Judges, Juveniles and In-Group Bias

We investigate the existence of in-group bias (preferential treatment of one's own group) in court decisions. Using the universe of juvenile court cases in a U.S. state between 1996 and 2012 and exploiting random assignment of juvenile defendants to judges, we find evidence for negative racial in-group bias in judicial decisions. All else the same, black (white) juveniles who are randomly assigned to black (white) judges are more likely to get incarcerated (as opposed to being placed on probation), and they receive longer sentences. Although observed in experimental settings, this is the first empirical evidence of negative in-group bias, based on a randomization design outside of the lab. Explanations for this finding are provided