Virtual Market Design Seminar
Centralized versus Decentralized Pricing Controls for Dynamic Matching Platforms
Ali Aouad (MIT Sloan School of Management)
Abstract:
Problem Definition: Amid intensifying regulatory and public scrutiny of the control exerted by digital platforms, we investigate how delegating pricing power from the platform to suppliers---hereafter 'decentralized pricing'---affects matching efficiency and social welfare in two-sided markets. In the benchmark centralized setting, the platform dictates the market price to maximize a chosen objective. In contrast, under full decentralization, each supplier chooses a price as a function of their private cost, and customers accept or reject matches based on their private valuations. Methodology/Results: Using a fluid model of dynamic two-sided matching with strategic customers and suppliers, we uncover structural properties of the stationary market equilibria. These properties allow us to characterize the unique equilibrium under centralized and decentralized pricing. By leveraging this characterization, we find that decentralized pricing typically reduces social welfare compared to welfare-maximizing centralized price control, mainly due to supply-demand imbalances, and may even leave suppliers worse off despite having more pricing power. However, decentralization may improve welfare against centralized platforms that prioritize revenue. Decentralized inefficiencies can also be mitigated by regulating market frictions (such as waiting costs and abandonment) through operational levers, and even achieve first-best outcomes in some cases. Managerial Implications: Decentralization trades supplier flexibility for added search/matching frictions due to supply-demand imbalance and competition inefficiencies. Regulators and suppliers pushing for pricing autonomy should therefore be mindful of such imbalances to avoid unintended consequences. Despite relinquishing price controls, platforms have alternative levers---such as utilization-based minimum earnings guarantees in ridehailing, demand-shaping incentives (e.g., "wait-and-save" delivery rebates), and bonuses that discourage multi-homing---to rebalance supply and demand, and nudge decentralized markets toward efficiency. Finally, since welfare outcomes critically depend on the platform’s pricing objective, transparency regarding how platforms set prices is essential for market stakeholders.
Joint with Ömer Sarıtaç and Chiwei Yan.
Paper available here.