Authors: Hui Kong*, Massachusetts Institute of Technology, Xiaohu Zhang, Massachusetts Institute of Technology, Jinhua Zhao, Massachusetts Institute of Technology
Topics: Transportation Geography, Geographic Information Science and Systems, Spatial Analysis & Modeling
Keywords: Vehicle Occupancy Rate, Ridesourcing, DiDi, Taxi
Session Type: Paper
Start / End Time: 4:00 PM / 5:15 PM
Room: Tower Court C, Sheraton, IM Pei Tower, Second Floor Level
Presentation File: No File Uploaded
The vehicle occupancy rate (VOR) – the percentage of time that a vehicle is occupied by one or more passengers – of ridesourcing is purported to be higher than that of sight-based street hailing taxi because of the instantaneous driver-passenger pairing through online platform. However, it remains unknown whether the higher VOR of ridesourcing is conditional on time and space. This paper, thus, investigates the DiDi and taxi services in Chengdu, China, and attempts to answer the following four questions: (1) What are the VORs of DiDi and taxi? (2) How do the VORs of the two modes vary among different drivers, over time, and across space? (3) What factors are associated with the VORs of the two modes? (4) What factors contribute to the difference of VOR between the two modes? By analyzing taxi trajectories and DiDi trips, we compared the overall VORs of the two modes. The exploratory spatiotemporal analysis was conducted to exhibit the VOR patterns of DiDi and taxi, and a regression model was applied to examine how the factors impact the VOR of DiDi and taxi and the difference in the two modes. Results reveal that DiDi achieves higher VOR, which is around 6% higher than taxi on the weekday and 12% higher on the weekend; however, the higher VOR of DiDi becomes less significant in city center where the taxi market is mature and the population density is high. The findings provide policy implications regarding matching algorithm, fare structure, and spatial regulations.