This study aims to examine how crime in Swedish municipalities is influenced by a social disruption. To investigate this, the storm Gudrun and the 2018 forest fires were selected. A difference-in-differences (DiD) model was used, with the treatment group consisting of affected municipalities and the control group of unaffected ones. Relative DiD, DiD regression with control variables, and DDD regressions were conducted. Data on the natural disasters were analysed separately in IBM SPSS Statistics.

The results show a complex, context-dependent effect of the disasters. Descriptive statistics indicate that reported crime decreases in both the treatment and control groups in the year following the disaster for several crime categories. A relative decrease in violent crime was observed for both disasters, though not at a significant level. A positive relationship was found for burglary and theft, with varying levels of significance. Socioeconomic factors modified the effects in the DiD regressions for several crime types and showed within-group variation. The most significant control variable for within-group variation was unemployment, where higher unemployment was associated with a reduced disaster effect.