Problem definition: Recently, there has been a rapid rise of on-demand ride-hailing platforms, such as Uber and Didi, which allow passengers with smartphones to submit trip requests and match them to drivers based on their locations and drivers’ availability. This increased demand has raised questions about how such a new matching mechanism will affect the efficiency of the transportation system—in particular, whether it will help reduce passengers’ average waiting time compared with traditional street-hailing systems. Academic/practical relevance: The on-demand ride-hailing problem has gained much academic interest recently. The results we find in the ride-hailing system have a significant deviation from classic queueing theory where en route time does not play a role. Methodology: In this paper, we shed light on this question by building a stylized model of a circular road and comparing the average waiting times of passengers under various matching mechanisms. Results: We discover the inefficiency in the on-demand ride-hailing system when the en route time is long, which may result in nonmonotonicity of passengers’ average waiting time as the passenger arrival rate increases. After identifying key trade-offs between different mechanisms, we find that the on-demand matching mechanism could result in lower efficiency than the traditional street-hailing mechanism when the system utilization level is medium and the road length is long. Managerial implications: To overcome the disadvantage of both systems, we further propose adding response caps to the on-demand ride-hailing mechanism and develop a heuristic method to calculate a near-optimal cap. We also examine the impact of passenger abandonments, idle time strategies of taxis, and traffic congestion on the performance of the ride-hailing systems. The results of this research would be instrumental for understanding the trade-offs of the new service paradigm and thus enable policy makers to make more informed decisions when enacting regulations for this emerging service paradigm.

In this paper, we study the relationship between several well known classes of discrete choice models, i.e., the random utility model (RUM), the representative agent model (RAM), and the semiparametric choice model (SCM). Using a welfare-based model as an intermediate, we show that the RAM and the SCM are equivalent. Furthermore, we show that both models as well as the welfare-based model strictly subsume the RUM when there are three or more alternatives, while the four are equivalent when there are only two alternatives. Thus, this paper presents a complete picture of the relationship between these choice models.

No abstract

No abstract