Exploring the effect of neighbouring built and demographic environment on station-level bike-sharing trips under COVID-19

Introduction As a non-motorized active transportation modes, bike-sharing can be resistant, resilient and sustainable in respond to COVID-19. Method This study aims to examine time-varying effects of neighbouring built environment and socio-demographic characteristics on station-level urban bike-sharing trips under the influence of COVID-19. Specifically, the bike-sharing data for Citi Bike stations in New York City were collected across three periods: pre-COVID-19, the early outbreak period, and the stabilization period. To completely measure the station-level bike-sharing travel, two travel demand indexes are used: amount of trips generated in a station (trip frequency) and average duration of trips (trip duration). The spatial regression approach was utilized for model building. Results The empirical results revealed that the estimated effects of neighbouring characteristics on station-level bike-sharing travel demand have changed during the three tested periods. For instance, the differences in the trip frequency of bike-sharing due to age decreased in light of COVID-19. The results also demonstrated that it is necessary to integrate trip frequency and trip duration to measure the bike-sharing travel because certain factors may have different impacts on these two indicators. Such as, during the stable period of COVID-19, the setting of bicycle lanes reduced the trip frequency but extended the trip duration. Conversely, the setting of bus stops or subway stations increased the trip frequency but reduced the trip duration. Conclusions Our findings provide practical implications for both government transportation planning and operators of bike-sharing company.