A Multi-Level Analysis of Bus Ridership in Buffalo, New York

It is essential to understand how the built environment affects transit ridership to prioritize public transit and make it more appealing, particularly in mid-sized cities on the Rust Belt due to the experience of population decrease and urban sprawl in the U.S. Although many studies have looked at factors that influence ridership, there is still a need for a methodological design that considers both route and environment characteristics for bus ridership. This study examined the daily ridership of 3794 bus stops across 57 routes in the Buffalo area of New York State and used random coefficients models to account for different levels of characteristics (bus stop level, route level, and transportation analysis zone (TAZ) level). The study found that bus frequency and bus stop centrality were positively correlated with ridership, while total route stops had a negative effect. By controlling the impact of bus routes, the study showed that the multi-level design using random coefficients models was more effective than traditional OLS and spatial lag models in quantifying the impact of bus routes and TAZs. These findings provide local policy implications for route design, bus operation, and transit resource allocation.