Time saved, but for whom? An intersectional analysis of travel-time differences for on-demand transit users in five countries

On-demand transit (ODT) has the potential to both cannibalize and complement public transit networks. Users switching to ODT away from alternative transportation modes may benefit from travel time savings, but uncertainty remains about how these savings vary by mode and by population group. We present a multi-country, multi-city analysis of ODT’s travel-time performance compared with car, public transit, cycling and walking modes in 33 locations, across the USA, Canada, Norway, Germany and Spain. Leveraging a unique dataset of 6,963 digitized trips that pairs actual ODT trip characteristics with travel survey responses, we derived realistic counterfactual alternative travel times using a publicly available routing algorithm. Our aim was to investigate how travel-time differences (TTD) between ODT and alternative modes differ across various socioeconomic and trip characteristics. We paid particular attention to the intersectional nature of these characteristics, to bring empirical nuance to our findings. We found that ODT plays a dual role as a trip-level substitute: a convenience substitute for the car on short trips (especially in North America), and a time-saving substitute for public transit on long trips (especially in Europe). However, multilevel modelling revealed significant inequities in the distribution of TTD: in our pooled USA and Canada subsample, men experienced 1.6 min higher TTD (i.e. greater time savings) than women across all alternative modes, White users experienced 10.4 min higher TTD than Black/African American users when the alternative was public transit, and high-income users experienced 4.4 min higher TTD than low-income users when the alternative was a car. We also uncovered an intersectional ‘disability penalty’, whereby the benefits accrued by men and high-income users were significantly diminished when those individuals also reported a disability. Beyond these substitution patterns, we also found that disabled riders were 1.85 times and low-income riders 3.4 times more likely than their counterparts to report taking a trip that would not have been made without ODT, indicating that ODT generates net new mobility for the most disadvantaged users. While ODT is a promising solution for addressing many challenges in public transit, our findings challenge the assumption that it inherently promotes equity. This underscores the urgent need for policymakers to apply an intersectional lens to service design, to prevent the entrenchment of existing mobility disparities.