Identification and Estimation of Large Network Games with Private Link Information

ABSTRACT We study the identification and estimation of large network games in which individuals choose continuous actions while holding private information about their links and payoffs. Extending the framework of Galeotti et al., we build a tractable empirical model of such network games and show that the parameters in individual payoffs are identified under large‐market asymptotics in which the number of individuals increases to infinity on a single large network. We then propose a semiparametric two‐step M‐estimator for these individual payoffs and demonstrate its good finite‐sample performance in simulations.