Table 1.
Summary of data filtering statistics.
Table 2.
Summary of case-specific settings.
Fig 1.
Histogram of out-of-vehicle time (a), in-vehicle time (b), nominal journey time (c), generalized travel time (d) and generalized travel cost (e) by public transport and Uber for each of the six cities. Median values are denoted by the x-axis ticks.
Fig 2.
The cumulative density function of the ratio between the nominal journey time of public transport and ride-hailing (a), and the histograms of the Modal Accessibility Gap (positive for competitive PT and negative otherwise) between public transport and ride-hailing in terms of (b) nominal journey time and (c) generalized travel cost.
Fig 3.
Zonal average nominal journey time by public transport (a) and Uber (b); the Modal Accessibility Gap between public transport and ride-hailing for the nominal journey time (c) and generalized travel cost (d); and the increase in service accessibility associated with Uber (e) for Amsterdam, Warsaw and Stockholm (above), and Washington DC, Houston and New York City (below), urban rail corridors (tram, light rail, subway, commuter train) are marked in black.
Fig 4.
Zonal Modal Accessibility Gap between public transport and ride-hailing in New York City for generalized travel cost (left) and increase in service accessibility associated with Uber (right) with a high spatial resolution.
Fig 5.
Uber demand share as a percentage of Uber and public transport demand in Washington DC, based on trip origin.
Blue lines reflect the different metro lines, and darker red colours indicate a higher Uber demand share.
Fig 6.
Scatter plots relations between the share of Uber demand (as a fraction of Uber and public transport demand) and the modal accessibility gap for nominal journey time.
Each dot represents a hex in Washington DC and its size corresponds to the total demand for public transport and Uber originating from that zone.