FORUM on Corrections Research

Profiling federal offenders on conditional release

The Correctional Service of Canada must determine how it can best respond to the unique risk/needs of offenders under community supervision. In the 1980s, many correctional agencies tended to see offenders as generic and, therefore, made few distinctions based on gender or criminogenic needs but, rather, took a broad-brush approach to allocating correctional services. This article is intended to raise awareness about the nature of the conditional release population (gender, location, release type, phase, age, dynamic risk factors) and how far Canadian federal corrections has progressed in profiling this population.

Static and dynamic risk factors

Distinguishing static from dynamic risk factors had been a problem in criminal prediction. Although some corrections professionals might still believe that offenders do not change -- "once a criminal, always a criminal" -- more believe in rehabilitation, the idea that offenders can learn to live as law-abiding citizens. The underlying assumption of rehabilitation is that, if one does a good job identifying static and dynamic risk factors and assists offenders accordingly, then these offenders may become more law-abiding.

Little can be done to change static risk factors such as criminal history. These variables have considerable predictive power, however, especially the number and variety of criminal convictions and breaches of trust.

Dynamic risk factors refer to criminogenic or case needs that reflect change in an individual.² This is a critical component not only of offender risk/needs assessment, but also of risk management, because this is where intervention takes place. The goal is to reduce the likelihood of a criminal future by targeting these factors effectively and applying appropriate interventions.

In the 1980s, the corrections community was divided on the relative significance of static and dynamic risk factors. Some took the position that static factors were the most predictive and that decisions about community supervision should be based on them. This approach is very problematic, because frequency of contact, level of supervision and amount of service to be delivered cannot be varied if offenders do not change, and an assessment method based on these factors cannot detect or measure changes in an offender. The deficiencies of this approach resulted in a conceptual shift toward a thorough examination of offender needs as a sub-set of dynamic risk factors, allowing the introduction of flexibility in service delivery.

Past research

In 1988, Correctional Service of Canada parole officers were asked to review their cases and rate offenders using static risk factors on a continuum ranging from low to high risk. They based their ratings on criminal history, the Statistical Information on Recidivism Scale,³ National Parole Board decisions, their general experience with offenders and their knowledge of each specific offender.

The parole officers took a similarly systematic approach to assessing offenders' needs, reaching a global rating of low-, medium- or high-need for each offender. To conduct this assessment, they used the Community Risk/Needs Management Scale,⁴ which measures case needs on 12 separate dimensions, including academic/vocational skills, employment, financial management, marital/family relations, companions, accommodation, emotional/behavioural, alcohol use, drug use, mental ability, health and attitudes.

This review was important because it meant we were assessing offenders on truly dynamic risk factors -- the idea being that these needs should change over time. However, the static risk, namely the criminal history to date, remained, and it does not change over time.

Population profiling

An important distinction must be made between samples drawn from caseloads and samples drawn from newly released offenders when considering information such as the results of prediction studies. Caseload samples are drawn from the entire group under supervision, whereas new release samples are drawn only from offenders recently placed on conditional release. The latter group of offenders is considered to be at greater risk of returning to custody than those who have been in the community longer.

Phase of conditional release is an important concept. The first six months after release is often difficult for offenders, and interventions and levels of supervision may have to be adjusted accordingly. The second phase, between 6 and 12 months after release, is somewhat less difficult. When an offender remains on conditional release for more than 12 months, he or she is considerably more likely to succeed in the long term. The relationship between phase of conditional release and offenders' risk/needs profile is an important consideration.

The issue of offender needs is contentious: Should we assess offender needs? Do needs have predictive validity? Will intervening in these needs influence the likelihood of a criminal future?

The early 1990s research explored distributions of identified needs, which lists the 12 need dimensions of the Community Risk/Needs Management Scale. Parole officers rated offenders on each factor on a continuum ranging from "asset to community adjustment" to "considerable need for improvement." The purpose of the pilot tests was to learn more about each factor to improve management of supervision cases. The research indicated several statistically significant relationships between specific dynamic risk factors and the likelihood of suspension.

In August 1995 and 1997, the available Community Risk/Needs Management scales of the federal offender population on conditional release were extracted from the Offender Management System.

In the past, women were far outnumbered by men in the correctional population, limiting research in this area and making it very difficult to generalize or draw conclusions about female offenders. The current proportion of women in the community supervision population (297 or 3.9%) is a large enough sample to study.

Location

The Correctional Service of Canada has five regions: Atlantic, Quebec, Ontario, Prairies and Pacific (see Table 1). Female offenders in the Ontario region (45.1%) and Prairies region (29.3%) are highly represented in the conditional release group. Male offenders in Quebec (31.4%) and Ontario (24%) are also highly represented in the community supervision population.

Table 1

Regional Distribution of Offenders on Conditional Release
	Year	Males	Females
Atlantic	1995	776(13.0%)	26 (12.9%)
	1997	800(11.0%)	27(9.1%)
Quebec	1995	1,897(31.8%)	20(9.9%)
	1997	2,286(31.4%)	32(10.8%)
Ontario	1995	1,524(25.5%)	91(45.0%)
	1997	1,748(24.0%)	134(45.1%)
Prairies	1995	1,029(17.2%)	47(29.3%)
	1997	1,593(21.9%)	87(29.3%)
Pacific	1995	741(12.4%)	18(8.9%)
	1997	859(11.8%)	17(5.7%)
Total	1995	5,967(96.7% of total)	202(3.3% of total)
	1997	7,286(96.1% of total)	297(3.9% of total)
Note: As of 31 August 1995 and 1997

Release type

There are three major types of conditional release: day parole, full parole and statutory release (see Table 2). Both female and male offenders on conditional release are most likely to be on full parole (77.8% and 56.5%, respectively). However, male offenders on statutory release (32.6%) are also highly represented.

Table 2

Release Type Distribution of Conditional Release Population
	Year	Males	Females
Day Parole	1995	804(13.5%)	32 (15.1%)
	1997	795(10.9%)	36(12.1%)
Full Parole	1995	3 750 (62.8%)	152(75.2%)
	1997	4 114(56.5%)	231(77.8%)
Statutory release	1995	1 413(23.7%)	18(8.9%)
	1997	2 376(32.6%)	30(10.1%)
Note: As of 31 August 1995 and 1997

Phase

Offenders may be reintegrated more easily if the parole officer knows what phase of conditional release they are in (see Table 3). About one third of federal offenders on community release have been out less than six months. This latter group often requires more intensive supervision.

Table 3

Phase of Release Distribution of Conditional Release
Phase of Release	Year	Males	Females
0 to 6 months	1995	1 519(25.5%)	53 (26.2%)
	1997	2 669(37.5%)	105(36.0%)
6 to 12 months	1995	1 715(28.7%)	64(31.7%)
	1997	1 368(19.2%)	56(19.2%)
12 months or more	1995	2 734(45.8%)	85(42.1%)
	1997	3 088(43.3%)	131(4.9%)
Note: As of 31 August 1995 and 1997; numbers may very due to missing information

The largest percentage of federal offenders on conditional release has been out for 12 months or longer, a statistic that can be explained by the simple fact that the group of offenders serving longer or life sentences accumulates over time.

Demographics

Age is important to an offender's reintegration potential. Most federal offenders on community release are between 30 and 50 years of age. Almost as many federal offenders under community supervision are 50 and over as under age 30, a trend that has been increasing recently as our offender population ages (see Table 4), like the rest of the population of Canada. The implication of this age distribution for community supervision is that the offender population of the future will probably have a sizeable young group and an equally large older group.

Table 4

Age Distribution of Conditional Release Population
Age	Year	Males	Females
< 30	1995	1 434(24.0%)	44 (21.8%)
	1997	1 547(21.2%)	83(27.9%)
30-50	1995	3 622(60.7%)	125(61.9%)
	1997	4 262(58.5%)	170(57.2%)
50+	1995	912(15.3%)	33(16.3%)
	1997	1 477(20.3%)	44(14.8%)
Note: As of 31 August 1995 and 1997

Dynamic risk factors

To examine changes in dynamic risk factors across the phases of conditional release, we collapsed the caseload snapshot into three groups: 0 to 6 months, 6 to 12 months and 12 months or over (see Table 5). We found some interesting patterns. Generally, offenders who have been in the community 12 months or longer have much less need on the majority of dimensions than more recently released offenders.

Table 5

Dynamic Risk Factors by Phase of Release (Female and Males Offenders) 1997
		Phase of Conditional Release (months)
Dynamic risk factors		0 to 6	6 to 12	12+	P
Academic/vocational	Female(%)	79.4	71.4	51.5	***
	Male(%)	59.7	57.0	40.8	***
Employment	Female(%)	77.3	76.8	64.6	***
	Male(%)	70.0	67.6	55.2	***
Financial management	Female(%)	75.3	69.6	60.8	ns
	Male(%)	62.2	61.1	53.3	***
Marital and family relations	Female(%)	60.8	64.3	55.4	ns
	Male(%)	50.0	47.0	38.2	***
Companions	Female(%)	70.1	46.4	39.5	***
	Male(%)	58.9	53.0	37.3	***
Accommodation	Female(%)	20.6	25.0	30.0	ns
	Male(%)	16.6	17.6	22.2	***
Behavioural and emotional	Female(%)	83.5	82.1	63.1	***
	Male(%)	71.2	68.1	47.2	***
Alcohol use	Female(%)	23.7	17.9	10.8	*
	Male(%)	35.8	27.8	17.2	***
Drug use	Female(%)	29.9	26.8	16.2	*
	Male(%)	38.8	31.1	18.3	***
Mental ability	Female(%)	3.1	7.1	5.4	ns
	Male(%)	8.2	9.0	6.0	***
Health	Female(%)	20.6	26.8	43.3	**
	Male(%)	20.7	22.6	30.9	***
Attitude	Female(%)	11.3	5.4	10.1	ns
	Male(%)	17.3	17.8	14.5	**
Note: ns = non significant; *p<0.05, **p<0.01, ***p<0.001

Conclusion

We clearly have more dynamic risk information about offenders on conditional release than we ever had before. This information shows us who we are dealing with, where they are, what they are like and what kind of problems they have in the community. Our new challenge is to devise strategies to identify the dynamic risk factors that are most influential on future criminal behaviour, and to develop interventions that respond to those key factors.

---

1. 340 Laurier Avenue West, Ottawa, Ontario K1A 0P9.

2. D. A. Andrews, J. Bonta and R. D. Hoge, "Classification for effective rehabilitation: Rediscovering psychology," Criminal Justice and Behavior, 17 (1990): 19­52.

3. J. Nuffield. Parole decision-making in Canada: Research towards decision guidelines (Ottawa, ON: Communication Division, Secretariat, Department of the Solicitor General Canada, 1982).

4. L. L. Motiuk and F. J. Porporino, Field test of the Community Risk/Needs Management Scale: A study of offenders on caseload (Ottawa, ON: Correctional Service of Canada, 1989). And see L. L. Motiuk, "The Community Risk/Needs Management Scale: An effective supervision tool," Forum on Corrections Research, 9, 1 (1997): 8­12