Abstract
Network effects in ridesourcing suggest fragmented (multi-platform) markets produce worse rider-driver matches than monopolistic markets. This research addresses two key gaps: (i) understanding how fragmentation costs such as longer detours for travellers, decreased driver income, and increased vehicle kilometres travelled (VKT) vary with market features and user attributes, and (ii) exploring conditions under which fragmented markets sustain. To this end, we propose an agent-based model accounting for day-to-day processes and within-day dynamics in markets with service providers offering private or shared rides. Amsterdam-based experiments reveal that two platforms can co-exist when ride-pooling is offered by just one of them, driven by differences in target demographics, or by both, due to negative network effects from detouring under limited demand. Fragmentation of the ride-pooling market yields 6.4% additional VKT. When both platforms offer (only) private rides, co-existence is feasible when a fraction of travellers and/or potential drivers engages in multi-homing.