FORUM on Corrections Research

Sex offender treatment priority: An illustration of the risk/need principle

This article contrasts outcome research on two Saskatchewan-based sex offender treatment programs. One program deals with provincially-incarcerated offenders, while the other operates at the Correctional Service of Canada's maximum-security Regional Psychiatric Centre (Prairies).

The two programs are of similar intensity. However, the provincially incarcerated offenders arguably need less intensive treatment, as federally incarcerated (serving sentences of two years or longer) sex offenders tend to commit more serious crimes, have greater needs and present greater risk.

Both treatment programs focus on relapse prevention and operate in group form to help offenders recognize high-risk situations, overcome rationalizations and denial, and assume responsibility for their offence(s).⁽²⁾ This treatment approach has been shown to be most effective with offenders considered to present the highest risk and degree of need.⁽³⁾ The provincial program A total of 30 inmates were referred to the provincial sex offender treatment program during the study period. Two inmates did not complete the program, one was excluded from the study sample because of a mental disability, and one was excluded from the sample because he died shortly after release. Therefore, 26 offenders were considered for this study.

Approximately 62% of the sample were Caucasian, with the remaining 38% being either Aboriginal or Métis.

The average sample age was 38, and the average level of educational achievement was at approximately a Grade Nine level.

Most of the offenders were first-time sex offenders who had been convicted of molesting children known to them. Although 56% of the offenders had previous criminal convictions, just 15% had been previously convicted of a sex offence. The average sample sentence (including probation) was 27.5 months.

Given the relative absence of criminal history, the prevalence of intrafamilial offenders and the fact that all the offenders were serving sentences of less than two years, this sample was considered relatively low-risk.

A comparison group was created from a sample of men who had been incarcerated for similar numbers and types of sex offences in Saskatchewan during the study period.

After matching for age, race, previous sex and non-sex offences, length of sentence and "time at risk", we were left with 35 comparison offenders who received no treatment before release to the community.

"Time at risk" represents the number of months the offender spent in the community after release. This variable was measured from the parole release date listed in the Canadian Police Information Centre database or from the date that marked completion of two-thirds of the offender's sentence.

Questionnaires were used to assess the emotional state, sexual attitudes and sexual knowledge of the offenders before and after treatment. Social desirability, cognitive functioning, possession of sexual information, psychiatric symptoms and behavioural role-play performance were also measured. Paper-and-pencil tests were administered before and after treatment, and the role-plays were conducted and rated before treatment, immediately after treatment, and three months after treatment. Finally, offenders were rated for skills such as assertiveness and anxiety, with social skills given an overall rating.

When we looked at average offender anxiety ratings over the three role-plays, the independent ratings suggested a significant lessening of anxiety throughout the three tests (p<.01).⁽⁴⁾ The role-play ratings also suggested an improvement in overall social skills (p<.01).

Recidivism data were extracted from Canadian Police Information Centre database records. These data included arrests, convictions, and parole or probation violations after release. The average follow-up period for the treatment group was 31.2 months and the average follow-up for the comparison group was 28.8 months.

The comparison group had a slightly lower average sexual recidivism rate than the treatment group, although the groups did not differ significantly as to either non-sexual or overall recidivism (see Table 1).

Table 1

Recidivism Data for the Provincial Treatment (26 offenders) and Comparison (35 offenders) Groups
New offences	Average recidivism rates	p value (2-tailed)
Sex offences
Treatment group Comparison Group	0.11 0.03	0.181
Non-sex offences
Treatment group Comparison group	0.88 1.23	0.536
Total offences
Treatment group Comparison group	1.00 1.23	0.645

Taken together, these results suggest that, while treatment seemed to improve offender emotional states and social functioning, these changes were not associated with recidivism - at least not in this low-risk group. The federal program A 5.2-year (with a range of 0.4 to 148.5 months) follow-up of sex offenders treated in the Clearwater program at the Regional Psychiatric Centre (Prairies) yielded quite different results. This group of high-risk sex offenders (recidivist rapists and pedophiles serving federal sentences) was compared with a group of recidivist offenders released from federal prisons and followed up for three years.⁽⁵⁾

Consistent with an earlier comparison,⁽⁶⁾ we found that the treatment group had a 59% lower sexual recidivism rate than the comparison group - despite being followed up for more than two years longer (the sexual recidivism rate for the comparison group was 14.6%, as compared with 6% for the treatment group; p=0.022). Further, the treatment group returned to federal custody less often than the comparison group (48.8% versus 64.7%, respectively; p=0.013).

When the recidivist (federal) and non- recidivist (provincial) offenders were pooled, there were no outcome differences in terms of sexual recidivism. This is a clear illustration of the risk/need principle,⁽⁷⁾ in that the offenders at the highest initial risk level (recidivists) demonstrated the greatest treatment effects (as measured by readmission to federal institutions). Discussion Admittedly, the provincial sample findings were negative in that treatment did not appear to affect risk (as measured by recidivism). However, treatment can be expected to do little to reduce risk in an already low-risk group.

The risk presented by the provincially incarcerated offenders seems not to have been high enough to warrant the intensive intervention they received. Neither the treatment nor comparison samples had high rates of sexual acting out or other criminal behaviour after release, which may reflect their ability to manage their risk in the community - even without treatment. This is consistent with the risk/need principle, in that correctional treatment should not be expected to produce differential effects among low-risk offenders.⁽⁷⁾

It appears that, although sex offender programming produced changes in immediate treatment targets among low-risk offenders, these changes were not related to recidivism. However, treatment was effective when delivered to high-risk offenders. This has important policy and resource allocation implications. It seems that sex offenders in their thirties with no previous sex offences, a limited (if any) history of non-sex offences, and serving sentences of less than two years should be directed into low- intensity, low-cost programming. More intensive programming should clearly be reserved for the more difficult offenders who stand to benefit most from it.

---

⁽¹⁾2520 Central Avenue North, P.O. Box 9243, Saskatoon, Saskatchewan S7K 3X5.
⁽²⁾W. L. Marshall, A. Eccles and H. E. Barbaree, "A three tiered approach to the rehabilitation of incarcerated sex offenders," Behavioral Sciences and the Law, 11 (1993): 441-455.
⁽³⁾D. A. Andrews and J. Bonta, The psychology of criminal conduct (Cincinnati: Anderson, 1994). See also P. Gendreau, F. T. Cullen and J. Bonta, "Intensive rehabilitation supervision: The next generation in community corrections?" Federal Probation, 58 (1994): 72-78.
⁽⁴⁾Friedman's Block/Treatment Test.
⁽⁵⁾F. Porporino and D. Robinson, Unpublished report, Research and Statistics Branch, Correctional Service of Canada, 1991.
⁽⁶⁾A. Gordon, R. Holden and T. Leis, "Managing and treating sex offenders: Matching risk and need in programming,"Forum on Corrections Research, 3, 1 (1991): 7-11.
⁽⁷⁾Andrews and Bonta, The psychology of criminal conduct. See also Gendreau, Cullen and Bonta, "Intensive rehabilitation supervision: The next generation in community corrections?"