Incentivize Ride Matching via Dynamic Games
Other conference paper


Full Text

Other information
AbstractSurge pricing finds equilibrium price at which driver supply matches rider demand in periods of excessive demand or scarce supply. Demand serge is accommodated, however, at the cost of riders through price hikes. Fare price increase threatens a firm's market share. It allows cherry-picking drivers to exploit the abundance of demand, driving unwilling customers to competitors. The situation could be turned to an opportunity of market expansion if the firm motivates drivers at its own cost without changing fare prices. We take a game approach to analyze monetary incentives that are provided only to the supply side. By treating supply-side incentives as driver-targeted loyalty programs, we model the spatial matching problem as a Stackelberg game between drivers and the matching platform. We then connects the optimal outcomes of each round of matching game, triggered repetitively upon rider arrivals but constrained by the driver supply, to the number of cars that are vacant and the number of orders to be filled. This results in integer-valued non-Markovian dynamics governing the time evolution of the spatial aggregates of supply and demand. By characterizing the tri-party system’s equilibrium in steady-state, we show that the maximum throughput of the system is capped by the maximum matching ratio. It is achieved at the optimal amount of incentives, through which the global imbalance of the ``spatial-temporal attractiveness" across the order population is adjusted locally by assigning order-specific loyalty points. Beyond that amount, matching ratio will decline. It implies that if increasing fare price is not an option, a firm can tap the effect of supply multiplier to enlarge supply capacity in response to a demand surge. Incentivize excessively, however, would be disruptive. Subsidizing further means money is not used to promote matches but passed down directly to drivers as free income.
Acceptance Date15/05/2017
All Author(s) ListMa Shumin, Wu Qi
Name of Conference2017 INFORMS Annual Meeting Houston
Start Date of Conference22/10/2017
End Date of Conference25/10/2017
Place of ConferenceGeorge R. Brown Convention Center & Hilton Americas Houston
Country/Region of ConferenceUnited States of America
Year2017
LanguagesEnglish-United States

Last updated on 2018-23-01 at 02:48