Skip to content | Skip to institutional links

Common menu bar links

Institutional links

Research Reports

Warning This Web page has been archived on the Web.

Security Classification Using the Custody Rating Scale

Brian A. Grant and Fred Luciani

Research Branch

Correctional Service of Canada

February, 1998

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

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

Correctional Service Canada
National Parole Board
Correctional Investigator
Justice
Department of the Solicitor General

Research/Evaluation Reports:

Information about Offenders
Security Classification of Inmates
Judicial Determination
The Temporary Absence Program: A Descriptive Analysis
Personal Development Temporary Absences
Work Release Program: How it is used and for what purposes
Day Parole: effects of the CCRA (1992)
Case Management: Preparation for Release and Day Parole Outcome
Accelerated Parole Review
Statutory Release and Detention Provisions
Community Supervision Provisions
Provisions Relating to Victims
Observers at National Parole Board Hearings
The National Parole Board Registry of Decisions
CSC Human Resources
Administrative Segregation
Search, Seizure and Inmate Discipline
Offender Grievance System
Urinalysis Testing Program
Inmate's Input in Decision-making
Information to Offenders
Aboriginal Offenders
Health Services
Women Offenders

Executive Summary

The Corrections and Conditional Release Act (CCRA) requires the Correctional Service of Canada assign a security classification of minimum, medium or maximum to all offenders. Security classification of offenders takes place at admission and periodically during incarceration according to established guidelines that assess three risk dimensions: institutional adjustment, escape and public safety risk.

The CCRA also requires a review of the legislation be undertaken five years after it was enacted. This is one of a series of reports addressing changes mandated by the CCRA.

The CRS was developed on a sample of incarcerated offenders between 1987 to 1989, approved for national implementation following field testing in 1991 and reaffirmed in 1996 after an extensive, national validation study. The validation report examined the validity of the CRS against a variety of operational indices, the degree of concordance between the security classification ratings derived from the Scale (minimum, medium or maximum) and actual initial placement decisions, the effects of overriding the Scale, and the applicability of the Scale to aboriginal and female offenders. Results indicate that the classification rating determined by the CRS is associated with institutional misbehaviour, escape, discretionary release potential and suspension while on conditional release. Offenders rated maximum security, for example, were more likely to be charged with security incidents while in the institution, less likely to be granted a discretionary release and more likely to be suspended on conditional release than offenders rated medium security; while offenders rated minimum security recorded the lowest incident and suspension rates and highest discretionary release rates.

Recently the CRS was recognized as the sole tool for use in assessing security classification and in determining the initial penitentiary placement of offenders admitted to federal custody.

A snapshot of the stock population of incarcerated offenders taken in August 1997 indicates that 67% of offenders had a CRS on file which is an increase from 48% in March 1995. The CRS completion rate for new offenders admitted in 1997 was 69%. A completion rate of 100% is not anticipated because offenders admitted prior to the implementation of the CRS, and some offenders returning on revocation or temporary detention would not have a CRS completed. While completion rates are increasing, 100% completion rates are not achievable until the entire population turns-over.

Approximately three-quarters of inmates are placed in an institution with a security level consistent with their CRS rating. The most common form of override is to place an offender in a higher level of security than that indicated by the CRS. In most regions concordance have increased as a result of a reduction in the use of overrides.

There are variations in the concordance rates and the types of overrides across regions. These differences may require additional investigation to ensure that overrides are used in situations where they are effective in promoting the correctional goals of reintegration and community safety. CRS results for Aboriginal offenders are consistent with those found for the full inmate population and overrides are actually lower for Aboriginal offenders. The results for female offenders indicate that the CRS rates a higher percentage of female offenders as minimum security risk than male offenders.

Overall, the use of the CRS has increased, the concordance rates have improved in most regions and the instrument has been shown to offer a valid method of assigning a security level to offenders.

EXECUTIVE SUMMARY
LIST OF TABLES
LIST OF FIGURES
CHAPTER 1: INTRODUCTION
RESULTS AND METHODOLOGY
SUMMARY AND DISCUSSION
REFERENCES
APPENDIX

LIST OF TABLES

Table 1: Rates of institutional and release performance
Table 2: National and Region CRS Stock Population Completion Rates
Table 3: National and Region CRS 1996 Completion Rates
Table 4: National and Region Concordance and Override Rates
Table 5: Security level distribution for Aboriginal offenders by CRS designation and actual penitentiary placement
Table 6: Security level distribution for female offenders by CRS designation and actual penitentiary placement

LIST OF FIGURES

Figure 1: Annual number of CRS's Completed 1991 to 1996 by Region

Chapter 1: Introduction

Accurate security classification of offenders and placement of offenders at admission to the appropriate level of security within the penitentiary system is a foundation of effective correctional management. Security classification has been part of the penitentiary regulations since 1889. The Corrections and Conditional Release Act (1992) recognized the importance of security classification and mandated the Correctional Service of Canada to "assign a security classification of maximum, medium or minimum to each inmate in accordance with the Regulations of the Act (Sec. 30). Corrections and Conditional Release Act (CCRA), legislated guidelines for offender security classification and ensures that the privileges associated with security level are based on an assessment of individual risk and needs. As a result of the CCRA, a system was implemented to assign a security level to all offenders as required by the law.

The core principles for both initial and reclassification require that offenders be rated on three risk dimensions: institutional adjustment, escape and public safety risk (Case Management Manual, 1997). Case management officers review the offender's criminal history and institutional behaviour using between three and five criteria for each dimension. The security level for the offender is then determined by assigning a risk rating (low, moderate, high) on each of these dimension and applying the decision-making rules established by the CCRA regulations. The security classification guidelines established for each dimension and security level decision assisted case management officers, however, a more objective and empirically valid system of security classification was needed.

In 1991, prior to the CCRA, the Custody Rating Scale (CRS) was introduced to provide objective criteria and standard protocols for initial penitentiary placement for offenders. The CRS provides numerical values which are used to determine the level of security for each offender. Development on the CRS continued after its implementation and in 1995 the scale was extensively evaluated (Luciani, Motiuk & Nafekh, 1996). Use of the CRS was reaffirmed by the CSC in February, 1996. The validation report provided evidence as to the effectiveness of the CRS in classifying offenders as either minimum, medium or maximum security.

CCRA Review

The CCRA requires Parliament to undertake a review of the provisions of the Act five years after it came into effect. This report is part of that review process and is one of a number of reports prepared for the review.

The intent of the security classification provisions of the CCRA is to ensure that a security classification assessment is prepared for all inmates based on objective criteria and standardized decision rules and to limit the potential inconsistencies related to the application of the security classification principles stated in the regulations and the CCM. In addition, the Act included various restrictions on conditional release for offenders in maximum and medium security institutions. For example, unescorted temporary absences are not permitted for maximum security inmates and the amount of time medium security offenders may be released on unescorted temporary absences is less than for minimum security offenders.

The purpose of this study is to address four main issues:

Is the security classification provided by the Custody Rating Scale associated with offenders' institutional and release behaviour?
What percentage of inmates have a Custody Rating Scale completed and available for reference in the Offender Management System (OMS), and is the percentage increasing?
How consistently does the inmate security level match the institutional security level?
Are there differences in inmate security classification levels across regions, ethnic groups and gender?

The Custody Rating Scale (CRS)

The Custody Rating Scale (see Appendix 1) consists of two independently scored sub-scales: a five-item Institutional Adjustment scale and a seven-item Security Risk scale. In most cases, item scores increase according to the frequency of incidents and, as scores escalate on either sub-scale, higher security classification is predicted. Security classification is based on the total sub-scale scores, in accordance with predetermined decision rules that specify cut-off values for minimum and maximum security. In the event of disagreement between the sub-scales, the Custody Rating Scale security level designation is determined by the sub-scale assigning the higher classification rating.

Potential scores range from 0 to 186 points on the Institutional Adjustment scale, and from 17 to 190 points (and higher when open-ended scores for release failures are considered) on the Security Risk scale. Item weight and classification cut-off values are, for the most part, empirically derived from a 1987 sample of federal offenders. In some cases, item weights are defined by policy priorities. For example, Offence Severity is weighted to prevent the initial placement of an offender serving a life sentence in minimum-security.

Results and Methodology

Methodology

This report presents results from three main sources. First, are results from a review of Custody Rating Scale (CRS) conducted in 1995 (Luciani, Motiuk and Nafekh, 1996). These results rely on data from a follow-up of offenders who had the CRS completed. The next section of the report presents comparisons between offenders with a CRS completed in 1995 and 1997. These results make it possible to determine if there has been change in CRS use. The third set of results is based on data for all offenders who had a CRS completed by 1995.

Results

CRS Relationship to Institutional and Release Performance

The report prepared by Luciani, Motiuk and Nafekh (1996) presented data on the relationship between the security rating on the CRS and performance of the offender while in custody, release type, and on conditional release. The authors concluded:

In all tests the scale [CRS] performed as expected. Offenders rated by the CRS as lower security were better adjusted (as indicated by lower frequencies of incidents) and lower risk (as indicated by higher frequencies of discretionary release and post-release success), than offenders rated as higher risk. (p.30)

Some of the more salient results of these analyses are summarized in Table 1. While 51% of the offenders rated by the CRS as maximum security had an institutional incident (ranging from murder and assault to disobeying a correctional officer) the rate declined to 35% for medium rated offenders and 16% for minimum rated offenders. In terms of violent incidents, the pattern was the same, with 14% of maximum rated offenders having a violent incident while only 3% of minimum rated offenders had similar types of incidents.

Table 1: Rates of institutional and release performance
Measure	Security rating from the Custody Rating Scale (CRS)
Measure	Maximum	Medium	Minimum
All institutional incidents	51%	35%	16%
Violent institutional incidents	14%	8%	3%
Discretionary release rate (day or full parole)	34%	65%	80%
Suspension of conditional release	84%	61%	41%

Most offenders are released either by a decision of the National Parole Board or at their statutory release date, normally after serving two-thirds of their sentence. A release by the National Parole Board is termed discretionary because the Board decides on factors such as the offender's history and behaviour in prison, whether or not he/she can be safely released into the community. However, at the statutory release date, the release is automatic; it does not require an NPB review except in the case of the most serious offenders who may be detained. The accuracy of the CRS in predicting discretionary release potential is demonstrated by the fact that while only one-third of the maximum rated offenders received a discretionary release, over four-fifths of the minimum rated offenders received a discretionary release by the NPB and were released prior to their statutory release date.

If an offender presents an undue safety risk to the community, the conditional release (day parole, full parole statutory release) can be suspended and/or revoked and the offender is returned to custody. Almost all of the offenders rated as maximum security (84%) by the CRS had their conditional release suspended while only 61% of the medium rated offenders and 41% of the minimum rated offenders had been suspended during the period under supervision on conditional release.

Another way of demonstrating the validity of the CRS is to see how it correlates with other measures of offender risk currently in use by the Correctional Service. In general, the CRS correlates about .50 with other measures such as the Statistical Information on Recidivism Scale, and assessments of offenders' criminal history risk, criminogenic needs (factors associated with the offenders criminal behaviour) and the combination of both criminal history risk and criminogenic needs.

Taken together, these results indicate that risk ratings on CRS are associated with institutional behaviour, conditional release decisions and performance on conditional release. Further details on the reliability and validity of the CRS can be found in the Luciani, Motiuk and Nafekh (1996) report.

Having demonstrated that the CRS is a good indicator of offender performance, the next issue is the use of the scale by the Service. The following sections provide information on the number of offenders who had the CRS completed in 1995 and in 1997 and shows the relationship between the CRS rating and the actual placement of offenders in different levels of security.

CRS Completion Rates Based on the Inmate Population - March 1995

The Offender Management System (OMS) files of all incarcerated offenders in March of 1995 and again in August 1997 were reviewed to determine whether the CRS was completed upon admission to federal custody.

The population consisted of all active cases including those on temporary absence, but excluding temporary detainees, offenders who were unlawfully at large, those in provincial custody and offenders awaiting assessment in the regional reception units. CRS completion rates were computed for each region and these are presented in Table 2.

Table 2: National and Region CRS Stock Population Completion Rates
Region	August 1997			March 1995 Completion Rate
Region	Inmate Population*	Files with a CRS	Completion Rate	March 1995 Completion Rate
Atlantic	1,240	695	56%	3%
Quebec	3,437	2,990**	87%**	87%
Ontario	3,151	2,456	78%	64%
Prairie	3,047	1,426	47%	21%
Pacific	1,615	835	52%	18%
National	9,053	5,412	67%	48%

* Based on EIS data of August 1997.

** These are estimates based on the 1995 data. Data for 1997 were not accurate as a result of the way CRS data were entered into OMS in 1997. The data we were able to obtain indicated a huge and unrealistic drop in number of cases of completed CRSs. Given the recent evidence of data reporting problems it was decided to use the 1995 results to estimate the 1997 results for Quebec.

The results indicate that in all regions, for which current data were available, there was a sharp increase in the rate at which the CRS was completed from 1995 to 1997. Overall, 67% of the inmate population had the CRS completed as of August 1997, while regional completion rates ranged from 78% in the Ontario region to 47% in the Prairie region. Regions which had the lowest completion rates in 1995 showed the largest increases with all regions having almost 50%, or more of their population with a completed CRS. It should be noted, that it will be some time before all offenders have a CRS completed because the scale is completed on admissions, and only completed for new admissions or offenders who had been released and then readmitted. Offenders admitted prior to the implementation of the scale would not be expected to have a completed CRS on their OMS file and it is unlikely that a 100% completion rate could be achieved until a complete population turn-over takes place.

Results presented for Quebec Region in 1997 are estimates based on the 1995 rate. The observed rate in 1997 was only 37%, and not deemed an accurate reflection of initial classification practice.

CRS Completion Rates Based on Annual Admissions - 1996.

Following the decision to reinforce the mandate to use the CRS in 1996, all regions agreed to pursue its full implementation. CRS completion rates, based on the number of admissions in each region for the year 1996, are presented in Table 3. The data provide an indication of the recent efforts at achieving this objective. By the end of 1996, 69% of all admissions for the year were administered the CRS. Regionally, the completion rates ranged from 44% to 87% with the larger regions having higher completion rates.

Table 3: National and Region CRS 1996 Completion Rates
Region	Admissions 1996*	Files with a CRS	Admission Completion Rate
Atlantic	933	509	55%
Quebec	2,284	1,987**	87**%
Ontario	1,845	1,311	71%
Prairie	2,014	1,289	64%
Pacific	877	388	44%
National	7,953	5,484	69%

* Admissions data drawn from Adult Correctional Service in Canada, 1995-95, Statistics Canada. All admission types included.

** These are estimates from 1995 data.

Annual CRS Completion Rates 1991 - 1996.

The use of the CRS has increased steadily over the years as indicated by the data in Figure 1. The figure shows the number of offenders each year for whom a CRS was completed. The 1996 results presented for Quebec are not considered accurate.

Figure 1: Annual number of CRS's Completed 1991 to 1996 by Region

CRS - Initial Placement Concordance Rates.

Given that almost 70% of the offenders admitted to federal penitentiaries in 1996 had a CRS completed, it is instructive to look at the percentage who were actually placed in an institution with the same level of security that the CRS indicated was appropriate. The degree of agreement between CRS rating and actual placement is referred to as the concordance rate. It is expected that case management officers will, on occasion, override the security rating of the offender and place offenders to higher or lower levels of security than that indicated by the CRS. The concordance rates for offenders in the August 1997 sample as compared to the March 1995 sample are presented in Table 4.

Table 4: National and Region Concordance and Override Rates
Region	Concordance Rates		Overrides to Higher Security		Overrides to Lower Security
Region	1997	1995	1997	1995	1997	1995
Atlantic	75%	66%	14%	20%	11%	14%
Quebec	75%	77%	16%	11%	9%	12%
Ontario	72%	71%	14%	22%	14%	7%
Prairies	76%	74%	14%	17%	10%	9%
Pacific	77%	70%	15%	21%	8%	9%
National	74%	74%	14%	16%	12%	10%

Overall, 74% of the offenders classified using the CRS were placed in an institution with a security level consistent with the CRS rating in both 1995 and 1997. In general, most regions showed an increase in the concordance rate between CRS rating and initial institutional placement from 1995 to 1997 suggesting an increasing reliance on the scale in making security placement decisions.

Case management officers may not agree with the security rating indicated by the CRS. The CRS does not address all security classification issues and there are a number of legitimate reasons such as the need for protection, programs, medical treatment, etc., to override the scale rating. In these cases, they rely on their professional judgment and may place the offender in an institution with a higher security or lower level of security than that recommended by the scale. When this occurs, the case management officer must provide a rationale for the change, or override, and this usually takes the form of statements indicating that in the judgment of case management officer the offender's current behaviour, motivation, etc. should be given more weight in the placement decision than they are given in the CRS. It is expected that some percentage of CRS ratings will be overridden by case management, but there has not been sufficient experience with the CRS to determine what that rate should be.

The results presented in Table 4 provide percentages of overrides for both increases and decreases in security relative to the CRS rating. Overrides to higher security have the potential to reduce the likelihood of release for the offender since it is more difficult to obtain a discretionary release from a higher security institution, all other things being equal (Luciani, Motiuk & Nafekh, 1996). Overall, the trend has been to reduce the percentage of cases overridden to higher security with a decline from 16% in 1995 to 14% in 1997.

Overrides to lower security increase the likelihood of a discretionary release. The overall rate of overrides to lower security shown in Table 4 has increased from 10% to 12%. However, most of this increase is the result of a rather large change in the Ontario region where in 1995 only 7% of cases were overridden to lower security, but in 1997 this increased to 14%. Other regions showed either a decline in overrides to lower security or no change.

Aboriginal Offenders

Overall, for Aboriginal offenders who have had the CRS completed (between 1991 and 1997), the concordance rate is 78%, that is, over three-quarters of offenders are placed in an institution with the same security level as indicated by the CRS. Overrides to higher security account for 12% of the remaining cases while overrides to lower security account for 10% of cases. Relative to all cases with a CRS completed, Aboriginal offenders have a higher concordance rate (78% vs. 74%), are less likely to be overridden to higher security (12% vs. 16%), and are about equally likely to be overridden to lower security (10% vs. 11%).

In terms of the distribution of offenders within the various security levels, it is interesting to note that the actual placement of Aboriginal offenders matches the percentage of Aboriginal offenders in each risk category of the CRS. From the results in Table 5, it may be concluded that the distribution of offenders across CRS risk category is very similar to the actual initial placement distribution.

Table 5: Security level distribution for Aboriginal offenders by CRS designation and actual penitentiary placement
	Maximum	Medium	Minimum
CRS designation	8.2%	76.0%	15.8%
Actual placement	8.0%	78.7%	13.3%

Note: percentages are based on 1,463 Aboriginal offenders

Female Offenders

The number of female offenders who had the CRS completed was extremely low. Given that only 57 female offenders had a CRS completed, the results presented need to be considered with some caution. The CRS security ratings are presented in Table 6. The security level of penitentiary placements for female offender cannot be used for comparison to the CRS. Most female offenders were placed at the multi-level Prison for Women which is only allowed for a maximum security placement designation in our information systems. More detailed analyses would be needed to determine the actual security level placement of female offenders within this prison. Without proper placement data, it is not possible to calculate concordance rates.

Table 6: Security level distribution for female offenders by CRS designation and actual penitentiary placement
	Maximum	Medium	Minimum
CRS designation	7.0%	54.4%	38.6%

Note: percentages are based on 57 female offenders

The CRS rates almost 40% of the female sample as minimum security and only 7% of female offenders as maximum security. Overall, the CRS rates many more females as minimum security than males. The opening of new facilities for women in the past year has increased the placement options for women offenders and these results will need to be revisited once data from the new institutions for women offenders are available.

Summary and Discussion

The results presented in the report provide support for the use of Custody Rating Scale as a means of classifying offenders by security level. The data showed that there was a clear relationship between institutional and release performance and the rating received from the CRS. Comparisons between 1995 and 1997 showed increased use of the CRS and a trend towards fewer overrides, although this was not consistent for all regions.

In the introduction, four questions were proposed to guide the study and the answers to these are presented below.

1. Is the security classification provided by the Custody Rating Scale associated with offenders' institutional and release behaviour?

Results presented indicate that the security rating determined by the CRS is associated with institutional behaviour, discretionary release and performance on conditional release. Maximum security rated offenders are the most likely to have security incidents while in the institution, are less likely to be granted a discretionary release and are more likely to have adjustment problems while on conditional release. Medium and minimum rated offenders performed markedly better on all indices of adjustment and risk.

2. What percentage of inmates have a Custody Rating Scale completed and available for reference in the Offender Management System (OMS) and is the percentage increasing?

In August 1997, 67% of offenders had a CRS on file and this was an increase from 48% in 1995. Not all offenders will have a CRS on file until the entire population turns over, which will not occur for some time because of long determinate sentences and indeterminate sentences<

3. How consistently does the inmate security level match the institutional security level?

Approximately three-quarters of inmates are placed in an institution with a security level consistent with their CRS rating. In most regions this has been increasing with reduced use of overrides. The most common form of override is to place an offender in a higher level of security than that indicated by the CRS.

4. Are there differences in inmate security classification levels across regions, ethnic groups and gender?

There are variations in the concordance rates and the types of overrides across regions. These differences may require additional investigation to ensure that overrides are used in situations where they are effective in assisting the correctional goals of reintegration and community safety. CRS results for Aboriginal offenders are consistent with those found for the full inmate population and overrides are actually lower for Aboriginal offenders. The results for female offenders indicate that the CRS rates a higher percentage of female offenders as minimum security risk than male offenders.

Future Directions

The CRS provides an objective and valid method of assessing the security classification needs of federal offenders. The available evidence indicates a strong relationship between CRS ratings and a number of indices of adjustment and risk and that it is can play a useful role in the management of offenders and in meeting the objectives set out in the CCRA. The CRS has specific implications for promoting the placement of offenders to the least restrictive form of incarceration while ensuring consistent, accurate placement of offenders that does not compromise the safety of the public.

The CRS provides information for the initial placement decision. Work is continuing on the development of a classification tool which will provide for systematic re-evaluation of the security level requirements of offenders. This tool will take account of the offenders' participation in treatment and work programs, general institutional behaviour and other factors that are related to positive correctional results.

References

Correctional Service of Canada, Case management manual (February 1997 edition)

Luciani, F. P., Motiuk, L. L., & Nafekh, M. (1996). An operational review of the custody rating scale: Reliability, validity and practical utility (Report R-47). Research Division, Correctional Service of Canada, Ottawa, Canada.

Appendix 1: Custody Rating Scale - Instructions

MINIMUM-SECURITY

Inmates with scores less than 79.5 on the INSTITUTIONAL ADJUSTMENT dimension, and/or less than 58.5 on the SECURITY RISK dimension are recommended as minimum-security candidates.

MEDIUM-SECURITY

Inmates with scores between 79.5 and 94.5 on the INSTITUTIONAL ADJUSTMENT dimensions, and/or between 58.5 and 133.5 on the SECURITY RISK dimensions are recommended as medium-security candidates.

MAXIMUM-SECURITY

Inmates with scores greater than 94.5 on the INSTITUTIONAL ADJUSTMENT dimension, and/or greater than 133.5 on the SECURITY RISK dimension are recommended as maximum-security candidates.

CUSTODY RATING SCALE

FPS:
COMPLETED:

NAME:

DATE:

INSTITUTIONAL ADJUSTMENT SCORE
			TOTAL SCORE
History of Involvement in Institutional Incidents	a.	no prior involvement	0
	b.	any prior involvement	2
	c.	prior involvement in one or more incidents in "greatest" or "high" severity categories	2
	d.	prior involvement during last give years of incarceration;	1
		- In an assault (no weapon or serious injury)	1
		- In a riot or major disturbance	2
		- In an assault (using a weapon or causing serious injury)	2
	e.	Involvement in one or more serious incidents prior to sentencing and/or pending placement for current commitment	5
		8 X TOTAL of a. to e.
Escape History	a.	no escape or attempts	0
	b.	an escape or attempt from minimum or community custody with no actual or threatened violence:
		- over two years ago	4
		- in last two years	12
	c.	an escape or attempt from medium or maximum custody or an escape from minimum or community custody with actual or threatened violence:
		- over two years ago	20
		- in the last two years	28
		two or more escapes from any level within the last five years.	28
Street Stability	a.	above average	0
	b.	average	16
	c.	below average	32
Alcohol/Drug Use	a.	no identifiable problems	0
	b.	abuse affecting one or more life areas	3
	c.	serious abuse affecting several life areas	6