Essays on networks and markets

This thesis consists of three essays on the theory of networks and its applications to markets. The first essay develops a new model for studying repeated games in buyer-seller networks, and explores the connection between network structure and cooperation. The second essay introduces new techniques to analyze the effect of Word-Of-Mouth (WOM) on cooperation in networked markets. The third essay studies how network structure affects the timing of hiring in networked markets. Consider a large market with asymmetric information, in which sellers choose whether to cooperate or deviate and 'cheat' their buyers, and buyers decide whether to re-purchase from different sellers. In the first chapter I model active trade relationships as links in a buyer-seller network and suggest a framework for studying repeated games in such networks. I derive conditions that determine whether a network is a Steady State Cooperation Network (SSCN)--a network that is consistent with trade and trust between every buyer and seller that are connected. In particular, three network features increase the incentives for cooperation: sparseness, moderate competition, and segregation. In the second chapter, co-authored with David Goldberg, we ask how is the ability of buyers and sellers to cooperate in different market structures affected by WOM? We allow for the presence of networks that capture the transmission of information between buyers. We find that WOM facilitates dense buyer-seller SSCNs. Surprisingly, WOM may limit the competition in a market, leading to potential welfare losses. However, such losses disappear in large markets. The third essay studies how network structure affects the timing of hiring in networked markets. In a model of local unraveling (early hiring) in labor markets, information about workers' productivity is revealed over time and transmitted via a network of connections between firms and workers. Although employment begins after workers finish their training, employment contracts can be signed earlier. Consistent with existing evidence, unraveling reduces mobility of workers. Unraveling increases with the network's span, and with the quality heterogeneity of firms that compete locally, but decreases with concentration around firms. Surprisingly, unraveling increases and then decreases with network density.