FORUM on Corrections Research

Initiating safe reintegration: A decade of Custodial Rating Scale results

Fred Luciani¹
Research Branch, Correctional Service of Canada

We are accustomed to thinking of objective or actuarial measures of risk assessment as an assessment aid to classification officers confronted with making specific decisions about individual offenders. While this is their prime function, objective assessment also serves a number of other objectives that may not often come to the attention of field staff. These involve strategic issues that address organizational or macro level objectives and include standardizing criteria and decisions rules for risk assessment, establishing authority and accountability frameworks, ensuring consistency and fairness, and creating an information base to validate and refine tools. Objective measures can also serve to shape, promote and evaluate the values and objectives of the correctional agency. This article explores the last of these objectives. Specifically, it examines the Custody Rating Scale and how it’s introduction has influenced both the policy and practice of initial penitentiary placement and contributes to safe reintegration.

Background

The Custody Rating Scale (CRS) is a twelve item instrument that assesses institutional adjustment and public safety risk and yields an initial security classification rating.² The CRS was approved for system-wide application by the Correctional Service of Canada in 1991, and is applied to both newly sentenced and return to custody admissions. Validation results³ based on the first five years of operation demonstrated that the CRS is effective in assigning security classification ratings that are consistent with actual incidences of institutional offence, escape, release type and community release adjustment. Also, offenders rated minimum-security by the CRS recorded significantly lower rates of new offence and escape from minimum-security facilities than medium-rated offenders who were also initially placed to minimum-security facilities. These findings support the position that the CRS is useful in identifying minimum-security candidates at time of admission.

Modern correctional legislation embraces the principle of “least restrictive” measure of confinement for reasons of fairness, practicality and economy. In addition, there are a number of advantages that enhance the discretionary release potential of offenders placed to lower security level facilities. Offenders reviewed for early release or parole from medium-, and in particular minimum-security level facilities, enjoy greater opportunities to establish release potential through wider access to community programmes, temporary absences, institutional employment and recreational activities. Over and above these opportunities, minimum-security status alone seems to improve discretionary release potential.

Research⁴ has shown that offenders rated minimum-security by the CRS and initially placed at a minimum-security facility enjoyed much higher discretionary release rates and were released much earlier in their sentence than minimum-rated offender placed to a medium security facility. Both minimum-rated groups shared higher reintegration potential ratings as measured by the Statistical Information on Recidivism (SIR) scale.⁵ As well, medium-rated offenders placed to a minimum-security facility enjoyed much higher and earlier release rates than medium-rated, medium-placed offenders even though both medium-rated groups shared lower reintegration potential ratings. In fact, the medium-rated, minimum-placed group was more likely to be released and released earlier than even a minimum-rated, medium-placed group.

Notwithstanding, analyses of initial placements found 43% of offenders with minimum-security ratings were overridden to a higher security level.⁶ It was estimated that improving concordance with the CRS could result in an increase in releases and reduce incarceration days. Operational reviews demonstrate the utility of the CRS in identifying minimum-security candidates and how initial placement can influence both the likelihood of discretionary release and the length of incarceration prior to release.

While there is little argument with the concept of assigning offenders to the lowest level of security, over-classification remains a hindrance to safe reintegration. The introduction of offender classification scales⁷ at both the initial and reclassification stages exposed the often conservative nature of classification practice.⁸ The tendency to error on the side of caution, to over-estimate security risk and the tendency to place (and thus classify) offenders to institutions that meet protection, medical or programme needs, as opposed to security profile, are all potential contributors to over classification practice.

Strategies for increasing placement to lower security

Objective measures of security classification represent the established norms for classification and offer a convenient method of communicating corporate policy and direction. They are based on fixed, weighted criteria and decision rules, amendable within discrete, identifiable increments that allows their effects to be tracked or calculated. In an effort to influence initial classification practice an analysis⁹ of the potential effects on institutional incident and escapes rates from minimum-security was undertaken utilizing CRS data gathered for validation purposes. The object of the analysis was to determine what effect moving the minimum-security cut-off points had on incident and escape rates and whether revising the cut-offs values could increase initial placement to minimum-facility.

CRS data was available on 6,745 offenders admitted to federal institutions between 1991 and 1995. Offender institutional incident and escape information was gathered for a period of up to four years following the application of the CRS. Table 1 presents a summary of the effects on incident and escape rates at the original cut-off values and those associated with 30%, 35% and 40% cut-off options.

Table 1

The Relationship Between Theory and Offender Assessment
	Original	Cut-Off Options
Outcomes	26%	30%	35%	40%
Number of Incidents	288	329	420	497
Incident Rate	15.6	16.4	17.9	18.5
Incident Rate of New Candidates	-	24.6	25.9	25.9
Number Escapes	78	86	102	115
Escape Rate	4.2	4.3	4.3	4.3
Escape Rate of New Candidates	-	4.8	4.7	4.6
CRS Candidate Pool	1,844	2,011	2,353	2,651
Distribution Gain	-	167	509	807

The original cut-off values of the CRS assigned 26% of admissions to the minimum-security category and produced a minimum-security candidate pool of 1,844 offenders. The incident rate for offenders rated minimum-security was 15.6%, and the escape rate was 4.2%. By way of comparison, adjusting the cut-off values to produce a 35% distribution to the minimum-security category expands the candidate pool by approximately 509 offenders and resulted in a very slight increase in the overall escape rate to 4.3%. The escape rate for the new candidate group associated with all three cut-off options is essentially indistinguishable from the rate of the original candidates. The incident rate for the new candidate group resulting from all options is approximately 10% higher than that of the original candidates.

The differences in incident rates are more evident and suggest there may be distinctions between original and new candidates, however, any negative influence of the latter would be dispersed across minimum-security facilities.

The results suggest that expanding the candidate pool by moving the minimum-security cut-off values of the CRS would have little or no effect on the escape rate and cause only a marginal increase in the institutional incident rates. As a result the cut-off values were adjusted in 1998 to allow for a 35% distribution to minimum-security. This strategy, however, would have marginal effects if the concordance rate with the CRS remained unchanged. Accordingly, new standard operating procedures were developed, additional attempts were made to improve staff understanding of the CRS guideline, and encourage higher concordance. An indication of the influence of the CRS and the effects of these steps are illustrated in Table 2.

Table 2

CRS/Placement Concordance and Escape from Minimum Security
Concordance (%)	1992	1993	1994	1995	1996	1997	1998	1999
Concordance (%):
– Overall	63.0	59.0	56.6	57.1	57.4	71.5	79.4	76.7
– W/Min. Rated	75.7	73.1	73.5	73.5	75.2	75.6	80.4	80.7
Distribution to Minimum (%)	12.0	24.4	27.3	27.0	25.1	26.6	32.7	37.5
Min. Sec. Escape Rate (%)	131	102	104	53	28	24	30	45

Initial direct placement to minimum-security is an important source for minimum-security offenders contributing about thirty-five percent of the stock population. As noted in the table the overall concordance between CRS ratings and final initial classification decisions improve from the low seventy percent to over eighty percent range. The improvement in overall concordance is attributed to increase in the concordance rate of minimum-rated offenders that rose from the high fifty percent to over seventy-five percent range in the recent years. The effect of improving concordance and the revisions to the minimum-security cut-off values can be seen in recent years where distribution to minimum security exceeds thirty percent of the admission population. Finally, the escape rate from minimum-security facilities (expressed as a percent of the annual, average minimum-security population) drops precipitously in the middle years under review and increases slightly in recent years. Closer examination of escapees suggests that the recent rise in the escape rate is largely attributed to offenders transferred from higher security facilities rather than those initially placed at admission. The escape rate, although slightly higher in recent years, remains below the ten-year average (6.8%) while during the same period the minimum-security population doubled. (Incident data is not available for later years and made it impossible to conduct similar analyses).

Conclusion

Objective measures of risk assessment provide a benchmark or standard that is useful in evaluating decisions and the decision-making process. They offer strategic insight to correctional practice and represent a convenient vehicle to both develop and implement policy.

Operational reviews over the last ten years demonstrate that the CRS serves as an effective aid to initial security placement and provide insight to the potential of offender classification to promote broader organizational goals.

In our analyses of the first five years of operation it was demonstrated that the CRS ratings were more accurate than subjective decisions in identifying offenders who would engage in incidents or escape from minimum security. Recent observations suggest that the gap is closing. In part, we would like to believe because of the growing confidence in the CRS (as evidenced by concordance gains) and its contribution to improving the classification assessment process. In addition, the initial placement trends over the last ten years suggest that the conservative nature of classification practice is gradually abating, without serious increase to the escape rate, and perhaps due in part to the support derived from corporately established classification norms.

Where an offender is initially placed effects discretionary release potential. Over classification of high reintegration potential offenders can erode and delay release while placing low reintegration potential offenders at lower security levels, in the long run, does not always serve the goal of safe reintegration. Measures that improve selection and initial placement of appropriate offenders to lower security will contribute to both maximizing reintegration and help ensure it is safe.

---

1. Regional Headquarters, P.O. Box 1174, Kingston, Ontario K7L 4Y8.

2. Solicitor General of Canada. (1987). Development of a Security Classification Model for Canadian Federal Offenders: A Report to the Offender Management Division. Ottawa, ON: Correctional Service of Canada.

3. Luciani, F. P., Motiuk, L. L., & Nafekh, M. (1996). An Operational Review of the Custody Rating Scale: Reliability, Validity and Practical Utility. Ottawa, ON: Correctional Service of Canada.

4. Luciani, F. P. (1998). Exploring reintegration potential Impacts of initial placement practice. Forum on Corrections Research, 10(1), 23-27.

5. Nuffield, J. (1982). Parole Decision-Making in Canada: Research Towards Decision Guidelines. Ottawa, ON: Solicitor General of Canada.

6. Luciani, Motiuk, & Nafekh (1996).

7. Buchanan, R. A., Whitlow, K. L., & Austin, J. (1986). National evaluation of objective prison classification systems: The current state of the art.
Crime and Delinquency, 32(3), 272-290.

8. Austin, J. (1983). Assessing the new generation of prison classification models. Crime and Delinquency. 29(1), 561-577. See Solicitor General of Canada (1987), and Luciani, Motiuk, & Nafekh (1996). And see Bonta, J., & Motiuk, L. L. (1990). Classification to correctional halfway houses: A quasi-experimental evaluation. Criminology, 28, 497-506. See also Van Voorhis, P. (1988). A cross classification of five offender typologies: Issues of construct and predictive validity. Criminal Justice and Behavior, 15(1), 109-124.

9. Correctional Services of Canada (1998). Effects of Adjusting the CRS Minimum Security Cut-Off Values: An Illustration of the Impact on Escape and Incident Rates. Ottawa, ON: Internal document.