Finding the Subway Disruption Regimes of Switching Subway to Uber in Toronto

Document Type

Journal Article

Publication Date


Subject Area

mode - subway/metro, mode - demand responsive transit, place - north america, place - urban, ridership - mode choice, ridership - behaviour, ridership - modelling


Ridesourcing, Service disruption, Subways, Travel behavior


The evolving relationship between public transit and transportation network companies (TNCs), such as Uber and Lyft, is of great interest to government agencies and has seen much critical attention in the academic literature. In this paper, we focus on the demand for TNC trips (also known as ride-hailing trips) during the disruption of the subway service. We combined a detailed dataset of Uber trips made in Toronto, Canada during the period September 2016 to August 2018 and subway disruption data provided by the Toronto Transit Commission. These data were used to examine the question: how long are subway users willing to wait during a disruption before switching modes? This question was framed as a threshold point, and an innovative structural threshold regression model was used to obtain an answer. Controlling for environmental and location-specific factors in the model, it was revealed that subway users in Toronto tend to switch to Uber after a service delay of as little as 3 min, with an average result of 7 min and an upper bound of 12 min.


Permission to publish the abstract has been given by SAGE, copyright remains with them.