How Justice Reinvestment States are Improving the Quality of Risk Assessment

More states than ever before are using actuarial risk assessment to determine the likelihood that people involved with the criminal justice system will reoffend. This information is critically important for developing case management plans for people in prison and on supervision, as well as to inform parole release decision making and determine the intensity of supervision and programming for people upon release from prison. While ensuring that jurisdictions use risk assessment is one cornerstone of the justice reinvestment approach, simply having an assessment tool is not enough. Assuring the effective use of a risk assessment instrument requires ongoing quality improvement steps to ensure that the tool is administered and scored consistently and that it continues to perform as designed. High-quality assessments are the result of a well-trained staff, clear scoring guidelines, regular validation studies, and ongoing quality improvement exercises, including evaluations of assessors’ scoring, random audits to ensure consistent practices, and booster training to freshen staff knowledge.

Whether using a risk assessment is new to a state or a long-established practice, justice reinvestment allows states to “get under the hood” and evaluate how well their risk assessment process is working. Following are examples of how the justice reinvestment approach has enriched states’ use of risk assessment:

Hawaii: Ensuring up-do-date risk assessment scoring and staff training

Hawaii has a long history of using risk assessment to inform case planning and supervision; however, the state’s scoring guide (instructions for how each item should be scored based on the information collected) had not been updated in years. States should check with the assessment tool developers every two to three years to learn if any scoring guidance has changed. Staff skills should also be assessed on a regular basis for training needs. And when new state practices are initiated, such as Hawaii’s first-of-its-kind swift and certain HOPE probation model, scoring rules need to be reviewed and updated as necessary to account for new practices.

After discovering inconsistencies in the scoring and administration of Hawaii’s risk assessments, the CSG Justice Center worked closely with the state to improve staff skills and enhance the risk assessment’s accuracy. Hawaii worked with the creators of their risk assessment tool to adopt updated scoring rules that reflected state-specific circumstances, and to issue a revised scoring guide. Approximately 200 staff conducting assessments were retrained statewide and successfully passed a state-based recertification process that required demonstrated proficiency on an audited assessment. The state also has trainers continually performing ongoing coaching and training sessions to improve staff assessment skills.

North Carolina: Using risk assessment to focus resources on people who are most likely to reoffend

Prior to justice reinvestment in North Carolina, supervision officers completed risk assessment on their probationers, but did not use the results to inform supervision practices. Despite research that demonstrated probationers in the state who were identified as meeting the criteria to be at high-risk of reoffense were three times more likely as a group to be rearrested within a year than low-risk probationers, probation officers supervised everyone with the same level of intensity, regardless of their identified risk level. To ensure that supervision officers not only assess an individual for risk of reoffense, but also use that information to inform supervision practices, North Carolina’s Justice Reinvestment Act requires supervision officers to assess for risk of reoffense and to provide an intensity of supervision for probationers based on their risk level. Supervision officers now focus the majority of their attention on those in the highest reoffense risk group and the state adopted a supervision grid to guide case planning and violation responses.

Idaho: Validating the risk assessment tool to ensure its continued efficacy

Before undertaking justice reinvestment, Idaho’s risk assessment had not been validated in over a decade. Local validation of an assessment tool ensures that the tool predicts reoffending in a particular jurisdiction as accurately as possible. Validation studies should be conducted regularly to monitor the continuing effectiveness of the tool, as changing demographics for the local population can reduce predictive accuracy. During the justice reinvestment process, Idaho addressed this need by requiring in statute that the risk assessment tool be validated every five years.

The proper use of validated risk assessment tools produces valuable information to help decision makers determine the appropriate level of supervision and type of programming for people under correctional control. While actuarial risk assessment should not be the sole factor in making these decisions, it is currently the best available tool for ensuring that research-based data helps inform the decision-making process. Just as regular service and maintenance enables the maximum performance of a car, regular maintenance of a risk assessment tool enables states to get the maximum benefit of risk assessment—that is, is the most effective application of resources to reduce recidivism.