The same, but different: The interactive effect of attributes of judges on judicial decision making and sentencing outcomes

Abstract

In this article, I consider whether certain attributes of a judge and, most importantly, the interaction effects of these characteristics affect the way judges decide cases and can explain disparities in sentencing outcomes. Using a supervised machine learning technique, the analysis uncovers some important effects not previously discussed in the extant literature. The results suggest that judges appointed by Republican Governors are most likely to give harsher sentences for all crimes but especially for felony and violent offences. Concerning crimes involving a high level of moral turpitude, judges get harsher as they age and men much more so than women.