On Scientific Method

Abstract

problem facing economists amounts to nothing short of explaining the state of knowledge or expectations. Economists may jib at Rutherford's (1984) assertion, but for those who define their central concern as the logic of choice it is no more than the truth. The failure of theorists to explain how economic agents can know what they are presumed to know has been at the heart of Shackle's criticisms of conventional theory; and it is to the credit of rational expectations theorists that they have taken up the challenge. Unfortunately, as Rutherford points out, their claim that agents learn the relevant true model through experience relies on an inductive theory of knowledge, the inadequacy of which was demonstrated by Hume more than two centuries ago. Post Keynesians are unlikely to lament the logical discomfiture of rational expectations theorists; yet Keynes's own attempt to establish a basis for degrees of rational belief depends on a probabilistic version of the same unjustifiable principle. Rutherford therefore concludes that all inductive approaches to knowledge be discarded and recommends Popper as a better guide. Now as Rutherford himself indicates, Popper has given much critical attention to Keynes's attempt to provide a foundation for partial knowledge (there are 25 entries under Keynes in the Index to the Logic of Scientific Discovery); but rather than examining that critique, it may be more helpful to consider where Popper's guidance might lead us.