Supplementary Evaluation of the Integrated Correctional Program Model: Community Program

Original signed by

Anne Kelly
Commissioner

Date
September 22 2021

Authors:^{Footnote 1}

Nicholas Chadwick, A/Evaluation Officer

Danielle Guarino, Junior Evaluation Analyst

Sara Johnson, Evaluation Manager

Kayla Wanamaker, A/Evaluation Officer

The evaluation team is grateful for the assistance provided by the Correctional Operations and Programs Sector. The ongoing support and continued collaboration with the Reintegration Programs Division was instrumental in the completion of this evaluation. The evaluation team would like to thank Shanna Farrell-MacDonald from the Research Branch for her assistance with the data extraction, and the members of the Consultative Working group for their invaluable feedback, including representatives from Indigenous Initiatives Directorate and Reintegration Operations Division.

Finally, the evaluation team would also like to thank everyone else who contributed to this report, whose names do not appear here.

This report acts as a supplement to the Evaluation of Correctional Reintegration Programs (2020). Analyses focused on the relevance and effectiveness of the Community Program (CP), a component of the Integrated Correctional Program Model (ICPM) offered to men offenders who have a need for correctional programming that was not met through the successful completion of a main correctional program while incarcerated.

The CP is an integrated program for moderate and high risk offenders, released into the community, who did not complete main correctional programming while in the institution, despite having a need for it. It provides an opportunity to gain exposure to correctional program content while in the community (CSC, 2014b; CSC, 2019b). Although the CP does not address the need for correctional programming that should have been received in the institution, it accomplishes the important task of providing program exposure to offenders who demonstrate a need for it, which supports successful reintegration (CSC, 2014b). As a result, participating in the CP is expected to be related to reductions in the rate of returns to custody while supervised in the community. The CP also prepares men offenders with the basic insights and skills required to effectively participate in the Community Maintenance Program (CMP), which provides offenders an opportunity to gain, rehearse, and maintain skills.

The evaluation included men offenders who were released to the community between April 1, 2015 and December 31, 2017 in Atlantic, Québec, and Pacific regions; April 1, 2016 to December 31, 2017 in Ontario region; and June 1, 2017 to December 31, 2017 in the Prairie region. Participation in the CP and CMP, as well as first release outcomes, were examined from the date of first release to either the first release return to custody date, sentence expiry date, or the end of data collection (February 29, 2020).

Three groups of offenders were identified for the purpose of this evaluation: 1) CP successful completers; 2) CP incompleters; and 3) eligible non-participants (i.e., men offenders with an unmet correctional programming need who did not participate in the CP).

The scope of the evaluation included the relevancy and need for the CP, and the effectiveness of the CP. To assess the effectiveness of the CP, the following were examined: 1) rates of participation (i.e., enrollment) in the CP; 2) characteristics of CP participants versus non-participants; 3) community outcomes for groups of offenders with various CP exposure during first release; 4) rates of participation in the CMP; and 5) community outcomes for groups of offenders with various CMP exposure during first release. The effectiveness of the CP was examined for White offenders, Black offenders, Indigenous offenders, and other Ethnocultural^{Footnote 2} offenders.

Data were collected through literature review, as well as extraction and analysis of administrative data from the Offender Management System (OMS). The key findings from the evaluation are highlighted below.

The CP was found to align with both the Correctional Service of Canada’s (CSC) and the federal government’s priorities, roles, and responsibilities. Results also indicated that there is a continued need for CSC to provide community programming to federal men offenders. Over recent fiscal years, approximately 17% to 24% of men offenders with an identified need for correctional programming did not complete the programming within the institution prior to their first release. These men serve as the target population for the CP.

More than half of offenders eligible to participate in the CP did so during their first release. Of those who participated, 58% successfully completed the program. The majority of offenders included in the evaluation sample were released on statutory release, indicating that they may have had limited time remaining on their sentence to participate in the CP. However, on average, those who enrolled in the CP did so within 55 days of their first release.

Key variables were explored to differentiate between those who enrolled in the CP versus those who did not. While accounting for risk relevant differences, Indigenous offenders and other Ethnocultural offenders (excluding Black offenders) were found to experience significantly lower odds of enrolling in the CP relative to White offenders. However, among those who enrolled in the CP, offender ethnicity was not related to the odds of completing the CP. In other words, White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders who had enrolled in the CP experienced comparable odds of completing the CP.

Regression analyses were conducted to account for risk relevant differences between the evaluation groups when examining the relationship with the likelihood of revocation. Findings indicated that those who completed the CP experienced significant reductions in the likelihood of being returned to custody, either for any reason or for a new offence, relative to CP incompleters and eligible non-participants. The effectiveness of the CP did not vary across offender subgroups, indicating that White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders all experienced comparable reductions in the likelihood of a revocation. In other words, the CP appeared to be equally effective for the Ethnocultural groups examined in the evaluation.

Participation in the CMP was examined for those offenders who successfully completed the CP. Most CP completers appeared to have sufficient time remaining on their sentence following the completion of the CP, indicating that they would have been able to participate in the CMP if they had the opportunity to do so and were interested. Approximately half of offenders (53%) eligible to participate in the CMP did so during their first release. Of those who did enroll in the CMP, 68% successfully completed it.

Key variables were explored to understand what differentiated those who enrolled in the CMP versus those who did not. The number of months remaining on supervision following the completion of the CP emerged as a significant predictor of enrolling in the CMP. As the number of months remaining on supervision increased, so did the odds of enrolling and completing the CMP, underscoring the importance of ensuring that the CMP is offered to offenders when they have sufficient time remaining on their sentence. Importantly, White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders enrolled in, and completed, the CMP at comparable rates.

The relationship between exposure to the CMP and the likelihood of returning to custody for any revocation was the only community outcome examined due to limited sample size for this analysis. Results indicated that completing the CMP was associated with reductions in the likelihood of any revocation, relative to those who did not complete the CMP after enrolling in it, and those with no participation in the CMP. The reduced sample size eligible for inclusion in this analysis prevented the examination of the CMP’s effectiveness across offender groups defined by ethnicity.

The evaluation findings highlighted two aspects that warrant further attention. First, it is recommended that CSC determines if the rates of enrollment in the CP for Indigenous offenders and other Ethnocultural offenders are discrepant from White offenders. Second, it is recommended that CSC determines whether there are opportunities to enhance the timely delivery of the CP and CMP to ensure offenders with limited time remaining on their sentence have an opportunity to participate and benefit from exposure to the program content.

CSC’s ICPM CP aligns with both CSC’s and the federal government’s priorities, roles, and responsibilities. The delivery of effective correctional programs, including the CP, contributes to the overall priority of a just, peaceful, and safe society.

There is a continued need for CSC to provide community programming to federal men offenders. Over recent fiscal years, approximately 17% to 24% of men offenders with an identified need for correctional programming did not complete the programming within the institution prior to their first release. These men would then be targeted to receive the CP during their release to ensure that they are exposed to programming and able to connect to the CMP.

More than half of offenders eligible to participate in the CP did so during their first release. The majority of those who enrolled in the CP successfully completed it. Most offenders appeared to have had sufficient time under sentence in the community to participate in the CP, which on average, began 55 days after first being released to the community.

Indigenous offenders and other Ethnocultural offenders experienced lower odds of enrolling in the CP relative to White offenders. Black offenders experienced comparable odds of enrolling in the CP relative to White offenders.

However, Black offenders, Indigenous offenders, and other Ethnocultural offenders experienced comparable odds of completing the CP relative to White offenders.

CP completers demonstrated significant reductions in the likelihood of a revocation to custody (both for any reason and for a new offence) relative to CP incompleters and eligible non-participants.

The reduction in the likelihood of a revocation for any reason and for a new offence is expected to be consistent across White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders.

More than half of offenders eligible to participate in the CMP did so after completing the CP. The majority of those who enrolled in the CMP successfully completed it. A portion of offenders likely would not have had sufficient time remaining on their sentence following the completion of the CP to participate in CMP, underscoring the importance of efficiently delivering both programs to remove potential barriers to CMP participation.

Offenders with more time remaining on their sentence following the completion of the CP experienced greater odds of participating in, and completing, the CMP. White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders experienced comparable odds of enrolling in, and completing, the CMP.

It is beneficial for offenders to complete the CMP following the successful completion of the CP. CMP completers experienced significantly lower likelihoods of a revocation for any reason relative to both CMP incompleters and those with no participation in CMP.

It is recommended that the CSC determines if the rates of enrollment in the ICPM-CP for relevant offender subgroups, including Indigenous offenders and other Ethnocultural offenders, are discrepant. Should it be determined that there are discrepancies in the enrollment rates to ICPM-CP for Indigenous offenders and other Ethnocultural offenders, CSC will address any unequal access to promote comparable enrollment rates.

It is recommended that CSC determines whether there are opportunities to enhance the timely delivery of the CP and CMP to ensure offenders with limited time remaining on their sentence have an opportunity to participate and benefit from exposure to the program content.

• Table 1. Key characteristics of evaluation sample (n = 1,040)
• Table 2. Key characteristics of evaluation groups (n = 1,040)
• Table 3. Rate of participation in Community Program and average time to enrollment and completion
• Table 4. Predictors of enrollment (regardless of completion status) in Community Program (n = 997)
• Table 5. Predictors of Community Program completion (n = 580)
• Table 6. Relationship between Community Program participation and community outcomes
• Table 7. Time on supervision following the successful completion of the community program (n = 355)
• Table 8. Community maintenance exposure following completion of community program (n = 355)
• Table 9. Predictors of enrollment (regardless of completion status) in community maintenance (n = 338)
• Table 10. Predictors of successful completion of community maintenance (n = 175)
• Table 11. Relationship between Community Maintenance Program participation and return to custody (n = 312)

In Fiscal Year (FY) 2019/20, the Evaluation Division of Correctional Service Canada (CSC) completed the Evaluation of Correctional Reintegration Programs (CRP), which explored the delivery of correctional programs within institutional settings and, in part, focused on the Integrated Correctional Program Model (ICPM). Overall, this evaluation found that men program completers were less likely to return to custody for any reason during their first release, compared to offenders who had a need for correctional programming, but did not participate (CSC, 2020). Generally, this finding held across each program stream examined, and for Indigenous and non-Indigenous men offenders. The current report acts as a supplementary evaluation and will explore the ICPM Community Program (CP), an integrated program for moderate to high risk offenders, released to the community, who did not complete main correctional programming while in the institution despite having a need for it (CSC, 2014b). Specifically, this report will focus on the CP’s relevancy and effectiveness (i.e., the extent to which the program is achieving its expected results).

CSC offers educational programs, social programs, vocational programs, and correctional programs. Educational programs teach literacy, academic, and personal development skills; social programs promote positive changes in offenders’ personal and social lives; vocational programs teach relevant job training to promote post-release employment; and correctional programs use evidence-based practices to target risk factors that lead to criminal behaviour in order to reduce reoffending. This report focuses on correctional programs.

In 2010, CSC began implementing a correctional model called ICPM for all federally sentenced men offenders. Compared to previous correctional programming that offered multiple different programs to address specific offender needs (multi-program model), ICPM takes a more holistic (multi-target model) approach to correctional programming (CSC, 2014c). The ICPM addresses a variety of needs within the same program with streams existing for sex offenders (SO) and Indigenous offenders, and intensity levels available for high, moderate, and adapted moderate offenders, which includes those with unique needs, such as mental health concerns (CSC, 2014c). ICPM offers offenders a comprehensive and efficient way to address risk factors that contribute to their criminal behaviour while teaching them how to cope using prosocial tools (CSC, 2014c). There are three main components that make up ICPM: Primers, Main Program, and Maintenance program. In addition to these three components, ICPM offers a Motivational Module and the CP to help further engage offenders in correctional programming at the right time (CSC, 2019b).

This report focuses on the ICPM CP, an integrated program for moderate to high risk offenders, released to the community, who did not complete main correctional programming while in the institution despite having a need for it (CSC, 2014b; CSC, 2019b).^{Footnote 3} Although the CP is not designed to replace correctional programming that should have been received in the institution, it helps offenders identify their risk factors and personal targets as well as develop basic skills that contribute to reducing their risk for future criminal behaviour. In addition to supporting the safe transition of offenders into the community, the CP provides offenders with the basic insights and skills needed to effectively participate in the Community Maintenance Program (CMP).

The CP begins with a pre-program interview, which gives facilitators the opportunity to uncover the offender’s crime process and help them develop a self-management plan (CSC, 2014b). Next, the offenders move on to the four modules; Module 1: Relationship and Support (5 sessions); Module 2: Feeling Good (4 sessions); Module 3: Think (4 sessions); and Module 4: Do (4 sessions) (CSC, 2014b).^{Footnote 4} Each session begins with a roundtable and introduction where the facilitator explains the importance of that session’s topic (CSC, 2014b). Next, the majority of sessions are comprised of activities (e.g., worksheets); these lessons are later applied when updating self-management plans (CSC, 2014b). Once an offender completes all the modules and its allocated sessions, a post-program interview is conducted to explore the offender’s self-management plans, finalize his crime process and personal targets, provide feedback, and gather any remaining information required for the final program report (CSC, 2014b). SOs who are assigned to the CP are required to attend one additional session per module (four extra in total) that addresses their sexual offending, which includes discussion around healthy intimate relationships, managing sexual fantasies and arousal, and managing thinking that supports harmful sexual behaviour (CSC, 2014b).

One goal of the ICPM (both in the institution and community) is to have offenders enroll (i.e., participate) in the maintenance program, which is a form of aftercare to help offenders maintain the skills they learned in prior programming. Therefore, those who complete a CP will ideally enroll in a CMP. CMPs differ from Institutional Maintenance Programs (IMPs) in that they are held in the community, but both reinforce important lessons and skills learned in original programming (that is, either Main or CP) and help offenders implement what they learned in a practical sense (CSC, 2014a).

CMPs were designed as a way to offer maintenance programming in the community that targets all types of criminal re-offending in order to ensure that programs are flexible and can maintain the continuum of care for a variety of offenders (CSC, 2010; 2014a). Like the Main Programs and the CP, the CMP focuses on four main areas of antisocial behaviour that are strongly related to recidivism, including anti-social associates, anti-social personality traits, anti-social cognitions, and anti-social behaviours (CSC, 2014a).

The CMP is delivered in a 12 session cycle. Participants may be referred to additional cycles at the discretion of the facilitator (CSC, 2014a). Throughout the entire cycle, the facilitator maintains a high degree of flexibility, allowing participants to address any immediate or relevant issue that may arise. All sessions follow a similar format, which include introductions and roundtable, a structured lesson (based on the theme of the session), updating offender self-management plans, and a discussion to wrap up the session (CSC, 2014a).

Correctional programs are designed to reduce recidivism and in turn, promote public safety in Canadian communities (CSC, 2019a). Correctional programs that target offenders’ risk factors for criminal behaviour while promoting pro-social ways of addressing these concerns (based on their personal responsivities) have shown to significantly reduce recidivism (Andrews & Bonta, 2010; CSC, 2009; 2019a). In order to ensure that programs align with the risk profiles of offenders, CSC employs the Risk-Need-Responsivity (RNR) model (Andrews & Bonta, 2010; CSC, 2019a). The RNR model is comprised of three core principles; the risk principle states that program intensity must match the offender’s level of risk for re-offending. Matching an offender to the correct intensity level is imperative because inappropriate designations not only lead to wasted resources but can also have the opposite of the desired effect, increasing the likelihood of recidivism (CSC, 2007). Second is the need principle, which states that programs must address the offender’s criminogenic needs (also referred to as dynamic risk factors) which includes antisocial attitudes, criminal peers, antisocial personality, substance use, family concerns, employability concerns, and poor use of leisure time (Andrews & Bonta, 2010). Finally, the responsivity principle states that programs must be tailored to the learning styles and strengths of the offender and should be grounded in cognitive social learning methods (Andrews & Bonta, 2010). Therefore, determining the most appropriate correctional goal, plan, and program need(s) for each offender under CSC jurisdiction is guided primarily by RNR concepts (CSC, 2007-06). Previous research has supported the importance of adhering to the RNR principles, with findings suggesting that increased adherence to RNR is related to significant reductions in re-offending for a variety of offender sub-groups, including violent offenders (Stewart, Gabora, Kropp & Lee, 2014) and offenders with substance abuse needs (Prendergast, Pearson, Podus, Hamilton & Greenwell, 2013).

Programming offered in the community has proven effective for a variety of offender subgroups, including child sex offenders (Lambie & Stewart, 2012), sex offenders in general (Abracen & Looman, 2016), and those with mental health concerns (Hodgins et al., 2007). One study conducted by the Research Branch at CSC (Wilton, Nolan & Stewart, 2015) explored the additive effects that key services and interventions have on the rates of revocations of conditional release for federally sentenced men. The study included 12,273 offenders who were admitted to custody between September 2009 and August 2013 and released on day parole, full parole, or statutory release before April 13, 2014. Overall, Wilton and colleagues (2015) found that additional services such as employment, visits, educational, vocational, maintenance, and correctional programs delivered in the community (e.g., CP, CMP, Community Employment Center services) improved offender outcomes. In particular, they saw reductions in revocations of conditional release with an offence, over and above the effect of completing correctional programming in the institution. In fact, the completion of correctional programs in the community provided some of the most positive findings (Wilton et al., 2015). For both non-Indigenous and Indigenous men, Wilton and colleagues (2015) found that the completion of correctional programs in the community contributed to lower rates of revocation with an offence. Those who failed to complete correctional programs in the community were 4 times more likely to have revocations of conditional release compared to offenders who did complete the intervention (Wilton et al., 2015). Although not speaking solely to community correctional programing, the findings highlight the importance of providing services in the community both in the capacity of the CP and the CMP; two programs that work together to reduce recidivism and promote public safety and security in Canada.

According to Zamble and Quinsey (2001), recidivism is a result of offenders’ inability to cope with their post-release life, despite having intentions to be prosocial citizens. The inability to cope with small issues creates larger issues and eventually the situation becomes uncontrollable for the offender. This can lead to a buildup of negative emotions (e.g., anxiety) which can lead to antisocial coping mechanisms (e.g., substance abuse), and may result in criminal behaviour (Zamble & Quinsey, 2001). The CMP addresses the typical recidivism process by reinforcing what the offenders learned during the CP (e.g., how to cope with negative emotions), while helping offenders implement adaptive strategies in real time as they reintegrate into the community (CSC, 2010).

Houser, Salvatore, and Welsh (2012) explored the role of community aftercare effects on offender’s post-treatment success. Houser and colleagues (2012) found that despite the effectiveness of drug treatment, many offenders with substance abuse needs relapse, especially in the first three months post-treatment. Aftercare programs in the community that maintained the continuum of care and provided support during substance abusers’ most vulnerable phases contributed to reducing the risk for relapse (Houser et al., 2012). Houser and colleagues (2012) referred to CMPs as the "hallmark to successful treatment" (p. 118) because they counteract the factors that lead to recidivism by reinforcing important lessons from original programming, while providing support to offenders throughout reintegration.

In line with this, a 2009 evaluation on Correctional Programs found that CMP exposure for both non-Indigenous and Indigenous men was associated with lower levels of readmission to custody for technical revocations, new offences, and/or new violent offences (CSC, 2009). Similarly, in a subsequent evaluation, CSC (2010) established that engaging in a CMP following completion of a Main Program in the institution played an important role in reintegration because it maintained the continuum of care, reinforced skills learned in Main Programming, and as a result, maximized the benefits of treatment. CSC (2010) found that both non-Indigenous and Indigenous men who did not participate in the CMP were 1.41 and 1.72 times more likely to return to custody compared to those who did. Although these evaluations focused on offenders who completed Main Programming within institutions, it nonetheless speaks to the importance of maintaining the continuum of care and supporting offenders throughout reintegration to contribute to reductions in recidivism. Given that the CP is delivered to a unique population of offenders who have not completed their Main correctional program, there is a need to conduct an evaluation to determine whether the CP is achieving its stated objectives.

Given that the CP is embedded within a sub-program of the overall model for Correctional Programs, a visual representation of the logic model would not adequately capture the complex interactions with the remaining program components. As such, a written description of the various logic model components, including key activities, outputs, and immediate and longer-term strategic outcomes for CSC’s CP is provided below.

Key activities:

• Deliver the CP
• Identify risk factors, crime process, and develop a self-management plan in order to prepare offenders to address their risk factors.

Outputs:

• CP completions
• Completed final program report
• Completed offender self-management plans

Immediate outcome:

• Offenders gain familiarity with ICPM terminology and concepts⁵
• Offenders gain knowledge and basic skills to begin to address their risk factors⁵
• Offenders are prepared to participate in the CMP

Intermediate outcomes:

• Decreased rates of returns to custody during release
• Should the offender return to custody, they are prepared to participate in their identified correctional Main Program need^{Footnote 5}

Impact:

• Correctional programs contribute to reduced re-offending and facilitate the safe and timely reintegration of offenders in the community

The Correctional Programs - Supplementary Evaluation of the Integrated Correctional Program Model: Community Program builds upon the Evaluation of CSC’s Correctional Reintegration Programs (2020). Specifically, the evaluation of the CP introduces additional information to the Correctional Reintegration Programs Evaluation (2020), which falls under Core Responsibility 2 (Correctional Interventions). It assesses program relevancy and measures achievements and outcomes of offenders who participated in the CP. The efficiency of ICPM was examined in the Evaluation of CSC’s Correctional Reintegration Programs (CSC, 2020), therefore, it is not examined in this evaluation of the CP.

For the purpose of this evaluation, the Primers, as well as the IMP and CMP components are out of scope and will not be examined. However, the CMP is discussed within the context of the CP, given that one of the goals of the CP is to have offenders enroll in and complete the CMP. Offenders who complete the CP and who also enroll and complete the CMP will be examined.

The following questions are addressed in this evaluation:

Relevance:

• Does the CP, a component of the ICPM, address a demonstrable need within federal corrections?
• How do CP objectives align with departmental and governmental priorities?
• Are the CP objectives consistent with the roles and responsibilities of CSC and the federal government?

Effectiveness (achievement of expected outcomes):

• What are the key characteristics of offenders who participate in (i.e., enroll) the CP versus those who are eligible but do not participate?
• Is the completion of the CP related to a reduction in the likelihood of returning to custody?
  • Is the CP equally effective for White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders?
• To what extent do CP completers participate in the CMP?
  • Do rates of enrollment and completion in the CMP following the completion of the CP differ among White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders?
• Among those who complete the CP, is there a relationship between participation in the CMP and the likelihood of returning to custody?

The evaluation used a mixed-method approach comprised of both qualitative and quantitative methodologies. The lines of evidence used to address the evaluation questions are listed below, along with a description of the evaluation sample.

An extensive examination of relevant peer-reviewed literature and internal reports was conducted to address questions related to the relevancy of community programming. This included:

• CSC and other governmental policies and legislation;
• CSC and other governmental reports and documentation (e.g., evaluation reports, research reports, operational documents);
• A review of the literature on correctional programming, community programs, community maintenance programs, and recidivism.

2.1.2 Quantitative data from the Offender Management System (OMS)

Quantitative methodologies were used to address the evaluation questions related to effectiveness (e.g., CP enrollments and completions). A portion of the release cohort used for the Evaluation of CSC’s Correctional Reintegration Programs (2020) was utilized to examine the CP. Data from the OMS were extracted to obtain all offender characteristics, first release information, program participation in the community, and first release community outcomes.

The evaluation sample consisted of men offenders classified as eligible non-participants that were included in the Evaluation of Correctional Reintegration Programs (2020), as well as men offenders who had enrolled in a Main Program during the same time period, but did not complete it. Eligible non-participants were identified as those offenders who had a need for correctional programming, but did not participate in a main correctional program prior to their first release.^{Footnote 6}Correctional program need and involvement in correctional programs was examined for all men offenders who had a first term release between April 1, 2015 and December 31, 2017 in Atlantic, Québec, and Pacific regions; April 1, 2016 to December 31, 2017 in Ontario region; and June 1, 2017 to December 31, 2017 in the Prairie region. The timelines for inclusion differed across regions according to the dates of ICPM implementation.

In total, 1,040 men offenders were identified as being eligible for inclusion in the evaluation sample. Characteristics of the evaluation sample are provided in Table 1. Notably, due to the staggered implementation of ICPM, over half of the sample was made up of offenders from the Québec region, while only 6% of the sample were from the Prairie region. The majority of offenders were White (67%), followed by Indigenous (20%), Black (8%), and those classified as Other Ethnocultural (5%). The sample tended to score low on Reintegration Potential (RP) at release, low on Motivation at release, and high on the overall dynamic factor score, which is based on the Dynamic Factor Identification and Analysis - Revised (DFIA-R). The vast majority of the sample were classified as either moderate (42%) or high (49%) according to the Criminal Risk Index (CRI), and most (80%) were released on statutory release for their first term.

To classify offenders into the evaluation groups, involvement in the CP during the first release was examined. Three evaluation groups were identified: 1) CP completers; 2) Incomplete CP; and 3) No Participation in the CP. Participation in the CP and any subsequent involvement in CMP was examined any time after the offender’s first release until either a return to custody, Sentence Expiry Date (SED), or the end of data collection (February 29, 2020).

Offenders were excluded if they had participated in a correctional program under the previous model of Nationally Recognized Correctional Programs or participated in the ICPM pilot, were admitted prior to June 30, 2009, were released prior to implementation of ICPM within their region, or were released on a long-term supervision order.

Rates of revocation to custody during the first term release were examined for the three evaluation groups to determine the effectiveness of the CP. Revocation for any reason included both with an offence and without an offence, and revocation with an offence included all offence types. Follow-up time began once the offender was released on their first term^{Footnote 7} and ended when they either were returned to custody for the first time, reached SED, or reached the end of data collection (February 29, 2020).

The community outcomes during first release were compared across the three evaluation groups: 1) CP completers, 2) Incomplete CP participants, and 3) eligible non-participants. In an effort to isolate the relationship between CP exposure and community outcomes, analyses controlled for the effects of the following covariates: CRI score, age at release, the number of DFIA-R domains that were scored as requiring an intervention^{Footnote 8}, Motivation level at release, and RP at release. Controlling for these risk-relevant differences enhanced the confidence that any observed relationship between the evaluation group and outcome was truly attributed to the program, rather than the result of pre-existing differences on the covariates. Where sample size allowed, the relationship between ethnicity, defined broadly as White, Black, Indigenous, and other Ethnocultural, and community outcomes was explored, including whether the relationship between CP participation and community outcomes was equivalent across the various ethnicities.

In order to answer the primary evaluation question: "Is the completion of the Community Program related to a reduction in the likelihood of returning to custody?", Cox regression survival analysis was conducted. Cox regression (or proportional hazards regression) is a method for investigating the effect of several independent variables (i.e., covariates) on the time to a specified event. The method assumes that the effects of the predictor variables upon survival are constant over time. As such, Cox regression predicting time from first release to offenders’ first outcomes following that release was conducted to identify the relationship between program participation and the following outcomes:

• Revocation for any reason (with or without offence);
• Revocation with offence.

The key statistic interpreted in the Cox regression results is the hazard ratio (HR), which is a comparison of the probability of events in an intervention group to the probability of events in a comparison group. In order to examine the likelihood of experiencing each community outcome for CP completers, relative to those who did not complete the CP or eligible non-participants, the program completer group was used as the reference group. However for ease of interpretation, the inverse of the hazard ratio (IHR; 1/HR) was calculated. A IHR of 1 indicates that the two groups experienced an equivalent likelihood of the community outcome, whereas a IHR above 1 suggests that those who complete the CP are more likely to experience the outcome, while IHRs below 1 indicate that CP completers are less likely to experience the outcome. A IHR of 0.5 indicates that the treatment group experienced half the likelihood of experiencing the outcome relative to the comparison group. Conversely, the percentage change in the likelihood of the outcome can be determined by computing IHR - 1 (e.g., 0.5 - 1 = -.5 * 100 = -50%), which is then interpreted as a 50% reduction in the likelihood of experiencing the outcome.

To examine predictors of participation in both the CP, as well as the CMP following completion of the CP, logistic regressions were conducted. Logistic regression models the relationship between independent variables and a dichotomous outcome variable. The key statistic interpreted in the results is an odds ratio (OR), which compares the odds of experiencing the outcome for one group (or level of a variable) relative to another. Similar to HRs, an OR of 1 indicates that the two groups had equivalent odds of experiencing the outcome. Multiple variables can be included in the model simultaneously to examine the effects of participation in programming on community outcomes, for example, while holding the effects of other relevant variables (e.g., age, risk information) constant. Similar to the OR, the percentage change in the odds of experiencing the outcome can be calculated by subtracting 1 from the OR.

Although the methodology used throughout this evaluation allowed for a rigorous analysis of the evaluation questions that mitigated as many limitations as possible, there are a few core limitations that should be considered alongside the conclusions drawn in this report.

Given the rolling regional implementation of ICPM, the regional representativeness of the sample was unable to be maintained. Offenders from the Prairie region had the shortest window of opportunity to be included in the evaluation sample. This means that the findings that are generated in this evaluation may not generalize to CSC’s offender population. Specifically, the lack of representation from the Prairie region meant that there was a lower proportion of Indigenous offenders included in the sample. Although Indigenous offenders were included as a subgroup in all analyses, it is possible that having a representative sample of Indigenous offenders would result in different findings. Small sample sizes among Black offenders and other Ethnocultural offenders also limited the ability to examine whether the effectiveness of the CMP was comparable across the relevant offender subgroups. As more data become available (e.g., as time passes and more offenders participate in the CP), it will be important to continually examine whether the program objectives are being achieved.

This evaluation focused solely on an offender’s involvement in the CP, and subsequent CMP, if applicable. Exposure to additional services (e.g., cultural services and supports for Indigenous offenders, substance abuse treatment, mental health services, employment programs, etc.) while in the community was unable to be accounted for. Additional services and supports that offenders receive can also promote successful reintegration, so it is important to acknowledge that the findings presented in this evaluation are not attributed solely to the participation (or lack of participation) in the CP and CMP. The results presented throughout the report present the relationship between various levels of exposure to the CP and CMP and first release outcomes, while considering the effects of other risk relevant variables. The lack of information surrounding each offender’s involvement in additional services represents one example of unobserved or unaccounted information that could be relevant to further understand the findings. Further examination of the effectiveness of the CP would benefit from including the exposure to other additional services and interventions to more fully depict the relationship between these elements of the ICPM and successful reintegration.

It is also important to note that the reasons for non-participation in the CP were unable to be explored. The administrative data that were utilized for this evaluation indicated whether an offender was assigned to the CP, began their participation in the CP (e.g., enrolled in the program), and whether they successfully completed it or failed to complete it successfully. Although all offenders who have a need for correctional programming that is not met while in the institution would be targeted for the CP, some may have refused to participate in the CP, may not have had the opportunity to participate, or were involved in other interventions that were unable to be accounted for. As discussed previously, in an effort to ensure that offenders classified as eligible non-participants would have reasonably had the opportunity to participate in the CP, an analysis of the time on supervision prior to a return to custody, if applicable, was examined. Results from this examination indicated that offenders in the eligible non-participant group tended to remain in the community for at least as long as the average time to the start of the program for those with exposure to the CP, indicating that it was unlikely the case that eligible non-participants had insufficient time in the community to access the program. Future examination would benefit from developing a greater understanding of why those classified as eligible non-participants were not accessing the CP during their first release.

Additionally, the administrative data did not allow for an examination of why an offender failed to complete the CP successfully. It is possible that a suspension of community supervision and subsequent return to custody prompted the incomplete CP, which would have influenced the findings on the effectiveness of the CP. In addition to considering return to custody as an index of program effectiveness, future examinations of the CP should also consider the rate of suspension of community supervision among the evaluation groups. Specifically, it would be worthwhile to determine whether eventual CP completers are more likely to have their release to the community maintained following a suspension, if a suspension occurs.

Taken together, these limitations underscore the importance of being cautious when drawing conclusions from the findings presented in the evaluation. However, the findings do provide an initial foundation of evidence pertaining to whether the CP is achieving the desired program outcomes and, as such, the findings can be used to identify areas in need of monitoring or improvement.

The key findings of the evaluation of the CP are presented under the following two sections:

• Relevance of the CP
• Effectiveness of the CP: Program Outcomes

This section summarizes the evaluation findings related to the relevance of CSC’s ICPM CP. It does this by exploring its alignment with governmental and departmental priorities, its consistency with the roles and responsibilities of the federal government, as well as whether there is a demonstrable need for the program. Analysis of the expected outcomes of the CP will be conducted in subsequent sections of the evaluation.

CSC’s ICPM CP aligns with both CSC’s and the federal government’s priorities, roles, and responsibilities. The delivery of effective correctional programs, including the CP, contributes to the overall priority of a just, peaceful, and safe society.

The ICPM CP aligns with, and supports the federal government’s priorities, roles and responsibilities. As stated in the Prime Minister’s 2019 Mandate Letter to the Minister of Public Safety and Emergency Preparedness, the Minister is directed to "advance measures to keep our communities safe". As noted in the literature review, one way to keep communities safe, within the context of correctional services, is through the rehabilitation of offenders. The CP, specifically targets offenders with a need for correctional programming who did not receive programming in the institution, assisting with the rehabilitation of offenders, and thus, enhancing the safety of communities.

According to Section 3(b) of the Corrections and Conditional Release Act (CCRA, 1992) "the purpose of the federal correctional system is to contribute to the maintenance of a just, peaceful and safe society by… assisting the rehabilitation of offenders and their reintegration into the community as law-abiding citizens through the provision of programs in penitentiaries and the community". CSC is mandated by law to provide correctional programs (CCRA, 1992, s. 5(b)). The CCRA (1992), section 3, paragraph 5(b), and sections 76-80 provide the legislative framework guiding the development, implementation, and maintenance of correctional programming for CSC. Particularly, CSC is mandated to "…provide a range of programs designed to address the needs of offenders and contribute to their successful reintegration into the community" (s. 76). The CP provides exposure to correctional programming for offenders released to the community who did not complete Main Programming in the institution, despite having a need for it.

The CP is also aligned with the Minister of Public Safety and Emergency Preparedness’ 2018 Mandate Letter to the Commissioner of CSC. The Commissioner is mandated to ensure that CSC protects Canadian communities and ensure that individuals in the custody of CSC "receive the programming, interventions and treatment they require, in an environment that is safe, secure and humane". The CP provides offenders with the opportunity to receive correctional programming in a community setting in cases where offenders were unable to do so in the institution. Although not a replacement for Main correctional programming, the CP’s ability to expose offenders to program content aims to contribute to reductions in reoffending and in turn, promotes the safety and security of Canadians.

The priorities of CSC’s ICPM CP are aligned with CSC’s corporate priorities, roles and responsibilities. For example, CSC’s mission statement is "the Correctional Service of Canada (CSC), as part of the criminal justice system and respecting the rule of law, contributes to public safety by actively encouraging and assisting offenders to become law-abiding citizens, while exercising reasonable, safe, secure and humane control". The CP contributes to assisting offenders to become law-abiding citizens by helping identify offenders’ crime process and risk factors, and encourage and motivate offenders to work through their risk factors and lead crime free lives (CSC, 2014b).

In addition, Commissioner’s Directive (CD) 726: Correctional Programs provides a policy framework from which CSC can implement correctional programming. The policy’s main objectives are to:

• ensure that offenders receive the most effective correctional programs at the appropriate time in their sentence;
• rehabilitate and prepare offenders for reintegration into the community;
• maximize the effectiveness of correctional programs identified in the offender’s correctional plan;
• ensure integrity in the program management and delivery; and
• ensure that correctional programs respect gender, ethnic, cultural and linguistic differences, and are responsive to the special needs of women, Indigenous offenders, offenders requiring mental health care, and other groups.

As previously indicated, the primary goal of correctional programming is to assist offenders in becoming law-abiding citizens. That is, the CP assists with providing correctional programming to offenders who have a need but did not receive programming while in the institution, providing offenders with an opportunity to gain program exposure, become law-abiding citizens, ultimately contributing to public safety. This is directly related to two of CSC’s (2016) six corporate priorities, which stem from CSC’s mission, legislative framework, and risk profile. These two priorities include: 1) the safe management of eligible offenders during their transition from the institution to the community, and while on supervision; and 2) the safety and security of members of the public, victims, staff, and offenders in our institutions and in the community. Together the CP along with the CMP provide offenders re-entering the community with an opportunity to learn and implement adaptive strategies that will contribute to their rehabilitation and support their transition into the community. Overall, the support provided to offenders with an unmet need for correctional programming through the CP and the CMP promotes the safety and security of all community members (public, victims, staff, and offenders).

Finally, as noted in the Departmental Results Report (DRR), CSC has three core responsibilities: (1) Care and Custody, (2) Correctional Interventions, and (3) Community Supervision. The CP best aligns with Core Responsibility 2: Correctional Interventions. It states, "CSC conducts assessment activities and program interventions to support federal offenders' rehabilitation and facilitate their reintegration into the community as law-abiding citizens" (CSC, 2018). The CP aligns with this responsibility as it provides offenders who did not complete their ICPM Main Program in the institution an opportunity to gain program exposure. The CP supports successful reintegration into the community, by encouraging offender accountability and full participation in the rehabilitation process.

There is a continued need for CSC to provide community programming to federal men offenders. Over recent fiscal years, approximately 17% to 24% of men offenders with an identified need for correctional programming did not complete the programming within the institution prior to their first release. These men would then be targeted to receive the CP during their release to ensure that they are exposed to programming and able to connect to the CMP.

There is a continued need to provide community programming to federal men offenders. CSC is mandated to develop and maintain intervention strategies that aim to facilitate the successful reintegration and rehabilitation of each offender. Offender program need is identified during the intake process based on offenders’ risk assessment results and criminal history. As reported in the Evaluation of Correctional Reintegration Programs (CSC, 2020), there is a large body of research indicating that effective correctional programs facilitate successful reintegration and reduce the likelihood of future reoffending. In 2018, approximately 72% of all federal offenders (in custody and in the community) were identified as having a met or an unmet need for a main correctional program (Performance Measurement and Management Reporting, 2018). Of the men offenders with an identified need for correctional programming over the last four fiscal years (FY 2016-17 to FY 2019-20), approximately 80% completed a correctional program prior to first release. Specifically, the percentage of offenders who completed correctional programming prior to first release was: 83% in FY 2016-17; 82% in FY 2017-18; 78% in FY 2018-19; and 76% in FY 2019-20.^{Footnote 9} This indicates that from FY 2016-17 to FY 2019-20, between 17% and 24% of men offenders with an identified correctional program need did not complete programming within the institution prior to their first release, demonstrating a need for programming offered in the community.

Overall, the abovementioned findings support the conclusion that the CP is a relevant program for CSC to offer.

This section speaks to the effectiveness of the CP, that is, whether it is achieving the objective of reducing the rates of revocation to custody during a first term release. First, key characteristics of the men included in the evaluation sample are presented to describe what differentiates those who participate in the CP upon release from those who do not, despite being eligible. The rate of participation in the CP and likelihood of community outcomes for the evaluation groups serve as the primary indices of program effectiveness. Subsequent participation in the CMP is also examined, given that one of the objectives of the CP is to enroll offenders into the CMP.

In total, 355 men offenders who completed the CP, 253 men offenders who did not complete the CP after enrolling (i.e., CP incompleters), and 432 men offenders considered to be eligible non-participants, were included in the evaluation sample. The characteristics of the three groups in the evaluation sample are included in Table 2.

It is noteworthy that a higher proportion of Indigenous men were classified as CP incompleters (24%) or eligible non-participants (22%) compared to CP completers (14%). The regional distribution was consistent across the three evaluation groups, with the majority of offenders from the Québec Region. CP completers were less likely to be released on statutory release (67%) compared to CP incompleters (92%) and eligible non-participants (85%). When examining a variety of risk relevant measures scored at release, the CP completers tended to be lower risk than those who did not complete the CP and eligible non-participants. For example, 38% of CP completers scored high on the CRI, whereas 58% of those who did not complete the CP and 53% of eligible non-participants scored high on the CRI. CP completers tended to be slightly older at release (M = 39) than those who did not complete the CP (M = 36) and eligible non-participants (M = 37). The average sentence length of slightly more than three years was similar across the evaluation groups. Importantly, all three groups appeared to have adequate time to complete the CP following their first release, as evidenced by the average number of days from first release to SED (M = 475 for CP completers, M = 415 for CP incompleters, and M = 482 for eligible non-participants). It is important to consider that the time between first release and SED considers the potential time that the offender could have had to participate in the CP, without considering revocations to custody. Among eligible non-participants, even when the time to first revocation is considered (for those who were revoked from their first term release), it appears that most would have had sufficient time to at least enroll in the CP prior to the revocation (M days to revocation from first release = 160, SD = 93, median = 131).

Given that some offenders could have had exposure to Main correctional programs during their first term of incarceration, it was important to examine this and determine whether there was the potential to impact the results for the effectiveness of the CP. The proportion of offenders within each evaluation group that had exposure to Main Programming during their first term of incarceration was relatively consistent. Among those who completed the CP, 82 (23%) had exposure to a Main Program during their first term of incarceration, compared to 66 (26%) CP incompleters and 90 (21%) eligible non-participants. On average, those who had exposure to a Main correctional program but did not complete it were enrolled for approximately 65 to 75 days (CP completers = 65 days (SD = 45), CP incompleters = 66 days (SD = 52), and eligible non-participants = 74 days (SD = 49)). The effect of previous Main correctional program exposure was examined in relation to CP participation and first release outcomes. It was not found be a relevant predictor of first release outcomes, after controlling for other risk relevant factors and CP participation. As a result, prior exposure to Main Programming was not included in the final models, as discussed below.

More than half of offenders eligible to participate in the CP did so during their first release. The majority of those who enrolled in the CP successfully completed it. Most offenders appeared to have had sufficient time under sentence in the community to participate in the CP, which on average, began 55 days after first being released to the community.

The rate of participation in the CP was examined to determine the extent to which offenders who would be eligible to participate in the program were in fact accessing it during their first release. More than half (59%) of offenders eligible to participate in the CP did so during their first release (see Table 3). Of those who participated (i.e., enrolled) in the CP (n = 608), 58% successfully completed it. One factor that may limit an offender’s ability to participate in the CP is the time that they will remain on sentence once first released to the community. As discussed previously, the majority of offenders were released on statutory release, meaning they would have a smaller proportion of their sentence to serve in the community. However, on average, the evaluation sample tended to have more than one year remaining on their sentence at the time of first release, indicating that there would have likely been sufficient time to participate in the CP for most offenders (see average time to CP enrollment below). For eligible non-participants, it was also important to ensure that the majority of offenders were not returned to custody before they would have had the opportunity to participate in the CP. On average, among eligible non-participants who were returned to custody (n = 227), the average time on supervision prior to the return was 160 days (SD = 93, median = 131).

For those that did participate in the CP, their enrollment in the program began after approximately 55 days in the community, although it started as soon as four days following release, to as long as 499 days after first release. On average, the CP was successfully completed after 185 days, with the shortest time to successful completion being within 35 days of release, and extended to as long as 554 days. Taken together, these results highlight the variation in the amount of time that passes from release to involvement in the CP, but indicates that enrollment in the CP tends to occur within the first two months of release.

Indigenous offenders and other Ethnocultural offenders experienced lower odds of enrolling in the CP relative to White offenders. Black offenders experienced comparable odds of enrolling in the CP relative to White offenders.

However, Black offenders, Indigenous offenders, and other Ethnocultural offenders experienced comparable odds of completing the CP relative to White offenders.

To understand what differentiated those who enrolled in the CP compared to those who did not enroll, the relationships between key characteristics and the odds of enrolling in the CP were examined. Eventual CP completers and CP incompleters were merged to create a group of offenders who enrolled in the CP compared to those who did not. Results presented in Table 4 highlight the relationship between the given predictor and the odds of enrolling in the CP. Results indicated that risk relevant characteristics (e.g., age at release, CRI score, RP) were not significantly related to the odds of enrolling in the CP. Results did suggest that, while holding the effects of all other predictors constant, Indigenous offenders and other Ethnocultural offenders experienced significantly lower odds of enrolling in the CP relative to White offenders. Specifically, Indigenous offenders experienced a 31% decrease in the odds of enrolling in the CP relative to White offenders. Other Ethnocultural offenders experienced a 50% reduction in the odds of enrolling in the CP compared to White offenders. It is important to highlight that the regional representation of the evaluation was skewed due to the regional rollout of the ICPM. As a result, few offenders from the Prairie region contributed to these results, where the proportion of Indigenous offenders is the highest. Additionally, the number of other Ethnocultural offenders in the evaluation sample is small (n = 25 enrollments, n = 30 eligible non-participants). Both points underscore the importance of replicating these findings with larger sample sizes when possible.

Next, the relationships between key characteristics and the odds of completing the CP were examined. Results from the analysis are presented in Table 5. The odds of completing the CP were not associated with the offender’s ethnicity, indicating that all four subgroups experienced equivalent odds of completing the CP if they had enrolled. Motivation level was significantly related to the odds of completing the CP, such that those who scored as having moderate motivation at release experienced an 87% reduction in the odds of completing the CP relative to those who scored as having high motivation. Similarly, offenders who were assessed as having low motivation at release experienced an 89% reduction in the odds of completing the CP. Age at release also emerged as a significant predictor of the odds of completing the CP, such that older offenders experienced an increase in the odds of completing the CP. Identifying factors that are related to the odds of completing the CP is important for targeting offenders who may be at a heightened risk of failing to complete the CP. These results suggest that strategies aimed at retaining offenders in the CP may best be directed to younger offenders and those with lower levels of motivation.

The relationship between the evaluation groups and first release outcomes of any revocation (i.e., with or without a new offence) and revocation due to a new offence was explored to determine the effectiveness of the CP. Cox regression survival analyses were conducted to determine the hazard of experiencing a revocation during the first release for CP completers relative to CP incompleters and eligible non-participants. To isolate the relationship between participating in the CP and the likelihood of first release outcomes, the effects of the following factors were considered: CRI score, the number of dynamic factor domains requiring intervention, RP at release, motivation level at release, age at release, and the offender’s ethnicity or racial background.

CP completers demonstrated significant reductions in the likelihood of a revocation to custody (both for any reason and for a new offence) relative to CP incompleters and eligible non-participants.

The reduction in the likelihood of a revocation for any reason and for a new offence is expected to be consistent across White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders.

Table 6 presents the relationship between evaluation groups and any revocation (i.e., with or without an offence) for men offenders during their first release. Out of 997 men, 420 had a revocation for any reason during their first release. When examining the relationship between evaluation group and any revocation (while holding the effects of all other variables constant) the results indicated that CP incompleters and eligible non-participants experienced significantly higher likelihoods of being returned to custody relative to CP completers. Said another way, CP completers experienced significantly lower likelihoods of a revocation for any reason relative to both CP incompleters and eligible non-participants. Specifically, CP completers experienced a 79% decrease in the likelihood of a revocation for any reason compared to CP incompleters (IHR = 1/4.87 = 0.21). Similarly, CP completers experienced an 81% decrease in the likelihood of a revocation for any reason compared to eligible non-participants (IHR = 1/5.30 = 0.19).

Results also indicated that the relationship between evaluation groups and the likelihood of any revocation during first release was not significantly different across White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders (as evidenced by the non-significant interaction terms included in the model). In other words, all men offenders who complete the CP (regardless of ethnicity) are expected to experience an equivalent reduction in the likelihood of a revocation, relative to CP incompleters and eligible non-participants.

Table 6 also presents the results for the relationship between evaluation groups and revocation due to a new offence. Out of 996 men eligible for inclusion in the analysis, only 65 had their first release end with a revocation due to a new offence. Due to the low occurrence of the event, the parameter estimates are less stable than the results for any revocation. That being said, results were consistent in that CP completers experienced a significantly lower likelihood of receiving a revocation to custody due to a new offence than both CP incompleters and eligible non-participants. While holding the effects of all other risk relevant variables constant, CP completers experienced a 73% decrease in the likelihood of a revocation due to a new offence compared to CP incompleters (IHR = 1/3.76 = 0.27). Relative to eligible non-participants, CP completers experienced a 77% decrease in the likelihood of a revocation due to a new offence (IHR = 1/4.41 = 0.23).

Consistent with the results for any revocation, there was no evidence that the relationship between evaluation groups and the likelihood of a revocation due to a new offence was different for White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders.

One of the objectives of the CP is to have offenders participate in the CMP. Participation in the CMP was explored for all CP completers in the evaluation sample. Before examining the rate of participation in the CMP, the amount of time remaining on supervision following the completion of the CP up until SED was examined to ensure that it would be reasonable to expect that offenders would have the opportunity to participate in the CMP. Table 7 presents the amount of time remaining on sentence following the completion of the CP. Based on the time remaining on their sentence, it appears that the majority of offenders would have had sufficient time to participate in at least one cycle of the CMP^{Footnote 10}, although it is possible that an offender was revoked to custody prior to their SED.

More than half of offenders eligible to participate in the CMP did so after completing the CP. The majority of those who enrolled in the CMP successfully completed it. A portion of offenders likely would not have had sufficient time remaining on their sentence following the completion of the CP to participate in CMP, underscoring the importance of efficiently delivering both programs to remove potential barriers to CMP participation.

An examination of the rate of participation in CMP following the successful completion of the CP revealed that 187 offenders (53%) enrolled in CMP during their first release (see Table 8). Of those who enrolled, 68% successfully completed it. However, it is worth noting that among the evaluation sample of CP completers, only 36% of offenders went on to successfully complete the CMP. It is important to remember that some may not have had sufficient time to participate in the CMP following the completion of the CP. The next section presents results from logistic regressions to better understand what differentiates those who enroll in the CMP versus those who do not.

Offenders with more time remaining on their sentence following the completion of the CP experienced greater odds of participating in, and completing, the CMP. White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders experienced comparable odds of enrolling in, and completing, the CMP.

To understand what differentiated those who enrolled in the CMP compared to those who did not enroll after successfully completing the CP, the relationships between key characteristics and the odds of enrolling in the CMP were examined. Eventual CMP completers and CMP incompleters were merged to create a group of offenders who enrolled in the CMP compared to those who did not. Results presented in Table 9 highlight the relationship between the given predictor and the odds of enrolling in the CMP. The number of months remaining on sentence for each offender following the completion of the CP was included in the model. This allowed for an examination of whether those with more time under sentence were more likely to participate in the CMP. Results indicated that for every one month increase in the remaining time under sentence following the completion of the CP, there was a corresponding 9% increase in the odds of enrolling in the CMP. Those with a greater number of DFIA-R domains scored as requiring an intervention also experienced increased odds of enrolling in the CMP. Specifically, for each additional DFIA-R domain requiring intervention, the odds of enrolling in the CMP are expected to increase by 32%. All other risk relevant variables did not distinguish between those who enrolled in the CMP versus those who did not. It is important to highlight that Black offenders, Indigenous offenders, and other Ethnocultural offenders all experienced comparable odds as White offenders for enrolling in the CMP.

Next, the predictors of successful completion of the CMP were examined. This analysis only included those who had enrolled in a CMP following the completion of the CP (n = 175). Similar to predicting the odds of enrolling in the CMP, the number of months remaining on sentence following the CP was significantly related to the odds of completing the CMP (see Table 10). For every one month increase in the amount of time between CP completion and SED, the odds of completing the CMP increased by 29%. The number of DFIA-R domains that were scored as requiring an intervention also emerged as a significant predictor of the odds of completing the CMP. Those with a greater number of the DFIA-R domains were less likely to complete the CMP. For each additional DFIA-R domain that was scored as requiring intervention, there was a corresponding 39% decrease in the odds of completing the CMP after having enrolled. Results indicated that the odds of completing the CMP were not related to ethnicity. In other words, Black offenders, Indigenous offenders, and other Ethnocultural offenders, experienced comparable odds of completing the CMP, relative to White offenders.

First release outcomes were examined to determine whether there was an added benefit associated with participating in the CMP following the successful completion of the CP. The likelihood of any revocation during the first release was examined for those who successfully completed the CMP, enrolled in the CMP but did not successfully complete it, and those who did not enroll in the CMP following the successful completion of the CP. The analysis was restricted to examining any revocation (i.e., with or without a new offence) due to sample size and low occurrence of revocations due solely to a new offence. Cox regression survival analysis was conducted to determine the hazard of experiencing a revocation during the first release for these groups. To isolate the relationship between participating in the CMP and the likelihood of first release outcomes, the effects of the following factors were considered: CRI score, the number of dynamic factor domains requiring intervention, RP at release, motivation level at release, age at release, and the offender’s ethnicity or racial background. Given that all offenders included in the analysis had to have successfully completed the CP, the follow-up time began at the time of CP completion, and ended either at the first release return to custody date for those with a revocation, SED, or the end of the data collection period. Those who had less than 35 days remaining on sentence following the completion of the CP were excluded from the analysis, as no revocations to custody occurred during this period and Cox regression requires that all censored cases (i.e., those who did not experience the event but ended their follow-up) are followed at least as long as the time to the first failure. This resulted in removing 26 offenders from the analysis (one with an incomplete CMP, and 25 with no enrollment in the CMP). Whether the effectiveness of the CMP was comparable across White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders could not be explored due to small sample size.

It is beneficial for offenders to complete the CMP following the successful completion of the CP. CMP completers experienced significantly lower likelihoods of a revocation for any reason relative to both CMP incompleters and those with no participation in CMP.

Table 11 presents the relationship between evaluation groups, defined by their exposure to CMP, and any revocation (with or without a new offence) for men offenders during their first release. Out of 312 men included in the analysis, 55 had a revocation for any reason during their first release. When examining the relationship between evaluation group and any revocation (while holding the effects of all other variables constant) the results indicated that CMP incompleters and those with no exposure to CMP experienced significantly higher likelihoods of being returned to custody relative to CMP completers. Said another way, CMP completers experienced significantly lower likelihoods of a revocation for any reason relative to both CMP incompleters and those with no participation in CMP. Specifically, CMP completers experienced a 91% decrease in the likelihood of a revocation for any reason compared to CMP incompleters (IHR = 1/11.15 = 0.09). Similarly, CMP completers experienced an 86% decrease in the likelihood of a revocation for any reason compared to those with no exposure to CMP (IHR = 1/7.14 = 0.14). It is important to note that although these findings are statistically significant, estimates of the parameter are imprecise (as evidenced by the wide confidence intervals). As a result, it is appropriate to conclude that there appears to be a relationship between the completion of the CMP and reduced rates of revocation for any reason during the first release, but future examination of this relationship is warranted to derive a more robust estimate of the magnitude of the effect.

It should be highlighted that the findings above do not suggest that CMP incompleters experience a higher likelihood of revocation during the first release relative to those with no exposure to CMP. A follow-up analysis that directly compared the likelihood of revocation for these two groups (while holding the effects of all other variables constant) indicated that they experienced comparable likelihoods (i.e., the difference was not statistically significant).

This evaluation examined the relevance and effectiveness of the CP, a component of the ICPM aimed at providing exposure to core correctional program concepts to offenders with a need for correctional programming who did not complete a Main Program in the institution prior to their first release. Results indicated that the CP is relevant and it is addressing a demonstrable need among the offender population supervised in the community. Positive findings were obtained when considering the effectiveness of the CP. Those who successfully completed the CP experienced a lower likelihood of being revoked to custody during their first release, compared to those who failed to complete the CP and those who were eligible but did not participate. The effectiveness of the CP was comparable for White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders.

The CP also aims to connect offenders with the CMP. Results indicated that approximately half of eligible offenders went on to enroll in the CMP, and those that did complete the CMP demonstrated reductions in the likelihood of a revocation to custody for any reason, relative to those who enrolled in the CMP but did not complete it, or those who did not enroll in a CMP following the successful completion of the CP.

Two recommendations are put forward in an effort to enhance the level of participation in the CP and the CMP so that offenders can benefit from exposure to these program components. Namely, the enrollment rate to the CP should be examined across relevant offender subgroups to determine if Indigenous offenders and other Ethnocultural offenders demonstrate discrepant enrollment rates, relative to White offenders, and opportunities to enhance the timely delivery of both the CP and CMP should be examined to ensure that offenders are not limited in their ability to participate in either program due to the time remaining on their sentence.

Lastly, it is important to note that the findings for this evaluation are based on a cohort of offenders released shortly after the implementation of the ICPM. Although the findings contribute to establishing a foundation for the effectiveness of the CP, there is the potential that a subsequent examination with a representative sample could yield different results.

Based on the abovementioned findings, the following recommendations are put forward in an effort to enhance the level of participation in the CP and the CMP so that more offenders can benefit from exposure to these program components.

Predictors of enrolling in the CP were explored to determine what differentiates between those who enroll in the CP during their first release versus those who do not. Results indicated that, while holding the effects of other risk relevant variables constant, Indigenous offenders, and other Ethnocultural offenders experienced significantly lower odds of enrolling in the CP, relative to White offenders. Given the finding that the CP appears to be equally effective for White offenders, Black offenders, Indigenous offenders, and other Ethnocultural offenders, it is important to ensure that all offenders are accessing the CP at comparable rates.

It is recommended that the CSC determines if the rates of enrollment in the ICPM-CP for relevant offender subgroups, including Indigenous offenders and other Ethnocultural offenders, are discrepant. Should it be determined that there are discrepancies in the enrollment rates to ICPM-CP for Indigenous offenders and other Ethnocultural offenders, CSC will address any unequal access to promote comparable enrollment rates.

Results indicated that those who successfully completed both the CP and the CMP had a significantly lower likelihood of a revocation to custody for any reason during their first release, relative to those who enrolled in the CMP and those who did not enroll in the CMP following successful completion of the CP. Although on average, offenders in the evaluation sample had more than a year remaining under sentence between when they were first released and when they were to reach SED, it is possible that some offenders did not have sufficient time remaining on their sentence to complete both the CP and the CMP. Most offenders who completed the CP had three months or more remaining on their sentence following the program, but results examining predictors of enrollment in the CMP indicated that those with more time remaining on supervision experienced greater odds of enrolling and subsequently completing the CMP. As a result, it would be beneficial to ensure that both the CP and the CMP are being delivered in a timely manner to offenders who have a need for correctional programming but did not complete their Main Program prior to their first release.

It is recommended that CSC determines whether there are opportunities to enhance the timely delivery of the CP and CMP to ensure offenders with limited time remaining on their sentence have an opportunity to participate and benefit from exposure to the program content.

Abracen, J., & Looman, J. (2016). Treatment of high-risk sexual offenders: an integrated Approach. Wiley/Blackwell.

Andrews, D. A., & Bonta, J. (2010). The psychology of criminal conduct (5th ed.). Cincinnati, OH: Anderson Press.

Conditions and Conditional Release Act. 1992, c. 20.Retrieved from http://laws.justice.gc.ca/eng/C-44.6/index.html

Correctional Service of Canada. (2009) Correctional Service Canada’s Correctional Programs. Retrieved from https://www.csc-scc.gc.ca/text/pa/cop-prog/cp-eval-eng.shtml

Correctional Service of Canada. (2010). Evaluation Report: Community Maintenance Program. Retrieved from https://www.csc-scc.gc.ca/publications/092/005007-2010-eng.pdf

Correctional Service of Canada. (2014a). ICPM Community Maintenance Program: Program Module.

Correctional Service of Canada. (2014b). ICPM Community Multi-Target Program: Program Manual.

Correctional Service of Canada. (2014c). Integrated Correctional Program Model. Retrieved from https://www.csc-scc.gc.ca/correctional-process/002001-2011-eng.shtml

Correctional Service of Canada. (2019a). Correctional Programs. Retrieved from https://www.csc-scc.gc.ca/002/002-0001-en.shtml

Correctional Service of Canada. (2019b). Reintegration Programs: Correctional and Education Program Descriptions. Ottawa, ON

Correctional Service of Canada. (2016). Report on Plans and Priorities. Ottawa. ON.

Correctional Service of Canada. (2007). Risk-Need-Responsivity Model for Offender Assessment and Rehabilitation. Retrieved from https://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/rsk-nd-rspnsvty/rsk-nd-rspnsvty-eng.pdf

Corrections and Conditional Release Act, Revised Statutes of Canada (1992, c. C-20). Retrieved from the Justice Laws website: https://www.laws-lois.justice.gc.ca/eng/acts/C-44.6/FullText.html

Hodgins, S., Tengström, A., Eriksson, A., Österman, R., Kronstrand, R., Eaves, D., Hart, S., Webster, C., Ross, D., Levin, A., Levander, S., Tuninger, E., Müller-Isberner, R., Freese, R., Tiihonen, J., & Kotilainen, I. (2007). A Multisite Study of Community Treatment Programs for Mentally Ill Offenders with Major Mental Disorders: Design, Measures, and the Forensic Sample. Criminal Justice and Behavior, 34(2), 211-228. https://doi.org/10.1177/0093854806291248

Houser, K., Salvatore, C., & Welsh, W. (2012). Individual-Level Predictors of Community Aftercare Completion. The Prison Journal, 92(1), 106-124. https://doi.org/10.1177/0032885511429261

Lambie, I., & Stewart, M. (2012). Community Solutions for the Community’s Problem: An Evaluation of Three New Zealand Community-Based Treatment Programs for Child Sexual Offenders. International Journal of Offender Therapy and Comparative Criminology, 56(7), 1022-1036. https://doi.org/10.1177/0306624X11420099

Prendergast, M., Pearson, F., Podus, D., Hamilton, Z., & Greenwell, L. (2013). The Andrews’ principles of risk, needs, and responsivity as applied in drug treatment programs: meta-analysis of crime and drug use outcomes. Journal of Experimental Criminology, 9(3), 275-300. https://doi.org/10.1007/s11292-013-9178-z

Stewart, L., Gabora, N., Kropp, P., & Lee, Z. (2014). Effectiveness of Risk-Needs-Responsivity-Based Family Violence Programs with Male Offenders. Journal of Family Violence, 29(2), 151-164. https://doi.org/10.1007/s10896-013-9575-0

Wilton, G., Nolan, A., & Stewart, L. (2015). The Additive Effects of Participation in Multiple Correctional Interventions and Services for Federally Sentenced Men. Correctional Service Canada.

Zamble, E., & Quinsey, V. (2001). The Criminal Recidivism Process. Cambridge University Press.