Did liberalising bar hours decrease traffic accidents?

Legal bar closing times in England and Wales have historically been early and uniform. Recent legislation liberalised closing times with the object of reducing social problems thought associated with drinking to "beat the clock." Indeed, using both difference in difference and synthetic control approaches we show that one consequence of this liberalisation was a decrease in traffic accidents. This decrease is heavily concentrated among younger drivers. Moreover, we provide evidence that the effect was most pronounced in the hours of the week directly affected by the liberalisation: late nights and early mornings on weekends. This evidence survives a series of robustness checks and suggests at least one socially positive consequence of extending bar hours.