Planning and Know-how: The Relationship between Knowledge and Calculation in Hayek's Case for Markets

Abstract

Ludwig von Mises' calculation argument against socialism is of fundamental importance to the modern-day case for the market. Yet it is to Hayek that some Austrian-influenced theorists turn when responding to the computational models for non-market price fixing proposed by some socialists. Their reading of Hayek's epistemological argument for markets as distinct from Mises' calculation argument needs to be questioned. Hayek's emphasis upon the dispersal of knowledge across space and time is consistent with Mises' position. In spite of his philosophical critique of rationalist constructivism and his treatment of tacit knowledge, Hayek's case for the market ultimately relies upon the Misean calculation argument. Hayek's work is therefore best understood as a shift in emphasis rather than as a philosophical departure from Mises' position.

Notes

¹A little-known, early version of the argument had been offered by Pierson in 1902 and versions by Weber and Brutzkus were published contemporaneously to Mises'. Yet Mises' paper is generally agreed to be the most comprehensive statement of the ‘economic calculation argument’ against socialism (Lavoie, 1985, p. 2n).

²It is important to note however that Dickinson's proposals for socialist calculation were never entirely reliant upon the equation-solving approach. He also suggests that there would be a role for marginal price adjustment (Dickinson, 1939, pp. 99–105).

³Essentially the same point is made in ‘The Use of Knowledge in Society’ when he states that ‘as long as things continue as before, or at least as they were expected to, there arise no new problems requiring a decision, no need to form a new plan’ (Hayek, 1935a, 82) . The point is also made in ‘The Meaning of Competition’ (Hayek, 1946, p. 101).

⁴For example, the Cockshott & Cottrell model adopts the assumption that technical coefficients are known.

⁵An interesting new area of research is that of multi-agent systems; see, for example, Luck et al. (2003).