A Criminal Career Study of the Continuity and Discontinuity of Sex Offending during the Adolescence-Adulthood Transition: A Prospective Longitudinal Study of Incarcerated Youth

Abstract

Longitudinal research has seriously challenged assumptions that juvenile sex offenders (JSO) are characterized by high level of dangerousness, mental health problems, and crime specialization in sex offenses. The current study examines the longitudinal pattern of offending among a sample of JSO and a sample of juvenile nonsex offenders. The research design includes longitudinal data over a nine-year period allowing the examination of offending patterns and the crime mix from age 12 to age 23. The findings highlight that, while JSO are prone to persist offending in adulthood, there is limited continuity of sex offending. Further, the findings stress the importance of taking into account nonsexual juvenile delinquency, more specifically, youth violence, to make a better assessment of early adult offending outcomes of JSO.

Keywords:

• sex offending
• longitudinal study
• juvenile offending
• criminal careers
• recidivism
• desistance
• adulthood

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Groth (1977) reported that over 20% of those in the study sample with a prior sex offense were either not charged or the charge for that initial sex offense was dismissed. It also indicated that another 34% received a suspended sentence for their offense. Indeed, the Groth data revealed that the initial offense frequently was dealt with through alternative sanctions, which may have diverted these first-time sex offenders from mental health facilities and clinical settings. Hence, these first-time JSO and those dealt through alternative means were not well represented in Groth’s analysis of the adolescent “sexual predator” which likely influenced the conclusion drawn by the researcher. In sum, the high level of persistence of sex offending found in early reports may have been a function of the way the court dealt with sexual recidivists (as opposed to first-time offenders). In effect, many of these initial studies may have relied on data concerning the persistence of sex offending mainly for sexual recidivists. The image of JSO as sexual predators, especially those youth adjudicated and sentenced to detention, remains among the public and policy-makers.

² Caldwell’s (2010) meta-analysis of 63 independent studies by showed that the weighted mean percentage of general recidivism was approximately 43% but only approximately 7% for sexual recidivism. Very importantly, the mean percentage of sexual recidivists across studies did not significantly vary across the measure of recidivism (arrest vs. conviction) and the population sampled (community, residential, and secured).

³ Note that the same data-set was used in the McCuish, Lussier, and Corrado (2015) study, which focused on the identification of offending trajectories using semi-parametric group-based modeling and whether the offending trajectories of JSO are different from those juvenile nonsex offenders.

⁴ Females (n = 66) were excluded from the current study for comparison purposes as contemporary studies typically analyze male and female sex offenders separately (e.g. Mathews, Hunter, & Vuz, 1997). Females were also excluded for methodological reasons as the analyses conducted in the current study cannot be conducted on the sample of female sex offenders (n = 4) due to their small number.

⁵ In total, 63.6% offended against females only while 30.8% offended against males only. Moreover, 31.9% of the sample sexually offended against children less than 13 years old. Also, 48.9% of the sample offended against adolescents and 12.8% against adults. Another 6.4% of the sample offended against multiple victims from multiple age groups. Finally, 20.0% of the sample offended against a family member, 15.6% against a stranger, whereas the bulk of the victims were either a friend or an acquaintance.

⁶ There is, however, a well-known overrepresentation of Aboriginals in detention, and the current study is no exception. Specifically, the percentage of Aboriginal offenders in the current study was substantively higher than within the general population of British Columbia (6.2%; Statistics Canada, 2013). Additionally, because this was a Canadian-based sample, the percentage of Aboriginal offenders was dissimilar from incarcerated adolescent offenders in the USA (e.g. Teplin et al., 2013) though the over-representation of Aboriginal offenders in Canada is similar to the over-representation of African American and Hispanic offenders in the USA The over-representation of Aboriginal offenders in custody is also present in states such as Alaska, North Dakota, South Dakota, and Montana, where Native Americans account for 29–42% of all youth in custody (Cross, 2008).

⁷ The recidivism probabilities for each X_t+1 conviction, where t is the first conviction, were plotted on a graph. A break in recidivism probabilities was observed at the seventh sentencing date during adolescence. The same pattern was observed in adulthood.

⁸ An odds ratio is a measure of association between an exposure variable (e.g. whether or not youth were convicted for a sex offense during adolescence) and an outcome variable (e.g. whether or not youth were convicted for a sex offense in adulthood). Essentially, the OR refers to the ratio of the odds of an event occurring (e.g. adult sex offending) in one group (e.g. JSO) to the odds of the event occurring in another reference group (e.g. JNSO). An OR of 1.0 indicates that the odds of one group relative to the odds of the reference group are identical (no association between the exposure variable and the criterion). An OR greater than 1.0 indicates that the odds are higher for the former group compared to the odds for the reference group, whereas an OR smaller than 1.0 indicates that the odds for the former group are lower than the odds for the reference group.

⁹ Models were ran with and without an interaction effect (JSO × juvenile violent offending) to determine whether the combination of these two crime types was associated with differential adult offending outcomes. Two key results emerged when the interaction effect between juvenile sex and violent offending was added to the model predicting the lambda of general offending: (1) the OR of juvenile sexual offending (main effect) and youth violence (main effect) became more strongly associated with a lower lambda; (2) the covariates remained relatively unaffected by the inclusion of the interaction effect.

¹⁰ The current study findings does not suggest that young incarcerated aboriginals should be singled out and labeled future adult sex offenders. Federal residential school policies instituted in the early 20th century and only terminated in the 2nd half of that century have been associated with the fundamental breakdown of many traditional First Nations and Aboriginal family structures and cultures along with various forms of systematic abuse within residential schools. High rates of alcohol use and misuse (usually explained in terms of self-medicating responses to trauma), the related within family violence, disproportionate levels of low social capital/ mother-only families, and high rates of geographic and residential mobility/instability reduce the likelihood of prosocial adolescents and, conversely, increase the likelihood of young offending and lengthy adult criminal trajectories for these young offenders (e.g. La Prairie, 1997). A related set of risk factors include Aboriginal young adults being vulnerable to employment and social interaction life stresses and difficulties linked to limited education and coping skills, high unemployment rate, and combined substance use and abuse issues (e.g. Bonta, LaPrairie, & Wallace-Capretta, 1997). These sets of risk factors may explain not only Aboriginal incarcerated young offenders’ higher risk for adult-onset sex offending, but also their higher risk for chronic offending, property offending as well as administrative violations, such as breaching their supervisory conditions.

Additional information

Notes on contributors

Patrick Lussier

Patrick Lussier is a criminology professor at the Faculty of Social Sciences at Laval University. He has over 100 publications on topics related to developmental life course criminology, offending trajectories, risk assessment/prediction and sex offending. His work has been published in journals such as Criminology, Criminal Justice and Behavior, Journal of Interpersonal Violence and Sexual Abuse: A Journal of Research and Treatment.

Raymond R. Corrado

Raymond R. Corrado is a professor in the School of Criminology at Simon Fraser University and an adjunct professor in the Faculty of Law at the University of Bergen. He has co-authored several edited books and published extensively on violent young offenders, mental health, and adolescent psychopathy.

Evan McCuish

Evan McCuish is a doctoral student in the School of Criminology, Simon Fraser University. He is the project director of the SSHRC-funded Study on Incarcerated Serious and Violent Young Offenders (ISVYO), the largest and longest running study on young offenders in Canada. His research interests include young offenders, psychopathy, developmental life course theory, sexual offending, and criminal trajectories. He has published on the assessment of adolescent psychopathy, criminal trajectories, and adolescent sexual offending.