Essays in Microeconomic Theory and the Economics of Networks PDF Download

Author: Yiqing Xing
Publisher:
ISBN:
Category :
Languages : en
Pages :

Get Book Here

Book Description
This thesis consists of three essays in microeconomic theory and the economics of networks. Chapter 1 establishes a model in which agents first form risk-sharing pairs, and then repeatedly share income risks under limited commitment. Agents of different occupations differ in income autocorrelations, i.e. how their current incomes correlate with past ones. I show that agents with high autocorrelation are hard to share risks with. With endogenous matching, two equilibrium outcomes can occur: either 1) agents match positive assortatively, or 2) agents from different occupations do match together, but in order to sustain such matches, agents share risks unevenly favoring the relatively less autocorrelated. Either equilibrium features substantial inequality across occupations and low total welfare, compared to what would happen if a social planner could impose an optimal matching to agents. The interplay between matching and risk sharing can change our views on policies. For instance, uniform increases in everyone's low income levels (minimal wages) may hurt some agents. Increases in occupation-specific common income shocks could improve overall risk sharing and reduce inequality. My results are also robust to forms of heterogeneity other than income autocorrelations, such as heterogeneous opportunities to rematch or migrate. This model is among the first attempts to consider both partnership formation and subsequent interactions. I highlight the point that how a pair of agents interact (share risks in this case) does not only depend on the two of them, but also on their potential links to others, and how others interact. Such a framework has important policy implications because a policy can change how agents interact as well as whom they interact with. Without considering these effects, out policy evaluations could be inadequate. Chapter 2 (coauthored with Matt Jackson) is an online experiment to justify homophily, the tendency of people to interact with others that are similar to themselves, by the ease of coordination among agents with a similar cultural background. In particular, we examine whether people are better at predicting how others with similar cultural backgrounds will behave, compared to others with different cultural backgrounds. We also explore whether this translates into better coordination. The more than a thousand participants in our experiment mainly reside in two countries: India and the United States. Participants are paired to act in a simple coordination environment with multiple coordination outcomes. Participants from India are much more likely than participants from the U.S. to choose actions that lead to very unequal payoffs across the two subjects. We also find that, although participants residing in different countries tend to choose different actions, they do not seem to adjust their actions according to their opponents' place of residence. One explanation for this pattern is that participants have no idea about what their opponents would do when the opponents are from a different cultural background from them, and wrongly believe that their opponents will behave similarly to themselves. This explanation is consistent with the data when we explicitly elicit participants' beliefs about the opponents' behaviors. In sum, due to the accuracy of predicting each others' behaviors, interactions between people who share a similar cultural background leads to a larger likelihood of coordination, and a higher payoff on average. Chapter 3 (coauthored with Matt Jackson and Hugo Sonnenschein) models negotiations that determine not only an agreement's price, but also its content, which typically has many aspects. We model such negotiations and provide conditions under which negotiation leads to efficient outcomes, even in the face of substantial asymmetric information regarding the value of each aspect. With sufficient information about the overall potential surplus, if the set of offers that agents can make when negotiating is sufficiently rich, then negotiation leads the agents to efficient agreements in all equilibria. Furthermore, the same negotiation game works regardless of the statistical structure of information - in this sense, no omniscient "planner" or "mechanism designer" is required. The theory and examples explore the anatomy of negotiation and may shed light on why many situations with significant asymmetric information exhibit little inefficiency. This chapter is within my research agenda of better understanding the social costs of asymmetric information without an omniscient and empowered "mechanism designer". Such a designer plays a key role in the mechanism design literature, but frequently is absent in applications. This chapter asks the question that whether two agents come about on their own, negotiating in "free-forms", can achieve (near) efficiency. Another paper of mine, "Intermediated Implementation", (with Anqi Li) is along the same line of research. There we ask the question whether a social planner can implement target allocations through market intermediaries (e.g., firms in the labor market) with simple policies such as per unit fee, labor income tax, or quota system.