The study analyzed pretrial risk assessment tools and found they reproduce bias. The report recommends they not be used in pretrial release decisions.
The policy memo, released May 2023 from The Science, Technology, and Public Policy program at the University of Michigan’s Ford School of Public Policy, found that although many jurisdictions assume pretrial risk assessments are a fairer and more objective alternative to using cash bail, they simply embed racial and socioeconomic bias in the technical decision-making.
Their key findings include:
- Pretrial risk assessment tools overstate certainty and objectivity.
- Pretrial risk assessment tools are based on flawed and subjective data.
- Pretrial risk assessment tools promote racial and socioeconomic bias.
- Pretrial risk assessment tools overestimate the likelihood of violence and of flight before a trial.
Their conclusion is that “Risk assessment tools should play no role in pretrial
administration” as they found “substantial evidence that pretrial risk assessment tools replicate the racial and socioeconomic disparities that bail reform seeks to address.”