FORUM on Corrections Research

Moderating segregation as a means to reintegration

Cherami Wichmann and Mark Nafekh¹
Research Branch, Correctional Service of Canada

The Correctional Service of Canada has developed a reintegration strategy to promote the safe release of offenders most likely to succeed in the community with appropriate safeguards. The probability and timing of release can be affected by many factors, not the least of which is institutional behaviours. One very important indicator of these factors is whether or not an offender is placed into segregation.

Placement in segregation can have wide ranging effects, and may even directly impact on the possibility of a discretionary release. Segregation involves voluntary or involuntary confinement of an offender to a cell, with limited movement out of that cell. These cells are primarily used for administrative or disciplinary reasons which are clearly outlined in the Corrections and Conditional Release Act (CCRA).²However, offenders may also be temporarily separated from the general population for other reasons (such as awaiting a disciplinary hearing).

Study design

A review of Correctional Service of Canada’s Offender Management System (OMS) revealed that 48,732 segregation placements had been logged for offenders in federal institutions between 1995 and 2000. While men offenders accounted for the majority of placements in segregation, women offenders accounted for 2.3% of these placements. As depicted in Table 1, the main reason for involuntary placements for men were the risk that the offender posed to others, the safe running of the institution and as a sanction for conviction of a serious institutional offence. Voluntary placements were mostly due to personal safety difficulties the offender would face by remaining in the general population.

Table 1

Percentage Distribution of Reasons for Segregation
	Men		Women
Reason For Segregation	Involuntary (37,484)	Voluntary (10,087)	Involuntary (926)	Voluntary (235)
Jeopardizing Institution/Others	58.4	4.4	64.2	8.5
Interference with Investigation	4.4	0.2	5.1	0.0
Personal Safety	6.8	93.8	21.0	77.0
Institutional Offences	20.7	0.4	8.3	0.0
Other	9.7	1.2	1.4	14.5

For women offenders, the distribution of reasons for segregation was different. About two thirds of the involuntary placements for women were due to the risk they posed to others. Interestingly, institutional offences accounted for 8% of involuntary placements for women compared to 21% for men. Although the predominant reason for voluntary segregation among women was their personal safety, 8.5% of these placements were requested by the women because the risk they potentially posed to others.

Characteristics of segregated and non-segregated offenders

Demographics

Overall, there were no significant differences between involuntarily and voluntarily segregated men in relation to selected demographic variables (see Table 2).3 This is consistent with previous research on the characteristics of segregated and non-segregated offenders.⁴ However, Aboriginal women offenders had a substantially higher percentage of involuntary placements to segregation than voluntary placements.

Table 2

Demographic Information
	Men			Women
Variable		Involuntary (10,248)	Voluntary (5,617)	Involuntary (234)	Voluntary (89)
Age at Admission	M	25.0	25.6	28.0	29.7
Sentence Length ( years)	M	5.3	3.6	3.6	4.1
Aboriginal	%	16.8	15.8	25.9	13.5
Serving Life/ Indeterminate	%	7.4	6.0	6.0	6.7
Note: M = arithmetic average.

Table 3

Static and Dynamic Risk Factors by Group for Men
	Involuntary		Voluntary
Risk Factors	No (3,596)	Yes (8,682)	No (5,306)	Yes (4,720)
Static ***/* * *
Low	16.5%	4.5%	19.3%	4.1%
Medium	44.9%	36.3%	45.7%	36.2%
High	38.6%	59.2%	35.0%	59.7%
Dynamic ***/***
Low	14.1%	3.2%	16.0%	2.3%
Medium	41.8%	30.2%	42.6%	26.6%
High	44.1%	66.6%	41.4%	71.2%
Note: statistical significance involuntary/ voluntary. *** = The difference is statistically significant at p< .001.

Criminal history

Between group differences were explored separately for men and women by comparing the two groups of segregated offenders with comparison groups not placed in segregation during their sentence. To control statistically for demographic differences between these two groups, they were matched on sentence length, age at first admission, and type of offence. An examination of past involvement with the criminal justice system revealed that compared to non-segregated men, both involuntary and voluntary groups were significantly more likely to have had prior contact with the criminal justice system as juveniles (p< .001). This finding did not hold true for the women. Nevertheless, the majority of segregated offenders (men and women) had a prior adult criminal record. Finally, both men and women offenders later segregated (involuntarily or voluntarily) were more likely to have experienced difficulties with institutional adjustment relative to their non-segregated counterparts.

Static and dynamic risk factors

Tables 3 and 4 depict a breakdown of the static (criminal history) and dynamic (criminogenic needs such as: employment, education, associates, substance abuse, community functioning, personal/emotional orientation, and attitudes) risk factors ratings from the Offender Intake Assessment (OIA) process completed when offenders enter into federal custody.⁵ The findings indicated that segregated offenders were more likely to be rated higher risk at intake than their matched counterparts.

Table 4

Static and Dynamic Risk Factors by Group for Women
	Involuntary		Voluntary
Risk Factors	No (131)	Yes (207)	No (43)	Yes (77)
Static ***/* * *
Low	37.4%	20.3%	34.9%	16.9%
Medium	40.5%	45.9%	46.5%	40.3%
High	22.1%	33.8%	18.6%	42.8%
Dynamic ***/***
Low	23.7%	8.7%	23.3%	3.9%
Medium	48.1%	37.7%	46.5%	36.4%
High	28.2%	53.6%	30.2%	59.7%
Note: statistical significance involuntary/ voluntary. *** = The difference is statistically significant at p< .001; ** p< .01.

Reintegration potential at intake

One of the measures employed to assist with reintegration efforts is the determination of each offender’s reintegration potential at intake.⁶ A three-point rating (high, moderate, or low) is determined by a combination of three standardized risk-based measures during OIA. For example, an offender rated as “high” reintegration potential could be viewed as requiring minimum security, as a very good release risk, and as having few or no criminogenic needs.

Figure 1 displays the reintegration potential for segregated and non-segregated men offenders.(see Figure 1 ). Figure 2 reflects the breakdown of reintegration potential ratings for women offenders. As the figures show, both men and women offenders placed in segregation were, at intake, rated as being significantly higher risk, higher need, and lower in reintegration potential than offenders who were not subsequently placed in segregation (see Figure 2).

Implications of segregation

Security reclassification

Institutional adjustment is an important consideration in determining whether offenders, men or women, cascade down to lower levels of security. Consequently, placement in segregation, and other indices of adjustment are major factors being considered during a review. In fact, placement in segregation is the strongest predictor of security reclassification, having been found to account for up to one-quarter of the variance in these decisions.⁷ Thus, standardized instruments developed to inform routine security reviews are heavily influenced by this factor.

Program participation

The nature of the segregation environment can impose limits on offender programming while in custody. As many offenders could be held there for significant periods of time, their ability to address the goals of their correctional plan may be impeded. In fact, research has shown that program completion is inextricably linked to an offender’s ability to cascade to lower levels of security.⁸

Discretionary release

Follow-up release and readmission information was gathered for the sample and statistical controls were introduced for level of reintegration potential. Due to the relatively small number of segregation placements for women, these analyses were conducted only for men offenders.

Within the high reintegration potential group, men offenders who were subsequently placed in segregation, were found to be less likely than their non-segregated counterparts to have been granted a discretionary release (63% versus 85%, p< .001).

Among the offenders assessed at intake to be “high” reintegration potential and subsequently released, significantly more of the segregated offenders were readmitted before their sentences ended.

In sum, it would appear that moderating the relative use of segregation would result in some important reintegration dividends. Notwithstanding, discovering creative and innovative alternatives to placement in segregation remains the task at hand.

---

1. 340 Laurier Avenue West, Ottawa, Ontario K4A 0P9.

2. The reasons for placement into segregation are described in Section 31-3 and 44-1(f) of the CCRA: 1) the inmate poses a danger to staff, other inmates or to the security of the institution (31-3a); 2) the inmate may interfere with an ongoing investigation (31-3b); 3) the inmate’s own safety is at risk (31-3c); or 4) as a sanction from conviction of a serious disciplinary offence (44 -1f).

3. There was some overlap between the involuntary and voluntary groups as some offenders had experienced both types of segregation.

4. Motiuk, L. L., & Blanchette, K. (1997). Case characteristics of segregated offenders in federal corrections. Research Report R-#57. Ottawa, ON: Correctional Service of Canada. See also Motiuk, L. L. & Blanchette, K. Characteristics of administratively segregated offenders in federal corrections, Canadian Journal of Criminology (in press).

5. Motiuk, L. L. (1997). Where are we in our ability to assess risk? Forum, on Corrections Research, 5(2), 18-22.

6. Motiuk, L. L. & Nafekh, M. (1999). Reintegration potential profiles for federally sentenced women, Forum on Corrections Research, 11(3), 13-17. See also Motiuk, L. L. & Serin, R. C. (1998) Situating risk assessment in the reintegration potential framework, Forum on Corrections Research,
10 (1), 19-22.

7. Luciani, F. (1997). The security classification of offenders: Development of security reclassification protocols. Ottawa, ON: Correctional Service of Canada.

8. Luciani, 1997.