Linking multi-anchor bus stop accessibility to bus ridership: Evidence from GPS and accelerometer data

Abstract

Public transport provides well-documented health, environmental, and economic benefits by promoting active travel, reducing car dependence, and enabling efficient land use. However, accessibility analyses often rely on measures defined at static spatial locations (e.g., around residential or zonal anchors), which may not capture the dynamic and context-dependent nature of accessibility across daily mobility. Incorporating accessibility across daily activity locations allows for a more behaviourally grounded assessment, reflecting the spatial and contextual conditions under which travel decisions are made. This study examines how accessibility to bus stops, measured at home, trip origin, and destination, relates to bus ridership in Auckland, New Zealand. Using high-resolution GPS and accelerometer data from 108 participants (835 trips), we applied mixed-effects logistic regression models and tested work status and neighbourhood residency duration as moderators. Results revealed an unexpected pattern: higher accessibility to bus stops around home (home-based accessibility) was negatively associated with bus ridership. Accessibility to bus stops around trip destinations (destination-based accessibility) showed a positive association with ridership, but only among newer residents. Employed participants were more sensitive to home-based accessibility than those who were unemployed. These findings suggest that accessibility–ridership relationships vary across key locations encountered along individual travel trajectories, likely indicating that measures defined at fixed, static locations may provide an incomplete representation of travel behaviour. Trip-specific, dynamic approaches may therefore offer a more behaviourally grounded basis for informing transport planning and policy.