Community Structure and Market Outcomes: A Repeated Games-in-Networks Approach

Consider a large market with asymmetric information, in which sellers have the option to "cheat" their buyers, and buyers decide whether to repurchase from different sellers. We model active trade relationships as links in a buyer-seller network and study repeated games in such networks. Endowing sellers with incomplete knowledge of the network, we derive conditions that determine whether a network is consistent with cooperation between every buyer and seller that are connected. Three network features reduce the minimal discount factor sufficient for cooperation: moderate and balanced competition, sparseness, and segregation. Incentive constraints are binding and rule out efficient networks. (JEL C73, D82, D85, Z13)