Website CV
Fields
Behavioral & Experimental, Theory
Job market paper
Coordination with Limited Shared Meaning: Experimental Evidence
How do people coordinate when they do not know which combinations of actions lead to success? We conduct a laboratory experiment in which pairs of participants try to connect by choosing among three telephones. We vary participants’ knowledge of the telephone connections and whether they can observe their partner’s past choices. First, when connections are unknown and past choices are unobserved, coordination is close to the independent random choice benchmark. Second, poor coordination arises in part because participants pursue role-dependent equilibria, which require asymmetric roles to be assigned without prior agreement. Third, knowing the connections or observing past choices improves coordination. Fourth, information acquired through a shared history improves coordination substantially more than the same information provided without a shared history. The findings suggest that participants attempt to coordinate by adopting complementary roles without first agreeing on who takes which role, contributing to poor coordination. The findings also suggest that a shared history can help participants coordinate.