Deterrent effects of demerit points and license sanctions on traffic violation using proportional hazard model

Abstract

The current traffic law enforcement is focused on reducing risk from human factors such as drunk driving and speeding. Among various implemented strategies, the demerit points and license sanction system are widely used as punitive and educational measure. This study has overcome the limitations of previous studies which they failed to estimate the separate effects of demerit points and license sanctions. The combined effect of both measures are assessed so it could produce more accurate and un-biased estimation and Cox proportional hazard model was used as a methodology. The used data was gained by Korea National Police Agency who mandatorily collects five years of drivers conviction records. The data included personal characteristics, demerit point accumulation and license sanction status. The regression results showed that accumulated demerit points had specific deterrent effects. Also, license revocation showed consistent and significant deterrent effects, and the effects were greater than that of suspension. The male drivers under 30s holding motorcycle were identified as the most violation-prone driver group.