Text

theses

Even when economic agents have common interests, it can be difficult for them to coordinate on a commonly-accepted outcome.
I use both human subject experimentation and theory to shed light on how coordination takes place and where and why it might be prone to fail.
In a human subject experiment, I find that the coordination problem leads to inertia in groups --- once coordination has taken place, agents avoid trying to change the outcome --- and that, with exogenous uncertainty and differing risk preferences, the less risk-averse agent tends to defer to the more risk-averse agent, though in a state of potential loss, conflicting notions of risk-aversion can make coordination even more difficult.
In the second chapter, I construct a theoretical model of storable consumption goods, in which markets for goods with high income-elasticity are shown to be prone to ``runs'' in which consumers, in the face of otherwise adequate supply, store up for a shortage, thereby creating the very shortage they feared; it would be in everybody's interest to avoid this outcome, and yet avoiding it requires a great deal of mutual trust and coordination.
In the final chapter, an information-theoretic elaboration of this idea is presented, in which agents form robust but diffuse beliefs about what other agents believe other agents believe, and I note that cutting off the chain of reasoning at the second level favors ``risk-dominant'' equilibria over ``risk-dominated'' equilibria, but can also favor non-equilibrium outcomes in situations in which equilibrium strategies are only best-responses to fairly precise actions by other agents.

ETD_9064

Ph.D.

Includes bibliographical references

by Dean Jens

doi:10.7282/t3-fpca-4b52

ETD doctoral

The author owns the copyright to this work.