Spaghetti economics: Shootout at Harvard Square
There were two letters about the poorly written (not the English, always impeccable, contrasting to my spaghetti English, which is always slightly off, like the Westerns) piece that The Economist had on Donald Harris. One by Robert Blecker, Steve Fazzari and Peter Ho, setting the record straight on the breadth and depth of Harris' contributions to economics. On this, they echo what the Post said about Harris' policy advice in his native Jamaica. The subtitle of the Post piece said: "An unconventional economist at Stanford, Donald J. Harris pushed strikingly nonideological economic solutions to the nation of his birth." Harris was (and still is, from what I can assume) a reasonable man, both as a scholar concerned with knowledge, and as a policy advisor. Nonideological being the key word.
The other one, that I reproduce in full, is by Professor Avinash Dixit, who says:
"You say that Mr Harris 'proposed that firms must choose from a 'book of blueprints', which need different capital goods' in his book published in 1978. Alas, that had been proposed more thoroughly and rigorously in 1953 in a brilliant paper by Edmond Malinvaud, for which he should have won a Nobel prize. And Robert Dorfman, Paul Samuelson, and yes, Robert Solow, whom you cite as author of the aggregate capital-growth model, covered it more comprehensively in a book from 1958. Perhaps the mathematics of all this were beyond the capacity of the so-called Marxists of Cambridge."
This, and The Economist piece itself, confirms the high degree of confusion about the meaning of the capital debates. The problem is not so much the lack of mathematical savvy of the Cambridge, UK, economists, Marxists or otherwise, but the lack of understanding of theory, by some very serious and important mainstream economists, and mainstream magazines (they do think they are a newspaper too, btw).
The Economist piece, commits the mistake that I suggested (see my old post on the Capital Controversies here) was typical of most incorrect views on the debates between the two Cambridges, and suggests, essentially, that it was an aggregation problem, and that:
"Mr Harris did not move on. In his 1978 book he developed a model of growth without an aggregate capital stock. Rather than the smooth 'production function' of Solow, in which the rate of saving and population growth determines capital per worker, Mr Harris instead proposed that firms must choose from a 'book of blueprints', which need different capital goods. Capitalists will compete to ensure the rate of profit is consistent across different industries, picking a blueprint based on the level of wages and profits in the economy."
That's why Dixit suggests that the choice of technique models in Malinvaud, and in Dorfman, Samuelson and Solow, using disaggregated notions of capital would avoid the circularity of the aggregative model. It is always the case that firms will choose, given distribution, the technique that minimizes costs. Of course, that's not the issue. The problem is that one cannot obtain a measure of capital that is independent of distribution, and hence, the idea that supply and demand in factor markets can determine the remuneration of capital and labor cannot stand. The classical concern with distributive conflict, both of Marx, but also of bourgeois economists, as he called them, Smith and Ricardo, becomes relevant again [meaning, if I need to clarify, the notion of conflict in the distribution arena was held both by critics of capitalism, and defenders of what Smith referred to as commercial societies].
Disaggregation does little to solve the problem. First, as noted by Garegnani, long ago, the abandonment of the notion of aggregate capital, and the notion of factors of production, came together with the view that means of production got their own rate of return, but also with the disappearance of the notion that the forces of competition lead to a normal and uniform rate of profit. The traditional long run method of economics, that was the theoretical foundation of the discipline, was abandoned. Further, it is still the case that, even with a disaggregated set of means of production, in the the aggregate, savings must be equalized to investment, and some process for that has to be used. That is why some notion of the quantity of capital was necessary for the mainstream [in recent times, mainstream economists and policy makers have started talking of the natural rate of interest again, a notion that had vanished more or less, and about very simplistic negative relations between investment and interest rates, w/o even knowing the logical problems associated with that, reinforcing problematic views about distribution].
In other words, forget the notion that firms actually face a book of blueprints, and that they can substitute and change the structure of production with relative ease, according to changes in relative prices, and that somehow that, by affecting the relative supply and demand for capital (many capital goods) and labor, would determine distribution, since that is obviously highly improbable in reality. Choices are more limited, and substitution irrelevant, at best. The Cambridge, UK, economists (and not only the Marxists; most were some kind of Post Keynesian, as The Economist notes) and Harris knew better. Distribution is determined by the relative bargaining power of workers and capitalists.
Samuelson knew enough, math and theory, to know when he was wrong. Mr. Dixit thinks his mathematical knowledge puts him above the rabble [the so-called Marxists of Cambridge, and Mr. Harris]. But what his letter [and the piece by The Economist] actually shows is lack of engagement with the theoretical argument, and, in contrast with Mr. Harris, an incredible amount of ideological bias. After all, the idea that distribution is according to effort, and that markets produce optimal outcomes is at play. It shouldn't be. He raised Malinvaud, I counter with Smith, which knew way less math than he does, for sure, but whose book is still worth reading.