Daylight saving time and crime: Does tiredness also affect criminal behavior?

Abstract

This paper explores the effect of Daylight Saving Time (DST) change on aggregate crime rates. In order to do so I run regressions with hourly data on crime and I implement a regression discontinuity design with a treatment variable centered on the beginning of DST in spring. After the DST change in spring, which implies the loss of an hour at 2 a.m., there is a significant decrease in crime. Further evidence suggests that the results are not driven by a temporal displacement of crime or changes in the level of darkness. The empirical evidence suggests that reduction in crime rates could be driven by the tiredness of potential offenders and interpreted under the lens of the Routine Activities Theory. However, the study is not able to identify the direct causal effect.