Friday, September 23, 2016

Equilibrium and Information Literacy

"Everybody except Joan Robinson agrees about capital theory." -- Robert Solow (as paraphrased by Robinson)
An essential text in my researches on mercantilism, usury and bills of exchange is Raymond de Roover's Gresham on Foreign Exchange, which just happens to be stored in part of SFU's library that is under construction and thus inaccessible. The immediate unavailability of that book, however, led to a fortuitous discovery.

I browsed in the call number section of the library's general collection where de Roover's book would have been and Robert Leeson's Ideology and the International Economy caught my eye. I flipped through the book and noticed on page 19 the delicious quote from Joan Robinson that, "the free-trade doctrine is just a more subtle form of mercantilism."

The quote is from a 1966 lecture, "The New Mercantilism" that is included in a collection of essays, Contributions to Modern Economics, which also contains "Capital Theory Up-to-Date," a 1970 review of C. E. Ferguson's The Neoclassical Theory of Production and Distribution, in which Robinson reprises her parody of neo-Walrasian, neo-neoclassical capital "leets." Leets is steel spelled backward and makes its debut in "Equilibrium Growth Models," Robinson's 1961 review of James Meade's Neo-Classical Theory of Economic Growth.

This allegedly ectoplasmic representation of capital is, in a nutshell, the crux of the "Cambridge capital controversy," which Robinson launched with her 1952 challenge, "I leave it to those who draw production functions to say what marginal productivity and the elasticity of substitution mean when labour and capital are the factors of production." Looking back, in 1978, on her 1952 essays and the "long struggle to escape... habitual modes of thought and expression," Robinson stressed that "it was precisely from the concept of equilibrium that Keynes was struggling to escape..." Contrarily, though:
"...textbook teaching in the department of so-called macro theory was an attempt to push Keynes into short-term equilibrium. ... The grand neoclassical synthesis (now known as bastard Keynesianism) was a more ambitious attempt to reduce the General Theory to a system of equilibrium."
In responding to Robinson's leets critique, Robert Solow began by acknowledging "much truth" to the objection that "the usual production functions, allowing for more or less substitutability between capital and  labor, attribute to 'capital' a degree of malleability which contradicts common observation." He then distinguished between the "econometrically-minded person" who would view the overly malleable capital as a "specification error" and others -- presumably including Robinson -- who judge it to be "a doctrinal error; and its consequence is a kind of Fall from Grace." Seven years later, Robinson had this to say about "doctrinal disputes":
Many economists, nowadays, who are interested in practical questions are impatient of doctrinal disputes. What does it matter, they are inclined to say, let him have his leets, what harm does it do? But the harm that the neo-neoclassicals have done is, precisely, to block off economic theory from any discussion of practical questions.
If one is concerned about actual unemployment in an actual economy, Robinson later explained, one "has to discuss it in terms of processes taking place in actual history. The concept of equilibrium is incompatible with history. It is a metaphor based on movements in space applied to processes taking place in time." In other words, it is not just some kind of ethereal affectation to object to the concept of equilibrium -- it is an argument with irrevocable real-world consequences.

The failure of what Robinson dismissed as "bastard Keynesianism" also had real-world doctrinal consequences. "In the era of stagflation, this notion [that equilibrium growth can be achieved through fiscal and monetary 'fine tuning'] has been discredited and the quantity theory of money is blossoming afresh amongst its ruins." This 'blossoming,' incidentally, was not something Robinson welcomed.

Well, my interlibrary loan of de Roover's Gresham on Foreign Exchange has arrived, so I'm off up the hill to pick it up. To be continued...



6 comments:

Thornton Hall said...

Could you pass on specific citations for the Joan Robinson quotations? Thanks in advance.

Sandwichman said...

"the free-trade doctrine is just a more subtle form of mercantilism."

"The New Mercantilism" in Contributions to Modern Economics, Basil Blackwell, Oxford 1978

***************
"I leave it to those who draw production functions to say what marginal productivity and the elasticity of substitution mean when labour and capital are the factors of production."

"Notes on the Economics of Technical Progress" (originally published 1952) reprinted in The Generalisation of the General Theory and other Essays, St Martin's Press, NY, 1979.

***************
"Everybody except Joan Robinson agrees about capital theory."

"Many economists, nowadays, who are interested in practical questions are impatient of doctrinal disputes. What does it matter, they are inclined to say, let him have his leets, what harm does it do? But the harm that the neo-neoclassicals have done is, precisely, to block off economic theory from any discussion of practical questions."

"Capital Theory Up-to-Date," in Contributions to Modern Economics, Basil Blackwell, Oxford 1978.

***************
"[the] long struggle to escape... habitual modes of thought and expression," Robinson stressed that "it was precisely from the concept of equilibrium that Keynes was struggling to escape..." "...textbook teaching in the department of so-called macro theory was an attempt to push Keynes into short-term equilibrium. ... The grand neoclassical synthesis (now known as bastard Keynesianism) was a more ambitious attempt to reduce the General Theory to a system of equilibrium."

"[one] has to discuss it in terms of processes taking place in actual history. The concept of equilibrium is incompatible with history. It is a metaphor based on movements in space applied to processes taking place in time."

"In the era of stagflation, this notion [that equilibrium growth can be achieved through fiscal and monetary 'fine tuning'] has been discredited and the quantity theory of money is blossoming afresh amongst its ruins."

Introduction to The Generalisation of the General Theory and other Essays, St Martin's Press, NY, 1979.

***************
No direct quotes from "Equilibrium Growth Models: A review article," The American Economic Review, Vol. 51, No. 3 (Jun., 1961), pp. 360-369

Thornton Hall said...

Awesome! Thanks

rosserjb@jmu.edu said...

Regarding production functions and standard neoclassical "factors" what we have is concepts formulated to try to explain national level income shares between social groups, identifying those groups by what they own: labor its own power and getting wages and salaries, capitalists owning produced means of production and getting interest and dividends, and aristocrats or landlords owning land and resources and getting rent. Robinson pulled the curtain aside to point out hard that these aggregated abstract factors really have nothing to do with real world engineering production functions where inputs are much more specficially defined and all this ownership stuff can get sort of fuzzy. This was a large part of the thrust of her "leets"("steel" backwards) critique of aggregates, that they really have nothing to do with production at all.

So we get these funny constructs like Cobb-Douglas, very widely used right now to model aggregate US production and used by Solow also, which was invented to "explain" what appeared in the late 1920s to be a several decade constancy of the shares of income going to workers and capitalists. It was all about income distribution and not remotely about actual production, with engineering production functions looking very different. A big joke is that those shares have been shifting a lot over the last few decades in favor of capital income shares, but we still have lots of economists using the Cobb-Doublas productgion function to model aggregate production in the US despite the total failure of its conditions to hold, one of the more prominent examples of blatant zombie economics going on out there, with this brought on by the complete breakdown of most economists taking the late Joan Robinson seriously.

One bunch that did try to do it right was those around Shepherd and his study of duality, where he actually did try to go to the engineering production function level (back in the 40s) with his studies of duality, not to mention some of the ag economists such as Earl O. Heady, although his corn production function went into the thousands of inputs. Current economists just cannot be bothered with this real world stuff.

Bruce Wilder said...

The classical factors were introduced to explain income distribution, but by way of the differing character of the factors and their corresponding characteristic incomes: rent to land, wages to labor and interest or profit to capital or stock. The important thing was the link between the character of the factor and its characteristic income.

The production function was a neoclassical innovation used to make the case for the idealized marginal product analysis of income distribution.

The classical concept does not confront the aggregation problem, because the classical factors are Types. The neoclassical concept wants variables for the operation of algebra and needs not just aggregation but homogeneity: character is lost.

Considered as a theory of production, rather than income distribution, there is a further problem: output is not a mathematical function of inputs. The production function is not actually a function. The dodge was to assume away any shortcomings in management or engineering: maximum output was a function of given inputs or a given output was a function of minimum inputs. Such an assumption makes efficiency exclusively a matter of allocative efficiency, but does not actually solve the problems of technical efficiency or organizational management; it just leaves management and technology as an unaccounted dimension.

rosserjb@jmu.edu said...

An engineering production function is a mathematical function that describes relations between specific inputs and specific outputs. The problem is that those inputs have little to no connection with the supposed "factors of production" that appear in standard neoclassical production functions and are tied to income distribution.