Stochastic Equilibrium in Hybrid Taxi Markets: Coupling E-hailing and Street-hailing under Information Friction and Bounded Rationality

Abstract The evolution of mobile internet technology has transformed urban taxi systems into hybrid markets where traditional street-hailing and app-based e-hailing coexist. However, existing literature often relies on assumptions of perfect rationality and perfect information, overlooking the perception errors and bounded rationality of drivers in dynamic search processes. This paper proposes a stochastic equilibrium model for hybrid taxi markets that explicitly incorporates information frictions and driver bounded rationality. First, we establish a coupled dual-market framework: the street-hailing market is characterized by a modified Cobb-Douglas matching function to capture spatial search frictions, while the e-hailing market is modeled using queueing theory and digital matching mechanisms. Second, we introduce a Logit-based stochastic perception model to endogenize drivers' dynamic choices between zones and operation modes (cruising vs. e-hailing), reflecting suboptimal decisions caused by information asymmetry. Finally, numerical simulations are conducted on a standard test network to investigate equilibrium properties. The results reveal that the introduction of e-hailing does not guarantee system-wide efficiency gains; intense competition from e-hailing can exacerbate search frictions in the street-hailing market through a "cannibalization effect." Furthermore, the study demonstrates diminishing marginal returns to information transparency, where excessive information availability may trigger "herding behavior" and localized congestion. These findings provide novel theoretical insights for regulation and pricing strategies in mixed-operation taxi ecosystems.