-Work Release	Post-Work Release
Day parole applications	430	677
Percent who applied for day parole¹	36.8%	58.0%
Percent granted day parole of those that applied	24.8%	73.1%
Percent of all work release cases granted day parole¹	9.2%	42.4%
Percent granted day parole of those who applied both pre and post work release²	19.6%	74.0%

¹ Percent is based on the total work release population (n = 1,167).

² Percent is based on number of offenders that applied for day parole both pre and post work release (n = 235).

Table 6 also presents the day parole application results for the one year after the work release to help determine the effect of work release on day parole grants. Prior to the work release, 37% of offenders applied for day parole, and this increased to 58% after the work release. Of the work release offenders who applied for day parole, 25% were granted a day parole before their work release while 73% were granted a day parole after their work release. These results suggest that work release is associated with a higher grant rate for day parole. Further evidence for this can be found with the 235 offenders who applied for day parole both before and after their work release. Only 20% of these offenders were granted a day parole before their work releases, but 74% were granted day parole after their work release.

A number of offenders applied for day parole in the year prior to their work release. Of these, 76% were denied a day parole release. The differences in the type of supervision for each type of release might explain the differential decision making.

The point at which the offenders receive a work release is of interest, especially in relation to day and full parole eligibility dates, as well as in relation to the time in the sentence. These dates indicate when an offender may apply to a formal release program (e.g., day parole, full parole). Table 7 presents the timing of work releases in relation to critical dates. Day parole eligibility is six months before parole eligibility for offenders with determinate sentences and three years before parole eligibility for offenders serving life sentences. For offenders serving determinate sentences, parole eligibility is normally at one-third of the sentence while for offenders serving indeterminate life sentences it is 25 years or a period between 10 and 25 years as set by the court.

Table 7: Percentage of offenders released at or before specific points in their sentence (offenders with determinate sentences)
	Percentage¹	Number
Before day parole eligibility date	15.6%	161
Between day parole and parole eligibility	22.5%	223
Parole eligibility date to 50% of sentence	23.8%	246
After 50% of sentence	38.1%	394

1 Percentage is based on number of offenders serving determinate sentences (n = 1,034). Problems with parole eligibility dates for offenders serving indeterminate sentences (n = 133) prevented them from being included in the table.

Approximately one-sixth of offenders granted work release have their first work release prior to the date they are eligible for day parole. Almost one-quarter of the offenders granted work releases have their first work release between the day parole eligibility date and the parole eligibility date and another quarter receive their work release prior to 50% of their sentence. Almost two-fifths of work release offenders received their release in the last half of their sentence, just prior to their statutory release date.

Temporary absences (TAs) received prior to the first work release are of interest because they provide an indication that an offender can be in the community and the risk involved to the community is at an acceptable level. Table 8 and Table 9 present the total number of reintegration escorted temporary absences (ETAs) and reintegration unescorted temporary absences (UTAs) that were given to each offender prior to receiving their first work release. Reintegration TAs do not include those granted for medical, compassionate or administrative reasons. The numbers include only TAs given during the same sentence that the offender received a work release. Overall, the percentage of offenders who received a work release and had no prior ETA/UTA experience was 26% (306).

Reintegration ETAs were granted to approximately 70% of the sample prior to their first work release. Among offenders who received a reintegration ETA, 17% (193) had only one to three previous ETAs, while almost 40% (466) had 10 or more ETAs prior to first work release.

Reintegration UTAs were granted to only 30% of offenders released on work release. Among offenders granted reintegration UTAs, 14% (158) had 1-3 previous UTAs, while just over 5% (61) had 10 or more.

Table 8: Percentage and number of reintegration ETAs per offender before first work release
Previous ETAs	Percentage	Number
None	29.6%	346
1-3	16.5%	193
4-9	13.9%	162
10-25	19.1%	223
26 or more	20.8%	243
Total	100.0%	1,167

Table 9: Percentage and number of reintegration UTAs per offender before first work release
Previous UTAs	Percentage	Number
None	69.7%	814
1-3	16.4%	191
4-9	8.7%	101
10-25	4.5%	53
26 or more	0.7%	8
Total	100.0%	1,167

For purposes of analysis, time spent incarcerated before receiving a work release was calculated separately for offenders serving life sentences and determinate sentences. Offender's serving determinate sentences spent an average of 3 years incarcerated before receiving a work release while offenders serving life sentences served an average of 12 years incarcerated before their work release.

Completion of Work Release

The completion rate for work releases departure for the years 1993-94 to 1995-96 are presented in Table 10. Less than one percent of work release departures resulted in offenders requiring additional intervention by the justice system. Most work release problems were the result of late returns and suspensions of the work release.

Table 10: Completion of Work Release Departures
	Successful	Late	UAL	Suspended	Total
1993-94	97.5%¹	0.2%	0.2%	2.1%	2,158
1994-95	96.7%	1.0%	0.3%	2.0%	737
1995-96	97.0%	0.4%	0.1%	2.5%	836

¹ Percentage represents the number of offenders for that year. For example, 97.5% (n = 2,104) of the offenders in 1993-94 successfully completed their work release.

Type of Release after Work Release

The type of release following work release is presented in Table 11. These releases do not necessarily follow directly after the work release, but may occur months or years later with additional time in custody. In addition, 31% of the work release offenders had not been released at the end of the study. A day parole release followed a work release for 51% of the population and was the most common release. Full parole followed the work release in 14% of cases and statutory release followed work release in 33% of cases. Further analyses indicated that, of offenders who received a day parole (410) after their last work release, approximately 67% (272) received a full release (full parole or statutory release) after their day parole. Therefore, in total, 43% of work release offenders were eventually released on full parole and 57% were released at their statutory release date.

Table 11: Release type after work release
	First Release		Full Release
	Percentage¹	Number	Percentage²	Number
Day parole	51.2%	410
Full parole	14.4%	115	42.6%	278
Statutory release	32.7%	262	57.4%	374
Other	1.7%	14
Total	100.0%	801	100.0%	652

¹ Percentage is based on work release offenders who were granted a release after completion of work release (n = 801). In 366 cases (31.4%) offenders had not been released by the end of the study period.

² Percentage is based on number of offenders who received a full parole or statutory release after the work release (n = 652). 138 of the offenders released on day parole did not have a full release at the end of the study period.

Follow-up

A two year follow-up period was used for offenders who were granted a full release after the completion of a work release. Of the 652 offenders who received either a full parole or statutory release after the completion of their work release, 20% (197) of the releases were granted before October 1, 1994, which permitted a minimum two year follow-up. The average follow-up period was 31 months. The follow-up period examined the number of offenders that had a re-admission to a federal institution after their full release. Table 12 presents the percentage of offenders returned to custody after completion of work release.

Overall, 65% of the work release offenders were still in the community two years after their release. In 24% of the cases re-admissions were for technical violations of parole or statutory release conditions. Technical violations occur for reasons such as; failure to report to parole officer, failure to meet release conditions (e.g., abstinence from alcohol), etc. New offences were committed by 20% of offenders and violent offences were committed by 6% of offenders. The readmission rate, technical violation and offence rates are slightly higher than for offenders who were followed up after a day parole release (Grant and Gillis, 1997).

Table 12: Admission type after completion of work release
Outcome	Percentage¹	Number	Day Parole²
No readmission	65.0%	128	77.3%
Any readmission	35.0%	69	22.7%
Technical violation	24.4%	48	11.4%
New offence	19.8%	39	14.9%
Violent offence	6.1%	12	4.5%

¹ Percentage represents the percentage of the follow-up group (n = 197).

² Average follow-up was 21 months. From Grant and Gillis, 1997.

Summary

Since the implementation of the CCRA a total of 4,659 work releases were granted to 1,167 offenders. An anomaly appears in the data during the year 1993-94 when 2,165 (almost 50% of the total) work releases were granted, with 1,608 coming from the Pacific region. Aside from the fluctuation in the number of work releases that were granted, the number of offenders that received a work release was consistently around 300 per year. Regional differences were found in the average number of work releases granted and the proportion of on-register offenders granted a work release.

Both female and Aboriginal offenders were underrepresented in regards to their respective proportions of the on-register population. Female offenders received approximately 1% of work releases, but account for 2% to 3% of the offender population, while Aboriginals accounted for 8% of work releases, but account for 11% to 12% of the offender population.

The average age at which each offender received their first work release was 37 years, which is five years later than the age at admission.

The average number of work release days for an offender is 60, but there is variation across regions and years. Most offenders receive multiple day work releases with 9.5% extended to over 130 days. This might also indicate that offenders successfully complete their work release and are renewed. Work releases have a 60 day maximum but may be renewed.

The criminal history results for the work release population indicated that they likely committed a violent offence in the past. However, while 66% of the work release population had been convicted of a violent offence, 76% (Basic Facts, 1994) of the on-register population had been convicted of a violent offence.

Day parole releases were granted to 24% of the work release population that applied for day parole within one year prior to first work release, whereas 73% of those that applied were granted a day parole within one year after their first work release. It appears that work release increases offenders' chances of being granted a day parole.

The point in the sentence at which offenders received their first work release varied considerably. Offenders serving determinate sentences received a work release, on average, after serving three years, while offenders serving life sentences received a work release after serving twelve years. Thirty-eight percent of work release offenders had served over 50% of their sentence before being granted a work release. Fifteen percent were released prior to their day parole eligibility date, while 23% were released prior to their parole eligibility date.

Very few (less than 0.5%) work releases resulted in offenders being charged by the police or absconding. About 2% of work releases were suspended.

Twenty-six percent of the work release population had no prior TA experience before their first work release and only 30% had a prior UTA.

At some point after the completion of work release, 51% of those who received a release were granted a day parole. Eventually 43% were released on full parole and 57% were released on statutory release.

Approximately 20% (197) of the work release offenders could be followed for two years after release on full parole or statutory release. Of these, 65% had not been readmitted after two years. The majority of the full release readmissions were due to technical violations. These readmission rates are slightly higher than for a group of offenders followed up after a day parole release.

Chapter 4: Review of File Information

Documentation

A random sample was drawn from the 1994-95 work release population for a detailed file review. Information was coded from the offender's correctional plan and progress summary reports using the manual presented in Appendix A. All information was obtained from the Offender Management System (OMS). Due to time constraints, paper files from the institution were not reviewed. The sample included 75% of offenders granted work releases in the year 1994-95, which represents about 20% of offenders granted work release since the CCRA was implemented.

As outlined in the Case Management Manual (1997), "upon receipt of an application for work release, a correctional officer reviews the application and conducts an interview with the applicant to discuss the proposed work release". This information is included in the progress summary report and is the responsibility of the correctional officer assigned to the case. The progress summary prepared in relation to the work release application should assess the following:

The criteria for granting a work release;
The immediate and long term risk to reoffend;
The appropriateness of the work release in view of the offender's correctional plan;
The offender's progress in relation to the correctional plan;
Any continued involvement in criminal activities (includes information/concerns from Preventive Security);
The supervision requirements, including:
1. The type of supervision (institutional staff, community staff, or private organization or individual),
2. The type of supervision contacts (specify: telephone, visit at the work site or place of accommodation), and
3. The minimum frequency of each type of supervision contact;
4. The requirement for any conditions; and
5. The details regarding the accompaniment of the inmate to and from the work site, where applicable.

Of the 223 offenders selected for a file review in OMS, 100% of the cases had a progress summary report (PSR) associated with their work release. However, not all the required information was available in the PSR as shown in Table 13. Most PSRs reported on risk to offend, appropriateness of the work release and progress in relation to the correctional plan. Special conditions and supervision requirements were specified in approximately three quarters of the cases reviewed. Details on involvement in criminal activity, criteria for granting work release and arrangements for accompanying offenders to the work site were addressed in 60% or less of the cases.

Table 13: Percentage of offenders with criteria mentioned in progress summary report
	Percentage	Number
Offender's progress in relation to Correctional Plan	95.1%	212
Immediate and long term risk to reoffend	83.4%	186
Appropriateness of WR in view of Correctional Plan	83.4%	186
Supervision Requirements	75.8%	169
Requirement for any conditions	72.6%	162
Continued involvement in criminal activity	60.5%	135
Criteria for granting a work release	46.6%	104
Accompaniment of inmate to and from work site	46.6%	104

The specific criteria for granting a work release are as follows (CCRA, 1992, Sec 18(2)):

1. The inmate will not, by reoffending, present an undue risk to society during a work release;

2. It is desirable for the inmate to participate in a structured program of work or community service in the community;

3. The inmate's behaviour while under sentence does not preclude authorizing the work release; and

4. A structured plan for the work release, including specific objectives to be achieved by the inmate during the course of the release, has been prepared.

Each of the above was examined in the offender's file and the results are presented in Table 14. The criteria, that it is desirable to participate in work releases and inmates behaviour does not preclude work releases, were addressed in 95% of the cases. Less frequently addressed were the issues of risk to society and a structured plan for the release.

Table 14: Specific Criteria mentioned in Progress Summary Report
	Percentage	Number
Desirable to participate in structured program of work	95.1%	212
Inmate's behaviour does not preclude authorizing WR	95.1%	212
Inmate will not present undue risk to society	83.0%	185
Structured plan for WR has been prepared	59.6%	133

An objective of the Correctional Service of Canada is to ensure that offenders are productively occupied and have access to a variety of opportunities to develop work skills and abilities which will serve them on release (Mission Document, 1996). Certain goals for an offender are outlined in their correctional plan and these goals may be accomplished through the implementation of a work release. As shown in Table 15, 5% of work releases were designed to achieve all of the goals outlined in the correctional plan, and for a further 84% the work release was designed to achieve some of the goals outlined in the offender's correctional plan. In 11% of the cases, the work release was not designed to meet any of the correctional goals.

Table 15: Percentage of work releases designed to achieve goals in correctional plan
	Percentage	Number
All	5.3%	12
Some	83.9%	187
None	10.8%	24
Total	100.0%	223

Of the correctional goals that were outlined for each offender, the most frequent was preparation for release, noted in 66% of cases. Developing work skills, ties to the community, and ensuring public safety were stated as goals in about one-third of cases. Table 16 presents a breakdown of the correctional goals that were outlined for each offender.

Table 16: Correctional goals outlined for each offender
	Percentage¹	Number
Preparation for Release	65.5%	146
Develop Work Skills	29.1%	65
Develop Ties in the Community	33.2%	74
Ensure Public Safety	29.6%	66
Other	9.4%	21

¹ Percent on based on those cases included in the sample (n = 223).

The type of work placement was also examined from file information contained in OMS. Types of placement included; community work, Corcan, education etc. As shown in Table 17 the most frequent type of placements were community work (26%), which included community groups, such as the boy scouts and church groups, and manual labour (25%) which included working in recycling depositories, and construction work. CORCAN accounted for almost 10% of the work releases while farming and forestry activities accounted for 14% of the releases.

Table 17: Type of work placement
Work Placement	Percentage	Number
Community	26.4	59
Manual Labour	24.7	55
Maintenance	15.7	35
CORCAN	9.0	20
Other	8.5	19
Fruit Picking	6.7	15
Farm Work	3.6	8
Forestry	3.6	8
Education	1.8	4
Total	100.0%	223

Work releases require a sponsoring agent and the types of sponsoring agencies for work releases are presented in Figure 2. Government (federal, provincial and municipal) sponsored work releases represented just over one third of all work releases. Non-profit organizations, such as charitable and religious organizations, sponsored 29% of work releases. Private businesses and individuals sponsored one quarter of work releases.

While on a work release offenders need to be supervised. Private or non-profit organizations provided supervision for one-third (32%) of work release. Institutional staff supervised about 20% of work releases while private individuals supervised 21% of the releases (see Figure 3).

Figure 2: Sponsoring agencies for work release

Figure 3: Supervising agencies for work releases

Post Work Release Evaluation

After a work release has ended, a formal evaluation is required to be completed. Of the 223 offenders in the sample, only 17% (37) had a formal evaluation completed in OMS. Of these 37 evaluations, 68% (25) were completed within 10 days of the termination of the work release. However, within this small percentage of completed formal evaluations, the information regarding the work release was quite informative. Table 18 presents the breakdown of comments that were mentioned in the formal evaluation. Most of the evaluations which were completed address the success of the work release, while about three-quarters provided an overview of the work release and identified problems. The percentage in the first column should be considered in relation to those in the second column showing the percentage of all cases that had follow-up information.

Table 18: Areas addressed in work release evaluation reports
	Percentage with evaluation completed¹	Percentage of sample²	Number
Overview of Work Release	78.4%	13.0%	29
Identification of Problems Related to Work Release	75.7%	12.6%	28
Work Release was Successfully completed	83.8%	13.9%	31

¹ Percentage is based on number of offenders with completed formal evaluations (n = 37).

² Percentage is based on number of offenders included in coded data (n = 223)

Close to 80% of the post work release evaluations addressed one or more important factor related to the work release. However, these account for only about 13% of all the work releases. The work releases with evaluations were considered successful in 84% of cases although problems were noted in 76%. The problems noted were not necessarily related to the success of the work release.

Other Information

Analyses were conducted to determine the number of offenders granted work releases who applied for day parole after the completion of their work release. Of the offenders who were included in the coded data, 109 (49%) applied for day parole within 1 year after the completion of their work release. Information on the outcome of the day parole decision indicated that only 38% of applications were successful. These percentages are somewhat lower than for the overall data which showed a 73% day parole grant rate after the work release. Since the coded cases were from 1994, shortly after work release was implemented, the difference may be the result of changes in perceptions about the work release program by the National Parole Board as they became more familiar with it.

Most offenders granted work releases were in minimum security institutions (67%), while the remaining 33% were in medium security institutions.

The results indicate that some offenders receive multiple work releases. Within the file review data, 44% (99) of the cases had more than one work release within one year and 11% (25) had more than five work releases within a year. These results are based on work release projects which include multiple departures. Results presented in the previous sections refer to work release departures.

For each work release (i.e., full or part time) the amount of time spent working each day should be specified, but only 66% of cases had this information recorded in OMS. Results which follow include only those cases with information. For 82% (118) of the cases, the work release was essentially full time (8 hours/day, 5 days/week). Twelve percent (18) of the cases were allotted a time period of less than 5 days per week. The remaining work releases were part time, either fewer hours per day or less than five days per week.

Required supervision to and from the work site provides an indication of the potential risk that is associated with each offender. Percentages are based on the number of cases that had information recorded in OMS (56%). For 68% (85) of the cases, constant supervision was required to and from the work site. The remaining cases either required supervision at the drop off and pick up point (17%) or required no supervision at all during travel to and from the work site (13%).

Contact between the inmate and the institution should be specified for each work release. Again, only 53% of cases had this information recorded in OMS. For 49% (57) of cases supervision at the work site by CSC staff consisted of dropping off and picking up the offender. For 44% (51) of the work releases, a CSC staff member visited the work site.

Additional condition(s) that were imposed on offenders granted a work release are presented in Table 19. The most frequent condition was for the offender to return directly to the institution following each work day (32%). Abstinence from all intoxicants was next frequent with 29% (64) of cases having this imposed. Reporting directly to the worksite was required for 23% (52) of cases. No additional conditions were imposed for 23% (52) of cases.

Table 19: Additional conditions imposed on work release
	Percentage	Number
Return directly to institution following each work day	32.3%	72
Abstain from all intoxicants	28.7%	64
Report directly to the worksite on each work day	23.3%	52
Not indicated	23.3%	52
Other	16.6%	37
Refrain from contact with individuals with criminal record	15.2%	34
Abstain from alcohol	11.2%	25
Avoid contact with children under 16	9.4%	21
Submit to urinalysis testing, if requested	8.1%	18
Not bring anything into the institution without approval	6.7%	15
Carry CSC ID card and WR permit at all times	3.1%	7
Not carry more than $35	2.7%	6
Disclose finances to management, if requested	2.2%	5
Not return with any unauthorized property	1.8%	4
Advise institution if arrested	1.3%	3

Summary

During the course of the file review, it was found that some of the most important elements of work release (i.e., work release in relation to correctional plan, risk to reoffend and appropriateness to correctional plan) were noted in the offenders' progress summary reports, but much of the other documentation required was not present in OMS.

Although not required, most work releases did address some of the correctional goals with 70% of releases addressing preparations for full parole or statutory release.

One quarter of work releases were for work with or for community groups while 40% were for manual labour and maintenance work. Corcan accounted for 10% of work releases.

Work releases were sponsored by government, non-profit groups and private individuals. Private or non-profit organizations, CSC staff and private individuals provided most of the work release supervision.

The majority of work releases required the offender's to work full time, 8 hours per day, 5 days a week.

Chapter 5: Discussion

Work release as a type of temporary release from prison was implemented in 1992. It is unique in that work releases do not require approval from the National Parole Board so responsibility rests solely with the Correctional Service of Canada. Prior to the CCRA work release activities were generally handled through the use of day parole. Changes to the CCRA eliminated the use of day parole for any reason other than preparation for full release.

Between November 1992 (the introduction of the CCRA) and September 1996 there were 4,659 work release departures by 1,167 offenders. The number of work releases in the fiscal year 1993-94 distorts the summary statistics for work releases, but based on the years 1994-95 and 1995-96 there were about 800 work release departures by about 300 offenders each year. Prior to the CCRA, there were about 350 day paroles for special projects which were used for purposes similar to those for work release. In relation to other release programs, work release affects a relatively small number of offenders, about 2% to 3% of the on-register offenders in each region. In addition, the number of offenders granted work releases has not changed over the past two years, indicating that increased familiarity with the type of release option has not increased its use.

There is considerable variability in the use of work release across the regions. The Atlantic and Pacific regions provide work release opportunities for the largest percentage of their offenders. In the Ontario and Pacific regions the number of days an offender is on work release tends to be less than in other regions, but paradoxically the Pacific region also has the highest percentage of offenders with 130, or more, work release days. The regional variation in the number of days on work release suggests that the regions may be using work releases for different types of activities. The median number of work release days for offenders is about 60 days, the maximum allowed for any single work release without the approval of the deputy commissioner in the region.

Female offenders accounted for only 1% of offenders on work release and Aboriginal offenders accounted for only 8% of work release. These results suggest the underuse of work release for these groups of offenders, although the difference is not great. Offenders who received a work release were approximately 37 years old at the time of their first work release.

Offenders given work releases are less likely to have committed a violent offence than offenders in the general population. About 65% of the offenders released on work release have committed a violent offence as compared to 76% of the offender population. In addition, about 12% of the work releases offenders were serving life sentences for murder.

Less than 40% of the work release population applied for a day parole within one year prior to their first work release. Of those that applied only 24% were granted day parole. When looking at the number of day parole applications within one year following first work release there is a notable difference. Among those that applied for day parole after work release participation, 73% were granted. Work release appears to increase offenders' chances for day parole and may provide an indication of the potential risk that is associated with each offender on other forms of release.

While 70% of the offenders who were given work releases had at least one escorted temporary absence for reintegration purposes, only 30% had an unescorted temporary absence. Overall, there were 26% of work release offenders who had not received any type of temporary absence prior to the their work release. These results are a bit surprising given that work releases may have less supervision than is required for escorted temporary absences.

About one-sixth of the work release offenders received their first work release before the date they were eligible for day parole, while a total of almost 40% received their work release before their parole eligibility date. Work releases were also commonly used later in the sentence with almost 40% of the offenders having their first work release after serving half of their sentence in custody.

The most common form of release after the work release was day parole, suggesting that the work release may serve to assist offenders is obtaining a day parole. However, more than 50% of the work release offenders were released at their statutory release date rather than on full parole.

A minimum two year follow-up period revealed that 65% of the offenders included in the follow-up had not been readmitted by the end of the study period. The most common form of readmission was for technical violations. Twenty percent of released offenders had committed a new offence, with 6% committing a violent offence. These rates are slightly higher compared to a group of offenders released on day parole.

Documentation for a sample of 223 work releases was reviewed to determine how and for what reasons work release is being used. The review indicated that some of the work release criteria were not addressed in the progress summary report prior to the release. In addition, approximately 20% of cases did not address the criteria of risk to society. Similarly, post-work release documentation was limited, with only 17% of the cases having this information in the file. For the cases that had a post-work release report, about 80% indicated that the work release had been successful at meeting the expectations of the work release. Other results indicated that less than 0.5% of work releases resulted in the offender failing to return to the institution.

The file review indicated that the work release, although not required to meet correctional goals, did in fact address a number of correctional goals for offenders.

Most work releases provided opportunities for low skilled labour in a variety of settings. Assisting in community projects was the most common work project while other activities involved manual labour and agricultural work.

Work releases provide an opportunity for offenders to be engaged in productive activities outside the prison. The results suggest work release meets the general goals of the CCRA, including providing preparation to offenders for their eventual release. In addition, experience on a work release is associated with increased likelihood of a day parole release.

References

Correctional Service of Canada. (1995). Basic facts about corrections in Canada. (4th ed.) Ottawa: The Communications Branch, Correctional Service of Canada.

Correctional Service of Canada. (1996). Mission of the Correctional Service of Canada. Ottawa: Ministry of Supply and Services Canada.

Grant, B.A., & Gal, M. (1998). Case Management Preparation for Release and Day Parole Outcome. Ottawa: Research Branch, Correctional Services Canada.

Grant, B.A., & Gillis, C.A. (1998). Day Parole Outcome, Criminal History and Other Predictors of Successful Sentence Completion. Ottawa: Research Branch, Correctional Service Canada.

Grant, B.A., & Johnson, S.L. (1998). Personal Development Temporary Absences. Ottawa: Research Branch, Correctional Service Canada.

Grant, B.A., & Millson, W.A. (1998). The Temporary Absence Program: A Descriptive Analysis. Ottawa: Research Branch, Correctional Service Canada.

Katz, J.F., & Decker, S.H. (1982). An analysis of work release: The institutionalization of unsubstantiated reforms. Criminal Justice and Behavior, 9(2), 229-250.

Marlette, M. (1990). Furloughs tightenned: Success rates high. Corrections Compendium, 14(1), 1-21.

Waldo, G.P., & Chircos, T.G. (1974a). Work release as a rehabilitative tool: An evaluation of two state programs. Tallahassee, Florida: Southeastern Correctional and Criminological Research Center, Florida State University.

Waldo, G.P., & Chircos, T.G. (1974b). Work release as a rehabilitative tool: An evaluation of two state programs (Appendices and References). Tallahassee, Florida: Southeastern Correctional and Criminological Research Center.

Waldo, G.P., & Chircos, T.G. (1977). Work release and recidivism: An empirical evaluation of a social policy. Evaluation Quarterly, 1(1), 87-108.

Appendix A Coding Manual for Work Release Study

Region _____________________ Institution ____________________

FPS # ______________________

Admission Date ______________ (yy/mm/dd)

Electronic Index Data:

OMS Data:

WR Permit Number⁴ #1 _______________