Disruptions as catalysts to sustainability? Long-term responses in bike-sharing demand to disruptions during the pandemic

Abstract

Understanding the implications of large-scale, prolonged disruptions on travel demand is important for informing the future design of resilient, efficient, and sustainable transport systems. Major disruptions during the COVID-19 pandemic provide an opportunity to shed light on this issue. While contemporaneous responses in such demand amidst these disruptions have been well documented, insights into long-term post-disruption responses remain limited. This research gap challenges the development of a transport policy agenda capable of adapting to and mitigating the enduring consequences of disruptions. This research contributes to this topic by scrutinising long-term responses in bike-sharing demand to major disruptions during the pandemic. It investigates (1) the characteristics of these long-term responses; (2) the discrepancies between the long-term and contemporaneous responses to these disruptions; and (3) the associations of the long-term responses with docking stations’ contextual characteristics. We use 57-month bike-sharing demand data from London, spanning the pre-, amidst-, and post-disruption phases. Utilising pre-disruption data as a baseline and data in subsequent phases as comparisons, we apply Bayesian time-series models for counterfactual analysis to assess bike-sharing demand’s responses.

We find that major disruptions during the pandemic contribute, in the long term, to a more than 20% rise in bike-sharing demand in the post-disruption phase, compared to a counterfactual scenario absent such disruptions. The increase in off-peak hour demand is greater than in peak hour demand. Demand for short- and medium-duration trips increases, whilst that for long-duration trips decreases slightly. However, despite the overall increase in demand post-disruption, the magnitude of this increase flattens over time. Moreover, bike-sharing demand’s long-term responses surpass its contemporaneous responses. Finally, docking stations located in areas with a more diverse land-use mix, higher intersection density, better accessibility to public transport, and a lower percentage of minority population show a larger long-term response in demand. Our findings remain robust while accounting for the confounding impacts of COVID-19 cases post-disruption and the implementation of active travel interventions during the pandemic. We suggest that prolonged disruptions like those during the pandemic may have functioned as catalysts for the uptake of sustainable transport, such as bike-sharing. Yet, our evidence of a diminishing long-term response over time underscores a need for persistent, proactive actions to support sustainable transport after disruptions subside, if the positive response is to be sustained.