Monday, May 2, 2016

The Legacy of Joan Robinson

Over a week ago Joan Robinson came up in connection with the ongoing controversy over Gerald Friedman's analysis of Bernie Sanders's economic plan.  When criticized for apparent problems with it he declared that those who understand what he is doing are "Joan Robinson economists."  In going after him, Justin Wolfers ridiculed him for this, referring to such people as a "sub-tribe of Keynesians," which is not inaccurate, but somehow comes across as very sneery.  Noah Smith took this up, declaring that this was the coup de grace by Wolfers, and a commenter on his thread, "Britonomist," dismissed Robinson as someone who admired the North Korean economy, "maybe the worst economy ever in the world," with Noah thanking him for this piece of information.  I made some comments on this there, with pgl picking up on this for some further discussion.  Other than to note the reasonableness of  Peter Dorman's request for  Friedman to provide his model publicly for people to figure out what he is doing (which is becoming less and less important as Bernie's chances of getting the nomination seem to be approaching epsilon), I have no comments on all of that contretemps.  Rather, on thinking about it I shall  return to Joan Robinson and talk about her and her legacy for modern economics, given that various people have been using her name in vain.

Joan Violet Maurice Robinson (1903-1983) was without doubt the most important woman economist born before 1930 and maybe still the most important woman economist ever.  While she would end her days as a radical leftist, she came out of an elite background, her father a baronet and a major general in the British army in WW I, with her maternal grandfather a famous surgeon who taught at Cambridge University and one of maternal uncles a polymath who advised Winston Churchill.  She was close to her father, who was forced to resign from the British army near the end of WW I for publicly reporting on misconduct by the government in managing the war.  In his final years until his death in 1951 he lived with his daughter in her half of the house she shared with her husband, E.A.G. (Austin) Robinson, whom she married in 1925, who himself was an important Keynesian economist, adviser of the British government, and founder of the International Economic Association, with them producing two daughters.  A story I have from a primary witness is that she had a graduate seminar in her half of the house, and one time there was an unpleasant odor there.  The participants realized that her father was sitting in a chair in the room dead.

While she nearly got a Nobel Prize, and certainly deserved one, she suffered professionally from being a woman.  She was only appointed a Lecturer at Cambridge in 1937, well after she had already published several highly innovative and influential  works.  She was only made a Full Professor at Girton College at Cambridge University (which she had attended) in 1965, the year her husband retired from his professorship.  Rumor has it that she came closest to receiving the Nobel  Prize in 1975, the year Kantorovich and Koopmans got it for linear programming (which created a major stink among mathematicians who said that at a minimum George Dantzig should have shared it).  I do not know if she was thought of as a possible third for them or a replacement for them, perhaps with Piero Sraffa sharing, who also never got one while arguably deserving it, who could have shared it credibly with Leontief in 1973 for input-ouput analysis.  The Encyclopedia Britannica reports that her leftwing political views may have played a role in her not getting it.  I also heard it from a primary source that Assar Lindbeck, the committee's dominant figure then, once said that if either Joan Robinson or James Buchanan got it, it would be over his dead body, although  Buchanan did get it in 1986, with Lindbeck still on the committee and not dead, although Joan Robinson had  been dead for three years by then.

As evidence that she was clearly in contention in the mid-70s, I shall report something I observed on an elevator in the New York Hilton during the 1973 AEA meetings (the first I ever attended).  Lionel McKenzie, another who never got the prize but should have, was talking to somebody else.  McKenzie told this other person that "they are going to give it to Joan Robinson next for her Economics of Imperfect Competition, but she will refuse it."  As it was, she never got the chance to do so.

 Speaking of that 1933 book, that was her first major publication and remains one of her most important, indeed worthy of a trip to  Stockholm in and of itself.  Among other things in it, she invented the word "monopsony."  While she later wrote less about monopolistic competition, one can see that it remained very much on her mind if one reads her excellent 1977 article in the JEL, "What are the Questions?" a good overview of how she viewed economics near the end of her life.  She spends quite a bit of it going on about the issue of monopoly power and its importance.  I note that this is one area where her concerns are very relevant to current economics, with many now posing that increased monopoly power in the US economy may be playing a role in secular stagnation.

She was indeed a core Keynesian, one of the three people thanked by Keynes himself in the Preface to his 1936 General Theory.  She also supported Kalecki, whom Keynes had in to  Cambridge, but by all accounts did not like.  In 1937 she wrote her influential essay on "Beggar thy neighbour policies," which made the concept associated with competitive devaluations widely known, although the term had appeared before previously, used once by Adam Smith and also by a British economist named Gower in 1932.

In 1941 she published her famous Essay on Marxian Economics, in which she rejected the labor theory of value and basically supported redoing Marx along Keynesian and Sraffian lines.  She would indeed later praise both Maoist China and North Korea, but saw China in particular  as possibly offering another way of modifying Marx along useful lines.  However, Robinson was always known for her pithy remarks, and one from that era was "There is only one thing worse than being exploited, and that is not being exploited" (that is, unemployed).

The 1950s may have seen the high water mark of her work.  She set off the Cambridge  capital theory debates with her 1954 paper in the Review of Economic Studies, "The production function and the theory of capital," in which she took apart the idea of aggregate capital, with Paul Samuelson in 1966 agreeing that she was right.  The first time I ever met Samuelson (in the early 70s) I gave him a hard time about this issue, and he just completely agreed with her and said that capital must be modeled as being heterogeneous.  One of the more hidden but very important roles she played in the 1950s was to work on Piero Sraffa to finally complete his short, but important, 1960 book, Production of Commodities by Commodities: A Prelude to a Critique of Economic Theory.  He had been working on it for 35 years, but it was still only a prelude to a critique, not a critique itselfn.  Samuelson claimed that if he had published it in 1930, he would indeed have shared the Nobel Prize with Leontief.

In 1956 she published what is probably her magnum opus, although now widely ignored, The Accumulation of Capital, which in contrast to her later critiques of analytical equilibrium analysis in favor of looking at "historical time," was in fact a study of various equilibrium growth models, many of  which she provided amusing names  for such as "bastard golden age" and "creeping platinum age."  She did not generally use formal equations but rather favored figures and graphs backed up by clear verbal descriptions and discussions.  Apparently in 1949 Koopmans asked her to be on the board of the Econometric Society, but she refused on the grounds that she did not want to be part of something that produced things she could not read.  After 1960 her work increasingly moved towards more methodological issues, such as her 1962 Economic Philosophy, as well as work looking at development issues, especially in India, but also her highly controversial work on China and North Korea.

I commented on the China and North Korea matters in the linked-to threads from Noah and pgl, but I shall repeat some major points here.  Regarding North Korea, she visited there in 1964.  She accurately reported that its economy was performing much better than that of South Korea, which many now may not believe, but was true,with a real per capita income probably twice that in the South.  What happened was that the South did not grow after the Korean War in the 1950s under the corrupt Syngman Rhee.  It took off after he was overthrown by the military dictator Park Chung Hee (whose daughter is currently the president in the South), who instituted a strong indicative planning drive run through state-owned banks until  he was assassinated in 1979.  During the 1950s the North Koreans had  successfully followed a Stalin-style command centrally planned forced industrialization, which had resulted in impressive results by 1964.  While it continued to grow for some time after that, the South began to catch up, surpassed it in the early 1970s, and, of course, today is far ahead of the now seriously impoverished North, although it has been able to produce nuclear  weapons.  In 1977, in a single paragraph, she predicted the North would absorb the South. She may not yet have become aware that the South was ahead of the North in per capita income, but she may have been more influenced by the absorption of South Vietnam by North Vietnam only two years earlier in 1975.  She was not completely out of it on her observations of the Koreas.

She can be more sharply criticized regarding her views of China, regarding which she wrote a book defending the cultural revolution.  Apparently before  she died she pulled back on some of her admiration for Mao, but this is clearly an area that one can criticize her political and economic views more than on her observations on the Koreas, where indeed North Korea really was ahead of South Korea when she made her most detailed study of them. But she wrote much more about China than about the Koreas, visiting there on several occasions, although she visited India many more times.

The final, and maybe most important, influence of Joan Robinson today is on Post Keynesian economics, or post-Keynesian economics, with her preferring this latter spelling, the British version.  Indeed, although I cannot  prove it, I have heard it that she was the one who  coined this term, as she did so many others.  That there are two spellings is due to Paul Samuelson also using this term with her spelling to describe what we now call his "neoclassical synthesis," with him referring to "post-Keynesian eclecticism," which supposedly represented his position.  When American Post Keynesians got seriously going in the 1970s, led by Paul Davidson who founded the Journal of Post Keynesian Economics,they favored that spelling to distinguish themselves from Samuelson's formulation, which the British and other non-Americans were never bothered by. In any case, Joan Robinson is widely viewed as perhaps the founder of the movement, varied and eclectic as it is today, or at least the godmother and main inspiration of the movement.  But I  note that it has many sub-varieties, with the Wikipedia entry on it having a curious "family tree" that has 8 different boxes: Keynes's inner circle (which included  Robinson and her husband), Cambridge Keynesians (which also  included Robinson), early North American Post Keynesians, Kaleckians, Sraffians, Fundamentalists, Kaldorians (I show up on that list), and Modern Monetary, with some other lists adding Institutionalists, and with some people loosely on the tree but not in any box or founding one such as Keynes himself, Richard Goodwin, G.L.S. Shackle, and Hyman Minsky (who always claimed he was not a a "Post Keynesian"), along with some others.

So when Gerald Friedman wrote of "Joan Robinson economists," I think that he had in mind the general current set of Post (or post-) Keynesian schools, even if he may or may not have been thinking about any particular one.  I note that Bernie Sanders's top economic adviser on his Senate committee is Stephanie Kelton, on leave from UMKC, who is allied with the modern monetary (not to be confused with the "new monetarist") school, one of those listed in the Wikipedia family tree (and she is in that box there).  But I know that many others like  Bernie Sanders, and most think well of Joan Robinosn, even if they may not all necessarily agree with Gerald Friedman's analysis.  But then again, as Peter Dorman has noted, we have not yet actually seen his model, so one can think what one wants, I guess.  But clearly various aspects of Joan Robinson's thought and career are both relevant and currently influencing many economists today, including many who have never heard of  her through some of her ideas simply entering into basic textbooks, such as "monopsony."

Barkley Rosser

Note on 5/3/16:  Upon further checking while Joan Robinson did discuss golden ages in her The Accumulation of Capital, it was only in the later Essays in the Theory of Economic Growth, 1965, that she expounded most fully on the theories of bastard golden ages and creeping platinum ages, although apparently she initially wrote about them prior to that book.  Also, just for the record, she did use equations from time to time, if only sparingly.


Peter Dorman said...

Thanks for putting in a word on behalf of Saint Joan, Barkley. I'd add that there are unfinished fragments in her work pointing toward a reformulation of international trade theory. (I wrote an article on this some time ago.) Essentially, she wanted to integrate the micro side with the macro, in which national income is endogenous and the neat comparative advantage results do not hold. I'm afraid her nonmathematical resources weren't sufficient to get the job done, but in fairness it's not like anyone else has done it either. My view is that this is one of the great unfinished tasks in economic theory, and anyone embarking on it can do worse than read JR's various fulminations on the topic. said...

The matter of her mathiness or the lack thereof is an interesting one. The fact is that several of her better known arguments involved things that were deeply mathematical that she managed to express in her way using figures and the English language. One finds this going on even in her first book where for example she presented the theory of monopoly with segmented markets such that the demand curve facing the monopolist has non-monotonic elasticities associated with varying stretched of flatness separated by steeper portions. It has since been shown that this can lead to chaotic dynamics by a monopolist facing such a demand curve, although she never pushed it that far. But she clearly understood what the model was about and so on.

Needless to say the capital theory debates involved a lot of mathematical issues, and she pretty much followed what the issues involved were, even if again she did not use a higher powered mathematical language herself in dealing with it. Indeed, one of the more impressive things about her work is how mathematically sophisticated much of it is without her expressing it explicitly in math per se. In this she harked back to some earlier figures, although some of these had the math but chose not to use it in their main expositions, with Marshall a leading example of this with his "put the math in the footnote and appendices" approach, and I would throw in Mill and Marx in this regard as well.

As it is, I am not surprised by your remarks about her views on trade theory, where I think something like that may be the case, but I have not seen her work on that specfically.

ProGrowthLiberal said...

One of the key issues in this Presidential debate will be where to set the minimum wage. The argument that a higher minimum wage necessarily ignores the prospect of monopsony power. Dr. Robinson's coined the term that progressives should not is one possible explanation why the Econ 101ers model of labor markets does not fit the empirical data.

MaxSpeak said...

FYI, Stephanie moved from the Budget Cttee to the campaign, and I believe she is now back at UMKC.

kevin quinn said...

Barkley: I agree that Robinson's prose was wonderfully precise - she did with words what others needed math to do. Interestingly, a close rival to her in this respect was Milton Friedman!

Someone needs to a good history of the reception of monopolistic competition theory. I remember textbooks from my undergrad years that presented it with barely concealed contempt. They would present Chamberlin's excess capacity result and then wave their hands and say something like "well, that's the cost of variety" - despite existing literature that actually did the welfare analysis and found no presumption that the zero-profit equilibrium produced the optimal amount of variety. This was before Dixit-Striglitz -

jcb said...

A question about the Cambridge Capital Controversy.

BR points out that Samuelson eventually accepted Robinson's point that capital must be modeled as heterogeneous. Did this have any practical effect on the way that capital was actually modeled in post-Samuelson American economics? Or was heterogeneity relegated to the status of a complication that did not fundamentally affect the simple model? If the latter, do you agree?

The Jobbing Leftie Historian said...

I have a recollection of a popular review essay by Solow- perhaps in the 1980s? - where he says that Robinson and Cambridge UK won the capital controversy but it didn't matter because no one listened. My memory is probably tricking me as his precise formulation - does anyone know the paper and the argument

Peter Dorman said...

The capital controversy has re-emerged in debates around Piketty. See Jamie Galbraith's rather angry riposte, for instance. (I referred in an EconoSpeak post to Piketty's "pages of shame" about this topic.) If you accept the notion of a composite capital good that earns its marginal product in a competitive economy, Piketty's parameters are dubious. But P measures financial wealth, which is entirely appropriate but a different animal altogether. An intro econ textbook of which I'm fond introduces the issue, points to Tobin's q as a rough measure of the difference between "real" and "financial" capital, and points out that even at the aggregate level q is not fixed -- it's conjunctural. said...

jcb, mh, and pd.


Luis Enrique said...

Barkley, how would you compare Robinsons behaviour towards NK and China with Milton F and Chile? I have read those sympathetic to Milton setting record straught, as they see it, as you do here. But also those who are hostile condemning in strongest terms and using Milt as evidence of how despicable his team is. Is that mirror to Joan and supporting and opposing teams? said...


Not a simple matter, and I confess that I have read only a little of what Joan Robinson wrote about China, which was pretty extensive, whereas Milton Friedman wrote very little about the Pinochet regime. That said, he may have had more influence on its economic policies than she did on Chinese policies, albeit through some of his students who advised the Pinochet regime in its early years, the "Chicago boys." I think in both cases they justified themselves by saying that they did not support political repression and executions, but sought to defend and improve economic policies in those nations. They may both have been misguided in their efforts.

Unknown said...
This comment has been removed by the author.
Unknown said...

"A summing up" by Samuelson in QJE, (1966, Vol 80 No 4) accepts one of the main criticisms of Robinson and others (Garegnani, Pasinetti etc) with regards to re-switching, or the fact that it is possible to have the existence of points where lower profit rates might be associated with lower capital-output ratios. This goes against one of the fundamental premises of the production function. I think Robinson credits Ruth Cohen for first alerting her to the possibility.

Robinson later said that the re-switching argument was not the central point of objection to the production function, simply because the production function makes more sense as a method to compare two different economies at a given point of time, rather than using a given production function to trace the movement of a single economy through time, which is what the neo-classicals tended to do. It took me a long time to fully comprehend it, but her device of thinking of each point on the production function as one where time runs perpendicular to the page on which it is drawn is superb.

The fact that she could provide an internal and external critique of the production function grounded in logic and rigour without the use of mathematical logic that carried the day against Samuelson and Solow is a measure of her contribution and intellectual prowess. Which makes it a pity that she can be dismissed out of hand solely based on her admittedly problematic support for the economic experiments in China and North Korea.

Robert Vienneau said...

I like Robinson's Essays in the Theory of Economic Growth. I get the feeling, in reading it, that she first did the math and then translated it into english. I know that that is not the case.

I finally think I understand that Ruth Cohen story. I think Robinson wanted to acknowledge Sraffa for capital-reversing. They had many talks. But Sraffa would not agree to the acknowledgement. But Robinson would not say it was her original idea. So Sraffa suggested that she credit Ruth Cohen.

AXEC / E.K-H said...

Joan Robinson and the ‘throng of superfluous economists’
Comment on Barkley Rosser on ‘The Legacy of Joan Robinson’

For Keynes it was obvious that economists had messed up economics: “The classical theorists resemble Euclidean geometers in a non-Euclidean world ...” (1973, p. 16). And he knew that the fault was in what today is called microfoundations: “For if orthodox economics is at fault, the error is to be found not in the superstructure, ...but ... in the premises. (1973, p. xxi)

Consequently, Keynes started the macrofoundations research program in the General Theory formally as follows: “Income = value of output = consumption + investment. Saving = income - consumption. Therefore saving = investment.” (1973, p. 63)

These formal foundations are conceptually and logically defective because Keynes never came to grips with profit and therefore “discarded the draft chapter dealing with it.” (Tómasson et al., 2010, p. 12).

Keynes’s original blunder kicked off a chain reaction of errors/mistakes:
• All I=S/IS-LM models are are false since Keynes and Hicks.
• Keynes’s profit conundrum has not been solved by After-Keynesians.
• Keynes got the employment equation/Phillips curve wrong.

Joan Robinson shared Keynes’s assessment of Orthodoxy but did not, just like the rest of After-Keynesians, spot the fatal mistake/error in Keynes’s approach. She was well aware that there was something wrong with profit but missed the solution by a hair’s breadth (Robinson 1956, p. 402, for details see 2011, eq. (32)). At this analytical peak things went inexorably into reverse.

After-Keynesians can until this day not tell the difference between income and profit. For want of the pivotal concept of their subject matter, economists fail to capture the essence of the market economy. The actual situation is this: the familiar profit theories are provably false (2014). Therefore, ALL models that have been built and are still being built on either the Walrasian or the Keynesian axioms are false.

This is why ‘economic science’ does not exist until this day. The gigantic heap of peer-reviewed incoherent and inconsistent models is what Feynman called cargo cult science (Wikipedia). Scientific ethic demands that this is clearly communicated to the general public. Therefore, as an all-important first step the word ‘sciences’ has to be deleted from the “Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel”. It is important to clearly understand that ALL economic policy guidance lacks sound scientific foundations. This is not at all a matter of political left or right.

Joan Robinson opened the Ely-Lecture in 1972: “When I see this throng of superfluous economists ... I am reminded how much the profession has grown since the thirties and how many more there are now to suffer from the second crisis than there were to be discredited in the first.”

So, who is discredited in 2016? All those who have still not realized that the pivotal economic concept profit is ill-defined and that all I=S/IS-LM model are false since Keynes. Featuring in the “throng of superfluous economists” are:* Krugman, Wren-Lewis, Rowe, Farmer, Syll, Glasner, Radford, DeLong, Douglas, Mitchell, Keen, Smith and, last but not least, Barkley Rosser.

It is high time to fulfill Joan Robinson’s scientific legacy: “Scrap the lot and start again.”

Egmont Kakarot-Handtke

References said...

Oh my. Egmont has come over from Economistsview with this rant. I responded at some length there. Given that he did not provide all his references here, I shall note that I went into his 2011 paper. His big claim, as can be seen here, is that up until he came along, economists have not understood what profit is. This failure means that all economics has been no good, as you can see.

So, when I looked at his 2011 paper (not specifically listed here), he defined profits at a certain point as being "sales revenue...less costs." This most certainly is not in at all different from just about how all the rest of the known universe, including both economists and accountants, defines profits as. Of course most others have gotten into big debates over how one measures costs, with the ongoing differences between how economists and accountants measure profits focusing on this matter of what counts as part of costs and what does not.

But this old chestnut is not what has Egmont all worked up. It turns out that his great discovery is that firms retain some of the profits, while distributing some others. His modeling and description of this only applies to corporations rather than partnerships or single proprietorships, as he defines distributed profits as being dividends times the number of shares. In any case, it has long been known that firms make investments out of the earnings they retain, not out of those distributed. There are indeed reasonably complicated models that take account of this distinction. It is not a big deal. Those old structural equations models that get used in the basement of the Fed, in competition with the VAR and DSGE ones, have enough equations that they do this sort of thing, modeling how different parts of firm earnings get distributed and used. There is nothing else to this. It has long been known and does not involve any overturning of all economics. Not even close.

Egmont has read a lot and seems to be fairly smart. But this stuff is just gonzo silly. He has been posting extended stuff on Noahpinion, now on Economistsview, and now here. But this is just trivial silly stuff, not remotely worth all the wild rhetoric he has engaged in.

Oh, and I have no idea exactly who the "After-Keynesians," are, but Egmont, I can assure you that this will not catch on any more than your obsession with modeling retained versus distributed earnings as the fundamental driving force of the economy.

kevin quinn said...

Barkley: We After-Keynesians are a small group, in number if not in spirit. It is our belief that after Keynes, nothing matters anymore. Our food has lost its taste, our once rapier-like wits are dull as butter knives, our appetite for life vanished like the morning mist. We are no bother to anyone; please tell Egmont to forbear from his insults! said...

I can do that, Kevin.

Egmont, please forbear from further insulting the miniscule, but very loyal and seriously depressed, group of After-Keynesians! They cannot take it any more.

Anonymous said...

A few irrelevant comments---

1. I had to look it up on wikipedia since i forgot it, but to go out on a limb, i'd say the 'cambridge capital controversy' ---robinson vs samuelson ---actually looks to be trivial (assuming i understand it). Of course this is in 'retrospect'--alot of trivial, common sense things nowadays were major puzzles at one time (even numbers zero, -1, and imaginaries were---though imaginaries i think still are a bit, or quite a bit, of a puzzle when one gets to the difference between classical and quantum mechanics---eg why the schrodinger equation is not the fokker-planck, assuming it isn't. ).

I looked at the part of the wikip article about the math---cobbs-douglas function. Yup. I guess i always reduce this to saying the 'whole is not the sum or its parts'---ie the issue is nonlinearity.

The comment about Piketty is 'spot on'. (I actually think Piketty really formalated his theory in a totally wrong way, but i guess this packaging 'sells'. Its like analyzing the economy by going to a store and buying stuff, and then working backward to see where it comes from. Also, its basically a timeless model, like alot of econ---pointed out by georges rogescue.)

You cant sum individual cobbs-douglas or other utility functions and get another one except in the one case cited. You can't even trade 2 haves of a baby for one baby, for some reason (probably culture; eg suppose 2 families each have a baby but one has a boy and the other has a transgender, but each family wanted both a boy and transgender. It might seem simple but if you cut each in half and trade, you dont reach the 'optimum' without some fairly drastic surgery). Interesting permutations arise if you add the 3rd sex---eg girls--and allow larger families to trade. See Krugman 'babysitter problem'.

As they say 'a dollar is not a dollar'. (Even 100 pennies are not a dollar---they dont accept pennies many places, due to friction or transaction costs, or prejudice).

The issue with cobbs-douglas nonadditivity is analogous to that in the group selection problem in biology (eg D S Wilson, who i once met, even stayed at his house, but i never mailed in my completed application ---i figured i only had a 50% chance to get in due to the competition--2 people, one spot---so i decided i could use the 50$ application fee for something else. Also, he told me basically my job would be to run his computer---he had some very nice computer simulations---but i had already done 2 computer simulation projects and was tired of it and wanted to do math. I showed him a paper by Joel E Cohen of Rockefellor U in Proc Royal Soc which had very similar results but was more math oriented. These were all generalization of Lotka-Volterra equations to more than 2 species, plus alot of other interactions. S Smale sort of wrote on this in the 70's and a few people have followed up on it). (next step is to unify this with statistical mechanics; one paper is halfway there).

(one question i've always had is besides its intuitive sense, why cobbs-douglas is so commonly used. The log of it has interesting properties, a la entropy.)

Nonadditivity is also the issue in Born's rule in quantum mechanics.
The wave equation is linear, but reality isn't.

2. Egmont has many posts over at 'real world economic review'. Always says the same thing, and has a long set of papers on SSRN, most of which appear to be the same paper. He reminds me of Chomsky---egmont rejects the idea of utility function on favor objective reality, chosky rejects semantics since reality is only syntax. After awhile, they change their mind, and say they invented the idea of utility and semantics.

3.I wonder what the value of all this writing is. One could do an energy based theory of value---so academic writing is most valueable, then finance, campaigns, talk radio, films, and finally infrastructure and food. said...

The funny thing about the Cobb-Douglas production function is that Cobb and Douglas invented it to fit the stylized fact from the early 20th century had fixed shares of labor and capital income in the US, a function that would explain this if the US were an optimizing micro firm. However, those shares have not been at all constant in more recent years, with, as we all know, the share of capital income rising noticeably compared to labor income. But somehow lots of economists continue to use C-D supposedly on empirical grounds, that it fits the data, even those who know about the CCC. Weird.

I have coauthored with David Sloan Wilson, very interesting guy.

Anonymous said...

ps. just saw the JEBO issue I've also read richerson , gowdy, etc.

if you havent seen it there's free online issue of edge magazine debate between s pinker , d s wilson, etc on 'the false allure of group selection'. Pinker is 'old guard' in his views. (Condensed matter physics i see to be a basic physical analog.) (S A Frank has good discussions/papers on group selection though he appears to favor Pinker's terminology---biology is chemistry and physics, as is economics, but its easier to work with larger units like organisms and species.)

I see D S Wilson's 2016 essay about evolution and economics, and how out of touch econ is with that. Nelson and Winter did use analogies to population genetics, H Peyton Young may have, R goodwin used ecological analogies, and there are more (based on stochastic, brith and death processes). I guess this would be considered out of the mainstream. R Brian Arthur's 'path dependence' model used the same math used in synergetics (Haken, Weidleich), as did E O Wilson in Genes Mind and Culture. Black-Schole equation is itself basically a fokker-planck equation used in all of these models (also by Joseph McCauley). Ormerod ('worrying trends in econophysics' if i recall used a more explicitly written brith-death process).

But this stuff looks different from what one sees in standard economics (is-lm, and similar ones ---what appear to be General equilibrium inspired optimizations).

People using evolution and physics analogies generally aren't doing things like analyzing possible impications of things changes in interest rates or trade deals; they work on a more theoretical level or 'ideal gas' type cases (eg Yakovenko).

Evolutionary game theory may come closest to unifying all the approaches but still tends to be abstract.

I sometimes wonder why for utility or production functions some people choose linear, or standard nonlinear, or instead cobbs-douglas function, which seem all applicable to both an individual's budget allocation (or production) and an entire economy when viewed as a social welfare function. One could include forms of capital besides labor and financial capital.

Intutitively it seems its describing 'making a perfect recipe'---one needs a combination of ingrediants each in perfect porportions. Society needs people distributed over occupations optimally, though likely there are multiple optima, with no way to choose among them (due to ideologicaly, politics, history--general equilibrium assures there is a unique one, if everything is infinitely dividable--a continuum---money and products are represented by real numbers, people have perfect information and rationality, and identical--no heterogeneous preferences---if i recall ).

(As a related aside, one problem i and a few others have noted with Yakovenko's income model ---i saw your joint review paper, which seemed to contain some possible fixes) --is that while it fits stylized facts about income distribution , it also implies using the analogy of the ideal gas model, that while society will always have invariant maxwell-boltzmann distribution of income (or some mixture)it implies that over time individuals all have the same average income---colliding individuals will gain and lose income like gas molecules exchange energy---they all ergodically pass through same microstates or income classes. (if one introduces saving this changes).

Anonymous said...

«I guess i always reduce this to saying the 'whole is not the sum or its parts'---ie the issue is nonlinearity.»

Nonlinearity? Does not seem to be stretchable that far. Capital (like GDP) is a vector of incommensurable quantities (from "oil refineries" to "stock of rearview mirrors"). One can build an index of capital (just like for GDP) using a vector of dollar prices for each quantity, and then say that nominal capital is N dollars. Then you can build an aggregate production function that takes X dollars of nominal capital, Y of nominal wages, and gives Z dollars of nominal output.

The problem that J Robinson pointed out, which was not unknown before her, is that the price vector is not independent of the vector of incommensurable quantities, and that therefore the aggregate production function is meaningless (under-determined) without a detailed theory of the price of each element of the capital vector, which I think only K Marx has produced ("Das Kapital" as the title is a bit of a giveaway really), and there is a bit of controversy about that :-).

The JB Clarkian and neoclassical schools claim to have a detailed theory of capital, which can be summarised pithily as "capital prices emerge during [intertemporal] tatonnement", which can be translated accurately into english as "lalala I don't hear you!" :-). P Sraffa anyhow showed that even in very simple neoclassical models *without money*, just using vectors of incommensurable quantities, capital prices can be all over the place; something that also was known as "Wicksell effects". His real major contribution is that he solved the biggest problem of theoretical political economy for centuries, how to define a "numeraire" that is independent of the distribution of income/production.

The big issue here is that the mere *assumption* that there is an aggregate production function denominated in dollars and which has certain further restrictiion is vital to proving that (in a certain broken and inconsistent family of models) marginal productivity (plus initial endowments) determine income, which can be translated into english as "the rich deserve everything, the poor, maybe, only subsistence". Which is the central truthiness of microfounded Economics, as even J Stiglitz pointed out recently.

Anonymous said...

«theory of capital, which can be summarised pithily as "capital prices emerge during [intertemporal] tatonnement"»

I have to add here that in the more sophisticated versions the handwaving is slightly camouflaged by saying that such "tatonnement" involves the demand and supply schedules of capital, as defined using the aggregate production function, so it all squares.

This was Samuelson's original position, except it was proven to him that this works only if there is a single type of capital, and does not work mathematically (or logically) if there are two or more types of capital, and he had to accept that.

The other neoclassicals just added "single type of capital good" to the assumptions of the model, and carried on singing the lyrics of "The methodology of positive economics" by fraudster M Friedman.

Some of the more recent and most advanced research shows that the central truthiness holds in neoclassical models even if there are multiple types of consumer goods, but I think that holds only other restrictive assumptions.