What question did this study set out to answer?

This research aims to evaluate the accuracy and fairness of recidivism prediction algorithms compared to human judgment.

January 5, 2018Open Access

The accuracy, fairness, and limits of predicting recidivism

Key Points

This research aims to evaluate the accuracy and fairness of recidivism prediction algorithms compared to human judgment.
Analyzed predictions from the COMPAS software and compared it to predictions made by individuals with minimal criminal justice expertise.
Developed a simple linear predictor using only two features to compare accuracy with COMPAS.
COMPAS did not show improved accuracy or fairness compared to predictions made by non-experts.
The simple linear predictor was found to be nearly equivalent to COMPAS despite utilizing only two features.

Abstract

Algorithms for predicting recidivism are commonly used to assess a criminal defendant's likelihood of committing a crime. These predictions are used in pretrial, parole, and sentencing decisions. Proponents of these systems argue that big data and advanced machine learning make these analyses more accurate and less biased than humans. We show, however, that the widely used commercial risk assessment software COMPAS is no more accurate or fair than predictions made by people with little or no criminal justice expertise. We further show that a simple linear predictor provided with only two features is nearly equivalent to COMPAS with its 137 features.

Bookmark

View Full Paper