Why individuals choose dockless bike sharing services?

Abstract

This study investigates dockless bike sharing service (DBS) users’ behavior; particularly, why individuals choose DBS. A latent segmentation-based logit (LSL) model is developed using data from a DBS user survey conducted in Kelowna, Canada. The model is developed considering the following reasons for choosing DBS: cheapest option, fastest option, exercise purposes, recreational purposes, parking constraint, and unavailability of other modes or other reasons. The LSL model captures unobserved heterogeneity by assigning individuals into discrete latent segments. The model is estimated for two segments. Results suggest that segment 1 can be identified to include older, lower-income, frequent bike rider, females; whereas, segment 2 includes higher income, younger, non-frequent bike user, males. The parameter estimation results suggest that built environment attributes such as bike index, land use diversity index, transit accessibility, density of destinations, and length of bike infrastructure might influence the choice of DBS. The model confirms significant heterogeneity across the segments. Individuals residing in mixed land use areas with longer active transportation infrastructure are more likely to use DBS for recreational purposes in segment 1. In contrast, higher-income individuals in segment 2 show a negative relationship. The elasticity effects suggest that transit accessibility, length of bike lanes and cycle tracks, dwelling density, and vehicle ownership reveal substantial impact in segment 1. In contrast, bike index, land use diversity index, and personal vehicle ownership reveal significant impact in segment 2. This study offers important behavioral insights; specifically, the heterogeneity addressed in this research needs to be accommodated within the policy-making for efficient operation and expansion of DBS.