Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper studies the link between natural disasters and crime, drawing on a case study of Kenya and employing a mixed method approach. Matching data on 412 disaster locations from the Geocoded Disasters Dataset with data on over 9,500 individuals from the Kenyan Afrobarometer survey over the period from 2003 to 2014, we conduct a difference-in-differences analysis of the link between disaster exposure and experiences of crime, documenting a strong positive relationship. These findings are complemented by an analysis of primary data collected through 75 semi-structured interviews in Baringo region of Kenya which was severely impacted by the 2020 East Valley Rift flooding. The analysis of primary data confirms the positive relationship between disaster exposure and crime. To understand the possible mechanisms, we combine a descriptive quantitative analysis with qualitative content analysis. Both approaches suggest that cost-benefit considerations as well as stress induced by a disaster are likely at play in the observed patterns of increased crime in a post-disaster setting.