Disaggregate Propensity Study on Red Light Running Crashes Using Quasi-Induced Exposure Method

Abstract

Red light running (RLR) is a serious traffic hazard at signalized intersections. A number of studies have been undertaken to understand the characteristics and prevalence of highway users involved in RLR crashes. However, most of the studies did not consider crash exposure systematically. To overcome the difficulties in obtaining direct exposure information, this study made use of the quasi-induced exposure method to identify the significant driver-vehicle characteristics that influence RLR crash propensity. This was done by developing models of relative exposure for intersections and relative crash involvement ratio for RLR crashes. Using crash data in Singapore from 1993 to 2002, a disaggregate analysis identified a number of significant factors contributing to RLR crash propensity for different driver age and gender groups, and vehicle types under various road environments. The results show that the problem related to RLR crashes is not just related to driver or vehicle or the road environment but a combination of these factors. Arising from the results of this study, potential countermeasures in reducing the high RLR crash risk for specific highway user cohorts are discussed and recommended.