Residential development patterns and child pedestrian injuries

Abstract

A source of physical activity and independence, walking during childhood has many health and developmental benefits. Unfortunately, children are often not able to safely walk to their destinations, traffic safety being a primary barrier. The purpose of this paper is to understand how land use development patterns are related to safety outcomes for child pedestrians. Assuming that most child pedestrian trips start at home, this work explores the relationship between residential land use types and child pedestrian injuries. Are roadways in low-density residential areas more or less safe for child pedestrians than similar roadways in high-density residential areas or in mixed-use residential areas? We explore child pedestrian injury rates in Portland, Oregon through GIS spatial analysis and Bayesian spatial autoregressive Tobit models that use Markov chain Monte Carlo simulation with a Gibbs sampling scheme. Findings suggest that, with exposure and other socio-demographic factors being controlled for, roadways in mixed-use residential neighborhoods experience lower child pedestrian injury rates than high-density single-use neighborhoods or low-density single-use neighborhoods. Findings inform the Safe Systems approach and provide insight for the development of cities that ensure safety for vulnerable road users.