| dc.description.abstract | My dissertation focuses on the decisions making of prosecutors, police officers, and judges. Understanding how these criminal justice actors affect crime and contribute to disparities is particularly important given the high and racially disproportionate costs of the criminal justice system.
First, I consider the effect of prosecutor bias on defendant outcomes in Chapter 1. There is little evidence on the extent to which racial bias exhibited by prosecutors is responsible for racial disparities in the criminal justice system. This paper tests for racial bias in convictions by prosecutors. To identify effects, I leverage as-good-as-random variation in prosecutor race using detailed administrative data on the case assignment process and case outcomes in a major US city. I show that the assignment of an opposite-race prosecutor leads to an increase in conviction for property crimes. I find no evidence of effects for other types of crimes.
Second, I estimate the effect of a new technology, risk assessment scores, on judicial and defendant decisions in Chapter 2. Using administrative data from a large county in Texas, we estimate the effect of a risk assessment score policy on judge bond decisions, defendant pretrial detention, and pretrial crime. We identify short-term effects by exploiting a large, sudden policy change using a regression discontinuity design. Results show that adopting a risk assessment score leads to increased release on non-financial bond and decreased pretrial detention. Additionally, we find risk assessment scores did not increase violent pretrial crime, however, there is suggestive evidence of increases in non-violent pretrial crime.
Finally, in Chapter 3 I consider how increased risk to officer safety changes police behavior. Residents of high-violence neighborhoods are concerned that police officers are failing to reduce crime, but also that when police officers do engage, their tactics are too severe. This paper examines whether risks to officer safety drive both phenomena. To do so, I exploit variation in unprovoked ambushes on police within and across beats using administrative 911 call data from a large American city. Results show that ambushes lead to a decline in arrests, but no increases in policing severity. | en |
