Who should pick the meeting point? Efficiency–fairness trade-offs between operator and users in ridepooling

Meeting points (MPs) consolidate passenger pick-up and drop-off (PUDO) in ridepooling at pre-designated locations, improving fleet efficiency while shifting part of the access and egress burden to passenger walking. This raises a question: who selects each trip’s PUDO MPs? In the operator-centric regime, the platform dynamically selects both MPs using vehicle schedules and fleet state to minimize a candidate vehicle’s remaining-schedule completion time. In the user-centric regime, each passenger selects an MP pair that minimizes weighted walking and estimated travel time, thereby controlling where they walk. These differing objectives and information create a trade-off between routing and service availability and the burden on served passengers. Although prior work establishes the operational benefits of MPs, it has not systematically quantified disagreement between operator- and user-preferred selections, its spatial distribution, or its cost consequences. We address this gap by formalizing party-specific cost functions and pick-up and drop-off disagreement, quantifying system-level trade-offs, and measuring cost discrepancies through counterfactual analysis that holds the realized vehicle schedule fixed while substituting one party’s preferred MPs with the other’s. We evaluate the framework across three structurally distinct cities under three MP placement strategies: grid-based, greedy coverage, and multi-objective optimization. Operator-centric selection improves service coverage, whereas user-centric selection reduces average walking distance per PUDO leg by roughly a factor of three. A decomposition of MP selection and the matching objective shows that who selects the MPs, rather than whose cost drives matching, primarily determines this trade-off. Across cities and placements, drop-off disagreement exceeds pick-up disagreement, disagreement is spatially associated with larger cost discrepancies, and replacing user-preferred MPs imposes a substantially larger relative burden on users than the converse imposes on operators. Low-willingness-to-walk passengers bear the largest user-side discrepancy. These findings establish a pervasive, spatially structured, and asymmetric conflict that supports zone-level regulatory screening and targeted service design.