Safety in numbers - Data from Oakland, California

This paper adds material to the ongoing body of literature exploring the effect of pedestrian volume on safety for pedestrians. Observed and existing data for 247 intersections in Oakland, California, were used to model the effect of pedestrian volume on the number and rate (number of collisions per pedestrians) of vehicle-pedestrian collisions at or near these intersections. The study used a data set containing pedestrian collisions, average annual pedestrian volume, average annual vehicle volume, and other intersection characteristics. Estimates of the model parameters show that the number of pedestrian collisions increases more slowly than the number of pedestrians; that is, the collision rate decreases as the number of pedestrians increases, consistent with previous studies by Leden and Jacobsen. Specifically, a doubling of the number of pedestrians (increase of 100%) is associated with only a 52% increase in the number of vehicle-pedestrian collisions, with the corresponding rate decreasing by about 24%. If further research shows this association to be causal, and not due to variation in intersection characteristics or selection factors involving pedestrian choice, the implications for policy and planning involving pedestrians are substantial.