Reducing Racial Disparities in Bail Decisions: An Experiment on Public Accountability and Feedback in Texas

Can measures that increase accountability and transparency reduce racial inequality in bail judges’ pretrial decisions?

Racial disparities exist at every stage of the U.S. criminal justice system and are particularly prominent in the setting of bail, a critical early decision point. These racial disparities in bail are even larger for the youth population. Prior research shows that racial disparities in the bail system are largely driven by racial bias and inaccurate stereotypes. In other settings, providing information has been shown to be highly effective at improving the provision of public services, but there is very limited practice or evidence of public accountability of criminal justice actors. Similarly, individualized feedback is a standard practice in fields like education and medicine but is mostly missing from the current pretrial system. Thus, this study tests the effectiveness of a two-pronged intervention to reduce racial disparities at the pretrial stage, specifically for the population of offenders under the age of 25. The first part of the intervention consists of public report cards providing statistics on each judge, such as the number of individuals detained and average cash bail amounts, and racial gaps in these metrics. In collaboration with the Texas Criminal Justice Coalition (TCJC), Dobbie and colleagues will conduct a randomized controlled trial with the approximately 50 bail judges in felony cases in Harris, Bexar, Travis, and Dallas counties in Texas. Each 1-2 months, they will send judges in the treatment group an email with a link to a TCJC website where their public reports are posted and a download including individualized feedback and personalized tips. The team will track whether the judge clicks on each link and downloads the materials to measure take-up of these materials. Both during and after the intervention, the team will obtain publicly available administrative data on pretrial release decisions and misconduct outcomes at the case level. These data will also include judge and defendant identifiers, defendant demographics, and additional case and defendant details such as the crime type and prior criminal history.