Can learnabilitysavenew-Keynesianmodels

McCallum(2009) argues that‘‘learnability’’cansavenew-Keynesianmodelsfrom indeterminacies.Heclaimstheuniqueboundedequilibriumislearnable,andthe explosiveequilibriaarenot.However,heassumesthatagentscandirectlyobservethe monetarypolicyshock.Reversingthisassumption,Ifindtheopposite:thebounded equilibriumisnotlearnableandtheunboundedequilibriaarelearnable.More generally,IarguethatathreatbytheFedtomovetoan‘‘unlearnable’’equilibrium for allbutonevalueofinflationisapoorfoundationforchoosingthebounded equilibriumofanew-Keynesianmodel