Reducing the Harm: Identifying Appropriate Programming for Low-Risk Offenders
Lowenkamp, Christopher T., Smith, Paula, Bechtel, Kristin, Corrections Today
Much of the focus of programming evaluation research has been on the impact of recidivism rates for the high-risk offender. With limited funding and resources, it is necessary to direct services to the group that can potentially demonstrate the largest percentage decline in recidivism. Although this is certainly appropriate, there is a need to identify the types and amounts of treatment and programming, if any, that may benefit the lower-risk offender. Specifically, the basis for this decision-making follows the risk principle.
More than 15 years ago, Andrews, Bonta and Hoge presented the concept of the risk principle. (1) Since then, there have been multiple studies and meta-analyses that have demonstrated support for the risk principle. Simply put, this principle suggests that an offender's risk level should dictate the types of services he or she receives, the dosage needed, and the amount of supervision required to reduce the likelihood or risk of recidivism. (2) Ideally, an offender's risk level should be determined by an actuarial risk and needs assessment that has been validated and normed on the targeted population. In addition, treatment target areas should be identified based on the criminogenic needs that are indicated through an actuarial risk assessment.
Programs that implement such practices have begun to recognize that lower-risk offenders have either been referred to or court-ordered to correctional treatment programming for services that may be more appropriately developed for a higher-risk individual. Subsequently, research has indicated that intensive treatment and supervision for low-risk offenders has increased this population's recidivism rates. (3) Given this negative implication, the following study offers some preliminary findings based on a meta-analysis of the existing research that has examined how programming has impacted the lower-risk offending population, and it identifies which services, if any, would minimize the harm.
Why Use the Risk Principle?
As previously stated, there is empirical evidence to support the risk principle. In particular, the overall finding of a meta-analysis conducted by Andrews and Dowden demonstrated a 19 percent decrease in recidivism when programs adhered to the risk principle; yet, when programs deviated from the risk principle, the recidivism rate increased 4 percent. (4) When examining the intensity of services, one study's findings showed that intensive rehabilitation supervision resulted in a 17 percent increase in the recidivism rates of the lower-risk offenders. However, the higher-risk offenders in this same program experienced a 20 percent reduction in recidivism. (5) Findings from a large halfway house study suggested that intensive programming for higher-risk offenders decreased recidivism by 10 percent to 30 percent. Yet, these same programs consistently increased recidivism for the lower-risk offenders. (6) To summarize, these findings suggest that intensive programming and supervision may be appropriate for a high-risk offender but not a low-risk offender. Further, these results may indicate that combining the different risk levels in programming could potentially increase the recidivism rates for the lower-risk group.
Evaluating Research With Meta-analysis
There are several benefits of choosing to conduct a meta-analysis to address this topic and to synthesize the existing research. First, this technique standardizes the review process through the use of a coding guide that is completed for each eligible study. Each item on the coding guide is intended to capture a study's important features that potentially could impact the overall effect size. For example, in the current study, three important features, or variables, were coded--the risk level of the group being examined, the types of services being evaluated, and the dosage of treatment and programming. Second, the final result of the meta-analysis is calculated into one number, called the effect size. …