We investigate the impact of job loss due to firm closure on criminal behaviour. We do this using matched administrative data on the quarterly employment, income, and arrest status of all prime age males in the Netherlands between 2005 and 2015. Focusing on bankruptcies enables us to obtain causal estimates of the average effect of loss of employment on crime which do not suffer from reverse causality and other individual self-selection biases. Such job destructions can influence criminality via shocks to income and/or time availability. To disentangle between these two channels, we compare the post-bankruptcy arrest rates of various groups of individuals categorized by income level and employment status. We conclude that it is only when the two factors are experienced simultaneously – i.e. when we observe both income drops and economic inactivity – that job destruction leads to crime. |
