Exploring Traffic Safety and Urban Form in Portland, Oregon

Abstract

Urban form affects community development, livability, sustainability, and traffic safety. Urban planners have long assumed a relationship between urban form and traffic safety. That relationship favors designs with fewer through streets because such designs are believed to improve safety. An empirical study to explore this assumed relationship used crash data and an extensive resource of other data to define the urban form. Total reported crashes (21,492) within the city limits of Portland, Oregon, from 2005 to 2007 were aggregated by using a uniform 0.1-mi grid for the spatial unit (n = 792 cells); the crashes were modeled by using negative binomial regression to study the effect of urban form, which was defined by variables that captured street layout, exposure, connectivity, transit accessibility, demographics, and trip making (origins and destinations). These relationships were modeled separately by mode (vehicle, pedestrian, and bicycle crashes), by crash type, and by severity of crash injury. The models found that urban-form variables of street connectivity and intersection density were not significant at the 95% confidence level for vehicle and pedestrian crashes or for different levels of crash severity, in contrast to results in earlier studies. Elasticity estimates for all models were dominated by increases in vehicle miles traveled. Business density, population, and transit stops were significant variables in many models; these results underlined the importance of the design and planning of streets in determining where growth in businesses, employment, and housing will occur so that added traffic volumes can be handled safely.