Keynes, Knight, and Fundamental Uncertainty: A Double Centenary 1921–2021

ABSTRACT

John Maynard Keynes’s Treatise on Probability (1921) and Frank Knight’s Risk, Uncertainty and Profit (1921) independently stressed the distinction between insurable risk and uninsurable fundamental uncertainty (in Knight’s terminology), inspiring two literatures that have engaged with each other only intermittently. I explore the relationship between what Keynes wrote about uncertainty in his Treatise and what Knight published the same year, and consider what contributions by Truman Bewley and by Kiyohiko Nishimura and Hiroyuki Ozaki on Knightian uncertainty and Knightian decision theory have to offer for furthering understanding of Keynes on uncertainty, particularly as suggesting a response to Alan Coddington’s critique of the risk/uncertainty distinction.

KEYWORDS:

• Keynes on probability
• Frank H. Knight
• fundamental uncertainty
• Knightian decision theory

JEL CLASSIFICATIONS:

• B31 history of economic thought: individuals
• D8 information
• knowledge
• and uncertainty

Acknowledgement

I am grateful for helpful comments to an anonymous referee and to Anna Carabelli, John Smithin and other participants in a Review of Political Economy webinar, February 26, 2021.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1 Moggridge (1992, p. 594) quotes Keynes on the two main conclusions of Knight’s review, ‘namely, that my book caused him intense irritation, and that he had great difficulty in understanding it.’ In his presidential address to the American Economic Association at the end of 1950, Knight (1956, p. 252) declared that ‘The latest “new economics,” and in my opinion rather the worst for fallacious doctrine and pernicious consequences, is that launched by the late John Maynard (Lord) Keynes, who for a decade succeeded in carrying economic thinking well back to the dark age; but of late this wave of the future has happily been passing.’ See Burgin (2009), Emmett (2009), Emmett’s entry on Knight in Emmett (ed., 2010), and Knight’s essays (Emmett, ed., 1999).

2 On these influences see Kyburg (1995) and the entries in Dimand and Hagemann (eds., 2019) by John Davis on Moore, Russell, Wittgenstein, and Ramsey, by Rod O’Donnell on A Treatise on Probability, and Sheila Dow on risk and uncertainty.

3 On Knight (1921), see McKinney (1977), Nash (1998), Rakow (2010) and especially Emmett (2009, 2020). Szipro (2020) provides a recent account of three centuries of economic analysis of risk and uncertainty since Daniel Bernoulli’s statement of the St. Petersburg paradox.

4 Although Keynes’s 1910–1913 lectures on money are published in Keynes (1971–89, Vol. XII), his 1910 lectures on company finance and the stock exchange, discussed by Carabelli (2002), are in an unpublished notebook in the Keynes manuscripts in King’s College Library, UA/6/3.

5 In keeping with his 1910 lectures, Keynes (1921, pp. 22–23) held that ‘underwriters themselves distinguish between risks that are properly insurable, either because their probability can be estimated between comparatively narrow numerical limits or because it is possible to make a “book” which covers all possibilities, and other risks which cannot be dealt with in this way and which cannot form the basis of a regular business of insurance, – although an occasional gamble may be indulged in.’ Cf. Keynes (1937, p. 214) on situations of fundamental uncertainty where ‘there is no scientific basis on which to form any calculable probability whatever.’

6 The philosopher Hugues Leblanc (1962) attempted to relate those two conceptions of probability to each other, arguing that the statistical probabilities of von Mises and Neyman and the inductive probabilities of Keynes and Carnap both ‘may be treated as sentence-theoretic measurements and that the latter qualify as estimates of the former’ (p. vii) but this argument has not found acceptance from either camp.

7 In contrast, Knight (1921, p. ix) opened his preface with the claim that ‘There is little that is fundamentally new in this book. It represents an attempt to state the essential principles of the conventional economic doctrine more accurately, and to show their implications more clearly, than has previously been done. That is, the object is refinement, not reconstruction.’ Notwithstanding such striking differences between the personalities of the authors, the central messages of the two books shared a pathbreaking emphasis on the limitations of knowledge that make uncertainty irreducible to risk.

8 See also Keynes to Townshend, 7 December 1938, in Keynes (1971–1989), Vol. XXIX, pp. 293–94.

9 G. L. S. Shackle’s work on limited knowledge and his advocacy of potential surprise as a replacement for probability was unusual in drawing inspiration explicitly from both Keynes and Hayek. Shackle (1938) was originally a 1937 PhD dissertation at the London School of Economics, where Hayek was teaching.

10 Although Nishimura (1992) and Bewley (1999) sought microeconomic foundations for New Keynesian macroeconomics, their writings on Knightian uncertainty do not mention Keynes (1921; 1936, Chapter 12; 1937) as talking about fundamental uncertainty but only Knight (1921). New Keynesian economists evidently do not read Keynes so that, for example, the relative wage analysis of Keynes (1936, Chapter 2) can be independently reinvented as the Taylor staggered wage-setting model.