Essays on social and economic networks

This dissertation can be divided into three essays. The first essay studies status competition in networked societies. In status competition studies, the utility of heterogeneous individuals typically depends on an economy-wide average indicator of status. In our model, emulative and jealous agents are embedded in an exogenous network where agent-specific reference group is determined by the direct link emanating from the agent. We show that individual consumption is proportional to the agent's ''outbound" Katz-Bonacich network centrality measure and equilibrium is in general inefficient. The negative externality associated with each agent depends on her ''inbound" centrality measure that we call the conspicuousness index. A tax based on this index combined with a uniform lump-sum transfer can attain efficiency. The second essay examines the stability of free trade network formation with asymmetric countries that are farsighted. We apply pairwise farsightedly stable network into an equilibrium theory of trade agreement. The farsighted lens offers new perspectives on a broad range of recent studies aimed at stable network structures on free trade agreements. While global free trade is stable when countries are relatively homogeneous, a star network with smaller countries as a hub is the unique pairwise stable set under high asymmetry among countries. Welfare analysis shows that the social welfare is higher when countries are farsighted than when countries are myopic, and highlights the advantage of farsightedness. In the third essay, we construct and present a conceptual framework for analyzing how collective reputation and individual reputation jointly influence sellers' behavior in the online marketplace. The presented results demonstrate how a two-equilibrium structure that comprises the expanded equilibrium and shrunken equilibrium emerges in repeated game scenarios. Comparing online marketplace equilibria with the traditional marketplace equilibrium, it yields three segments of online-offline channels as shown in previous empirical studies.