About me
The Missing Bus? School Bus and Student Transit Programs in California - The school bus remains an important part of providing children a ride to school in the United States. In most of the country, nearly four in ten students rely on the school bus. But in California, less than ten percent of students take the school bus, with parental chauffeuring making up most of the difference. This lack of use stems from relatively few school districts providing school bus service to the general student population (Kapphahn and Ehlers, 2014). Scholars and policymakers alike have looked to public transit as a means of bridging this gap (Vincent et al., 2014), but research about the feasibility and observed use of this option for K-12 students is limited.
Is transit a solution for California public school students’ access to education? To understand this broader question, we consider two research questions: (1) How do the transportation service options for students and families — both traditional school bus and public transit fare pass programs — vary between school district areas in California? And (2) what characteristics predict students’ travel mode choices for trips to and from school?
This paper conducts a three-part, mixed-methods analysis of school transportation and travel in California. We rely on original data collection, the 2017 National Household Travel Survey (NHTS) California Add-On, and the Google Directions API. The Federal Highway Administration conducts the NHTS semi-regularly and asks participating U.S. households from a stratified random sample about their travel and activity patterns. The California Add-On’s confidential spatial dataset includes the home and school locations of each child ages 5 through 17 in the sample. In all, we analyze 5,182 trips taken by 2,514 students.
First, we collect data on student-oriented transit fare programs and school-provided transportation eligibility by age and distance for a stratified sample of 117 public school districts in California (roughly 10% of all districts). Second, we use NHTS trip-level data to identify school trips’ origin and destination times and geocoordinates, and then use the Google Directions API to estimate distance, duration, and complexity of trips to and from school for all potential transportation modes (private car, walking, biking, and public transit). And third, we estimate a multinomial logistic regression to predict how families choose travel modes to school. Traditionally, such models have been limited by a lack of complete choice set data — in particular, whether a school bus was available to students (He, 2015; McDonald, 2008; Speroni, 2023). To address this, we use the results from our first two analyses to test how policy actions (e.g., if a student was offered a district-provided school bus or a public transit pass) and other trip options (e.g., trip duration on transit versus the observed travel mode) influence the mode choice decision.
Preliminary descriptive findings suggest that travel behavior to and from school is unequal for students by race/ethnicity. Black and Latino/a students are more likely to walk to school than their white and Asian counterparts, despite Black students living further from school on average. Accordingly, Black and Latino/a students spend 6 to 10 minutes longer per day traveling to/from school than white students, on average.
Results of this paper can inform planners both in their domain and in areas shared with education policy.
