According to Swedish penal code, there is a “rebate” on all prison sentences before the 21st birthday. We exploit this age discontinuity to investigate how individuals respond to harsher punishments. We use a large Swedish dataset, including dates for all crimes which led to convictions for cohorts born during the period 1973–1993.
We find evidence of “bunching” in the sense that more crimes were committed during the week prior to a 21st birthday, followed by a reduction in crime during the week after this birthday. We do not, however, find that harsher punishment reduces the crime rate permanently.
