Investigating the recovery of For-Hire-Vehicle, Taxi, and AirTrain at two New York City airports during the COVID-19 pandemic

Abstract

For operators of large-scale infrastructure systems such as airports, predicting landside access demand amid constant exogenous changes that modify travel volumes and preferences is immensely challenging. COVID-19 disrupted travel behavior and mode preferences, making it even more challenging for airport operators to plan infrastructure and design strategies to accommodate landside access. We use time series regression discontinuity to explore to what extent the volumes of airport ground access modes changed during the pandemic, how the recovery differed across modes, and how the pandemic changed the established relationship among access modes. We do so by examining the change in the three main airport access modes at two large New York City airports, John F. Kennedy International Airport and LaGuardia Airport (traditional taxis; for-hire ride-hailing services such as Uber; AirTrain, the public transit system, at JFK only). We find the mismatch between taxi pick-up and drop-off trips that predates the pandemic widened and persisted during the pandemic; the pandemic diminished taxis’ role and exacerbated the replacement of taxis by for-hire-vehicles. For-hire-vehicle trips recovered at both airports, possibly at the expense of taxis. AirTrain continued to function as an important and resilient airport access mode during the pandemic. Our research helps to inform the airport operators’ resource allocation and traffic management strategies while showcasing the usefulness of time series regression discontinuity in examining changes in travel behavior for future planning in the aftermath of a major shock.