Exploratory analysis of recent trends in school travel mode choices in the U.S. Abhay Lidbe a, ⁎ , Xiaobing Li a , Emmanuel Kofi Adanu a , Shashi Nambisan b , Steven Jones c a Alabama Transportation Institute, Box 870288, University of Alabama, Tuscaloosa, AL 35487, United States of America b Transportation Research Center, Box 454007, University of Nevada, Las Vegas, Las Vegas, NV 89154-4007, United States of America c Department of Civil, Construction & Environmental Engineering, Box 870205, University of Alabama, Tuscaloosa, AL 35487, United States of America ABSTRACT ARTICLE INFO Article history: Received 6 February 2020 Received in revised form 20 May 2020 Accepted 1 June 2020 Available online xxxx The study explores the recent trends in school travel using the 2017 National Household Travel Survey data. The study also investigates the exogenous factors affecting the school travel mode choice using random parameters multinomial logit (RPMNL) model. The results indicate that urban school trips range between 3 and 5 miles, whereas, average rural trips are longer than 6 miles. School commute times are higher among lower-income households. Further, the share of school bus and auto has declined while that of walking and biking has increased in 2017. This change is significant among high school students. Like other studies, the findings of the RPMNL model confirm that students within shorter distances from school are more likely to walk or bike to school. However, the likelihood of riding a school bus for dis- tances >15 miles is higher than that of auto, indicating a policy implication to support school transportation budgets, especially in rural school districts. Lower-income households have a higher likelihood of riding the school bus. Females are more likely to use a car and less likely to bike to school. Interestingly, households with more than three vehicles are more likely to use the school bus compared to no-vehicle households. Children living in rented houses are less likely to ride the school bus or car. Also, an increase in gas price is indirectly but positively linked with walking, biking, and auto use. The findings from this study will assist policymakers in formulating policies and planning decisions towards improvements in the current school travel trends. Keywords: 2017 NHTS data School travel Trip mode choice School bus Walk Bike 1. Introduction There is a growing interest in the school travel modes used by elemen- tary and secondary school children. In the 2014–15 academic year, about 50 million elementary and secondary students were reported to attend school daily with 55% of them transported by school buses. With approxi- mately $24 million spent on school transportation, this translates to about $932 per pupil in public expenditure (Snyder et al., 2019). The number of children that walked or biked (active modes) to school has been on a declin- ing trend over the past few decades (Ham et al., 2008; McDonald et al., 2011; Rothman et al., 2018) while children traveling to school in cars has been increasing (Sirard and Slater, 2008). Such shifts in school travel modes have raised many legitimate concerns for the policymakers, includ- ing its contribution to childhood obesity (Centers for Disease Control and Prevention, 2018; Larouche et al., 2014) and environmental problems such as pollution (EPA, 2003; Sirard and Slater, 2008). Besides, different school travel modes are likely to affect children's academic achievements differently (Yeung and Nguyen-Hoang, 2019). Further, school travel mode choices can also have longer-term implications for children as they transition into adulthood. Indeed, studies have shown that travel mode choices made by adults are likely influenced by the travel modes used in their childhood (Johansson, 2004; Schlossberg et al., 2006). National Household Travel Survey (NHTS) is one such data source that provides rich information on school travel modes used by children from households with different socio-demographic backgrounds. The current literature lacks a comprehensive study that explores and analyses the re- cently published 2017 NHTS data in ways similar to the 2001 and 2009 NHTS data (Ham et al., 2008; McDonald, 2012, 2007; McDonald et al., 2011). Considering the importance of understanding the trends and pat- terns of school travel in developing policies, this study seeks to explore and document the recent trends in school travel and travel modes (walk, bike, auto, school bus, and others) using the latest 2017 NHTS data (U.S. Department of Transportation; Federal Highway Administration, 2017). The study specifically focuses on differences among rural and urban areas and how they relate to household income levels. Whereas previous studies have used a range of traditional discrete choice modeling techniques to explore school travel mode choice from the NHTS data (Kontou et al., 2020; McDonald et al., 2011; Mehdizadeh et al., 2017; Mitra et al., 2010; Moudon and Lin, 2011; Stone et al., 2014; Sultana, 2019), the current Transportation Research Interdisciplinary Perspectives 6 (2020) 100146 ⁎ Corresponding author. E-mail addresses: adlidbe@ua.edu, (A. Lidbe), xli158@ua.edu, (X. Li), ekadanu@crimson.ua.edu, (E.K. Adanu), shashi@unlv.edu, (S. Nambisan), sjones@eng.ua.edu. (S. Jones). http://dx.doi.org/10.1016/j.trip.2020.100146 2590-1982/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). Contents lists available at ScienceDirect Transportation Research Interdisciplinary Perspectives journal homepage: https://www.journals.elsevier.com/transportation-research- interdisciplinary-perspectives