Association for Public Policy Analysis & Management

Millions of students in the U.S. begin and end their school days on the school bus. For these students, the availability, timing, and location of bus routes critically shape their school day by determining whether, or to what extent, they arrive at school “ready to learn” in the morning and/or are able to participate in activities before school – such as school breakfast. Afternoon school bus schedules affect students’ ability to participate in after school activities and the amount of time for homework, among other things. Despite this, little is known about how the bus shapes students’ educational outcomes. In this paper, we begin to close this gap, by estimating how the length of the bus ride affects students’ outcomes. We first paint a rich descriptive portrait of bus use in New York City (NYC), including who rides the bus and for how long. Next, we derive credibly causal estimates of the effect of commute times on academic performance in math and English Language Arts, exploiting the variation in commute times generated by idiosyncrasies in routing, traffic, and eligibility. More specifically, we condition on distance to school in order to compare the outcomes of students who live equally far from school but experience different commute times. Finally, we explore the impacts on attendance and participation in school breakfast to shed light on potential mechanisms. We find key differences in bus utilization by race/ethnicity. Our key result is that increasing commute time significantly reduces academic performance.