1987 Articles

Three Questions about Sunspot Equilibria as an Explanation of Economic Fluctuations

Woodford, Michael

It is by now well known that the sort of difference equations that characterize the equilibrium conditions of an infinite horizon competitive economy may have solutions in which the endogenous variables fluctuate in response to "sunspot" variables, that is, to random events that in fact have nothing to do with economic "fundamentals," and so do not directly affect the equilibrium conditions. It is possible to view such "sunspot equilibria" as a representation of an actual phenomenon economic fluctuations not caused by exogenous shocks to fundamentals, but rather by revisions of agents' expectations in response to some event, which revised expectations become self-fulfilling.
Early discussions of such solutions sometimes suggested that a more rigorous derivation of the requirements for equilibrium might yield additional restrictions that would eliminate the sunspot solutions from the set of true equilibria. The demonstration by Karl Shell (1977), David Cass (1981), and Costas Azariadis (1981) that sunspot equilibria can exist in a rigorously formulated intertemporal equilibrium model, namely the overlapping generations model of Samuelson, has shown that this is not always the case. Nevertheless, many economists remain skeptical about the reasonableness of the sunspot hypothesis as a possible explanation of actual economic fluctuations, and for quite general reasons, independent of judgments about the empirical plausibility of any particular models. I discuss here three such general reasons for skepticism.