UC Santa Barbara

This dissertation consists of three essays that each explore the United States criminal justice system. In the first chapter I evaluate the impact of "Ban the Box" Laws on rates of criminal recidivism. Despite their goal of increasing ex-offender employment and reducing recidivism, several recent studies of "Ban the Box" (BTB) policies have cast doubt on BTB's efficacy at improving ex-offender employment outcomes. Evidence of BTB's effect on criminal recidivism, however, remains limited. Using administrative prison data, this chapter examines the direct effect of BTB policies on rates of criminal recidivism. I find that, while BTB policies don't appear to reduce criminal recidivism in the aggregate, they may be exacerbating racial disparities. In particular, I show that being released into a labor market with a BTB policy is associated with higher rates of recidivism for black ex-offenders, with young black ex-offenders being particularly affected. In contrast, older white ex-offenders seem to benefit from the policies. In the second chapter I estimate the effect of electoral pressure on the sentencing behavior of prosecutors in California. Prosecutors in the United States wield immense discretionary power over the outcome of criminal cases. Despite this, there has been relatively little research concerning the effect that electoral cycles might have on their sentencing behavior. Conventional wisdom dictates that prosecutors will likely pursue harsher sentences on average, in an attempt to appear "tough on crime". To test this, I construct a novel dataset of California prosecutors and electoral outcomes. Using criminal sentencing data, I then estimate the impact of electoral pressure, as measured by electoral proximity and competition, on criminal sentencing. I find that electoral pressure is associated with a decrease in the average severity of criminal sentences for serious violent crimes. Then, using data from the San Francisco District Attorney's Office, I provide evidence that this effect can be explained, in part, by prosecutors engaging in charge bargaining. Finally, in the third chapter I estimate the impact of changes in local television broadcast news on criminal sentencing. The local television broadcasting industry in the United States has undergone significant consolidation over the past two decades, leading to the rise of large national media conglomerates. It is unclear, however, what impact consolidation will have on the quantity, quality, and content of local news broadcasts. This chapter uses the rapid expansion of one of the largest media conglomerates, Sinclair Broadcast Group, on the sentencing behavior of local criminal justice officials. I find that Sinclair's entrance into a media market is associated with 2.55% decrease in the average sentence length for serious violent crimes. Heterogeneity analyses show that this effect is concentrated among black defendants, and is primarily driven by sentences for robberies and aggravated assaults. Similarly, I show that this effect is almost entirely driven by counties that select their judges via elections rather than appointments. Taken together, these results suggest that a reduction in local crime coverage associated with Sinclair's acquisition limits voter information concerning local judges, reducing their incentive to appeal to voters by imposing harsher sentences.