Evidence-based decision-making is a priority for many justice systems, especially those in the Safety and Justice Challenge (SJC). One of the most popular of these techniques is actuarial risk assessment at the pretrial stage to help inform judicial bail decisions. According to proponents, an empirical assessment of individual defendant risk can help judicial officers identify those that can be released into the community pretrial with minimum system resources, others that require greater monitoring or supervision, and those whose release may unreasonably endanger public safety. The growing use of pretrial actuarial risk assessments has raised questions about these instruments’ potential to exacerbate racial disparities at bail setting. Further, the research still is developing on how effective these tools are at the pretrial stage. Nonetheless, many criminal justice researchers and practitioners view pretrial risk assessments as an effective way to minimize the use of jail resources.
Several justice systems have adopted public domain risk assessments such as the Virginia Pretrial Risk Assessment Instrument and the Ohio Risk Assessment System or the Arnold Foundation’s proprietary PSA-Court. There also is a mini trend within the field to create assessments based on local or state data. Localities such as Washington, D.C. and Salt Lake County, UT and states such as Virginia, Nevada, and Alaska have developed localized risk assessment instruments, a practice favored by many bail advocates.
In Shelby County, TN, officials have taken a more middle-of-the-road approach. The County contracted with Crime and Justice Institute (CJI) to validate the local pretrial services agency’s risk assessment procedure. (The agency uses a qualitative assessment based on defendant demographic and criminal history to formulate bail recommendations). Based on CJI’s findings, site officials either will choose a public domain/proprietary risk tool or develop a localized instrument. The final decision will include input from the county’s Justice System Coordinating Committee and become a strategy under SJC.
The risk validation project highlighted several challenges within the county’s collaboration effort, including participants’ ability to prioritize strategies, collect and share data, and participate in multi-agency initiatives. The project also has started local conversations about the definition of “pretrial risk,” how best to predict and manage it, and what it means to the balance of individual rights and community safety. This is a critical question, given the site’s goal under SJC of reducing the jail population—mostly pretrial detainees—by 24 percent.
CJI should be complete its validation by the end of March. The CJI team and JMI then will work with site officials to interpret findings and decide the next course of action, including how future work on risk assessment development will fit within SJC. We will report new developments through our newsletter and website.
 See Mamalian, C. A. (2011). State of the science of pretrial risk assessment. Washington, DC: Pretrial Justice Institute. Bechtel, K., Lowenkamp, C., & Holsinger, A. (2011). Identifying the predictors of pretrial failure: A meta-analysis. Final Report. Levin, D. (2011). Development of a validated pretrial risk assessment tool for Lee County, Florida. Washington, DC: Pretrial Justice Institute.