Opponent
Professor Giovanni Mastrobuoni, Department of Economics, Social Studies, Applied Mathematics and Statistics, University of Turin, Italien.
Grading committee
Professor Emily Owens, Department of Criminology, Law, and Society, University of California, Irvine, USA.
Associate Professor Arizo Karimi, Department of Ecnomics, Uppsala University.
Associate Professor Dawei Fang, Department of Economics, the School of Business, Economics and Law, University of Gothenburg.
Chair
Professor Randi Hjalmarsson, Department of Economics, the School of Business, Economics and Law, University of Gothenburg.
Abstract:
From Friends to ‘Family’: Schools, Neighborhoods, and Gang Recruitment
This paper studies how gang influence spreads through family and peer networks, and how exposure to peers with family connections to gangs contributes to gang expansion. I combine novel Swedish police data on gang members with administrative records on family links, schools, neighborhoods, crime, education, and health. I define gateway (GW) peers as students whose relatives have prior gang-related offending. Using male students in grade 7, I leverage cohort-to-cohort variation in exposure to GW peers in schools and neighborhoods. I find that exposure to GW peers increases the probability of being listed as a gang member. A 5 percentage point increase in exposure to GW peers with older brothers who have prior gang-related offending raises the probability of gang listing by about 14 percent relative to the mean. I find no corresponding effects for exposure to peers whose older brothers have prior criminal offending that is not gang-related. Additional evidence points to a recruitment-based mechanism. Co-offending patterns extend beyond GW peers to their older brothers and other gang members, and the effects weaken when the older brother is incarcerated, lives elsewhere, or is older than the typical recruiter age range. The effects are nonlinear and concentrated among disadvantaged students, while spillovers to female students appear in adolescent mental health rather than crime.
The Effect of Neighborhood Incarceration Rates on the Criminal Behavior of Neighborhood Peers
We use a reform that increased the time served in prison from one-half to two-thirds of a sentence to study the causal effect of neighborhood incarceration rates on the contemporaneous criminal behavior of non-incarcerated neighborhood male peers. Higher neighborhood incarceration rates have small, but significant, crime reducing spill- over effects. These crime reductions are seen for both serious and non-serious crimes and across neighborhoods of varying socioeconomic characteristics. Within the neighborhood, the effects are driven by males with a criminal history (especially co-offending histories) who are 25 or older. In contrast, crime rates for young men with no criminal history actually increase when neighborhoods are exposed to the reform. We demonstrate that the crime reducing neighborhood spillover effects of the reform are not just generated by those directly linked to individuals in prison, but also by non-linked neighborhood members solely affected by changes in the local neighborhood environment.
Criminal Background Checks Required: Firm Hiring Responses to Local Violence
This paper considers whether a firm's exposure to local violence contributes to the substantial heterogeneity across firms in their preferences and hiring policies towards workers with criminal records. Data on millions of Swedish job advertisements from 2016 to 2024 allow us to geolocate hiring establishments and systematically measure one such policy, namely whether the firm signals in the ad that a criminal background check is part of the hiring process. We use simple pre-post event and staggered difference-in-differences designs to study whether firms are more likely to signal background checks in the months after being first exposed to nearby (predominantly gang-related) shootings. For establishments within 500 meters of a violent event, there is a more than 80% increase in the propensity to screen criminal records; effects are not seen for firms outside this exposure radius. The effect on nearby firms: (i) is immediate, significant, and persistent, (ii) cannot be explained by overall reductions in hiring or changes in the types of advertised jobs, and (iii) is driven by (and even larger for) occupations in which criminal records are most prevalent among the workers. Finally, firms continue to increase their criminal record screening propensity when exposed to subsequent nearby violent events.
Link to the thesis: In the Path of Crime: Schools, Neighborhoods, and Firms
