Published online by Cambridge University Press: 04 January 2017
Lewis and Schultz (2003) develop a statistical signaling model to deal with international conflicts or bargaining situations in which states have private information about their payoffs. They claim that they can confirm that there always exists a unique equilibrium in their model. In this paper, I show that Lewis and Schultz's claim is not true and their model admits multiple equilibria under some parameter settings. Monte Carlo analysis shows that when there are multiple equilibria, the parameter estimates may not converge to their true values even if the number of observations increases.