Weight of Argument and Economic Decisions
The Treatise on Probability (Keynes, 1973a ; henceforth TP), published in 1921, is a crucial reference to understanding in depth the General Theory (Keynes, 1973b ; henceforth GT) published a few years later (1936). This nexus has been neglected or downplayed for many decades. Only recently has its importance been fully recognized (among the early contributions we mention Carabelli, 1988, and O'Donnell, 1989). However, its importance and scope remain quite controversial.
In this chapter I intend to clarify the controversial concept of 'weight of argument' as introduced in TP and explicitly resumed in crucial passages of GT, in order to assess its influence on the theoretical framework and methodological approach of GT. To this end, I will carry out a preliminary examination of whether, and for what reason, we should expect that the weight of argument has a significant impact on economic decisions. As is well known, Keynes himself admitted being puzzled over the importance to be attributed to this 'somewhat novel' concept (TP, p. 77); in particular, he admitted on a few occasions to be uncertain about its 'practical significance' (TP, pp. 83, 345, 348), also because 'it is difficult to think of any clear example of this' (TP, p. 83) that exemplifies and corroborates its importance. As a matter of fact the concept of weight of argument was not really new when Keynes wrote the TP. Keynes himself refers to writers on probability who at the end of the nineteenth century explicitly, although briefly, raised the question, in particular Meinong and Nitsche (TP, pp. 84-5). From the theoretical point of view, Keynes's contribution is not much more than a reappraisal of the concept within the framework of his own theory of probability. What is really new is his tentative application of the concept to the explanation of economic decisions in GT. In our opinion Keynes's innovative method suggested in GT cannot be properly understood without reference to the crucial role played in it by the weight of argument.
As is well known, in GT Keynes explicitly refers to the weight of argument in crucial passages of his reasoning showing its significant impact on economic decisions. This provides a crucial example of its practical role, which was missing in TP, and shows why the weight of argument should play a crucial role in any satisfactory account of how a monetary economy works. Unfortunately the passages of GT referring to the weight of argument have long been neglected by followers and interpreters, for many reasons. One reason is the difficulty of providing an operational definition of the weight of argument and integrating it within analytic or econometric models. Another has been the emergence since the early 1930s of decision theories under uncertainty that are characterized by rigour and operational power (Ramsey, 1931; de Finetti, 1937; von Neumann and Morgenstern, 1944; Savage, 1954); these theories have deeply influenced the foundations of economics, including mainstream Keynesian macroeconomics after the death of Keynes, in such a way to exclude any possible role for the weight of argument. A few recent developments in epistemology and decision theory under uncertainty have reopened the issue, providing at the same time new analytical instruments capable of translating Keynes's intuitions in rigorous and operational instruments.