Review of sentence management

In 2017, Correctional Service Canada (CSC) erroneously granted an inmate statutory release (SR) prior to his SR eligibility date. The Correctional Operations and Programs Sector (COPS) describes the sequence of events that led to the error as follows:

• In 2012, the inmate was sentenced to 12 years in custody less a credit of 4 years and 10 months for time spent in pre-trial custody. This credit left the inmate with 7 years and 2 months remaining to serve.
• In 2013, the court increased the sentence to 14 years by way of a decision from the provincial Court of Appeal. All ancillary orders remained the same, including the credit for time spent in pre-trial custody.
• CSC incorrectly interpreted and applied the 2013 appeal decision. The inmate's sentence remaining to be served was unchanged following receipt of this decision. This error led to the inmate being granted statutory release prior to his statutory release eligibility date.

The sentence management (SM) process includes collecting, analyzing and researching information to structure, amend and maintain the sentence timeline for an offender. A Warrant of Committal (WOC) is one of the key documents used by SM staff to create a sentence timeline that determines key release dates up to the expiry of the warrant. The information on the warrant as well as sentencing transcripts and other related information are used to ensure that orders from the courts comply with the Criminal Code of Canada (CCC) and other laws, as well as to ensure that the sentence being served by an offender is equal to the sentence that was imposed.^{Footnote 1}

In response to the issues discovered in the above-mentioned case, the Commissioner asked the Internal Audit Sector (IAS) to conduct a Review of Sentence Management as it was identified as a high priority and as an area of high risk to CSC; IAS did not complete a traditional risk assessment for this review; however, the Commissioner communicated specific areas of concern, which directed the lines of enquiry for this engagement (see section 2.2).

The objectives of this review were to ascertain that:

1. the process in place to ensure correct sentence calculation was well defined, understood and implemented given recent changes to the structure; and
2. sentence calculations were correct for a sample of high profile offenders.

To fulfill the review objectives, IAS used the following approach:

1. IAS assessed the extent to which the sentence management framework supports accurate calculation of sentences in compliance with applicable legislation and CSC policy. This work included all five regions and national headquarters.
2. IAS, in consultation with COPS and the Policy Sector, selected a sample of complex sentences for recently released high profile offenders, for high profile inmates that were incarcerated, and for offenders who were grieving their sentence calculations. National and regional sentence management resources provided the sentence calculation expertise, and re-performed the sentence calculations to ensure accuracy.

The review assessed:

• Sentence management process map(s) and organizational reporting lines;
• The extent to which the process(es) in place reflected the process(es) defined in CSC policy;
• The extent to which roles and responsibilities for key positions had been established and communicated;
• The extent to which sentence management tools, including a March 2018 update to the Offender Management System (OMS), were in place to support the achievement of objectives;
• The extent to which quality assurance was being performed on files; and
• Sentence calculation compliance with applicable legislation and CSC policy.

Information was gathered through a number of methods, including:

Interviews: discussions with staff and management at the institutional, regional, and national levels took place in order to support the lines of enquiry, and to gain an understanding of their perspective as it related to current sentence management practices.

Review of Documentation: applicable legislation, CSC policy (including desktop procedures, manuals, etc.), process maps, organizational charts, sentence management tools, and evidence of quality assurance work was reviewed for a sample of sentence calculations in order to support the lines of enquiry.

File Review: sentence management calculations were reviewed to assess compliance with applicable legislation and CSC policy. This work entailed re-performing some of the work already done by sentence management staff, and assessed:

• sentence management calculations for high profile incarcerated inmates (expected to be released in 2018/2019) who had an identified sentence complexity (i.e. revocation(s));
• high profile offenders with an identified sentence complexity (i.e. revocation(s)) who had been released from a penitentiary; and
• offenders grieving their sentence calculations.

IAS asked local and regional sentence management staff (independent of the initial person who completed the calculations) to complete these file reviews to assess accuracy. A high level of accuracy in the independent review would provide some assurance that CSC administered sentences in compliance with the law.

Source data obtained from Performance Measurement and Management Reports identified 894 high profile offenders at the time of our information request in March 2018. Using the above criteria, IAS identified 125 offenders as meeting the desired criteria. A total sample size of 49 files were selected for testing across the five regions. IAS selected files after a review of Offender Management System (OMS) data to identify those with specific complexities.

The first objective was to determine whether the process in place to ensure correct sentence calculation was well defined, understood and implemented given recent changes to the structure.

IAS conducted interviews with key positions within SM to obtain an understanding of the environment and its current operations. Interviews were conducted with two Sentence Management Officers (SMO) per region, one new to the position (less than two years), and one experienced (five+ years). IAS selected one Chief per region for an interview as well as the four Regional Managers currently in place for a total of 19 interviewees.

The Offender Programs and Reintegration Branch at NHQ directs the operations of SM. Up until April 2015, it was typical that, within each institution, one or more Chiefs managed Sentence Management Associates and their workload was restricted to the offenders within their institution. As such, in the event of temporary absences or departures, files were re-assigned to other employees onsite. The institutions provided administrative support, including staffing. Under this approach, there were significant inequalities across the institutions regarding workload, due to varying sizes, populations, and security classifications. In an effort to balance workload and create efficiencies, CSC implemented a caseload approach and eliminated and/or reduced SM positions.

The caseload approach divides offender files evenly across the SMOs in the regions. The average caseload at the time of our review was 348 files per SMO. SMOs considered this a high caseload to manage, especially considering the additional duties related to international transfers and Long Term Supervision Orders that were recently added across the board. Recruitment and retention issues have also contributed to high caseloads. Due to the resourcing shortages, new SMOs were immediately assigned a full caseload that included complex files. Feedback from new SMOs relayed their discomfort and challenges with managing a high caseload, and handling complex files, while still developing an understanding of applicable legislation. The learning curve for an SMO is estimated to be approximately 18 to 24 months. A new SMO not supported by a Chief onsite faced added challenges, this being the new reality following the elimination and reduction of positions.

The caseload approach introduced a significantly high movement of files between institutions and has created an unforeseen administrative burden on administrative services who are responsible for file transfers. Under the prior model, the offender's file would remain at the institution, and the employees onsite would provide coverage. The current model requires file reassignment to SMOs who are often not onsite and has resulted in hundreds of file transfers per month, having an adverse impact on file management activities which is likely cost inefficient to the process.

The restructure of SM to a centralized approach directed by NHQ versus the previous regional design sought to standardize the SM process and promote consistency.

For sentence calculations, SMO interviewees reported that NHQ had developed a standard structure for reporting the results of sentence calculations, including subsequent adjustments, into OMS. These standard templates are being used by each Region.

However, SM involves several other responsibilities that are outlined in job descriptions, but they are not documented in a detailed process map. Regions manage their own collections of guidelines that clarify steps within the process. Through interviews, SMOs that are new on the job communicated challenges in knowing how to manage arising issues, such as ambiguous court transcripts and judge's orders, and reported that the environment was "learn as you go". Given the reductions to the number of Chiefs, provision of coaching and mentoring was less available. That said, it is still considered a critical step in developing SMO competencies and judgment. The current workload pressures and organizational design do not allow Chiefs to fully support the SMOs.

At the time of our review, there was no step-by-step process outlining the general tasks a SMO is to perform. SMOs and Chiefs balance a variety of duties within a week, and priorities are typically re-evaluated on a daily basis. Several lists are consulted on a daily, weekly, and monthly basis to track offender movements to keep up-to-date on any developments that could amend the sentence timeline for an offender.

NHQ is executing its responsibility to communicate to the regions any changes to legislation that has an effect on sentence calculations. NHQ distributes information by electronic mail and stores the information communicated in a central repository that is accessible to all staff. SMO and Chief interview responses indicated satisfaction with the information flow from NHQ.

Each region has a Project Officer responsible for delivering training and ensuring data quality for reporting purposes. The materials for training purposes originated from NHQ. Additionally, NHQ is maintaining a tracking document that reports on all SMOs and Chiefs who have completed the five-day induction course as well as the yearly two-day mandatory training. For the most part, interviewees were satisfied with the yearly mandatory training received as well as the five-day induction training. Two issues identified with the delivery of training were that:

• Training materials provided to participants were not aligned with the Instructor Manual; and
• SMOs are not always trained on a timely basis. In some instances, SMOs have worked for months without formal training, while managing a full and complex caseload.

Staff and managers at RHQ confirmed these concerns, and regional managers indicated that there is no time to update the training materials due to the resource shortages. Specifically, in Quebec, interviewees reported that training materials were poorly translated and difficult to understand. Further, while the National Training Standard indicates that SM employees must complete their training within 12 months of appointment, this training is not mandatory prior to assuming a SM position, which increases the risk of errors. Moreover, the turnover of employees is high and has effected the stability required to routinely schedule training courses.

Interviews determined that NHQ is providing good direction over regional activities. Chiefs are aware of their oversight responsibilities and have measures in place to support their requirement to complete file reviews and audits and SM staff have a good understanding of the current process. However, clear concerns are evident regarding the pressures on SM staff due to large caseloads and increased administrative burden.

CSC could improve in relation to the management framework by:

• Reviewing existing file management practices to find efficiencies;
• Enhancing training materials provided to staff for added clarity;
• Making available a referral manual for SMOs and Chiefs; and
• Reporting any calculation errors to NHQ for information and follow-up, to support quality assurance activity and learning, and to identify any systemic issues.

The second objective was to ascertain that sentence calculations were correct for a sample of high profile offenders.

IAS enlisted the support of the SMOs in the regions to review quality assurance activities, such as sentence calculation verifications and audits to assess compliance with CD 703 Sentence Management.

IAS asked SM staff in the regions to review a selection of files to assert that the DPED, FPED, UTA, ERD, SRD and WED dates were calculated correctly. Reviewers were to determine if the verifications, six-month audits, and pre-release audits were completed as required. Table 2 presents the results of the 49 files reviewed.

CD 703 stipulates that the Chief authorizes the admission of an offender to federal custody and assigns the offender's case to a SMO. CD 703 requires that within 20 working days, the SMO performs a set of required duties, such as: structuring the offender sentence, including the determination of conditional release dates, and other eligibilities. Typically, SMOs determine relevant eligibility dates for new admissions using basic tools which include a standard calculator and multi-annual calendar. No evidence of how the calculation was derived is required to be kept on file.

SMOs interviewed had their own independent methods for prioritizing their daily and weekly activities. Typically, SMOs considered addressing new admissions a high priority and reported having no difficulties with meeting the 20 working day timeline. The SMOs performed analysis to assess the timeframe associated with sentence calculations by comparing the calculation dates recorded in OMS to admission dates. Results reflect the required 20 working day timeline.

In accordance with CD 703, the Chief is responsible for verifying sentence structure for an offender. A full verification of the hard-copy file is to be performed within six months of its creation to ensure all required documentation is on file and that the sentence structure is accurate. This policy requirement was implemented subsequent to the restructure. As such, not all files reviewed (30) were subject to the newly established requirement. See Table 3 for details.

The CD 703 requires that throughout an offender's sentence, the SMO is responsible for maintaining an offender's records in OMS to account for the impact that new events such as court appearances or additional charges may have on his/her sentence. A pre-release audit is a control in place to ensure that there are no outstanding charges or any other matters that may require further detention. See Table 4 for details.

At the time of our review, interviews with both NHQ and the regional employees advised that OMS upgrades to SM screens had not been fully implemented. As a result, there were system limitations towards its performance and usability to support Chiefs and Regional Managers. Upgrades were being made to facilitate more fixed field entries from the free text approach that had been in place to increase reporting and accountability.

Our testing results from the review of the hard-copy files determined that SM staff have been diligent in ensuring that verifications and audits are completed on a timely basis. However, review of OMS determined that the system did not reflect the actual state of file monitoring activities. In several cases, OMS did not reflect that a verification or six-month audit had been performed. See Table 5 below for details.

The tested sample produced errors in 2 of 49 high profile offender files reviewed. Additionally, for the tested sample, Chiefs were performing their six-month and pre-release audits as required.

CSC could improve in relation to its administration of sentence calculations by:

• Enhancing system functionality to better support data input, analysis, and monitoring; and
• Promoting the development of automated tools to support sentence calculation.

For the first objective, the results of interviews demonstrated that SM staff are generally complying with the processes in place and the requirements of CD 703 to ensure that offender sentences are well constructed and that ongoing reviews and audits are performed to effectively manage an offender's sentence up to their release. However, the following areas require further consideration by management to ensure the effective administration of sentence management activities at CSC:

• Reviewing existing file management practices to find efficiencies;
• Enhancing training materials provided to staff for added clarity;
• Making available a referral manual for SMOs and Chiefs; and
• Reporting any calculation errors to NHQ for information and follow-up, to support quality assurance activity and learning, and to identify any systemic issues.

With respect to the second objective, interviews, analysis and testing found that SM staff are completing required verifications and audits to support accurate sentence management administration. However, the following areas require further consideration to reduce the occurrence of errors:

• Enhancing system functionality to better support data input, analysis, and monitoring; and
• Promoting the development of automated tools to support sentence calculation.