Why Hayek Abandoned the Average Period of Production?

Download full paper

This paper is closed for comments.


Between 1933 and 1936, Frank Knight and Friedrich Hayek engaged in a spirited dispute about the Austrian Theory of Capital through an exchange of articles that revealed Knight’s criticism and Hayek’s defense of the concept of the Average Period of Production (APP). The controversy began with attacks by Knight (1933) in his Capitalistic Production, Time and the Rate of Return, to which Hayek (1934) responded with On the Relations between Investment and Output. Later, Knight (1934, 1935) continued with his “Capital, Time and the Interest Rate” and Professor Hayek, with his Theory of Investment. Finally, Hayek (1936) accepted in The Mythology of Capital problems that the concept of the APP caused the Austrian theory of capital, stating that he would never use it.

Posted for comments on 11 Aug 2017, 9:53 am.

Comments (3)

  • Avi J. Cohen says:

    The stated argument of the paper is that, “Unlike Cohen (2003), this work will show in detail how the arguments of Knight persuaded Hayek of the inappropriateness of the APP as an instrument of general equilibrium analysis of the process of aggregate capital accumulation, in both equilibrium and comparative statics.”

    But Cohen’s (2003) conclusion includes:

    Hayek also recognizes the measurability problems associated with any period of production, due to the influence of the interest rate on the calculation of the length of any production period. Outside of a model with a single homogeneous input and one-commodity output, Hayek (1941, 141–42) freely acknowledges that “all attempts to reduce the complex structure of waiting periods . . . are bound to fail, because the different waiting periods cannot be reduced to a common denominator in purely technical terms.”

    The author’s conclusions (pp 13-14) are all part of the conclusions of Cohen (2003).

    Another stated difference from Cohen (2003) is that the author excludes additional Knight and Hayek papers discussed in Cohen (2003). One would expect of original work to extend the scope of literature discussed, rather than limit it!

    I cannot find any original contribution in the paper.

    I also note much sloppiness suggesting the paper was not proofread – spelling errors “Bõhm-Bawer (fn 3); repeated sentences “The production period lacks a beginning and an end” (p 4); the wrong title of the Cohen (2003) article in the References; among others. When an author does not care to take the time to proofread his/her work, the reader loses motivation to take the time to read it carefully.

  • Julian Libreros says:

    Dear Professor David Cohen,

    Thank you a lot for your comments. I will integrate the remarks in the document.

  • Nicolas Cachanosky says:

    The paper revisits the Hayek-Knight controversy on capital theory, in particular, the issue of the “(average) period of production”. Even though this episode is well known, a few implications and comments are worth noting.

    First, as long as we agree that production takes time, then there has to be something like the “period of production”, regardless of how difficult it may be to precisely measure it. This is a reason why debates surrounding this issue are so conflicted. Rejecting a particular measure of the period of production is not the same than rejecting the concept. The role of time in production can be removed from the problem by assuming the market is *already* en equilibrium and somehow capital automatically produces output. However, the role of concepts such as the period of production is precisely to explain movements and changes in the market, not just to describe how equilibrium looks like.

    There is a case to be done that Knight pushed back on Hayek’s stance on the period of production as this paper shows. However, Hayek did not abandon the idea completely, nor its implications. Namely, for Hayek as the interest rate (the price of time) is reduced, then there will be “some sort of” overinvestment in the sense that production processes are taking too much time. One way Hayek tried to return to this issue was through the Ricardo Effect.

    Finally, I would assign more weight to the shift in the concept of the period of production from being a backward-looking concept to be a forward-looking concept. This change is important since leaves Hayek only one step away from offering the same solution that Hick does. Namely, define the period of production as (we call it today) the Macaulay duration of expected the cash-flow of a production process. It is not clear why this ends in a circular reasoning (p. 13). Recall that in the Austrian literature interest rate is the price of time, not the price of capital. An interesting question would be how and why supporters and critics of the period of production overlooked Hick’s contribution.