Nonlinear effects of fare discounts and built environment on ridesplitting adoption rates. (March 2023)

• Record Type: Journal Article
• Authors: Yang, Hongtai; Luo, Peng; Li, Chaojing; Zhai, Guocong; Yeh, Anthony G.O.
• Is Part Of: Transportation research. Volume 169(2023)
• Publication Date: 2023-03
• Subjects: Ride-sourcing -- Built environment -- Land use -- GAM -- Ridesplitting; Transportation -- Research -- Periodicals; 388.011
• DOI: 10.1016/j.tra.2022.103577
• Languages: English
• ISSNs: 0965-8564

Abstract

Highlights: Spatial and temporal variation of ridesplitting adoption rates was analyzed. Impacts of fare and built environment on ridesplitting adoption were explored. Nonlinear models (GAM) outperformed the linear models (linear regression). A middle-level fare discount 0.23 is effective in improving ridesplitting. Locations with high potential for ridesplitting improvement were determined. Abstract: As a new mode of shared mobility that allows users to share the same trip (vehicle) with others at a low travel cost, ridesplitting reduces environmental pollution and eases traffic congestion. Although the relationship between the built environment and the ridesplitting adoption rates has been explored before, few studies investigated the effect of fare discounts on the ridesplitting adoption rate (proportion of ridesplitting trips to ride-hailing trips) while controlling for the origin and destination characteristics. Thus, we explored this topic by analyzing the ride-hailing trip data of Chicago from January to May 2019. The generalized additive model was used to investigate the nonlinear impacts of built environment variables (e.g., population density and employment density) and travel attributes (fare discount and median trip distance) on ridesplitting adoption rates. One notable finding is that the fare discount is most effective in improving ridesplitting adoption rates when its value is around 0.23. In addition, because the trip fare is rounded to the nearest $2.50, a sensitivity analysis was performed to make sure that the approximation had a limited impact on the study results. Finally, the origin–destination (OD) pairs with a high potential for improving the ridesplitting adoption rate were identified. These OD pairs are the trips related to the airports and the trips from the north to downtown. The findings can help transportation planners and government agencies identify the areas for ridesplitting improvement and provide guidelines for transportation network companies to set appropriate fare discounts for ridesplitting.