Paths to prosperity: How transport networks and income accessibility shape retail location

Abstract

Retail location patterns in large cities are shaped by multiple factors, with transport accessibility playing a crucial role in commercial concentration. Traditional approaches often rely on proximity to transport infrastructure or network centrality, overlooking actual travel times. This study refines these methods by incorporating real travel-time data for both private and public transport to assess their influence on retail clustering in Rio de Janeiro. Using estimated travel times from Google API for private travel and GTFS data for transit networks, we analyze how retail density responds to network betweenness and gravity-based accessibility to income, given that income is likely a stronger predictor of retail activity than population density, as we show. Results from the XGBoost machine learning algorithm indicate that accessibility to income via public transport exhibits a stronger correlation with retail density than private transport, highlighting transit networks as a decisive factor in shaping commercial activity. Additionally, transit network centrality emerges as a key predictor of retail concentration, reinforcing the economic advantages of well-connected public transport services. These findings emphasize the importance of integrating transport accessibility metrics into urban planning, as they offer practical tools for guiding policy interventions. Enhancing transit coverage, frequency, and integration could support retail activity in underserved areas, reducing spatial inequalities and fostering balanced urban development. Future research should explore the role of informal retail and alternative modeling techniques to refine the understanding of transport-driven commercial patterns, particularly in cities where economic disparities and accessibility constraints pose significant challenges.