Rational Expectations in the Macro Model

THE ANTICIPATIONS of households and firms played a central role in Keynes' General Theory, and in the thinking of every macro theorist since. My purpose in this paper is to examine the major new issues about anticipations raised by the recent explosion of theoretical and empirical work based on the theory of rational expectations. In the General Theory, anticipations were taken, in general, as irrational in the sense to be defined below. Because they existed in the mind, anticipations were analyzed in psychological terms. They were determined by the "animal spirits" of businessmen, by speculators' guesses as to how other speculators would behave, by waves of optimism and pessimism. Changes in anticipations were held to be frequently or even usually selffulfilling. After World War II two developments led economists away from reliance on psychologically determined anticipations. First, the effort to build and estimate quantitative models of the business cycle involving expectational variables forced acceptance of the idea of anticipations functions formed on observable data, because without them such models could not be estimated and empirically tested. It was natural to argue that anticipations of