Automated record clearance corrected historical bias by removing procedural barriers

Evidence

Recently, two California counties teamed up with Code for America to automate the clearance of more than 50,000 marijuana convictions that had been made eligible for clearance under a newly enacted proposition (Ready, 2019). An algorithm combed through government files to identify eligible individuals and file the motions to clear. The automation was thorough: the person with the criminal record did not have to file a petition, hire an attorney, or attend a hearing—everything was completed by the machine (Ready, 2019). Because African Americans were overrepresented in this population, one can say that the technology was working to correct for historical bias. — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law, 1171 (#expands, via linked claim)

In a criminal justice context, a history of treating one demographic group differently creates redundant encodings all over the place. Policing a minority neighborhood more aggressively than others, for instance, means that all of a sudden something as seemingly neutral as length-of-rap-sheet, i.e., number of previous convictions, can become a redundant encoding of race. — Brian Christian, The Alignment Problem, 296 (#complicates, via linked claim)

Categorization

• XREF: connects to: Automated record-clearing algorithms corrected racial disparities from past convictions. — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law: A Definitive Look at…, 1171; Legal AI can be deployed to identify and correct existing bias in law — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law: A Definitive Look at…, 1223; Prior biased policing leaks into race-neutral variables like conviction history — Brian Christian, The Alignment Problem, 296 (+1 more)

Broader Connections

• Automated record-clearing algorithms corrected racial disparities from past convictions. #expands (0.93) — Joseph Avery, Jim Sidanius, Joel Cooper, Gregory Davis, Amanda N., Bias in the Law, 1171
• Prior biased policing leaks into race-neutral variables like conviction history #complicates (0.84) — Brian Christian, The Alignment Problem, 296