Effectiveness of Treatment Types
The question asked at the beginning of this Bulletin, "Can intervention programs reduce recidivism rates among serious delinquents?" has been answered. A review of the statistical findings of 200 studies found that the average intervention effect for these studies was positive, statistically significant, and equivalent to a recidivism reduction of about 6 percentage points from a 50-percent baseline, but variation in effects across studies was considerable.
Because there were relatively few studies of any one type of treatment and a range of influential variables, only tentative conclusions can be drawn from this meta-analysis. The first and most important finding is that sufficient research has yet to be conducted on the effects of intervention with serious offenders. Keeping this in mind, then, the question arises, "What types of programs are most effective for reducing recidivism?"
Again, the differences between interventions with institutionalized and noninstitutionalized offenders should be noted. For noninstitutional interventions, effects were most strongly related to the characteristics of the juveniles, especially those with a history of prior offenses. The influence of treatment type and amount was intermediate, and program characteristics were weakly related to effect size. This order was reversed for interventions with institutionalized juveniles. Program characteristics were most strongly related to the size of intervention effects; the type and amount of treatment were moderately related, and the characteristics of the juveniles were not especially important.
The specific program characteristics most closely connected with the reduction of reoffense rates of serious offenders were different for institutional programs for incarcerated offenders than for noninstitutional programs for offenders on probation or parole in the community. These characteristics did not necessarily have to do with the type of intervention; some were part of the administrative context or due to the characteristics of the juveniles treated. Therefore, a good match between program concept, host organization, and the targeted juvenile is essential.
Effective Interventions for Noninstitutionalized Offenders
The selection criteria for the 117 studies of noninstitutionalized offenders included in this meta-analysis were not highly restrictive, resulting in a range of programs in the study. The samples also varied considerably in terms of the severity of the juveniles' offense records. This allowed for some analysis of whether the interventions generally used with noninstitutionalized offenders would also be effective in reducing the recidivism of more serious offenders. The research directly addressing this question is limited, so there is no assurance that these interventions would be effective. However, this meta-analysis indicated that the intervention effects were larger for samples having greater numbers of serious offenders (with prior offenses). Also, there was little difference in the effects of interventions with respect to other characteristics of the samples (extent of aggressive history, gender, age, and ethnic mix). These two factors provide reason to believe that the interventions that are generally effective for noninstitutionalized delinquents would be equally effective with more serious offenders. Table 1 compares the effectiveness of different types of treatments for noninstitutionalized and institutionalized offenders. Treatment types are given in descending order of effectiveness.
In this meta-analysis, the types of treatment that were the most effective for noninstitutionalized offendersindividual counseling, interpersonal skills, and behavioral programswere shown to reduce recidivism by about 40 percent, a significant decrease. It is interesting to note that individual counseling appears to be an effective form of treatment for noninstitutionalized serious offenders but not for institutionalized offenders. Further examination of this discrepancy is warranted but was beyond the scope of the meta-analysis.
Following is a description of the most effective intervention programs for noninstitutionalized offenders, as represented in table 1.
Effective Interventions for Institutionalized Offenders
Of the 83 studies on interventions with institutionalized offenders examined in the meta-analysis, 74 involved juveniles in the custody of juvenile justice institutions and 9 involved residential institutions administered by mental health or private agencies. All juveniles had committed serious offenses warranting confinement or close supervision in an institutional facility.
Recidivism effect sizes for the different treatment types were most consistently positive for interpersonal skills interventions and teaching family homes. Recidivism effects for behavioral, community residential, and multiple service programs were somewhat less consistently positive. However, the small number of studies in each category makes it difficult to draw strong conclusions about the relative effectiveness of treatment types for institutionalized offenders. Using control group results from the available studies, the researchers estimated that the recidivism rate for these juveniles would be approximately 50 percent without treatment. Relative to that, the most effective treatments would reduce recidivism by 3035 percent, a significant decrease considering the seriousness of these juveniles' delinquency.
The following describes the most effective intervention programs for institutionalized offenders, as represented in table 1.
Teaching Family Homes
Community Residential Programs