Skip to content

Blind Justice RCT: Algorithmically Masking Race in Prosecutorial Charging Decisions

except text here

Grant Recipient: Harvard University

Principal Investigator(s): Sharad Goel, Julian Nyarko, Alex Chohlas-Wood, Todd Rogers, Charles Dorison, Joe Nudell

Term: 2023 – 2025

Funding: $635,358

Summary: A prosecutor’s decision to charge an individual with a crime can have severe consequences for that individual, their family, and their community. Despite these high stakes, there is evidence that prosecutorial decisions are racially biased, with Black, Hispanic, and other marginalized race and ethnic groups bearing the costs of this discrimination. 

As a result, it is critical that prosecutors make every effort to ensure that charging decisions are made impartially. To prevent race from influencing charging decisions, we created a blind charging” algorithm that automatically redacts race-related information from written crime reports. In response to two pilot deployments of our algorithm, the state of California passed a new law last year, AB-2778, that mandates the use of race-blind charging across the state by 2025

To evaluate the impacts of blind charging, we will run a pre-registered randomized controlled trial (RCT) with up to a dozen prosecutors’ offices in Missouri, Washington, and California. Our evaluation will study how blind charging affects charging rates for marginalized groups, including whether it causes unintended adverse impacts.