Research Highlights - Crime Prevention : Transitions From Juvenile Delinquency to Young Adult Offending: A Review of Canadian and International Evidence
Police-reported data shows that young adults aged 18-24 have the highest rates of criminal offending of any age group in Canada (Allen, 2014). Therefore, reducing offending in the young adult years should be a priority and could have a big impact on the total crime rate.
Young adult crime rates are lowest in Ontario, Quebec, Prince Edward Island, and British Columbia, and highest in Saskatchewan, Yukon, Nunavut, and Northwest Territories. It is important to investigate why crime rates are so high in the Territories.
The most extensive recent review of knowledge about young adult offending was published by a US National Institute of Justice Study Group chaired by Rolf Loeber and David Farrington, in a 2012 book entitled From Juvenile Delinquency to Adult Crime: Criminal Careers, Justice Policy, and Prevention (Loeber & Farrington, 2012). However, there is little mention of Canadian research in this book.
It is critical to understand why juvenile delinquents continue to offend in the young adult years and to make recommendations about how this transition to young adult offending can be most effectively prevented. There is a great need to review Canadian research on the transition from juvenile delinquency to young adult offending, and to compare the findings with conclusions from the United States (US), United Kingdom (UK), and other countries.
This document reviews longitudinal research on key questions concerning transitions from juvenile delinquency to young adult offending. Canadian research will be reviewed in particular, but some conclusions from other countries will also be mentioned (see Farrington, 2017b).
Longitudinal research, beginning in the juvenile years and continuing at least up to age 24, is needed to answer the key questions. In studying official offending in Canada, it is essential to begin in the juvenile years, because Canadian law dictates that youth records should be purged after a certain time period, depending on the severity of the crime and the behaviour of the offender. Thus, beginning with an adult sample would miss many juvenile offences.
Relevant longitudinal studies can either begin with community samples or with samples of official offenders. Community samples are preferable, so that representative samples of offenders emerge naturally from an initially non-delinquent sample, and the prevalence of offending can be determined. Incarcerated offender samples are informative, but they are unlikely to be representative of all offenders. It is also desirable to measure both self-reported and official offending, since it is known that officially recorded offending is only the “tip of the iceberg” of real offending, and both official and self-reported offending are biased measures of real offending.
Major Canadian longitudinal studies of offending have been carried out by: Leena Augimeri and Christopher Koegl in Toronto; Annie Yessine and James Bonta in Manitoba; Peter Carrington in 6 provinces; Ray Corrado and Evan McCuish in Vancouver; David Day in Toronto; Marc Le Blanc in Montreal; and Richard Tremblay in Montreal. More information about these projects is given in the text below.
The key questions are as follows:
What proportion of juvenile offenders up to age 17 become young adult offenders at age 18-24?
Koegl (2017) followed up high risk boys who received the SNAP early intervention program in Toronto at the average age of 9 and who were still alive at age 25. Of those who were convicted as juveniles, 74% were also convicted as young adults, compared with 31% of those who were not convicted as juveniles.
Le Blanc and Frechette (1989) compared juvenile and adult official offending for Montreal male wards of the court (delinquents or youth protection cases) who were first studied at the average age of 15 and followed up to their fifties. Of the convicted juvenile delinquents, 75% were convicted as young adults, compared with 59% of the non-delinquents.
Yessine and Bonta (2008) followed up male and female juvenile probationers in Manitoba from when they were first adjudicated in 1986-91 up to 2005. The data were reanalysed for the purposes of this report. The official offending data was grouped into ages 12-15, 16-20, 21-25, 26-30, and 31 or older. Of males who were convicted at age 16-20, 74% were convicted at age 21-25, compared with 51% of males who were not convicted at age 16-20.
Day (2017) studied 764 male juvenile offenders, who comprised the entire population of offenders who were sentenced to two open custody facilities operated by a children’s mental health centre in Toronto in 1985-96, and followed them in criminal records up to their thirties. Of the juvenile offenders, 90% were reconvicted as young adults (at age 18-24).
It is possible that official juvenile court processing may increase rather than decrease later offending. This was tested in the Montreal Longitudinal and Experimental Study, directed by Tremblay, in which boys were followed up from age 6 to the twenties. Petitclerc et al. (2013) found that 50% of boys who were processed in the juvenile court had an adult criminal record at age 18-25, compared with 24% of matched boys who were arrested but not taken to court, suggesting that juvenile court processing was associated with an increased likelihood of reoffending.
To what extent is there specific continuity in types of crimes, as opposed to versatility in offending?
It is important to know whether specific types of crimes committed in the juvenile years significantly predict specific types of crimes committed in the young adult years, over and above the general continuity in offending. International research usually shows that offenders are versatile, in committing several types of crimes, rather than specializing in only one type of crime. However, it is not clear whether this applies to the continuity in offending from the juvenile to the young adult years.
In his follow-up of Toronto high risk boys, Koegl (2017) reported that 55% of those who were convicted for a juvenile property offence were also convicted for a young adult property offence, compared with only 14% of those who were not convicted for a juvenile property offence. Similarly, 65% of those who were convicted for a juvenile violent offence were also convicted for a young adult violent offence, compared with only 20% of those who were not convicted for a juvenile violent offence. However, this observed continuity in types of offending may reflect the general continuity in offending rather than specific continuity in types of offending.
In the Incarcerated Serious and Violent Young Offender Study in Vancouver, McCuish and his colleagues (2015) followed up male juvenile sex offenders and male juvenile non-sex offenders up to age 23. They found that juvenile sex offenders were not more likely than other juvenile offenders to commit a sex offence as a young adult. Therefore, there was little evidence of specialization in sex offending from the juvenile to the young adult years.
In Bonta’s follow-up of Manitoba probationers, there was little evidence of specialization in violent compared with non-violent offending. Of the male violent offenders at age 16-20, 66% were violent offenders at age 21-25, compared with 60% of those who only committed non-violent offences at age 16-20.
To what extent can the probability of persistence into young adult offending be predicted, based on features of the previous criminal career?
International research shows that an early onset of offending predicts many offences and a long criminal career. Carrington (2017) studied all persons born in the fiscal year 1979-80 who appeared in youth or adult court in six Canadian provinces (which accounted for 78% of the national population) at least once between ages 12 and 21. Interestingly, the probability of persisting into the young adult years did not vary significantly with age of onset in the juvenile years. However, the average number of crimes in the criminal career decreased with increasing age of onset in the juvenile years. Also, the average duration of offending careers decreased with increasing age of onset.
In Bonta’s follow-up of Manitoba probationers, convictions at age 16-20 predicted convictions at age 21-25, but convictions at age 12-15 did not. Therefore, an early age of onset was less predictive of persistence than was a later age of onset in this project.
In Day’s (2017) follow-up of male juvenile offenders, an early age of onset predicted persistence into young adult offending. The juvenile-only offenders were more likely to have been convicted of a sex offence (33%) than the persistent offenders (23%). However, the juvenile-only offenders were less likely to have been convicted of a drug offence (8%) than the persistent offenders (43%).
Similarly, in the Incarcerated Serious and Violent Young Offender Study in Vancouver, McCuish (2017) found that juveniles who continued to offend between ages 18 and 23 had a significantly earlier average age of onset than those who were not convicted in young adulthood.
To what extent can different features of young adult criminal careers, such as the frequency of offending, the duration of criminal careers, and the monetary cost of offences, be predicted from earlier features of juvenile careers?
Le Blanc (2017) reported the relationship between juvenile (age 12-17) and young adult (age 18-24) official offending for Montreal male wards of the court. He found that the number of juvenile offences predicted the number of young adult offences.
Lussier and his colleagues (2015) reported the relationship between juvenile (age 12-17) and young adult (age 18-23) official offending for adolescent offenders in Vancouver. Once again, the number of juvenile offences predicted the number of adult offences.
In the Toronto research of Day (2017), the number of juvenile convictions predicted the number of adult convictions. Day (2017) also investigated the monetary cost of offences, based on Day et al. (2016). He found that the victim costs incurred as a juvenile were not correlated with the victim costs incurred in adulthood. However, the correctional disposition costs incurred as a juvenile were correlated with the correctional disposition costs incurred as an adult. The absence of a correlation for the victim costs may be attributable to the different types of crimes committed during the juvenile (more property crimes) compared with the adult (more violent crimes) years. There is a great need for more research on the monetary cost of criminal careers in Canada.
To what extent can the probability of persistence into young adult offending be predicted, based on earlier risk and protective factors?
It is important to know which risk and protective factors, measured in childhood and adolescence, are the strongest predictors of persistence into young adult offending. This knowledge would help in developing a risk assessment instrument to predict the probability of persistence. Also, methods of preventing persistence should target the most important risk and protective factors.
In his follow-up of Toronto high risk boys, Koegl (2017) investigated the ability of risk factors measured in childhood to predict persistence in offending from the juvenile to the young adult years. The most important childhood risk factor that predicted persistence was high impulsiveness. This result is impressive because impulsiveness was measured prospectively, before any boys appeared in the juvenile court.
In Bonta’s follow-up of Manitoba probationers, seven risk factors were recorded prospectively by probation officers (before offending at age 21-25 was known). Of these, two significantly predicted which offenders at age 16-20 would continue offending at age 21-25: financial problems and antisocial associates. Persistence was not predicted by negative attitudes, alcohol or drug abuse, family problems, educational problems, or accommodation problems.
In his follow-up of male juvenile offenders in Toronto, Day (2017) investigated the extent to which childhood (up to age 12) and adolescent (age 13-17) risk factors (derived from presentence reports by probation officers) predicted which juvenile offenders would become young adult offenders at age 18-24. Two childhood risk factors predicted persistence. Persistent offenders were significantly likely to come from broken homes, but they were significantly unlikely to have low intelligence or low achievement. Three adolescent risk factors predicted persistence. Persistent offenders tended to have received alternative care (e.g., foster care, child welfare, other institutional care), and tended to have poor peer relations (e.g., criminal peers, peer rejection). However, persistent offenders tended not to have health problems (e.g., traumatic head injury, obesity).
To what extent do findings vary with gender, ethnic/national origin, and different community or provincial contexts?
In his national research, Carrington (2017) found that males were twice as likely to persist in offending from the juvenile to the young adult years as were females (40% versus 21%). The highest rates of findings of guilt per 100 cohort members were in Saskatchewan, followed by Newfoundland and Labrador, Alberta, Prince Edward Island, Ontario, and lowest in Quebec (see Carrington et al., 2005). It is important to try to explain provincial differences in offending.
Yessine and Bonta (2009) compared the offending careers of 235 Indigenous and 204 non-Indigenous male offenders in Manitoba, who were originally aged 12-19 and on probation in 1986-91. Most (85%) were first time offenders. They were followed in criminal records up to 2005. Yessine and Bonta found that Indigenous offenders were more likely to reoffend: 39% of Indigenous offenders aged 12-15 reoffended, compared with 22% of non-Indigenous offenders, and 83% of Indigenous offenders aged 16-20 reoffended, compared with 65% of non-Indigenous offenders.
Yessine and Bonta’s data were also re-analysed here to investigate persistence in offending from age 16-20 to age 21-25. The results showed that 77% of Indigenous offenders persisted, compared with 55% of Indigenous non-offenders; and 70% of non-Indigenous offenders persisted, compared with 49% of non-Indigenous non-offenders. Therefore, there was only a slight tendency for Indigenous offenders to persist more than non-Indigenous offenders. The current re-analysis also investigated risk factors for persistence by Indigenous and non-Indigenous offenders, and found that the previously identified significant risk factors (financial problems and antisocial associates) were significant only for Indigenous offenders, not for non-Indigenous offenders. Therefore, risk factors might be different for Indigenous compared with non-Indigenous offenders.
What Do We Need to Know in Canada? And How Can We Find Out?
Official records often show that most juvenile offenders become young adult offenders. However, it is not clear that particular types of juvenile crimes predict the same types of young adult crimes. More research on the continuity of offending is needed in Canada. Since there is some evidence that juvenile court processing can increase rather than decrease offending, research is needed on how to make the juvenile court more rehabilitative and less damaging.
The existing Canadian evidence does not show conclusively that an early onset of offending predicts persistence from juvenile delinquency to young adult offending, although the number of juvenile offences does predict the number of young adult offences. It would be valuable to carry out a systematic review of the predictors of persistence into young adult offending and compare them with the predictors of juvenile offending.
The majority of findings that are relevant to this report are based on official records of offending. More research on self-reported offending is needed. Also, since the majority of research is based on male samples, more research is needed on female offending. Also, more research is needed on Indigenous compared with non-Indigenous samples. Official record studies rarely provide information on a wide range of risk and protective factors. It is clear that more survey research from adolescence to young adulthood is needed, collecting information on risk and protective factors and on self-reported offending. The only existing Canadian longitudinal studies containing self-report information are the Montreal studies by Le Blanc and Tremblay. These could usefully be reanalyzed to address the key questions using self-report data.
It may be possible to reanalyze other Canadian longitudinal studies to obtain relevant information. For example, self-reported offending was measured in the Canadian National Longitudinal Survey of Children and Youth (NLSCY). This began by surveying children aged 0-11 in 1994-95 and continued collecting data every two years until the final eighth survey of children and youth aged 14-25 in 2008-09. It is also possible to construct longitudinal datasets by linking records in administrative data (see Carrington, 2010). Ideally, a new longitudinal study on the development of offending in Canada (preferably in a large city) is needed. Such a study should begin with a sample of children under age 12 and follow them up at least until age 24. Ideally, high risk samples of about 500 boys and 500 girls should be followed up, with regular interviews.
Another possibility would be to add information about offending to an existing Canadian longitudinal study of children or youth (e.g., an education or health study). For example, official records of offending could in principle be collected although, if this information is collected retrospectively in adulthood, many juvenile offences may be missed, because of the purging of juvenile records.
Recommendations for Prevention and Intervention
The main recommendation is that programs that have proved to be effective should be implemented to prevent the transition from juvenile delinquency to young adult offending. These programs have been reviewed by Farrington (2017a). The most detailed review of effective interventions with juvenile offenders was completed by Lipsey (2009), who concluded that deterrent interventions such as “Scared Straight”, and disciplinary interventions such as boot camps, were ineffective or even harmful. Intensive probation or parole was moderately effective, while restorative and reparative programs were quite effective. The most effective programs were skill building programs such as cognitive-behavioural therapy and counselling programs such as mentoring.
Welsh and his colleagues (2012) reviewed effective prevention and intervention methods that have proved to reduce young adult offending. These included early interventions such as parent management training and preschool programs, as well as programs targeted on juvenile delinquents, such as Functional Family Therapy, Multisystemic Therapy, and Multidimensional Treatment Foster Care. The researchers also highlighted the effectiveness of multiple component programs including skills training, parent training, and teacher support, such as the Seattle Social Development Project and the Montreal Longitudinal and Experimental Study. Welsh and his colleagues also recommended Communities That Care, job training programs such as Job Corps, restorative justice and drug treatment programs, and they usefully included a table of benefit-to-cost ratios of intervention programs.
Lösel (2012) reviewed effective methods of correctional treatment for young adult offenders. He particularly focussed on accredited programs such as Reasoning and Rehabilitation, Enhanced Thinking Skills, and Aggression Replacement Training, and he also considered that therapeutic communities and milieu therapy were useful. Like Lipsey (2009), he concluded that boot camps and “Scared Straight” were harmful. Lösel also concluded that intensive community supervision was only effective when it was combined with some other kind of treatment, and that electronic monitoring was not effective. He recommended that effective programs should follow the R-N-R principles of Andrews and his colleagues (1990): match the offender’s risk of reoffending (R), address his/her criminogenic needs or risk factors (N), and use methods that fit his/her learning style (R). These various conclusions by Lipsey, Welsh, and Lösel should inform methods used in Canada to prevent juvenile delinquents becoming young adult offenders. However, it would be desirable to carry out a systematic review of the effects of programs specifically on young adult offending.
Implications for Justice Processing
Implications for the justice processing of young adult offenders have been reviewed by Farrington (2017a). Empirical evidence shows that there is no sharp change on the eighteenth birthday in cognitive functioning or in offending careers. Many aspects of higher executive functioning, including impulse control, planning ahead, reasoning, thinking before acting, emotion regulation, delay of gratification, abstract thinking and verbal memory, as well as resistance to peer influence, continue to mature through the mid-twenties. Most young offenders naturally “grow out” of offending in the early twenties, which is the peak period for desistance. However, adult court processing makes offenders worse; getting convicted tends to be followed by an increase in offending, juveniles who are dealt with in adult court are more likely to reoffend than other juveniles, and sending young people to adult prisons leads to an increase in their recidivism. The rehabilitative approach of the juvenile court is more successful than the retributive approach of the adult criminal court.
Currently, at least in most US states, the emphasis changes from rehabilitation to retribution and deterrence on the eighteenth birthday, and there are rarely risk/needs assessment or evidence-based programs from age 18 onwards. The book by Loeber and Farrington (2012) led to a series of recommendations about possible changes in the justice processing of young adult offenders, which were summarized by Farrington and colleagues (2012). More recently, Farrington (2017a, 2017b) reviewed later developments in the US, UK, and Europe. These possibilities include: (1) increasing the minimum age for adult court; (2) applying a “maturity discount” to young adult offenders, which involves giving them less severe dispositions if their judgment is immature; (3) establishing special courts for young adult offenders; (4) establishing special correctional facilities for young adult offenders; (5) using risk/needs assessment and screening of young adult offenders to assess risk and protective factors and executive functioning; and (6) using evidence-based programs for young adult offenders in correctional facilities and in the community after release. It would be desirable to review Canadian evidence on the development of young adult offenders, comparing males and females and Indigenous and non-Indigenous samples. It would also be desirable to pilot some of these possible changes in the justice processing of young adult offenders in Canada.
Information is needed in Canada about self-reported offending from the juvenile to the young adult years. Also, more information is needed about female offending and Indigenous offending. Canadian studies have identified a number of risk factors that predict persistence into young adult offending, including high impulsiveness, poor parental supervision, broken homes, and antisocial associates. Research in Manitoba by Bonta suggests that risk factors might differ for Indigenous compared with non-Indigenous offenders. Research on risk factors could inform the development of assessment instruments to identify which juvenile delinquents are likely to persist in offending as young adults.
It would be desirable to reanalyze Canadian longitudinal studies to advance knowledge about the transition from juvenile delinquency to young adult offending. Ideally, a new longitudinal study on the development of offending in Canada, preferably based in a Canadian city, should be mounted. Another possibility would be to collect information about offending in an existing Canadian longitudinal study of children and youth.
Effective programs should be implemented to prevent the transition from juvenile delinquency to young adult offending. New methods of dealing with young adult offenders should be piloted in Canada. Possibilities include: (1) increasing the minimum age for adult court; (2) applying a “maturity discount” to young adult offenders; (3) establishing special courts for young adult offenders; (4) establishing special correctional facilities for young adult offenders; and (5) using risk/needs assessment and screening of young adult offenders to assess risk and protective factors and the maturity of brain functioning.
Allen, M. (2016). Young Adult Offenders in Canada, 2014. Ottawa, ON: Statistics Canada.
Andrews, D.A., Bonta, J., & Hoge, R.D. (1990). Classification for effective rehabilitation. Criminal Justice and Behavior, 17, 19-52.
Carrington, P. J. (2010). Opportunities and Challenges in using Administrative Justice Data for Longitudinal and Network Research. (Proceedings of Statistics Canada Symposium 2010). Ottawa, ON: Statistics Canada.
Carrington, P. J. (2017). Personal communication, February 27.
Carrington, P.J., Matarazzo, A., & deSouza, P. (2005). Court Careers of a Canadian Birth Cohort. Ottawa, ON: Statistics Canada (Crime and Justice Research Paper No. 7).
Day, D. M. (2017). Personal communication, March 3.
Day, D.M., Koegl, C.J., Rossman, L., & Oziel, S. (2016). The Monetary Cost of Criminal Trajectories for an Ontario Sample of Offenders. Ottawa, ON: Research Division, Public Safety Canada (Research Report 2015-R011).
Farrington, D.P. (2017a). Preventing the transition from juvenile delinquency to adult crime. Journal of Community Corrections, 26(5-6), 15-16.
Farrington, D.P. (2017b). Transitions from Juvenile Delinquency to Adult Offending: A Review of Canadian and International Evidence. (Research Report 2017-R014). Ottawa, ON: Public Safety Canada.
Farrington, D. P., Loeber, R., & Howell, J. C. (2012). Young adult offenders: The need for more effective legislative options and justice processing. Criminology and Public Policy, 11, 727-750.
Koegl, C. J. (2017). Personal communication, February 27.
Le Blanc, M. (2017). Personal communication, February 28.
Le Blanc, M., & Frechette, M. (1989). Male Criminal Activity from Childhood Through Youth: Multilevel and Developmental Perspectives. New York: Springer-Verlag.
Lipsey, M. W. (2009). The primary factors that characterize effective interventions with juvenile offenders: A meta-analytic overview. Victims and Offenders, 4, 124–147.
Loeber, R., & Farrington, D. P. (Eds.) (2012). From Juvenile Delinquency to Adult Crime: Criminal Careers, Justice Policy and Prevention. Oxford, UK: Oxford University Press.
Lösel, F. (2012). What works in correctional treatment and rehabitation for young adults? In F. Lösel, A.E. Bottoms, & D.P. Farrington (Eds.), Young Adult Offenders: Lost in Transition? (pp. 74-112).Abingdon, UK: Routledge.
Lussier, P., McCuish, E., & Corrado, R. R. (2015). The adolescence–adulthood transition and desistance from crime: Examining the underlying structure of desistance. Journal of Developmental and Life-Course Criminology, 1, 87-117.
McCuish, E. (2017). Personal communication, March 21.
McCuish, E., Lussier, P., & Corrado, R. (2015). Criminal careers of juvenile sex and nonsex offenders: Evidence from a prospective longitudinal study. Youth Violence and Juvenile Justice, 14, 199-224.
Welsh, B.C., Lipsey, M.W., Rivara, F.P., Hawkins, H.D., Aos, S., & Hollis-Peel, M.E. (2012). Promoting change, changing lives: Effective prevention and intervention to reduce serious offending. In R. Loeber & D.P. Farrington (Eds.), From Juvenile Delinquency to Adult Crime: Criminal Careers, Justice Policy and Prevention (pp. 245-277).Oxford, UK: Oxford University Press.
Yessine, A. K., & Bonta, J. (2008). Pathways to Serious Offending. Ottawa, ON: Public Safety Canada (Research Report 2008-01).
Yessine, A. K., & Bonta, J. (2009). The offending trajectories of youthful Aboriginal offenders. Canadian Journal of Criminology and Criminal Justice, 51, 435-472.
For more information on research at the Community Safety and Countering Crime Branch, Public Safety Canada, to get a copy of the full research report, or to be placed on our distribution list, please contact:
Public Safety Canada
340 Laurier Avenue West
Ottawa, Ontario K1A 0P8
Research Highlights are produced for the Community Safety and Countering Crime Branch, Public Safety Canada. The summary herein reflects interpretations of the report authors’ findings and do not necessarily reflect those of the Department of Public Safety Canada.
This material may be freely reproduced for non-commercial purposes provided that the source is acknowledged.
- Date modified: