Inaccessibility in Public Transit Networks

TL;DR

The paper applies network-science methods to infrastructure accessibility, constructing separate accessible and full networks for London and New York City to quantify how step-free stations are distributed and how central they are within transit systems. By analyzing degree distributions, closeness, and betweenness centralities, the study reveals persistent gaps between accessible and total networks, with accessible stations often forming focal hubs in central areas but not achieving comprehensive coverage. The results show scale-free degree patterns and reveal correlations between station centrality, income, and tourism, underscoring equity and policy implications for transit planning. Overall, the work demonstrates how network-theoretic tools can inform design and prioritization of accessibility upgrades in major urban transit systems.

Abstract

The study of networks derived from infrastructure systems has received considerable attention, yet the accessibility of such systems, particularly within public transit networks, remains comparatively underexplored. Accessibility encompasses a broad range of considerations, from infrastructure-based features such as elevators and step-free access to spatial factors such as the geographic distribution of accessible stations. In this work, we investigate infrastructure-based accessibility in two major transit systems: the London Underground and the New York City Subway. We construct network models in which nodes represent accessible stations and edges represent adjacency along transit lines. Using tools from network analysis, we examine the structural properties of these accessibility networks, including clustering patterns and the spatial distribution of accessible nodes. We further employ centrality measures to identify stations that serve as major accessible hubs. Finally, we analyze socioeconomic and tourism-related variables to assess the influence of neighborhood wealth and popularity on the prevalence of accessible stations. Our findings highlight significant disparities in accessibility across both systems and demonstrate the utility of mathematical and network-theoretic methods in understanding and improving modern transit infrastructure.

Figures (12)

• Figure 1: Subgraphs of the ten nodes with the highest (a) betweenness and (b) closeness centralities. Colored pieces of the nodes represent the lines that stop at them and size of the node is based on the degree of the node in the whole accessible network.
• Figure 2: Number of stations in both the total network and the accessible network for the London Underground with (a) betweenness and (b) closeness centralities sorted.
• Figure 3: (a) Betweenness and (b) closeness centrality values compared to degree of the nodes for the London Underground network. Trendlines are added to visualize correlation between these centrality measures and degree.
• Figure 4: Degree distribution of the London accessible and total network in logarithmic scale with power-law trendlines.
• Figure 5: Median annual income, per thousand pounds (£), per borough and the number of accessible stations in each borough. The accessible stations in our top ten list of betweenness and closeness centralities and their median annual income are indicated in orange and yellow, respectively.