酒駕罰則是否有效抑制酒駕? 中斷點迴歸設計的應用

This study examines the effects of the statutory changes in drunk driving penalties on the incidence of alcohol-involved traffic accidents. During 2003-2013, several regulatory changes regarding the punishment for drunk driving were implemented, aiming to deter driving after drinking via raising the fine, extending the license suspension period and adding jail time, etc. Employing the regression discontinuity design, we empirically investigate whether the four statutory changes have effects on the deterrence of alcohol-related traffic accidents, leading to a lower ratio of drunk-driving traffic accidents to total road accidents. Our empirical results indicate that only the last statutory change of Article 185-3 of the Criminal Code, significantly reduced the ratio of drunk driving accidents. That statutory change also had spillover effects, lessening both the number of alcohol-related traffic accidents and the total number of road accidents. Two changes contributed to the drunk-driving reducing effects. One is the insertion of the BrAC/BAC legal level in the statute of the Criminal Code. The other is the lowering of the BrAC/BAC criteria for administrative penalties for drunk driving. Our empirical results provide implications for policy aiming to ban drunk driving.