While design solutions aimed at enhancing the safety of pedestrians are viewed as being incompatible with those intended to improve the safety of motorists, there has been little meaningful evaluation of the issue. Instead, this disagreement is based largely on the theoretical assertion that traffic crashes are the result of random driver error, and that the only certain means for addressing safety is to design roadways to be forgiving of these errors when they occur. This perspective overlooks the possibility that crashes may instead be the product of systematic patterns of behavior associated with the characteristics of the built environment. This study sought to discover whether urban crash incidence is the product of random error, or whether it may be influenced by characteristics of the built environment. The authors used negative binomial regression models to examine the relationship between several aspects of the built environment and the incidence of crashes involving motorists, pedestrians, and cyclists. They further subdivided motorist crashes into multiple-vehicle, fixed-object, and parked-car crashes to determine if these crash types had unique characteristics. They used vehicle miles of travel as a proxy for random error and found it to be positively, but weakly, associated with crashes involving motorists and pedestrians. We found stronger associations between crashes and characteristics of the built environment. They found miles of arterial roadways and numbers of four-leg intersections, strip commercial uses, and big box stores to be major crash risk factors, while pedestrian-scaled retail uses were associated with lower crash incidences. The results suggest that improvements to urban traffic safety require that designers balance the inherent tension between safety and traffic conflicts, rather than simply designing roadways to be forgiving. Most of the ongoing debate between pedestrian advocates and traffic engineers has focused on the relative desirability of designing urban roadways to be forgiving to random driver error. Such debates have led both groups to ignore the more salient issue of systematic error. This study finds that the factors associated with a vehicle crashing into a pedestrian and cyclist are largely the same as those resulting in a crash with another vehicle. Designs that balance the inherent tension between vehicle speeds and traffic conflicts can be used to enhance the safety of pedestrians, cyclists, and motorists alike. (Author/publisher)
Samenvatting