This paper investigates the effects of school starting age on crime while relying on variation in school starting age induced by administrative rules; we exploit that Danish children typically start first grade in the calendar year they turn seven, which gives rise to a discontinuity in...