Multi-Dimensional Equity Constraints

The increasing availability of transport customer data has led to the thriving of data-driven decision-making algorithms in the transport industry. Emerging mobility service providers (MSPs), such as ride-hailing and Mobility-on-Demand (MOD) operators, widely engage in price discrimination and uneven vehicle dispatching tactics, by leveraging customers’ historical trip data and personal characteristics. Empirical studies have found that neighborhoods with larger non-white populations and lower incomes are associated with higher ride-hailing service prices and fewer vehicles, indicating discriminatory treatment to travelers.

We propose to design effective policy instruments to balance efficiency and equity. The research activity will include:

1. To develop a joint discriminate pricing and mobility resource allocation model for multiclass travelers under multi-dimensional equity constraints, and evaluate the impact of equity constraints on MSP, travelers and society.

2. To develop a game-theoretical model to measure the impacts of equity constraints under asymmetric competition and design policy instruments tailored for heterogeneous MSPs.

3. To design an equity-aware dynamic pricing algorithm with limited information about demand and conduct numerical experiments to demonstrate its performance.