Wednesday, February 27, 2008

I Stared At This For A Good Five Minutes

Referee Report on "Some Capital-Theoretic Fallacies of Austrian Economics"
January 2008

Summary: This paper challenges some of the very premises of the Austrian theory of the business cycle, via a criticism of Mengerian/Böhm-Bawerkian/Hayekian capital theory. In particular, the author challenges the standard Austrian claim that a reduction in the rate of interest leads to a deeper, more roundabout capital structure. To prove that such a claim is false in the general case, the author draws on the reswitching debate from the Cambridge Capital Controversy (CCC).

Recommendation: I must advise that the current paper is not suitable for publication in the RAE in its present form. If I have understood it, the paper is basically saying, "The reswitching examples that surfaced in the CCC demonstrate that the starting point of ABCT is wrong. As Sraffa and his followers showed, entrepreneurs can switch from one technique to another, and then back again to the first, as interest rates fall. This means it is clearly impossible to rank objective capital structures in terms of their roundaboutness."

However, if this is indeed the author’s contribution (and if it isn’t, then s/he needs to rewrite the paper to make the true thesis much clearer!), then it is not novel. Paul Samuelson's famous "A Summing Up" (1966 Quarterly Journal of Economics) article - in which he comes as close to admitting he was wrong as Paul Samuelson ever does — presents a beautiful illustration of the flaws in the basic Böhm-Bawerkian "fable" (Samuelson’s term). The numbers in Samuelson's example are nice round ones, making the issue far easier to grasp than the complicated fractional numbers chosen by the current author. Moreover, historians of economic thought understood the implications as well: In his Economic Theory in Retrospect, Mark Blaug says that the possibility of reswitching proved the "final nail in the coffin" for Böhm-Bawerkian capital theory.

Therefore, economists have been well aware of the vulnerability of orthodox Böhm-Bawerkian capital theory to the reswitching examples. What's more, the Austrians themselves have responded. Ironically, the present author cites Garrison's 2006 work - which John Hicks apparently told Garrison was the best response to the reswitching issue—but doesn’t deal directly with Garrison's responses, except for a short quote questioning the empirical relevance of reswitching. A novice who read the present paper would have no idea that Garrison (or any other Austrian) had tried to grapple with the theoretical challenges posed by reswitching. In fact, when the author claims on page 6 that "this paper is the first to look at Hayekian triangles in light of the CCC," the novice reader would probably consider this entire critique as a new challenge to ABCT.

My recommendation is for the author to first, read the Samuelson paper. Then, if s/he is willing, a new paper could be written, demonstrating to the Austrian reader that reswitching is not merely a theoretical curiosity, but a real-world phenomenon. I personally have not read the empirical literature to which the author alludes on page 14 (in response to Garrison). So a new paper might be acceptable for the RAE, if it (a) started out with a very quick summary of the reswitching controversy, and used a simple example such as Samuelson's, then (b) presented the major Austrian responses, and finally (c) used the empirical literature and/or theoretical arguments to show why the responses quoted in (b) are unsatisfactory.

Let me end this section by disclosing my identity as Robert Murphy, because I am going to recommend some of my own work as background (and I don’t want to be coy about it). The following online article gives a quick summary of Samuelson’s piece (and my own response): http://mises.org/story/1148

Some other of my articles that may interest the writer are another online article about Sraffa: http://mises.org/story/1486. I also have two formal expositions of Böhm-Bawerkian capital theory in The Journal of the History of Economic Thought. The first is "Dangers of the One-Good Model: Böhm-Bawerk's Critique of the 'Naïve Productivity Theory of Interest'" (Vol. 27, No. 4 (December 2005), pp. 375-382), and "Interest and the Marginal Product of Capital: A Critique of Samuelson" (December 2007).

Please note that I am NOT saying the above articles need to be cited in a future RAE submission. It's just that it is very rare for economists to write articles that deal with both Sraffian and Austrian concerns, and so I want the author (who belongs to this select club) to be aware of them.

Specific Comments: Below I offer some minor comments on the draft for the author.

Page 2, bottom quote from Garrison: It seems that this particular statement is OK, even with reswitching.

Page 4, top: The citation says Lewin 1991, but no such work is in the bibliography. I think the year is wrong?

Page 4, first full paragraph: I didn't really understand the Lewin argument described here. Is he simply talking about learning by doing, or something else?

Page 7, table: I don't understand why these numbers are so complicated. Would it hurt anything to multiply through by 49? Or by some higher number in order to make it consistent with Table 3? If the fractions are used simply to allow for a surplus of one unit of corn, I’d recommend increasing the surplus in order to make the inputs integers.

Page 8, Table 2: I stared at this for a good five minutes and couldn’t figure it out - and I'm familiar with such diagrams from Hayek, Rothbard, Garrison, and even Samuelson. It should be explained more thoroughly if it is to be retained.

Page 9, middle: "Consistency…ensures…the existence of a maximum above which the interest rate cannot rise in the ERE." The author doesn't take this any further, but it seems s/he is saying that the standard Austrian story wouldn’t "work" in this case, because an increase in time preference couldn’t raise the interest rate above the ceiling set by physical constraints. However, this is only because the author has assumed that certain production processes will be carried out indefinitely, à la Sraffa. There is no reason that these processes would be carried out - thus "determining" the rate of interest - if consumer preferences changed in certain ways.

Page 15, Paul Davidson entry: I think the title is incorrect. Should it be "The Economics of Ignorance OR Ignorance of Economics?"?

4 comments:

Anonymous said...

"In his Economic Theory in Retrospect, Mark Blaug says that the possibility of reswitching proved the "final nail in the coffin" for Böhm-Bawerkian capital theory."

So, Austrians know that their theory of the business cycle is logically flawed but do not care? Partly because it is doubtful that it has any empirical validity?

Whatever happened to the Austrian argument that empirical evidence is irrelevant and only the logic of the theory counts?

YouNotSneaky! said...

Sorry about the negative referee report. Never a pleasant experience.

Robert Vienneau said...

Thanks for the sympathy.

I find Murphy's comments on small errors helpful. I suppose it is also helpful to see where a reader cannot follow my exposition.

I made the point in the paper that refusing to accept reswitching on empirical grounds was a change in the Austrian method. I found some other comments by Murphy odd.

While I don't reference Samuelson, I implicitly address his rejection of Bohm-Bawerk with my comments about Austrian's rejection of quantitative measures of the period of production. Contrary to the impression Murphy gives, some Austrians distinguish between Bohm-Bawerk and modern Austrian capital theory. Furthermore, Samuelson explicitly says he is not drawing the sort of triangles I organize my paper around.

I explicitly say Austrians have responded to the CCC. SO I don't see how one would think I imply Austrians haven't said anything about reswitching.

I don't understand where Murphy gets his interpretation of a maximum interest rate from. He should be familiar with a so-called factor price frontier and should have seen that it can intercept the interest rate axis for the sort of circular flows I analyze.

Nevertheless, I have rewritten the paper. I'm not sure what I'm going to do with this. It's getting to long to submit elsewhere. I suppose I could throw out section 2, which I dislike to do.

YouNotSneaky! said...

I wonder what percentage of all referee reports are of the form:

"This paper is too long for publication. Also the author needs to expand this section, elaborate on that section, discuss this research and include these citations".