Unveiling and Mitigating Bias in Ride-Hailing Pricing for Equitable Policy Making

• DOI: 10.48550/arxiv.2301.03489
• Publication Date: 2024-05-29
• Authors: Nripsuta Ani Saxena; Wenbin Zhang; Cyrus Shahabi
• Keywords: FOS: Computer and information sciences; Computer Science - Computers and Society; 11. Sustainability; Computers and Society (cs.CY); 0502 economics and business; 05 social sciences; 1. No poverty; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology

AbstractRide-hailing services have skyrocketed in popularity due to their convenience. However, recent research has shown that their pricing strategies can have a disparate impact on some riders, such as those living in disadvantaged neighborhoods with a greater share of residents of color or residents below the poverty line. Analyzing real-world data, we additionally show that these communities tend to be more dependent on ride-hailing services (e.g., for work commutes) due to a lack of adequate public transportation infrastructure. To this end, we present the first thorough study on fair pricing for ride-hailing services by first devising applicable fairness measures to quantify this bias and then proposing novel fair pricing mechanisms to alleviate this bias. We present two pricing mechanisms to provide flexibility and account for different platform needs. By taking affordability into account and potentially providing discounts that may be government-subsidized, our approaches result in an increased number and more affordable rides for the disadvantaged community. Experiments on real-world Chicago ride-hailing data demonstrate worse scores for the proposed fairness metrics for rides corresponding to disadvantaged neighborhoods than those of a control group (random mix of neighborhoods). Subsequently, the results show that our fair pricing mechanisms eliminate this inequality gap. Our mechanisms provide a basis for the government and the ride-hailing platforms to implement fair ride-hailing policies.