Keynes and Ramsey, once more
The centennial anniversary of the publication of A Treatise on Probability, John Maynard Keynes’s first investigation of uncertainty issues through the lens of probabilities, has been marked by a series of new critical assessments of this foundational work. Apart from the implications about what Keynesian uncertainty really means, the Treatise is fundamental also because it spurred, as a reaction to it, the twentieth century’s dominant view in probability theory. In his review of the Treatise, and the follow-up essay “Truth and Probability”, Frank Ramsey dissected Keynes’s logical approach proposing an alternative, subjective approach to probability that, after being elaborated by Leonard Savage, was endorsed by mainstream economic theorists. The episode is well known in the Keynesian literature, since the issue of the continuity between the Treatise and the General Theory centres mostly on Keynes’s comment on Ramsey’s criticism. But even the most recent celebratory contributions seems to show that the issue of whether Keynes conceded that Ramsey’s criticism was well placed or not still defies definite assessment. In taking into account the new recent assessments, this paper leaves most of philosophical concern aside and concentrates instead on the issues more directly related to the calculus of probability. The paper tries to show that the specific aspect of how to represent degrees of beliefs through probabilities makes it possible to highlight a difference between the two theories that could never be accommodated, neither in the Keynes-Ramsey “dispute” nor later in the works of their followers. Keynes named “non-numerical” his probabilities both in the Treatise and in his 1938 exchange with Hugh Townshend. In view of our discussion of how the representation of degrees of beliefs differs between Keynes and Ramsey, we take this as substantial evidence that he never really conceded to Ramsey.
Area: Eshet Conference
Keywords: uncertainty, decision-making, Keynes
Please Login in order to download this file