Samuelson on Hayek -- HT: Larry White and Barkley Rosser

|Peter Boettke|

Larry White and Barkley Rosser alerted me to this quote yesterday with posts on FaceBook.

The Coase-Samuelson generation were brought up witnessing the great debate between von Mises and Lerner-Lange concerning the feasibility of socialist rational pricing to produce Utopia. (That was a reprise of earlier Pareto-Barone-Wieser-Taylor debates.) Many contemporaries believed Lerner-Lange triumphed in the debate. I came to believe that Friedrich Hayek was the true victor.

Barkley stopped the quote there, but I think the next paragraph is also important to consider:

Under static conditions where all is known or knowable (to whom?), whatever optimal states laissez-faire might occasion, so could some computer solution or some algorithms of play the game of competition also achieve. But in the real world all is changing, even in the time it takes me to write this sentence. Hayek has been persuasive -- not in Whig ideology or in declaring that moderate reform of laissez-faire leads inevitably down the road to totalitarian socialism but -- in arguing that experience suggests that only with heavy dependence on market pricing mechanisms can there be realized quasi-efficient and quasi-progressive organization of societies involving humans as Darwinian history has bequeathed them. If a reader does not find the Hayek dynamic arguments persuasive, I will not here argue the matter further.

Here is the link to Samuelson's original article. The article's title is "Some Uneasiness with the Coase Theorem."  In some fundamental sense, Samuelson is arguing for the irrelevance of the zero transaction cost Coase Theorem, and demanding a more rigorous analysis of the Coase insights concerning the positive transaction cost world.  As he says at one point,

Ronald Coase is, of course, better than is Stigler on Coase. Many disciples cheerfully agree: There is no Coase Theorem. The case of Zero Transactions Cost is a polar parable, and the great value of the new revelation comes in grappling with various degrees of transaction costs, in analyzing their complexities, in devising property rights and legal procedures that help to minimize social deadweight loss. However, there is still needed - from the pen of Coase or anyone else - a coherent analysis of just what is afoot in the non-polar real world. Anecdotes do not a science make. If the polar cases, the easy cases, are defective in their purported analysis, then the subject is still in a primitive stage of development.

There is no doubt that more work must continually be done on the institutional level of analysis.  But to hear this demand come from Paul Samuelson is both exciting and extremely frustrating at the same time.  As I stated in my own obituary of Samuelson, he was the transformative figure in the discipline of economics in the 20th century.  However, I don't consider the transformation that he led a positive development for truth-tracking in economic science.  It is amazingly impressive no doubt, but it took us farther away from a real understanding the nature and signficance of economic science.

Economics in Samuelson's hands was transformed into social physics.  Pure theory sought to be institutionally antiseptic as a scientific goal, and this was in fact achieved.  It is against this backdrop that one must ultimately negotiate the intellectual terrain that folks like Alchian, Buchanan, and Coase travelled in the 1950s and 1960s as they sought -- to use Buchanan's phraseology -- to develop a "genuine institutional economics" that "becomes a significant and important part of fundamental economic theory." (See Buchanan, The Demand and Supply of Public Goods, 1968, p. 5)

As David Levy and Sandra Peart have revealed in their work on the Virginia School of Political Economy controversy, Samuelson was not particularly reserved in his private assessment of the work of these 3 thinkers when it came to funding proposals and their respective statures within the professional economists pecking order.  He considered them hard working, but biased and not as capable as those within his circle.  This was no doubt his honest assessment, but in retrospect reading this article on the Coase Theorem is very revealing.

First, let me get this off my chest --- Hayek wrote a letter to Samuelson complaining about the way Samuelson depicts Hayek's argument in The Road to Serfdom in his Economics textbook, he responded to Hayek he would not do so again.  Samuelson depicts Hayek as making an "inevitability" argument, Hayek does not make such an argument.  He makes an instability argument, or a law of unintended consequences argument, that means that the logic of the situation would require more and more direct control unless course is reversed.  And as Ulrich Witt has argued, Hayek's warning suffers the fate of the endogenous public choice theorist --- Hayek's very "prediction" will prove false precisely because politcal decision makers heeded the "prediction" and changed their behavior accordingly.  Anyway, my complaint here is that Samuelson against depicts Hayek's argument this way in order to dismiss it. So while accepting Hayek's major point about centralized economic planning and the nature of the dynamic market economy, he seeks to dismiss Hayek for his extremism. This leads me to my second point.

Second, if you read carefully here just as in the case of Hayek, I would argue that Samuelson does not get the import of the Coase (and I would add Buchanan [see Cost and Choice) contribution of the 'problem of social cost' and the zero transaction costs world.  Textbook public economics in the 1950s taught that markets fail due to the existence of externalities, and that state action through appropriate tax and subsidy schemes could correct this failure.  The textbook model employed to identify the externality was a zero transaction cost world.  This alone created a problem, and furthermore, the solution belied a failure to engage in behavioral symmetry in analysis.  The upshot of the Coase (and Buchanan) analysis was that on its own terms the sort of textbook Pigouvian public economics was either redundant -- individuals in a zero transaction cost world would negotiate away their conflicts -- or non-operational, public officials in a positive transaction cost world will be unable to identify the precise tax or subsidy scheme to internalize the externalities as well.

This basic and elementary point about the nature of assumptions and the analysis that follows from them is too often missed in these discussions.  Coase (and Buchanan) if anything were consistent and persistent proponents of the elementary principles of the economic way of thinking --- as of course was Alchian.  To the super sophisticated this sort of emphasis on basic reasoning seemed unsophisticated.  Yet, if we take Samuelson's remarks above seriously, this was also the case in the Mises-Hayek versus Lange-Lerner debate.  Slow and steady, in these episodes of economic controvery, wins the race.

Samuelson argues in the cited article here that we need more research in the institutional level of analysis.  Too bad he didn't see it in his time in the work of Alchian, Barzel, Buchanan, Cheung, Coase, Demsetz, Kirzner, North, Olson, Ostrom, Rothbard, Tullock, and Williamson.  The entire Coasean program is to identify the deals that are struck, to see entrepreneurial solutions, to see how situations of social conflict can be turned into opportunities for social cooperation.  Economics is about exchange and the institutions within which exchange takes place.

I think it is important to stress once again Kenneth Boudling's warning about the impact that Samuelson's Foundations would have on economic understanding.  "It may well be," Boulding wrote, "that the slovenly literary borderland between economics and sociology will be the most fruitful building ground during the years to come and that mathematical economics will remain to flawless in its perfection to be very helpful."

How prophetic was Boudling, and how sterile was Samuelson is something each economist must decide.  The interesting historical point is that Samuelson and Boulding were the first two John Bates Clark Medal winners, so the choice of going in one or the other direction was actually a live one at the time.  It seems in the piece above, Samuelson's own words seem to be suggesting that there are good reasons to believe the economics profession made the wrong choice, and that Boulding's literary borderland between economics and sociology --- what we now term New Institutionalism --- is indeed the more fruitful research program.

Comments

[Barkley Rosser]

Just two notes:

1) I agree that Samuelson missed the large new institutionalist lit focused on how varying tranactions costs (admittedly a fuzzy term including even the kitchen sink) alter how thins turn out, and that Coase was focused on this part of it, not the polar case emphasized by Stigler and textbooks. So, as Cheung noted in his "Fable of the Bees," we see orchard owners paying beekeepers to locate next to their orchards thereby internalizing positive externalities of pollination in that near-zero transactions cost case, whereas we are unable to get feasible global agreements about global warming where transactions costs are enormous.

2) I also agree that Samuelson responded inappropriately to Hayek's letter about Samuelson claiming he made a strong "slippery slope" argument in his Principles book. OTOH, it must be noted that Hayek never went out of his way to discourgage doggerel versions of RtS, such as the Readers' Digest version that essentially made such an argument, which is widely made by many who claim to be "Hayekians" even in current policy discussions (perhaps some of the "amateurs" that have had Pete upset recently).

[Roger McKinney]

Barkley, maybe Samuelson only read the Reader's Digest version.

[FriedrichHayek]

"Fruitful" in what sense?

Economics is driven by self-interest & the bureaucratic structure of academia. "Fruitful" must be understood from the methodological individualist point of view of incentives facing the academic.

"Fruitful" = having a simple & univocal formal metric of "good" which makes teaching simple, Ph.D dissertation evaluation simple, peer review, advancing ones career simple, picking Clark winners & Nobel Prize winners simple, etc.

"Fuitful" = having formal puzzles with "objective", fixed solutions.

"Fruitful" = having endless statistical projects to crank through and publish.

"Fruitful" = having endless statistical problems to teach from a textbook.

etc.

[FriedrichHayek]

"Hayek never went out of his way to discourgage doggerel versions of RtS" -- Rosser

Note that Hayek's permission was never asked nor granted before the Reader's Digest "version" was printed. That was the decision of Hayek's publisher, and Hayek wasn't involved in the decision.

[FriedrichHayek]

John K. Galbraith was intentionally mis-reporting the content of Hayek's The Road to Serfdom and other aspects of Hayek's work in official AEA publications, for no reason but partisan advantage against a "political opponent" -- left economists did this *routinely* to Hayek.

What Samuelson did was common practice among left economists -- intentionally misreport Hayek's perspective in order to marginalize Hayek as an economists.

Arrow did it also. As did Keynes.

It was accepted academic practice in the day.

[FriedrichHayek]

I should point out that Robert Solow is still using this strategy of misreporting Hayek's work as a technique in intentional marginalization of rival scientific work.

[Rafe Champion]

How does Samuelson rate as an economist in view of his commentary on the Soviet economy?

When will Stanley Wong get proper recognition for his demolition job on Samuelson on revealed preference? http://www3.nd.edu/~pmirowsk/pdf/Wong_Introduction.pdf

Interesting to note, despite Mirowski's dismissal of Wong's reference to Popper, the "little gem" is Popperian through and through, due to the influence of Larry Boland. It is an outstanding illustration of the power of Popperian/Austrian Situational/Institutional Analysis and the reconstruction of intellectual/scientific problem situations.

Incidentally Wong's experience, described in the preface to the new edition, also illustrated the dysfunctional intellectual climate of the economics profession, to the point where he quit and made a career in the law.

[Barkley Rosser]

I happen to own the collected scientific paper of Samuelson, 7 volumes, several of them over 1000 pages in length. I have strongly disagreed with him on a number of issues and did so to his face the first time I ever met him over 40 years ago. It could be argued that he is the ultimate whipping boy of much of my complexity economics work, which he took with good humor, more or less.

There are many things he did that were either outright wrong, such as his presentation of the Soviet economy growth prospects in a bunch of editions of his Principles text and his characterization of Hayek's "slippery slope" views, along with lots of other stuff.

But, I would agree with Pete that whether one thinks he is the most brilliant economist who ever lived or the ultimate demon of economics of the 20th century (or something in between, with being brilliant not necessarily ruling out also being demonic, as it were), he probably was the most influential economist of the 20th century with a massive influence running through many different sub-fields to the point that much of it people forget was originally from him in the field textbooks, to, yes, the foundations as his PhD thesis addressed. Whether you like him or not (and my feelings are and were very mixed), one cannot deny his enormous influence, for better or for worse.

[Rafe Champion]

Nobody can deny his influence, just look at the way his textbook was standard for many years. The result of that is rather alarming if you think that he scored a trifecta of errors (abuse of maths, Keynesianism and the welfare state). How long will it take to wash Keynesian thinking out of the system if so many generations of students (now citizens) grew up on Samuelson's standard text?

[FriedrichHayek]

Paul Samuelson straight up explained why his textbook economics took the form it did -- he constructed his textbook to fit the skills and interests of math-centric MIT engineers.

Think about that just a little bit. The textbook was no designed to capture and convey an understanding of the way the world works, it was designed to fit the math interests and math skill of undergraduate engineers.

I don't no were he explained that, but he was very up front about the genesis of his textbook economics.