Real world agents, REH agents and the econometrician

Abstract

This paper aims to explain from within mainstream theory why incorporating the rational expectations hypothesis in new classical macroeconomics leads to stable market equilibria, while other branches within the mainstream end up demonstrating the impossibility of such equilibria. It argues that Rational Learning describes a learning process where the object of learning, the set of equilibrium prices for example, is unwittingly taken as given data (and so is part of the decision-making kit for the agents from the beginning of the learning process), whereas real life decision making in the markets does not have the benefit of this knowledge.