Stay or flee? Hit-and-run accidents, darkness and probability of punishment

Empirical studies on the economic theory of crime have extensively analyzed the importance of the probability of punishment with regard to premeditated criminal activities. Unplanned crimes also occur, however, and this paper will focus on a very serious and widespread example: the hit-and-run road accident. Using police records for every road accident with injuries or mortalities that took place in Italy in the period 1996-2016, we rely on changes in daylight, both when switching between daylight saving time and winter time and across seasons, as an exogenous source of variation affecting the probability of apprehension and find that the likelihood of hit-and-run conditional on an accident taking place increases by around 20% with darkness. Our results suggest that policies increasing the likelihood of apprehension could be effective in reducing hit-and-run.