Network Governance in Relational Contracts (Job Market Paper)
This paper analyzes relational contracts in a principal/many-agent model where agents are connected to one another via a network. If the principal fails to pay an agreed-upon bonus, that information diffuses through the network, causing a larger number of agents to quit over time. This added consequence increases the sustainable level of production above what would be possible for each agent in isolation. However, this production remains lower than what is feasible with a multilateral contract (Levin 2002) where a single deviation results in the immediate loss of all agents' effort. As a result, this paper serves as a middle ground that reflects many real-world situations: employees naturally form friendships and cliques, suppliers of a large manufacturer may interact regularly with overlapping subsets of other suppliers, news in general may travel slowly, and workplace custom may discourage widespread information sharing. I analyze how the particular structure of a network affects incentive provision and profits, and detail specific examples in the two and three agent cases. My results indicate a desire to centralize unique agents (whether they be `strong' or `weak') and suggest a method for computing an optimal network.
Joint Decisions with Misaligned Information Revise and Resubmit at Theory and Decision
I analyze how the type of private information available to players affects outcomes in a joint decision making process. Information can either be aligned or misaligned. When it is aligned, it reveals to each player which alternative will lead to their highest personal utility. On the other hand, when information is misaligned, players learn which alternative would be preferred by the other player. I find that with aligned information, one player is always able to force their preferred outcome, leading to a first-mover (or second-mover) advantage that depends only on players’ patience relative to the strength of their preferences. When information is misaligned, players have conflicting incentives both to reveal and to hide their private information. I characterize all pure-strategy equilibria, and find that while equilibria are often unique when prior beliefs are extreme, under more uncertainty multiple equilibria are possible, with different levels of revelation and different welfare implications. In any given equilibrium, however, the preferences of both players never simultaneously impact the final decision.
“Introducing Strength-of-Preference into Deferred Acceptance Tiebreakers”