The evolution of trust under institutional moral hazard

We study the behavior of an institution that broadcasts reputational signals to facilitate trust in a population. Using an online marketplace as a motivating example, we develop a theoretical model in which buyers and sellers are matched on a platform to engage in transactions involving moral hazard: After receiving payment, sellers may either faithfully deliver goods or renege. Although buyers do not observe a seller's true strategy-good-faith or bad-faith-the platform broadcasts binary reputation signals about sellers. Buyers condition their purchase decisions on these signals, sellers adapt their strategies over time, and the resulting market composition determines the platform's commission revenue and players' welfare. Our analysis reveals a second layer of moral hazard at the institutional level. Because revenue depends on transaction volume, the platform has an incentive to inflate ratings, making good-faith and bad-faith sellers more difficult to distinguish. This distortion is self-limiting, however: Excessive inaccuracy erodes buyer trust and collapses trade. When signal accuracy is costless, the platform maximizes profit by perfectly identifying good sellers while tolerating some false positives. When accuracy is costly, the platform has an incentive to actively erode signal quality, even at a cost. If the platform can also set commission fees, higher fees are accompanied by stronger incentives to maintain accuracy. These results clarify when institutional incentives align with, or diverge from, the welfare of buyers and good-faith sellers who rely on reputational information.