FORUM on Corrections Research

An examination of release decisions for sexual offenders: Do biases exist?

Master Thesis, Carleton University¹
Sara L. Johnson²
Advisors: Brian Grant and Adelle Forth
Committee Members: William Petrusic, Ralph Serin, Tulio Caputo, and Joanna Pozullo

Research has demonstrated that sex offences are rated as very serious crimes³ and that sex offenders are judged to be very dangerous⁴as compared to other offenders and offences, including serious violent offences. Furthermore, certain victim, perpetrator and offence variables have been found to be related to attitudes towards sexual offenders,⁵ and to the severity of decisions made in the criminal justice system.⁶ This study examines the differences between sexual and (non-sexual) violent offenders on release decisions and recidivism following release. The study group of sexual offenders is compared to a group of violent offenders on first release type, and three types of recidivism.

Background

Researchers have concluded that sex offenders are targets of a disproportionate number of statutes for extra supervision and specialized sentencing.⁷ These include sex offender notification laws, specialized sentencing for sexual offenders, and specialized release practices. In Canada, Dangerous Offender legislation allows for the imposition of an indeterminate sentence to offenders considered dangerous. Although the criteria for a Dangerous Offender designation is the commission of a “serious personal injury offence” (such as sexual assault, manslaughter, aggravated assault), a large percentage of dangerous offender (approximately 92%)⁸ have committed a sexual offence. More recently, Canadian legislation has allowed for the imposition of a “Long Term Supervision Order” to sexual offenders who are judged to be in need of longer term supervision following release than normally provided. In addition, the application of detention legislation in Canada has tended to target sex offenders for extended time in custody.

In the Canadian federal correctional system, there are generally three options for release from prison: discretionary releases such as parole or day parole, statutory release, or release after the statutory release date following a period of detention. For discretionary releases, institutional parole officers recommend parole or day parole and a final decision is made by the National Parole Board (NPB). In general, offenders are eligible for full parole after they have served one-third of their sentence, and for day parole, six months prior to full parole eligibility. In the case of statutory release, offenders are automatically released after they have served two-thirds of their sentence, unless the offender is referred for detention. In certain circumstances, offenders may be detained for a period of time after the statutory release date to the end of their sentence if they meet the criteria set out in the Corrections and Conditional Release Act (CCRA).⁹ Offenders may be detained if (1) they have committed a violent or sexual offence as set out in Schedule I of the Act, that has caused serious harm to the victim(s) or they have committed a serious drug offence (Schedule II), and (2) they have been deemed likely to reoffend with a serious violent, sexual or drug offence before the completion of their sentence. In addition, in special circumstances, offenders may be detained if they are not currently serving a sentence for a serious violent, sexual or drug offence, but if they have been deemed likely to commit an offence causing the death of, or serious harm to another person or a serious drug offence prior to the expiration of the sentence.

Institutional parole officers decide which offenders to refer to the NPB for a detention hearing following a review of the offender’s criminal history and institutional behaviour. All offenders who have committed a violent or serious drug offence have their cases reviewed for detention prior to their normal release date. Once a referral has been made, the NPB must decide to detain or release the offender at the statutory release date. If the NPB decides to detain the offender, the Board must review the case every twelve months to determine if the offender can be safely released.

For serious violent offences (Schedule I), serious harm is determined by the existence of a number of factors, as set out in the Case Management Manual.¹⁰ However, recent amendments to the detention provisions of the CCRA¹¹ state that serious harm does not need to be proven where the offence is a sexual offence involving a child victim. The factors defining serious harm include: extent of injury of the victim; the nature of the offence and the circumstances surrounding it, such as use of brutality, excessive force, viciousness, or deviant sexual behaviour; the use of weapons to harm or threaten; prolonged or repeated abuse or terror; any particular vulnerability of the victim, such as being very young, aged, infirm, helpless, or handicapped. In addition, sexual offences are considered to be more likely to cause severe psychological damage than non-sexual offences.

Research suggests that a large percentage of detained offenders are sexual offenders.¹² However, these offenders are rated as lower risk and recidivate at an equivalent or lower rate than other offenders.¹³ Therefore, it appears that sexual offenders are being detained more often than would be expected given their risk to reoffend.

Current study

The current study compares the release type and rate of recidivism of sexual and non-sexual violent offenders serving federal sentences in 1991.

Study group: Sexual offenders

Using the Sex Offender Censusss¹⁴ conducted in 1991, a study group was selected who met the following criteria: men; were serving a sentence for a sexual offence or a sexually-related offence at the time of the census; were serving a determinate; had reached their statutory release date by January 1, 1999 and had a release date prior to March 30, 2000. This group consisted of 1,417 offenders.

Comparison group: Violent offenders

The comparison group was identified from a pool of offenders in custody on March 31, 1991. Among this pool of offenders, subjects meeting the following criteria were included in the comparison group: men; were serving a sentence in 1991 for a non-sexual violent offence (excluding homicide and robbery); were serving a determinate sentence; were not contained in the sex offender census; had reached their statutory release date by January 1, 1999 and had a release date prior to March 30, 2000. This group contained 1,777 offenders.

Outcome measures

Two sets of outcome measures were used in this study; release and recidivism. Release decision is the operationalization of correctional decision-making in the study, and is defined as the first release type granted to offenders after March 1991. The release decision variable has three levels: discretionary release (day parole or full parole), statutory release, and warrant expiry/ end of sentence (detention).

Recidivism was measured several ways in this study depending on the analysis used: presence or absence, first recidivism date, time to recidivism, and number and type of new offences. Three types of recidivism were examined: sexual recidivism, violent non-sexual recidivism and violent recidivism. These types of recidivism are defined below:

Sexual recidivism: includes offences classified as sexual offences in the criminal code such as sexual assault (all levels), sexual interference, indecent assault, incest, sexual exploitation, invite sexual touching, etc.

Violent non-sexual recidivism: includes all non-sexual violent offences such as assault (all levels), cause bodily harm, overcome resistance, attempt murder, and related person-based offences. In addition to these offences, all homicide (first degree murder, second degree murder, and manslaughter) offences are included. Robbery is not included.

Violent recidivism: includes both violent non-sexual and sexual recidivism (see above)

Results

As can be seen in Table 1, sexual and violent offenders significantly differed on release type (X² (2) = 282.87, p < 0.001). Furthermore, the Cramer’s Phi coefficient (0_c = 0.298) indicates that the relationship between offender type and release type is a strong one. In general, violent (non-sexual) offenders were more likely to have a discretionary release or a statutory release while sexual offenders were more likely to be released at warrant expiry after a period of detention. Overall, approximately 30% of the sexual offenders were released at warrant expiry, compared to approximately 8% of the violent offenders, amounting to more than three times as many sexual offenders being released at warrant expiry than violent offenders.

Table 1

Figure 1

In addition, significant differences were observed between sexual and violent offenders on violent non-sexual recidivism (X² (1) = 81.88, p < .0001), with approximately 10% fewer violent offenders surviving without committing a new violent non-sexual offence than sexual offenders. At 96 months (8 years), approximately 80% of the violent offenders compared to over 92% of the sexual offenders had not recidivated with a violent non-sexual offence. In terms of sexual reoffending, violent offenders were found to have significantly higher rate of surviving without a new sexual offence than sexual offenders (X² (1) = 108.1, p < .0001). Eighty-nine percent of the sexual offenders compared to 98% of the violent offenders did not have a sexual reoffence 8 years after release.

Discussion and conclusions

Sexual offenders were found to have more restrictive release types than violent (non-sexual) offenders, with approximately three times as many sexual offenders being released at the end of their sentence, after some period of time detention, as violent offenders. Although sexual offenders had higher rates of sexual reoffending, when considering both violent non-sexual and sexual reoffending (referred to as violent recidivism in this study), sexual offenders had lower rates of reoffending. This suggests that information unrelated to future risk of recidivism is being considered by the decision makers when making decisions about release.

---

1.  Extract of Johnson, S. L. (2001). An Examination of release decisions for sexual offenders: Do biases exist? Master’s Thesis, Carleton University.

2.  340 Laurier Avenue West, Ottawa, Ontario, K1A 0P9.

3.  Bannister, P., and Pordham, S. (1994). Public perceptions of crime seriousness.Issues in Criminological and Legal Psychology, 21, 4-10. For other authors, please refer to the thesis.

4.  Howe, E. S. (1988). Dimensional structure of judgments of crimes.
Journal of Applied Social Psychology, 18, 1371-1393. See also, Howe, E. S. (1994). Judged person dangerousness as weighted averaging. Journal of Applied Social Psychology, 24(14), 1270-1290.

5.  Hogue, T. E. (1993). Attitudes towards prisoners and sexual offenders. Issues in Criminological and Legal Psychology, 19, 27-32. For other authors, please refer to the thesis.

6.  Berliner, L., Shram, D., Miller, L. L., and Milloy, C. D. (1995). A Sentencing Alternative for Sex Offenders: Astudy of decision making and recidivism. Journal of Interpersonal Violence, 10(4), 487-502. For other authors, please refer to the thesis.

7.  Prentky, R. A., Lee, A. E. S., Knight, R. A., and Cerce, D. (1997). Recidivism rates among child molesters and rapists: A methodological analysis.Law and Human Behavior, 21(6), 635-659.

8.  Motiuk, L. L., and Bonta, J. (1996). High-Risk Violent Offenders in Canada. Research Report R-50. Ottawa, ON: Correctional Service of Canada.

9.  RCS, C-20, (1992). Corrections and Conditional Release Act

10.  Ottawa, ON: Correctional Service Case Management Manual (1998). of Canada.

11.  RSC, C-20, Corrections and Conditional Release Act, August 1997 Edition, (1997).

12.  Grant, B. A. (1996). Inmates Referred for Detention (1989-90 to 1993-94): A Comparative Analysis, Research Report R-45. Ottawa, ON: Correctional Service of Canada.

13.  Ibid. See also Nugent, P. (1999). The use of detention legislation: Factors affecting detention decisions and recidivism among high-risk federal offenders in Ontario. Unpublished doctoral dissertation, Kingston, ON: Queen’s University.

14.  Porporino, F. J., and Motiuk, L. L. (1991). Preliminary results of national sex offender census. Research Report R-29. Ottawa, ON: Correctional Service of Canada.