The DRAG (Demand for Road use, Accidents and their Gravity) philosophy aims to improve understanding of the demand for road usage, and the complex interactions affecting the occurrence and severity of road accidents. An application of the DRAG approach to the Stockholm region is presented, using three time-series models with monthly data for the period 1970-75. DRAG uses a three-step approach: risk exposure, accident rate and accident severity; it also uses the TRIO statistical software package for its analyses. The study used a database for the County of Stockholm, which has a wide range of explanatory variables and other information. The exposure model gave total vehicle-km for passenger cars, and used a mathematical demand function. The model's 20 factors explained 96.5% of the total variation in total vehicle-km during the period. The correspondence between observed and expected vehicle-km was quite good. The frequency model and the severity model respectively analysed total numbers of accidents and numbers of light injuries, severe injuries and fatalities per accident; they explained the monthly variations in terms of 30 factors. It was found that the number of accidents rose much more than proportionally to the traffic rise, but that this effect was balanced by the effects of safety improvements.
Abstract