Keynes, Harrod, and the Rational Expections Reveolution

Abstract

In the past fifteen years, economists' modeling of expectations formation processes has undergone what some have called a revolution. The rational expectations framework has replaced adaptive expectations in much theoretical and empirical work. In the minds of many economists, this revolution finally places macroeconomics on firm microfoundations, relegating the vestiges of Keynesian theory to a special case of equilibrium models with externally imposed nominal rigidities. However, the rational expectations revolution has not adequately considered the work of Keynes himself or the research of the post-Keynesians. While the rational expectations school emphasizes the ultimate stability of market processes, Keynes and the economists who claim to follow his work most closely conclude that expectation formation plays a destabilizing role in a market economy. This paper analyzes how rational expectation techniques affect the structure and results of a dynamic Keynesian output determination model based on Harrod's (1939) growth theory. I show that standard rational expectations arguments do not stabilize the model. In fact, some of the considerations suggested by the rational expectations approach make the model less stable. The assumption that agents form expectations rationally is not sufficient to overturn Keynesian results. The argument usually made in favor of rational versus adaptive