Non-linear relationships between built environment characteristics and electric-bike ownership in Zhongshan, China

Although electric bikes (e-bikes) proliferate in China and other developing countries, few studies examine the association between the built environment and e-bike ownership. Moreover, the association is often assumed linear in variables. This study adopts a semi-parametric model to relax the linearity assumption and explores the complex relationships between built environment variables and e-bike ownership in Zhongshan, China. We found that almost all built environment characteristics have non-linear associations with e-bike ownership and the non-linear patterns vary by variable. Within certain thresholds, distance to transit, employment density, and land use mix are positively associated with e-bike ownership whereas residential density has a negative association. Furthermore, distance to city center shows an inverse V relationship with e-bike ownership. We recommended planners to deliberate land use and transportation policies to promote e-bike growth, but not facilitate reliance on cars.