COVID-era built environment and travel: Insights from location-based services data

Abstract

This study revisits the linkage between land use interventions and travel behavior in the COVID era using increasingly available cell phone-based individual mobility data. Reducing the carbon emissions associated with personal vehicle travel is crucial to achieving climate targets; policies such as California's Senate Bill 375 require that land use planning achieve climate targets at the regional level. The implementation relies heavily on local placemaking efforts such as higher density infill development and walkable streets which have long been considered potential reducers of automobile travel. However, the rise of telework, decline of transit, and increase in pedestrian deaths following the pandemic have cast doubts on the efficacy and cost-effectiveness of strategies seeking to foster low-Vehicle Miles Traveled (VMT) location patterns. This study uses StreetLight Insight data on vehicle trip origination at the census tract-level before and after the emergence of COVID-19 (2019 and 2021) to assess the contribution of several built environment measures to VMT and to the share of short trips in the 6-county Southern California region. Despite concerns over COVID-induced changes, we find that several built environment measures remain solidly associated with travel efficiency in multivariate models investigating VMT levels, VMT rebound, and the share of trips that are shorter than two miles. While the prevalence of neighborhood-scale destinations is an activity generator, it also fosters shorter trips, and a region-level measure of job accessibility provides some evidence that more populated areas nearer the region's core did indeed struggle to return to pre-COVID activity levels. After the first year of the pandemic, VMT rebound was most pronounced in tracts with a high share of residents under the age of 18, suggesting that while many adults did not return to prior activity patters (e.g. due to telecommuting), children mostly did. Findings suggest that local policies and placemaking efforts, including 15-minute communities, may still be promising trip reducers, while near real-time data provides a mechanism for far faster performance evaluation.