Fair prediction with disparate impact: A study of bias in recidivism prediction instruments
This addresses fairness issues in criminal justice risk assessments, which is a critical societal problem, but it is incremental as it builds on existing fairness criteria without introducing a new solution.
The paper tackles the problem of discriminatory bias in recidivism prediction instruments by analyzing fairness criteria, demonstrating that these criteria cannot all be satisfied when recidivism prevalence differs across groups, and showing how disparate impact arises from failing to meet error rate balance.
Recidivism prediction instruments (RPI's) provide decision makers with an assessment of the likelihood that a criminal defendant will reoffend at a future point in time. While such instruments are gaining increasing popularity across the country, their use is attracting tremendous controversy. Much of the controversy concerns potential discriminatory bias in the risk assessments that are produced. This paper discusses several fairness criteria that have recently been applied to assess the fairness of recidivism prediction instruments. We demonstrate that the criteria cannot all be simultaneously satisfied when recidivism prevalence differs across groups. We then show how disparate impact can arise when a recidivism prediction instrument fails to satisfy the criterion of error rate balance.