Actuarial Tools and Professional Judgement in Offender Security Level Decisions: 2018-2019 to 2022-2023

Research highlights: Concordance rates between actuarial tools and professional judgement demonstrate consistency over the five year period.

Background

Within the Correctional Service of Canada, Offender Security Level (OSL) decisions are determined using a combination of actuarial tools and professional judgment. The actuarial tool employed upon offender admission or readmission is the Custody Rating Scale (CRS). The actuarial tools used to assist in security reclassification decisions are the Security Reclassification Scale (SRS) and Security Reclassification Scale for Women (SRS-W). The CRS and SRS/SRS-W are used in conjunction with the structured professional assessment of an offender’s institutional adjustment, escape risk, and public safety risk to determine an appropriate OSL.

What we did

All OSL decisions in Fiscal Years 2018-2019 to 2022-2023 were examined in relation to the corresponding CRS, SRS or SRS-W designation. Analysis was undertaken to explore the prevalence of concordant decisions (i.e., the OSL decision was consistent with results of the actuarial tool), OSL increase decisions (i.e., the OSL reflected a security increase from the actuarial tool), and OSL decrease decisions (i.e., the OSL reflected a security decrease from the actuarial tool).

What we found

Across the five year period, there was relative consistency in the concordance rates across the tools. Overall, the concordance rate between OSL and initial applications of the CRS was 73% (range: 72% to 73%). The concordance rate was slighter lower for subsequent CRS applications, i.e., 68% (range: 67% to 69%). OSL decreases were consistently more common than security increases.

The concordance rate was higher for the SRS, i.e., 79% (range: 78% to 80%), with greater similarity in the relative percentages of OSL increases (10%) and decreases (11%).  In the case of the SRS-W, the concordance rate was 69% (range: 67% to 74%). Over the five year period, the relative percentage of OSL increases declined somewhat.

Table 1. Initial CRS-OSL Relationship, 2018-2019 to 2022-2023.
FY Concordant Increase Decrease
2018-2019 73% 11% 17%
2019-2020 73% 11% 17%
2020-2021 73% 12% 15%
2021-2022 73% 11% 16%
2022-2023 72% 11% 18%
Total 73% 11% 17%
Table 2. Subsequent CRS-OSL Relationship, 2018-2019 to 2022-2023.
FY Concordant Increase Decrease
2018-2019 69% 15% 17%
2019-2020 68% 13% 19%
2020-2021 67% 14% 20%
2021-2022 69% 12% 19%
2022-2023 68% 14% 19%
Total 68% 13% 19%
Table 3. SRS-OSL Relationship, 2018-2019 to 2022-2023.
FY Concordant Increase Decrease
2018-2019 78% 13% 9%
2019-2020 79% 12% 10%
2020-2021 79% 10% 12%
2021-2022 80% 8% 12%
2022-2023 78% 10% 13%
Total 79% 10% 11%
Table 4. SRS-W-OSL Relationship, 2018-2019 to 2022-2023.
FY Concordant Increase Decrease
2018-2019 67% 18% 15%
2019-2020 68% 20% 13%
2020-2021 69% 16% 15%
2021-2022 74% 14% 12%
2022-2023 69% 13% 18%
Total 69% 16% 14%

Note: Percentages may not sum to 100 due to rounding.

What it means

In the past five years, there has been general consistency in the overall concordance rate, though the percentage of OSL increases in relation to security reclassifications has slightly decreased.

For more information

For questions and/or more information, please email Research Branch. You can visit the Research Publications section for a full list of reports and one-page summaries.

Prepared by: Laura McKendy and Kevin Lau

Page details

Date modified: