Carsharing usage and station-level factors: A tale of ten Chinese cities

Abstract

Understanding the factors influencing carsharing usage at the station level is crucial for promoting and managing carsharing services. Previous research primarily focused on single-city studies in western countries, which yielded inconsistent evidence and limited the transferability of findings to diverse urban contexts. This study investigated carsharing usage in ten Chinese cities and examined the relationships between station-level factors and usage using both the Mixed-effects Negative Binomial Model and the Generalized Additive Mixed Model. The results show the residential land use and restaurant POI density are positively associated with carsharing usage. The non-linear effects of the built environment, transportation infrastructure, and station facilities on usage exhibit substantial variations across cities. While the numbers of parking spaces and available vehicles have in general positive influences on usage, non-linear plots reveal diminishing marginal benefits and inverted-U shaped relationships caused by over-supply in certain cities. This study offers insights into carsharing service management and operational strategies.