Sunday, July 29, will be the second anniversary of the sudden passing of Earl Thompson, one of the truly original and creative minds that the economics profession has ever produced. For some personal recollections of Earl, see the webpage devoted to him on the UCLA website, where a list of his publications and working papers, most of which are downloadable, is available. Some appreciations and recollections of Earl are available on the web (e.g, from Tyler Cowen, Scott Sumner, Josh Wright, and Thomas Lifson). I attach a picture of Earl taken by a department secretary, Lorraine Grams, in 1974, when Earl was about 35 years old.
I first met Earl when I was an undergraduate at UCLA in the late 1960s, his reputation for brilliant, inconclastic, eccentricity already well established. My interactions with Earl as undergraduate were minimal, his other reputation as a disorganized and difficult-to-follow lecturer having deterred me, as a callow sophomore, from enrolling in his intermediate micro class. Subsequently as a first-year graduate student, I had the choice of taking either Axel Leijonhufvud’s macro-theory sequence or Earl’s. Having enjoyed Axel’s intermediate macro course, I never even considered not taking the graduate sequence from Axel, who had just achieved academic stardom with the publication of his wonderful book On Keynesian Economics and the Economics of Keynes. However, little by little over the years, I had started reading some of Earl’s papers on money, especially an early version of his paper “The Theory of Money and Income Consistent with Orthodox Value Theory,” which, containing an explicit model of a competitive supply of money, a notion that I had been exposed to when taking Ben Klein’s undergraduate money and banking course and his graduate monetary theory course, became enormously influential on my own thinking, providing the foundation for my paper, “A Reinterpretation of Classical Monetary Theory” and for much of my book Free Banking and Monetary Reform, and most of my subsequent work in monetary economics. So as a second-year grad student, I decided to attend Earl’s weekly 3-hour graduate macro theory lecture. Actually I think at least half of us in the class may have been there just to listen to Earl, not to take the class for credit. Despite his reputation as a disorganized and hard to follow lecturer, each lecture, which was just Earl at the blackboard with a piece of chalk drawing various supply and demand curves, and occasionally something more complicated, plus some math notation, but hardly ever any complicated math or formal proofs, and just explaining the basic economic intuition of whatever concept he was discussing. By this time he had already worked out just about all of the concepts, and he was not just making it up as he was going along, which he could also do when confronted with a question about something he hadn’t yet thought through. But by then, Earl had thought through the elements of his monetary theory so thoroughly and for so long, that everything just fit into place beautifully. And when you challenged him about some point, he almost always had already anticipated your objection and proceeded to explain why your objection wasn’t a problem or even supported his own position.
I didn’t take detailed notes of his lectures, preferring just to try to understand how Earl was thinking about the topics that he was discussing, so I don’t have a clear memory of the overall course outline. However his paper “A Reformulation of Macroeconomic Theory,” of which he had just produced an early draft, provides the outline of what he was covering. He started with a discussion of general equilibrium and its meaning, using Hicksian temporary equilibrium as his theoretical framework. Perhaps without realizing it, he developed many of the ideas in Hayek’s Economics and Knowledge paper, which may, in turn, have influenced Hicks, who was for a short time Hayek’s student and colleague at LSE — in particular the idea that intertemporal equilibrium means consistency of plans so that economic agents are able to execute their plans as intended and therefore do not regret their decisions ex post. From there I think he developed a search-theoretic explanation of involuntary unemployment in which mistaken worker expectations of wages, resulting from an inability to distinguish between sector-specific and economy-wide shocks, causes labor-supply curves to be highly elastic at the currently expected wage, implying large fluctuations in employment, in response to economy-wide shocks, rather than rapid adjustments in nominal wages . With this theoretical background, Earl constructed a simple aggregative model as an alternative to the Keynesian model, the difference being that Earl dispensed with the Keynesian expenditure functions and the savings equals investment equilibrium condition, replacing them with a capital-market equilibrium condition derived from neo-classical production theory — an inspired modeling choice.
Thus, in one fell swoop, Earl created a model fully consistent with individual optimizing behavior, market equilibrium and Keynesian unemployment. Doing so involved replacing the traditional downward-sloping IS curve with an upward-sloping, factor-market equilibrium curve. At this point, the model could be closed either with a traditional LM curve corresponding to an exogenously produced money supply or with a vertical LM curve associated with a competitively produced money supply. That discussion in turn led to a deep excursion into the foundations of monetary theory, the historical gold standard, fiat money, and a comparison of the static and dynamic efficiency of alternative monetary institutions, combined with a historical perspective on the Great Depression, and the evolution of modern monetary institutions. It was a terrific intellectual tour de force, and a highlight of my graduate training at UCLA.
Unfortunately, “A Reformulation of Macroeconomic Theory” has never been published, though a revised version of the paper (dated 1977) is available on Earl’s webpage. The paper is difficult to read, at least for me, because Earl was much too terse in his exposition – many propositions are just stated with insufficient motivation or explanation — with readers often left scratching their heads about the justification for what they have read or why they should care. So over the next week or so, I am going to write a series of posts summarizing the main points of the paper, and discussing why I think the argument is important, problems I have with his argument or ways in which the argument needs further elaboration or what not. I hope the discussions will lead people to read the original paper, as well as Earl’s other papers.
Uneasy Money 
Wauw David…I have tried to read “A Reformulation of Macroeconomic Theory” a number of times, but everything about that paper make it hard to grasp so I certainly look forward to your posts on the paper.
LikeLike
That´s blogging as “public service”! Great.
LikeLike
Great stuff!
To play the devil’s advocate, wouldn’t Leijonhufvud say that any unemployment left over when capital markets are in equilibrium (as opposed to simply market clearing) is not Keynesian involuntary unemployment in the first place? Further, that if capital markets are in equilibrium, then the economy is also in a savings-investment equilibrium. An economy where the average rate of interest is different (capital stock) from the marginal rate of interest (savings and investment) is not in a capital markets equilibrium (there is an obvious arbitrage), and if the two are equal, then it is also in S-I equilibrium.
It would seem that Earl Thompson’s model explains a generalized Friedman(’68)-Sumner-Selgin unemployment, and does a better job of explaining Modigliani-Solow-Samuelson unemployment than those worthies did themselves. But it does not explain Keynesian unemployment, at least as understood by Leijonhufvud.
Notice also that Nick Rowe achieved the other Sumner/ Friedman explanation (of real rates being procyclical) by simply postulating an upward sloping IS curve, using mpc + mpi > 1, rates held constant. This is analogous in many ways to the upward sloping factor-market equilibrium curve that you mention above.
Taken together, the Rowe/ Thompson formulation provides the coherent, consistent micro-foundations of the pure N/N (to use another Leijonhufvud-ism) case of Sumner/Friedman. It is one theory of unemployment and output fluctuations. But as far as I understand, it is not an explanation of Keynesian unemployment and output fluctuations, which necessitates capital market disequilibrium.
LikeLike
David,
What exactly is the difference between Thompson’s story here and that of Lucas? I recall you once said something to like “Lucas corrupted the UCLA model.”
Best regards,
JoeMac
LikeLike
JoeMac
As far as I understand, the UCLA (Thompson) model is a more solid version of Friedman ’68, which can itself be summarized as managers correct, workers stubborn. The Friedman story was one of adaptive expectations across sectors (nominal changes being mistaken for real changes), but the UCLA model was one of essentially rational expectations with imperfect information – absolute nominal changes being mistaken for relative nominal changes.
Lucas ’72 turned this around – workers correct, managers irrational. What Lucas has in common with UCLA is that it retains the story of absolute nominal changes being perceived as relative nominal changes – the Robinson Crusoe parable. Rational expectations with imperfect information. But Lucas for some reason chose to focus on labour demand rather than labour supply.
David, please do correct me if I’m misrepresenting something here.
LikeLike
While I did not know Earl Thompson,may he rest in peace. He seems like he’s had interesting contributions to make to the study of economics.
Out of curiosity though, David Glasner, have you heard of a character by the name of Dr. Michael Emmett Brady? Did you see his review of Axel Leijonhufvud’s book? If not, here it is.
http://www.amazon.com/review/R3OU6EAGUZ73YD/
If you have heard of him, what do you think of him? If you have seen his review of Leijonhufvud’s book, what did you make of it?
LikeLike
Sounds fascinating. I’m looking forward to the explanation, because I very much doubt I could follow the paper without a primer.
LikeLike
Lars, Especially in his younger days, Earl assumed that if he understood what he was writing, everyone else would also understand. As he got older, he began to make more concessions to his readers, but his Reformulation paper makes no concessions at all.
Marcus, I hope so.
Ritwik, It’s clear that Thompson and Leijonhufvud had very different views of macroeconomics and what constitutes Keynesian unemployment. In Thompson’s model, the market for capital services is always in equilibrium, so Keynesian unemployment can result only from inconsistent expectations of future wages on the part of workers (aka sticky wages). The real (or natural) rate of rate of interest depends, as Keynes explicitly argued in the GT, rises as employment increases because the marginal product of capital increases along with the quantity of labor used as an input. Nick Rowe’s argument for an upward sloping IS curve seems ad hoc, Thompson had a simple derivation based on neoclassical theory.
I agree that there may be capital market complexities that Thompson’s model is not capturing, because it assumes that the market for capital necessarily clears at a positive price, so that the real rate of interest must be unambiguously positive. Real interest rates are now negative, and I don’t see how that phenomenon can be reconciled with a single homogeneous output that can be consumed or used as a factor of production. When the marginal product of capital falls to zero, people just eat it.
JoeMac, I think in Thompson’s model, he pretty much subscribes to Lucas’s idea that workers cannot accurately distinguish between economy-wide and sector-specific demand shifts. He cites Lucas’s 1972 paper “Expectations and the Neutrality of Money” in his discussion of the labor market. I think that the difference between Thompson and Lucas is that Thompson had a more sophisticated understanding of temporary equilibrium as a consequence of incorrect expectations than Lucas and therefore was much more flexible than Lucas in postulating rational expectations as a necessary property of a model. A model has to have a sensible interpretation under rational expectations, otherwise there is an inconsistency in the model. That doesn’t mean that in applying the model empirically you have to assume that expectations are always formed rationally. My own view is that there are also very serious problems in applying models with this much aggregation to the real world, because it is hard to derive a persistent effect from a transitory shock. If there are multiple sectors, it is easier to get shocks that eco as they are propagated from sector to sector. I think that becomes an issue when there is trading at disequilibrium prices, and those effects are propagated through the operation of Say’s Law in the manner discussed by Hutt, as I explained in a recent post. I don’t know how Earl felt about that issue.
Ritwik, Didn’t you mean absolute nominal changes being mistaken for relative changes? A nominal change can’t be relative.
Could you spell out the basis for saying that Lucas assumed that workers’ expectations were correct and managers expectations were incorrect (incorrect = irrational?). I don’t disagree, but I don’t think I recall that distinction ever being made and would like to have it documented.
Blue Aurora, I only know of Brady through his numerous book reviews on Amazon.com. He seems to be an interesting guy, but pretty quirky, perhaps a touch OCD. I have not checked up on his assertions about Keynes and the rampant misinterpretation of him in the literature. However, when I read through the first 100 pages or so of Keynes’s Treatise on Probability, it did not seem to me that his characterization of Keynes’s understanding of probability and Frank Ramsey’s misrepresentation of Keynes’s position was accurate. However, I am very far from being an expert on probability theory, so I cannot be confident that my reading of Keynes is more reliable than Brady’s. But I do have my doubts about him.
W. Peden, I hope you will find it worthwhile to follow along.
LikeLike
It’s amazing what stories we can tell sometimes from our experience being a student at a University. I really wish that everyone one of us has the change to go to a University to pursue whatever they please.
LikeLike
David
Yes, I meant absolute nominal changes being mistaken for relative changes. I was just trying to emphasize the difference with Friedman’s economy-wide adaptive expectations.
I think it was Hy Minsky who made the manager/ worker distinction. I think it may be better stated as workers don’t matter, managers are confused and choose to produce more whenever there is a rain of external money, an vice-versa.
http://digitalcommons.bard.edu/cgi/viewcontent.cgi?filename=1&article=1335&context=hm_archive&type=additional
LikeLike
David Glasner: What do you think of Dr. Michael Emmett Brady’s review of Leijonhufvud’s book? Yes, I can see how others would perceive him as eccentric. However, I genuinely think he’s trying to help, and once you look past his poor use of punctuation, he seems to know what he’s talking about.
As for Dr. Michael Emmett Brady’s understanding of A Treatise on Probability…he has had an undergraduate and graduate-level training in mathematics before he went on to do his doctoral dissertation in economics. So I think he has a pretty good idea of what he’s talking about. I’ve actually asked a professional mathematician for his own opinion of Ramsey’s reviews of Keynes’s work on probability – the professional mathematician also came to the conclusion that Ramsey’s reviews were “based on a misunderstanding of Keynes, and are thus, worthless” (I’m paraphrasing, I don’t remember his exact words, but they were essentially on those lines).
I know that your field of specialty is monetary economics, and not decision theory, David Glasner. But I do think that you might want to take a look at the work of Dr. Michael Emmett Brady – particularly his 1993 article in The British Journal for the Philosophy of Science and his 1994 article in International Studies in the Philosophy of Science. They could be used to explain informational uncertainty very well. See the following links.
http://www.jstor.org/stable/687654
http://www.tandfonline.com/doi/abs/10.1080/02698599408573487
Also, out of curiosity, David Glasner, have you read the following book by Karl Brunner and Allan Meltzer?
Dr. Michael Emmett Brady has reviewed that book positively.
http://www.amazon.com/review/R3C73NUOWO0MF0/
LikeLike
Blue Aurora, Sorry, but I can’t give an adequate response without going back and rereading the relevant chapters of the General Theory, which I have not looked at for a long time. What he says is certainly interesting and worth pondering, but I can’t give you specific response to his claims.
I didn’t mean to disparage his mathematical abilities, I only said that from my reading of the Treatise on Probability, after seeing Brady’s critique of Frank Ramsay, who was a pretty smart guy, it was not obvious to me that Ramsay had it wrong. I make no claims for my deep understanding of the issues involved and I would not suggest that you or anyone else rely on my assessment. Thanks for the links to Brady’s papers. I will try to have a look at them. And no I have not read the book by Brunner and Meltzer. As you can see I have a lot of reading to catch up on.
LikeLike
Regarding your position on Ramsey’s review of A Treatise on Probability, I see. Also, I hope you get a chance to read that book by Brunner and Meltzer…perhaps it could help you in your advocacy of Market Monetarism.
And you’re welcome, David Glasner. If you want to find Dr. Michael Emmett Brady’s papers on the SSRN, follow this link.
http://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=1033456
LikeLike