Search

On March 22, 2020, the State of New York issued a "stay-at-home" policy, wherein all non-essential businesses were on pause until June 8, 2020. The bike-sharing system (BSS) and yellow taxi system (YTS) in Manhattan were substantially affected. This sudden drop in demand can impact not only short and long-term mobility but also the sustainability of transport network. Given that few empirical studies are focusing on the impacts of the "stay-at-home" policy on the BSS and YTS, this further substantiates the importance of analyzing how the policy affects the overall transportation system in New York City (NYC). This paper aims to fill this gap by quantifying the impacts of the "stay-at-home" policy on the two aforementioned transportation systems. Specifically, the following three research gaps are summarized in this study: I) The hidden biases in current "stay-at-home" policy estimation methods were not properly addressed; II) The policy impacts on BSS and YTS during different periods of the effective day were unclear; III) The sensitivity of uncontrolled confounders in long-term policy impact estimations was poorly discussed. We addressed these important research gaps by introducing robust statistical approaches like regression discontinuity design (RDD) and propensity score matching (PSM) methods, which can overcome methodological challenges such as counterfactual restoration, spatiotemporal heterogeneities, and unmeasured confounders. The BSS and YTS were studied at the aggregated neighborhood levels. Results demonstrate that the impacts on BSS have higher variations than YTS usage. The monthly average treatment effects on the treated (ATT) for BSS ranged from −72% to −28% respectively in March and June, while YTS ranged from −96% to −94%. Evidence suggests that demand for BSS surged on weekends in May and June. Understanding the impact of this short-term yet significant policy change on travel behavior will help optimize supply and demand management strategies, thereby improving the long-term sustainability should similar situations arise in the future. • Quantifying the short-term and long-term impacts of the "stay-at-home" policy. • Data-driven approaches using regression discontinuity design and propensity score matching method. • The bike-sharing system and yellow taxi system have different sensitivities to the policy. • Policy implications are derived from the robust analyses. [ABSTRACT FROM AUTHOR]