“An Optimal Test for Strategic Interaction in Social and Economic Network Formation between Heterogeneous Agents,” Bryan Graham, UC Berkeley

Event time: 
Thursday, February 6, 2020 - 12:00pm to 1:15pm
Institution for Social and Policy Studies (PROS77 ), A002 See map
77 Prospect Street
New Haven, CT 06511
Event description: 


Abstract: We propose a test for \textit{preference interdependencies} in the context of network formation. When preferences are interdependent the utility agent $i$ receives from directing a link to $j$ may vary with both the fixed attributes of $i$ and $j$ (some of which may be unobserved by the econometrician) as well as with the presence or absence of links elsewhere in the network (e.g., the number of indirect connections generated by an $i$-to-$j$ link). Interdependent preferences feature in many theoretical models of network formation and generate distinct policy implications \citep[e.g.,][]{Jackson_NetBook08}}.

Our null hypothesis is a composite one with a high dimensional nuisance parameter (reflecting the presence of agent-level unobserved preference heterogeneity). Under the alternative our model is incomplete due to multiple equilibria. These model features make constructing a test with good size and power properties non-trivial. Motivated by size control concerns we restrict ourselves to similar tests and derive the locally best test statistic. We characterize the exact distribution of this statistic under the null and formulate a novel algorithm for simulating from it. We present two illustrative applications. First, we test for whether nations behave strategically when choosing which countries to establish diplomatic missions in. Second, we test for whether firms prefer to sell to firms with richer customer bases (i.e., whether firms prefer more “indirect customers”. (Joint research with Andrin Pelican, University of St. Gallen)

Bryan S. Graham is a Professor of Economics at the University of California, Berkeley. He is an econometrician with research interests in network formation, the identification of peer group effects, panel data and missing data problems (including those related to causal inference). His research has appeared in a variety of journals, including Econometrica and the Review of Economic Studies. He is currently a co-editor at the Review of Economics and Statistics.

This workshop series is being sponsored by the ISPS Center for the Study of American Politics and The Whitney and Betty MacMillan Center for International and Area Studies at Yale with support from the Edward J. and Dorothy Clarke Kempf Fund.

Open to: