Treating Youth in Conflict with the Law: A New Meta-Analysis

1. Introduction

There is a substantial, and still growing, body of literature on the effectiveness of treatment in reducing delinquency in youth.   The results of these studies, however, are rather conflicting when examined individually.   Numerous studies demonstrate that treatment programs have significantly improved the recidivism rates of youth while others provide contradictory results.   Antonowicz and Ross (1994) argue that the literature attests to the fact that "some rehabilitation programs work with some offenders in somesettings when applied by some staff" (p. 1).   In an attempt to develop a more definitive understanding, previous researchers have utilized meta-analytic techniques as a method of aggregating the results of the numerous studies (Andrews et al., 1990; Cox, Davidson & Bynum, 1995; Dowden & Andrews, 1999; Garrett, 1985;   Izzo & Ross, 1990; Latimer, 2001; Lipsey, 1995; Lipsey & Wilson, 1988; Whitehead & Lab, 1989; Wilson & Lipsey, 2000).   With the exception of one study (Whitehead & Lab, 1989), these meta-analyses have concluded that treatment has a positive overall effect on reducing recidivism amongst youth.   

Researchers also sought to identify which types of programs were most effective at reducing recidivism.   Andrews et al. (1990), found that the most effective treatment programs were those that adhered to the principles of risk, need and responsivity.   That is to say, programs that structured their intensity based upon the risk level of the youth (risk), programs that targeted criminogenic factors related to recidivism (need), and programs that matched the mode of treatment to the learning styles of the clients (responsivity) yielded the highest mean reductions in recidivism.   Lipsey (1995) found that programs longer than six months or programs that provided lengthy exposure to treatment (more than 100 hours of direct participation) were more successful than shorter programs.   In addition, treatment was found to be more effective when delivered in the community versus an institutional/correctional setting (Andrews et al., 1990; Garret, 1985).

These findings, however, have been based upon a large body of literature that dates back to the 1960s and 1970s where youth delinquency included what has traditionally been labelled status offences (e.g., truancy, promiscuity).   In addition, a large proportion of the meta-analytic literature examined youth who were considered 'at-risk' to engage in delinquent behaviour.   As such, it is important to disentangle the potential effects of status offenders and at-risk youth from the results to determine what program characteristics are effective with youth who are already engaged in criminal behaviour.   Finally, existing meta-analyses have defined 'youth' quite liberally and have included programs that primarily targeted young adults (18 to 25 years of age).   Children under 18 have long been recognized, however, as having special needs, requirements and vulnerabilities related to their developmental maturity and therefore require a separate criminal justice system (Bala, 2003).   Including older youth in the results is not consistent with the underlying rational for separate youth and adult systems.

This paper provides a new meta-analytic synthesis of the literature to date in order to examine the effectiveness of treatment for youth.   The central goal of this project is to communicate to program developers, policy-makers and key decision-makers in the youth criminal justice system a clearer understanding of the program characteristics that decrease the likelihood of recidivism among young offenders.   The present meta-analysis also addresses the limitations of previous research by only including studies that examined youth under the age of 18 years who had committed what would constitute a criminaloffence using current adult standards.  

Date modified: