Data Sharing in Hybrid Marketplaces With Endogenous Fee Setting

We study a hybrid marketplace where a vertically integrated platform competes with a seller, who pays the platform a transaction fee, in a horizontally differentiated downstream market. The platform has a data advantage that allows it to price discriminate against consumers and can choose to share such data with the seller. First, in a setting where the transaction fee is exogenous, we show that the platform may opt to share data with the seller. However, although data sharing increases total welfare, such practice is never Pareto-improving: in particular, whenever data sharing increases the platform's profits, it also hurts consumers. Second, we show that, if the platform can endogenously choose the transaction fee, it always shares data, as it sets the fee to avoid competition with the seller, leading to higher profits and lower consumer surplus. This result is robust to a large family of consumer distributions. From a policy perspective, we show how a consumer-oriented policymaker would prefer to ban the practice of price discrimination altogether, as this policy would also level the playing field while simultaneously avoiding the surplus extraction stemming from price discrimination. Finally, we also discuss the welfare consequences of banning the hybrid business model, and we argue that the incentives of the platform to soften downstream competition to increase surplus extraction would remain qualitatively unchanged.