| If you believed something different, you wouldn't be sitting where you are sitting |
Suppose one wants to discuss capitalism versus socialism or some smaller matter. One might think the discipline of political economy, now known as economics would be helpful. But it is not.
What is taught in most universities in the United States was shown to be nonsense more than half a century ago. I find it hard to account for this except on the grounds of political ideology. I realize that most academic economists and their students that persist do not experience themselves as propagandists. And it does take some study to master the mathematical models, even if they are incoherent.
Obviously, exceptions exist. I am most aware of the economics departments at the University of Massachusetts at Amherst, the New School, the University of Missouri at Kansas City, and the University of Utah. And I think the situation might be different in some other countries. At least, they can list some prominent universities like the above. Furthermore, in taking courses in academic economics, one should learn something useful about how national products and income accounts are kept. Many economists might think they are doing measurement without theory, that these theoretical incoherences that I go on about do not matter to them. And there are many partial models that might be useful in a narrow context.
These sort of questions should have clear answers: For some model, what are the parameters and and what are the variables found in the solution? For each parameter or variable, what are the units of measure? Lately, I have been recommending a John Eatwell lecture on the bomb that Piero Sraffa placed at the foundations of economic theory. Working through Kurz and Salvadori's 1995 textbook is also a good way to understand my favorite devasting criticism of marginalist economics.
Smith's natural prices, Ricardo's prices of production, and Marshall's normal prices all characterize a long-period position or equilibrium, depending on the theory. Marginalist economics is about the allocation of given resources. The quantity and initial distribution of capital goods are among the givens, at least in Walras' formulation. Supply and demand are supposed to clear in all markets in equilibrium, and the capitalists obtain the same rate of profits in all markets. This model is ovedetermined and inconsistent. Walras was mistaken.
Taking the numeraire quantity of capital and its initial distribution as given was another incorrect marginalist approach. The physical composition of capital is supposed to be endogeneous. But prices of capital goods are found as solutions of the model. The quantity of capital is simultaneously inside and outside the model. Knut Wicksell realized this approach does not work. And waiting or abstinence cannot explain profits either.
So from about 1930 to the 1970s, marginalists abandoned long period theory in their most general models. The Arrow-Debreu-McKenzie model of intertemporal equilibrium is the cumulation of this trend. In the model, commodities are distinguished by physical properties, when they are available, and the state of the world. Prices are established in forward markets, found at the start of time.
This is a model of supply and demand in some sense. Households maximize utility subject to constraints. Plans are precoordinated, and all markets clear for all time. On the other hand, one can not draw well-behaved supply and demand schedules at the level of the market, as is shown by the Sonnenschein-Debreu-Mantel theorem.
Economists cannot explain how any economy would get in or approach such an equlibrium. Franklin Fisher investigated this question. Fabio Petri notes that the givens of initial quantities of capital equipment would change if production goes on while the economy is in disequilibrium. The equilibria consistent with the givens are not the equilibria that would be approached. The model does not depict tendencies in any possible capitalist economy.
Given an equilibrium, however, the forward prices embody predictions of what spot prices would be. Mainstream economics, when talking about dynamics, often mean the time paths of these spot prices. A conceptual problem arises here. If markets can open and close later, the model is not the Arrow-Debreu-McKenzie model. Anyways, the rate of profits is not the same among industries at any time period, since prices are typically not stationary.
Mainstream economists have basically given up, as I understand it, on trying to develop any general approach to explaining prices and distribution in a capitalist economy. I think the textbooks are not clear on this point. I like some of the bits of mathematics, such as game theory, in some of these textbooks.
Why study this stuff? Even though academic economists are mostly trapped in an intellectual ghetto, they still have a connection to what ideas are hegemonic. And the disciple of economics provides a puzzle for the sociology of 'knowledge' and the philosophy of science. If academic economists were merely useless, the world would be improved.
Hi Robert, I was wondering what you thought of the Austrian economist Rob Thorpe's response to your paper "Some Capital-Theoretic Fallacies of Austrian Economics," I saw that he said the following:
ReplyDelete"The idea that a good can have more than one "order" is correct. The concept is not exact in the sense that each good has only one order. General purpose goods like steel and oil occupy many orders. I don't think this is a big problem, maybe Dabchick or someone here disagrees. Notice that Vienneau's example isn't really useful here because it's a conceptual example with only two very simple processes in it. The real issue of "orders" is whether the concept works in the real world.
I more of less agree with the points Vienneau quotes from Lewin. I don't think that any of these are really fatal to ABCT. Again, I'd be interested in what Dab and thunderbbx think here.
Vienneau's numerical example in table 1 and table 2 is very limited. There is no potential to change production technique. There are no fixed capital goods, only circulating capital goods. Although a good is called "steel" it is consumed each year. This is why interest rates in his model just change the distribution of incomes between wages and interest. It is because of this extreme simplicity that all of the prices in the model are found once the interest rate is decided.
In table 1 & table 2 all of Vienneau's talk about what entrepreneurs think is irrelevant. There is one process in each industry. Entrepreneurs either take it or leave it, it's a binary choice. If they take it they expand until all labour is employed. If they leave it nothing is produced and there is no economy. Notice if the steel industry refuses to make steel then nothing can happen, and if the corn industry refuses to make corn nothing can happen. So, to get any economy at all we must assume that entrepreneurs in both industries decide to operate, that is all.
-
Vienneau discusses how varying the interest rate produces differences in the value of inputs. Notice that Vienneau presents his Hayekian triangle rotated by 90 degrees, that with the x-axis and y-axis swapped compared to the normal convention. Vienneau claims that the shape of the Hayekian triangle changes even if the constituent capital goods used does not. The prices change, with the interest rate, so the shape changes. This is not wrong and certainly happens in practice. It's relevant in recent property bubbles especially. Due to zoning and planning limits it's not possible to build more housing in certain areas. So, as the interest falls and financing becomes cheaper houses in such areas become more expensive but it is rarely possible to build more of them. This price-Wicksell effect is real, and Hayek actually makes this point himself in Prices and Production. However, in the majority of cases it is actually possible to build more capital when capital prices are high! The real world offers many different processes to select from, unlike Vienneau's simple models.
Later in section 3.4 Vienneau introduces choice-of-technique. Here we have reswitching. We have all of the usual problems of the pro-reswitching side. They constantly present extremely simple models. They provide no evidence that reswitching actually occurs in practice and have no theory to explain why we should expect it to occur. To get reswitching in any sort of economy which has real complexity you need fixed capital costs to be paid after goods have been sold, which makes no sense. Vienneau said he was going to sort all this stuff out years ago, but he never has. I won't go into Vienneau's note 35 here since I don't have the time."
Would be curious on your thoughts
Where does Thorpe express this muddle? I found a conversation on Reddit from two years ago.
ReplyDeleteI see that Thorpe accepts that commodities can be simultaneously in several orders at once and that the existence of price Wicksell effects is valid. Mayhaps he could explain this to Mateusz Machaj, since this is enough to make some of his work pointless.
I think it of interest to show that some of the processes operating today are going to produce commodities for immediate consumption, for use as capital goods in producing consumption goods in the next periods, for use as capital goods in producing capital goods for use in producing consumption goods two periods hence, and so on.
Maybe Thorpe has forgotten some of what he reads. Which axis is which is not consistent in presentations of Hayekian triangles. No point of substance is involved.
Anyways, Thorpe's comments on the limitations of the simple counterexamples is confused in logic. You do not require generality in counter-examples.
Furthermore, reswitching, capital-reversing, reverse labor substitution, recurrence of processes, and so on are consistent with the existence of many sectors, alternative processes in all of them, the existence of fixed capital, and so on.
I do not know what I promised to "sort out" years ago. I think my work with fluke cases demonstrates that, say, capital-reversing is not a fluke case. Bertram Schefold's work on random matrices implies that the choice of technique is hardly likely to work out as exploded theorists imagined. Schefold's work was inspired by empirical work.
The more confusion I see from Austrian school economists, the more impressed I am that Fratini convinced some Austrian school economists to accept one of his publications.
It was on Discord, I'm not sure how familiar you are with the usage of it, but there is a (relatively popular) discord server dedicated to the study of Austrian Economics, they even have their own website.
ReplyDeleteThere is that Reddit post, there is also this from Thorpe, which seems to me to not be particularly compelling, a bundle of misinterpretations and misunderstandings that most likely could be soundly resolved by a reading of Kurz & Salvadori (1995). Interestingly, in the one you linked the user 'Eleutheria' is mentioned who was a former Austrian who "converted" to Sraffian economics, I believe there's a post somewhere in one of your Elsewheres where you cover a blog post he wrote criticising Murphy's CCC-commentary.
On a more general topic, this seems to mention that you were in the midst of writing a book in 2020. I'm curious, is it on the way soon, or is it stuck in development hell like, for instance, the English translation of Michael Heinrich's book the Science of Value? And if it is soon on the way, will it be largely inspired by the contents of your blog posts, or will it be entirely new?
I have a draft book which I have been drawing on for blog posts and article submissions. As I do this I am increasingly dissatisfied. So I am not anywhere near submission to a publisher.
ReplyDeleteThorpe could also read Steedman's Marx after Sraffa. Yes, I agree he does not understand.