Tests of the Rational Expectations Hypothesis

Abstract

This paper reviews evidence on the structure of anticipation from a number of empirical studies. Jack Muth's "rational expectations hypothesis" is compared with a variety of alternative models, including Ferber's Law, the "implicit expectations" model of Edwin Mills, and Muth's new "errors in the variables" model. The cumulative evidence is of such strength as to compel the suspension of belief in the concept of rational expectations. This implies that it is a mistake to proceed under the maintained hypothesis that expectations are rational; instead, it is necessary to test the sensitivity of empirical and theoretical results to alternative assumptions about the structure by which anticipations are generated. Copyright 1986 by American Economic Association.