An examination of children’s school travel: A focus on active travel and parental effects

Abstract Concerns regarding children’s health and physical activity levels have prompted a growing interest in understanding the school mode choice decision. Walking and biking to school can help school-age children achieve recommended daily physical activity levels. The growing literature in this area has provided valuable insights but has often failed to adequately consider the influence of parents on children’s decision making. This study aims to provide additional insights by examining the school mode choice with a focus on active travel and the role of parental effects. A multinomial logit model was estimated for children’s school mode choice using the 2009 National Household Travel Survey with the add-on sample for Texas. A comprehensive set of variables, including individual and household characteristics, built and travel environment measures, and parental characteristics and attitudes, were found to be influential on the decision of active travel to school. Although the focus of this study is on children’s school mode, a model examining parent’s active travel was also estimated to provide a picture of parents and their own active travel decisions through a Poisson count regression model for the parents’ non-motorized travel frequency (weekly). Findings strongly suggested that mothers’ concerns regarding traffic conditions in the neighborhood can discourage children’s active mode use. In contrast, fathers’ work flexibility seemed to facilitate more non-motorized school travel. Similarly, many factors were important in determining non-motorized trip frequency. A flexible work environment, particularly for the mothers, was observed to increase the overall trip frequency of parents. The results from this study corroborate the findings from several earlier studies while also indicating new factors not previously identified. The work done in this study can be extended in several directions. One particularly interesting avenue of future research is to model the two decisions in a joint model to identify the causal direction while accounting for unobserved factors.