Criminal offenses are often associated with lower or drops in income. In this thesis, I identify the causal effect of income shocks on crime. As criminal offenses might also harm income, leading to reversed causality, an instrumental variable model is used and income shocks are instrumented by termination of employment. Using an extensive panel dataset of males in The Netherlands, I study several crime types, various sources of income and multiple job termination reasons. I find that property crimes and violent and sexual offenses are the most sensitive to shocks in earnings. By employing Bayesian techniques, I verify that the inference remains valid even if the instrument is not strictly but instead plausibly exogenous.
Keywords: economics of crime, income shocks, termination of employment, register data, instrumental variables, Bayesian econometrics