Saturday, May 7, 2016

The Revenge Of Joan Robinson: Capital Theory Controversies Revive

It is easy to argue about what is the most important or influential thing that the late Joan Robinson did in economics.  More conventional types would probably point to her widely accepted and even textbookified Economics of Imperfect Competition.  Some would point to her later more philosophical and methodological writings about historical time versus analytical time.  Many Post (or post-) Keynesians revere her closenss to Keynes and Kalecki and see her as the godmother-founder of their movement  who could see the unity among their various bickering factions.  But others would look to her work on capital theory in the 1950s, her 1954 paper in the RES, "The production function and the theory of capital," along with the appendices to her 1956 The Accumulation of Capital, as well as her influence on Piero Sraffa to publish his famous book in 1960.  This was the trigger for the famous Cambridge controversies in the theory of capital that reached a culmination a half century ago when Paul Samuelson agreed that the "parable" he and some of his students had set forth was not true in general, given such paradoxes as reswitching (called the "Ruth Cohen curiosum" by Joan Robinson), with him declaring in 1966 in the QJE that, "the foundations of economic theory are built on sand."

As discussed elsewhere, we know that although the Cambridge, England side won an intellectual victory at that point, it turned out to be pyrhhic as those following the Cambridge, MA side simply ignored these results and proceeded to act as if nothing had happened, merrily estimating aggregate neoclassical production functions with well-behaved capital in them, with never a note or caveat about any potential problems, as well as building growth theory models, as well as general equilibrium macro models on such assumptions.  Joan Robinson may have drawn figures showing non-monotonic relations between "real capital" and "output," in the appendices of her 1956 book, as well as non-monotonic relations between interest rates and employment rates elsewhere.  But none of that mattered.  It was all to be ignored as oddities or whatever, things not to tell the children lest they become upset and lose sleep and cry when fed their porridge.

So it is a curious matter that the question of  capital theory has re-arisen recently.  It was just two years ago that Thomas Piketty's book, Capital, made the best seller lists.  Right now considerable attention is being paid to  Anwar Shaikh's voluminous magnum opus, Capitalism.  Both of these books take as their central issue that of the underlying forces driving secular trends in income distribution, particularly the division between wage incomes and profit or interest-based incomes.

Curiously, Piketty's theory remains firmly in the neoclassical camp regarding the questions raised by the old Cambridge, England school.  He briefly notes those controversies, but more  or less dismisses them, perhaps reflecting the influence of being at MIT for a long period of  time, even as he mocks excessive mathematical abstraction of much of modern growth theory.  Jamie Galbraith and others, including Shaikh, have taken Piketty to task for his dismissal of the issues raised by those old controversies, and many have criticized him for his super-aggregated view of capital that includes not  only land but also such things as collectibles, while others have taken him to task for  his arguments regarding r >g.  Where he has been most praised has been in his presentation of data supporting his broader arguments, as well as his ability to bring in broader historical and cultural support.  In any case, he is bringing the issue of capital and the foundations of income distribution front and center, even if he is not  doing so along British Cantabridgian lines.

Shaikh is an old fan of Sraffa's and a participant in the original  debates.  While he also does not present most of his theory as drawing on these old arguments, his approach is much closer to that approach in flavor and discusses it at some length sympathetically, even if he eventually draws more heavily on modern econophysics methods.  These fit nicely into his more Marxist approach,  even as he downplays Marx. But, of course, it was Marx who more sharply posed these questions regarding the nature of capital and how it affects income distribution, as well as power distribution, within societies.

I  note, of course, that the ever increasing emphasis  on financial theory in economics is itself an offshoot ultimately of capital theory, even if  it has gone off into its own niche with more emphasis on problems related to risk and uncertainty, while ignoring the broader issues of growth and income distribution.

I do  not know what Joan Robinson would think about this recent revival, but given what has been going on with income distribution trends in the higher income nations, I suspect she would not  be surprised.  And unlike many others, she always thought deeply about the problems of capital theory, taking account of historical and philosophical issues well beyond her peculiar graphs showing supposedly paradoxical relationships.

I close this by quoting myself from my 1991 book (p. 125):

"What really is capital and what does it mean for value, growth, and distribution?  Is it a pile of produced means of production?  Is it dated labor?  Is it waiting?  Is it roundaboutness?  Is it an accumulated pile of finance?  Is it a social relation?  Is it an independent source of value? The answers to these questions are probably matters of belief."

Barkley Rosser. 


Anonymous said...

I admit i am so peripheral to these dialogues and debates, sort of like someone way in the back at a campaign speech whose views and questions can be safely ignored, as well as pragmatically--- a presidnetial candidate is not some equilibrium in an n person game, though perhaps they should be.

(One could use all the heavy machinery of machanism desing, d g saari, s brams, and others on voting systems (post-Arrow), or any of the myriad formalisms developed to find social consensus (Hahnel and Albert had another one in their book on participatory economics/socialism , there's range voting...and one can't forget simulated annealing in physics---a solution to social frustration, poverty traps (perhaps even of the intellect, or marx's proverty of philosophy), Ramsey theroy, and 'haider balance' in sociology).

Perhaps at every university there is someone with their own formalism.

These could also be seen as forms of 'intellectual' capital. I learned somewhere there's like 6 kinds of capital---financial (money), labor, machines or technology, nature, human (knowledge and information), and social.

One could probably break them down into more elements (a la the periodic table, or alchemy---my academic field---my university now has a PhD program in this, and the Noble committee has put this field on the short list for new Noble prizes---basically Noble winners get 1 million$, and the winner in the new fields will be selected by merit, and contribution to inHumanity, by lottery---invented by shirley jackson---nice, uplifting, spiritual short film about farming in the midwest on youtube. (My favorite in this genre is 'a tour of streater north dakota' also on youtube---this is a recently discovered metropolis not too far from williston---its one of the megacities of the world, but somehow had escaped notice, like the continent of atlantis. I caught my first (green) snake in streater. Later i moved on to vipers and rattlesnakes. Because prices have gone up, a million is basically nothing, so the noble committee (described in the 'notorious' (and shallow) french book by 'the invisible committee') have partnerd with publishing clearing house lottery and powerball so one gets 30 million. Life will be one big lottery. While some advocate for a universal basic income (paine, friedman, bien, qatar, switzerland, uk ) instead everyone at birth will be given a lotto ticket).

Of course, definitions of capital (like god) are based on belief, and some say belief are a product of ideology, which may be due to 'class consiousness', or 'alienation', or a 'social construction' (pierre bourdiaeu (sic) though i prefer 'bo(u)nded rationality'. What people know or believe basically can often be fiarly easily be figured out or approximated by what they don't know. For academic papers, one reads the bstract, maybe the intorduction, glance through the text, and go straight to the references---see what is in , and what is not there. This can be formalized via calculus of variations and maximum entropy or its newtonian equivalent (eg verlinde's entropic gravity). What is, is the sum of its constraints and/or boundary conditions--see schrodinger 'what is life' on 'quantum jumps and states'---- p~Cexp-(Beta sum (E- u (ij)) 'grand canonical ensemble' or something like that--i'd have to look it up in wikip or derive it).

It may be nearly impossible to assign values to forms of capital. As they say the strength of a chain is as strong as its weakest link. said...

Well, media, you bring up some interesting points, and indeed I did not provide a full list of definitions or uses of the concept of capital. As you noted, there has long been discussion of human capital, and among ecological economists we have that very important natural capital.

I note for the record that the etymological origin of the word is from Latin, where is meant a herd of cattle. I leave it to you all to figure that one out and how it relates to all those other definitions.

marcel said...

Apropos Piketty: The title of his book was a very poor choice, at least for making it easy for economists to understand what it is about. (For the goal of maximizing sales, it was probably a good choice because it attracted many who were taken in by the thought that Piketty was a modern day analogue to Marx). The title would have been entirely accurate had it been Wealth in the 21st Century. For a study of the evolution or dynamics of income and wealth distribution, it is entirely reasonable and appropriate to aggregate all these types of assets, both real and financial, as he did.

To the extent that he dismissed JR and the CC controversy, IIRC my sense 2 years ago was that Piketty was hippie punching more than anything else, making certain criticisms to demonstrate his own legitimacy.* Both the title and the hippie punching were rhetorical choices, choices about how best to persuade and convince. I believe that he was mistaken in both choices but that that neither affects the substance of his argument.

*Paul Romer engaged in similar hippie punching against JR last year in his campaign against mathiness. This too was a rhetorical choice that had little to do with the substance of his argument. Delong used to engage in this tactic with some regularity, as did Krugman longer ago, although I have not noticed it much recently on the part of either.

marcel said...

PS. I had originally meant to include the following in the 1st paragraph of my previous comment, about why Piketty's title was a poor choice for the purpose of economists' understanding.

I recall many objections from economists when the book came out about how ridiculous it was to use a definition of "capital" that included all the asset classes in Piketty's definition, because (e.g.) clearly art is not productive and does not belong in productive function. This objection was clearly beside the point of Piketty's study, but it was very difficult to get some people to move past this objection rather than to dismiss the analysis altogether on the basis of that single point. said...

Another word derived from a herd of cattle that has capital-financial meaning is "pecurniary," which comes from the proto-Indo-European "peku," which in Lithuanian is "pekus," and "pecus" in Old German.


I think you are on to something with the "hippie punching" bit. I liked Piketty's book, but it is a mass of contradictions, with him going back and forth from being conventional orthodox MIT (hippie punching JR and all that) to his own heterodoxies. Why Romer thought he was scoring points by his particular hippie punching when he was denoucning "mathiness," I do not know, although maybe both of these guys are cocking an eye to those dour folks in Stockholm who never did give Joan Robinson what she deserved.

marcel said...

A major reason for hippie punching is to establish one's bona fides with the audience you desire, so as to be heard rather than dismissed out of hand (as one of them). I imagine that, despite being to some extent a Lucasian apostate, Romer was hoping to be respectfully heard by other Lucasians, whose work he wanted to influence.

I imagine something similar for Piketty. Ed Wolfe seems to be having a bit of his 15 minutes of fame since the financial crisis or perhaps since OWS; however, for many years, he labored on the fringes of the profession in an area that many thought not of especially good odor, i.e., the study of income and wealth distribution (Lucas and others have suggested that concern with these issues is indicative of ressentiment and thus disreputable). Earlier, both Piketty and Saez seem to have managed to avoid this, perhaps because they had demonstrated that they could sling math with the best of them (or at least Piketty had, I am less familiar Saez's history). Nevertheless, I imagine it remained a concern lurking in the back of his mind.

Bewley is one of the foremost mathematical economists of his generation, but he seems to have fallen into the abyss, at least in so far as the broad profession is concerned, since he developed an interest in reality.

stewarjt said...

Isn't one of the conclusions of the Cambridge capital controversy that there's no logically consistent foundation for the proposition that equilibrium prices and quantities are determined by supply and demand? Paraphrasing Lord Eatwell (if not quoting him verbatim), "That proposition is false!" Isn't this the conclusion that neoclassical economists avoid like the zika virus?

Add the foregoing to the fact that there isn't any empirical evidence showing that equilibrium prices and quantities are determined by supply and demand curve. To summarize, the theoretical foundation for the fundamental proposition of neoclassical economics is logically inconsistent. Plus, there's no empirical evidence supporting it!

In what other science is such a fundamental and devastating conclusion ignored by the subscribers to the dominant paradigm? This reveals the main reason why neoclassical economists ignore the capital controversy, it's theoretical implications and the lack of any empirical evidence. They do it for purely ideological reasons. They are, in fact, the "scientific" representatives of the capitalist class. Their vocation is presenting a theory that concludes capitalism is the best of all possible worlds. There's no exploitation in capitalism! Neoclassicals don't even ask the question of profit's origin. They use double think to avoid this issue by redefining accounting profit as a cost. Further, the unemployed choose to be unemployed! And unrestrained greed of individuals, not the social relations they inhabit and the laws of motion to which they inevitably give rise, are responsible for recurring crises!

Unknown said...

I'm glad you posted this follow-up. I happen to be reading J. Felipe and J.S.L. Mc Combie's new book, "The Aggregate Production Function and The Measurement of Technical Change: 'Not Even Wrong,'" which provides a meticulous critique of the aggregation problem in general and the Neoclassical conception of an aggregate production function in particular. The authors draw heavily on Franklin M. Fisher's work in this area. I can understand the resistance to J. Robinson's critique on ideological, if not psychological grounds, but the new book is so carefully argued, it should have some impact. On the other hand, F. Fisher, no bomb thrower by any stretch, made many of these arguments decades ago, but to no avail.

Anonymous said...

I wonder if a herd of sheep would count as capital, and if these are say 'heterogeneous goods' exchangeable for cattles (presumably different equilibria in a 'path dependent process').

Also 'capital as power' (bnarhives ; had debates with people like cockshott and kliman --LTV, falling rates of prophets, tempests in teapots---a model for climate . Power is a form of social capital, though its 'networked' with financial and human capital (IT), and 'cultural capital' (eg corn and potatoes from indians, ancient math from africa,religious traditions, tv reruns...

Also DC--the nation's capital. With sleight of hand waving, and math pyrotechnics, likely one could show all of these forms are equivalent.

(pecuniary sounded like an animal---peccarary. also called javelina---some dog walker got mauled by these recently. i used to see wild boars when i was in north carolina and hawaii. dont want to get close to them.)

i was in the dc jacobin reading group ( younger people's version of the nation mag, seemingly divided between 'curious no nothings in search of knowledge and a social group', liberal democrats, and vulgar marxists. They also discuss local issues like gentrification---it seems the idea is to gentrify neighborhoods, and then hold meetings on it, and point out they are the only legitimate experts on the issue---the people whose houses they live in, who were pushed out, are discouraged or banned from showing up. 'the new class' (bakunin, djilas, burnham). They had a study group reading all of marx's capital. I suggested they read a bit of more recent stuff---even paul samulson/ r goodwin. you mention anarchism in that group, u may be banned (as i am). They discourage anyone from reading anything outside the 'canon'---jacobin . first person i met who identified himself as a marxist could cite marx chapter and verse--- also belived for example that 'homosexuality is a disease of capitalism'. I mentioned the indian berdaches but he ignored that.)

(some jacobin people are trotskyites---tools of stalin against anarchists, went to mexico, where stalin sent them a gift packet. trotsky's secreatary moved to harvard, and wrote 'from frege to godel'.)

the idea that 'art' (one person's art is another's garbage) doesn't belong in the 'productive function' may be wack---in my area 'the creative class' idea is big. Art is an essential element of the IT and finance sector---having a 'club scene' is just like having a bathroom.

regarding beweler, i thought old papers from the 80's---eg grandmont (chaos theory, french) and many others (including some of JBR's) pretty mcuh were more state of the art, along with np-incomleteness and undecidability of n-person games.

i wonder what the relevance of variants of marxism are to economics. i consider marxism as a different interpretation or dialect of adam smith's capitalism, or even GET. set profit = exploitation. good=bad. similar to physics, where all the constants h-bar, c, G, k, are set = to 1 to cut down symboic clutter.

Haven't read Piketty's book, tho quite a few of the original papers, some with Szaz of UCB----mostly a collection of data---certainly more complete than earlier versions but nothing conceptually new. I sortuh agreed with Galbraith's critique, as well as Vanoufakis (sic, greek finance minister) tho i think they also used a clumsy formalism.

Regarding equivalence, i found this old book by M Kac in DC 's mlk library---it had a path integral formulation of classical diffusion and fokker planck equations, tho i gather it goes back to N weiner---who had his own (failed) attempt to model quantum theory as a classical diffusion process. . said...


I note that Saez got the John Bates Clark Award a few years ago, prior to Piketty pubbing his book. The JBC is even narrower and more rarely given than the Nobel, long dominated by the MIT/Harvard crowd. This probably explains at least some of the hippie punching, given that the underground buzz there now that Samuelson has been in the grave for some time is more along the lines of Solow, who seems to be fine with people running around econometrically estimating aggregate Cobb-Douglas production functions.

I am sorry that Bewley seems to have disappeared from sight. said...


The matter you raise has long been one of those issues that rages between the various factions of Post post- Keynesian economics. Indeed, Joan Robinson herself was of two minds on this. Her 1950s papers and appendices and revelation to the world of the tell-tale Ruth Cohen curiosum was all done within a comparative long-run equilibrium analysis, which the Sraffians would take up.

But then she turned against all this later as she emphasized the greater importance of historical time over analytic time, although this was not so much posed in terms of micro partial equilibrium supply and demand analysis. said...

Greg Hill,

The book by Felipe and McCombie is well done. McCombie in particular has been the person at Cambridge U. in England continuing the effort deriving from Joan Robinson's 1954 article. Indeed Franklin Fisher followed up excellently, with him ironically at MIT. But, for better or worse, his work was largely ignored, and I fear the same will happen to that of Felipe and McCombie.

I note that while Samuelson agreed with the Cambridge (UK) critique, his top student Robert Solow always took a harder line, and while I think he knows better, he does not denounce or question openly the scads of empirical estimates of Cobb-Douglas (and variations) aggregate production functions for all kinds of models and even policy analysis. said...


Lot of stuff there, and I shall not comment on all of it. So...

Indeed sheep and cattle were early heterogeneous capital. I think that cattle beat sheep for the etymological origin of the word because they are bigger than sheep and worth more in any market where both are available. Power relations were there in the beginning.

You have brought up yet another form of capital not yet mentioned, the more recent and very fashionable "social capital." So we have the Robert Putnam version, which is positive for growth and inclusive, except when it is not (bridging vs bonding soccap), and then we have the Bourdieu version, which involves more seriously problems of social debt and power, more intensely drawing on Continental sociology traditions.

Robert Vienneau said...

Ian Steedman and Arrigo Opocher's 2015 book, Full Industry Equilibrium, is good for thinking about the implications of the CCC for theories of partial equilibrium.

Many mainstream economists seem to "think" that the CCC is subsumed by General Equilibrium Theory.

Anonymous said...

to add more verbiage, one could ask if equating apples with origins is ok.

its beyond my 'pay grade' but it almost appears that the SMD theorem is itself possibly equivalent to the CCC problem of heterogeneous goods---ie the fact that there is no unique assignment of values or utilities on the preference (or consumption , rather than production) side. There are heterogeneous preferences, and interdependent utility functions.
Also, many things are path dependent or 'non-commutative'---giving someone a job first ---eg child labor---and then education, ) may not be equivalent to giving them an education first, and then a job.

(This relates to another topic currently discussed by another group i follow--the relation between syntax and semantics. A particular syntactic construction, sentence or pattern can be interpreted by many people differently---ie it is not mapped one to one into a unique meaning, just as people value different 'buncles of goods' differently.
(This is also related to 'whorfian economics' discussed by keith chen (yale)---the effect of language on economic behavior).

Bourdiue also discussed 'linguistic capital' (a variant of intellectual plus social capital---eg the stories of 'wild children' raised by wolves who dont speak; they don't have much, though possibly a different kind or currency.
Presumably or ideally, yen and dollars are equivalent, but not in reality due to transaction costs, and the finite speed of communication.

Under the us constition 'all men are created equal, but some are more equal than others'---a nice math problem to formalize.

Myrtle Blackwood said...

I found this definition of 'capital' on the web tonight:

"wealth in the form of money or other assets owned by a person or organization or available for a purpose such as starting a company or investing." [Where 'investment' is defined as "the action or process of investing money for profit."]

Which effectively makes 'capital' a function of (monetary)'demand' and the profit motive.

I'd settle for wealth and abundance and forget the rest.

"Whoever loves money will never have enough money. Whoever loves luxury will not be content with abundance. This also is pointless."
Ecclesiastes 5:10 said...




Oh, and I forgot about "cultural capital," yet another one.


I included financial capital in my list.

Myrtle Blackwood said...

re: "Myrtle, I included financial capital in my list."

Yes. But 'financial capital', while not the most critical form -'natural capital' is - it's the one that is the most problematic.

Barkley, without terms such as 'capital' being clearly defined, and we work in the knowledge of which particular definition is being referred to at any given time, then it is easy to conclude that "the foundations of economic theory are built on sand."

Myrtle Blackwood said...

""What really is capital and what does it mean for value, growth, and distribution? Is it a pile of produced means of production? Is it dated labor? Is it waiting? Is it roundaboutness? Is it an accumulated pile of finance? Is it a social relation? Is it an independent source of value? The answers to these questions are probably matters of belief."

A fair summation.

Except that 'economics' may be more of an 'ideology' than a 'belief'.

Myrtle Blackwood said...

PS: I see 'media' has summed it up: "Of course, definitions of capital (like god) are based on belief, and some say belief are a product of ideology, which may be due to 'class consiousness', or 'alienation', or a 'social construction' (pierre bourdiaeu (sic) though i prefer 'bo(u)nded rationality'. What people know or believe basically can often be fiarly easily be figured out or approximated by what they don't know..."