The Death of Paul Samuelson and Selection Bias

by Mario Rizzo  

Paul Samuelson has died at the age of 94. There is already a big New York Times obituary lauding his many contributions and more will inevitably follow. Many economists will want to use this occasion to demonstrate how much they appreciate economics as a science and how this appreciation transcends ideological divides. This will reassure them that all is well in the queen of the social sciences.

Of course, I’d like to strike a discordant note. I did not know the man personally. He may have been a very good person indeed. I can only think about him as an economist who had an enormous influence on the character of economics in the second half of the twentieth century.  

So many people learned economics from his Principles text. Although I was a college student during the heyday of his text, my teachers never used it. Many people learned about the tradeoff between equity (=income equality) and efficiency from him (although he did not invent the idea). I didn’t accept his simplification of equity. Many people learned that the Soviet economy was simply one in which a different point on the equity tradeoff “was chosen” – although here too he believed the Soviets would one day exceed us in output. He believed their statistics. I never believed these claims.  

He did not understand the Austrian critique of the rationality of socialist economic calculation. On more than one occasion he made fun of the Austrians. I didn’t share his sense of humor. 

But what he is loved most for, by many professional economists, is his contribution to formalizing and mathematizing economic theory. His Foundations of Economic Analysis was used in many graduate courses until it was superseded by more complex and extensive mathematical economics produced by his students and their students. 

Was this for the good? I don’t want to go over that territory in any detail. I believe that economics would be far better with more philosophy and less mathematics. I believe that the profession has, because of the formalistic direction in which it has traveled, more than its share of idiot savants.   

The interesting thing for me is that when a man so influences the course of economics as Samuelson did those who enter the discipline after him are largely (though not entirely) those who like the kind of intellectual constructs, methods, attitudes he promoted – those who have a comparative advantage in this sort of thing. So in a way the whole phenomenon is self-congratulatory. “That Samuelson was great. He made economics into something I am really good at.” Selection bias at work.

The problem for Samuelson’s theoretical contributions is that one cannot easily point to any way in which economics is better rather than just different because of them. But even the differences are in a self-referential world. It is not as if because of his contributions we put a man on the moon or cured venereal disease. 

What he did accomplish was a kind of methodological exclusivism. Economists more and more looked down on the “literary economists” who came before them. An idea plainly in Adam Smith became a “new idea” because someone was able to oversimplify it and put in mathematics. Economists overestimated their contributions to knowledge because they confused knowledge about the world with refinements in theory. The fact that Keynes and Hayek became so relevant (once again) to current events and policy made clear, I believe, that refinements of theory were of limited value in understanding the world.  

He also contributed to the arrogance of economics vis a vis the other social sciences. Economists of many stripes laugh when the word “sociology” is mentioned (although I would admit that sometimes sociologists didn’t help matters).  

So, all in all, Paul Samuelson’s influence on economics was something I could have lived very well indeed without.

UPDATE: I was rather startled to read in today’s Financial Times that prior to Samuelson economics was simply “scattered thoughts.” (Note that the print edition uses the term “scattered  thoughts” in its headline while the web edition does not.) I was also surprised to learn that Samuelson “received a rigorous mathematical grounding in classical economics from Frank Knight, Jacob Viner…” I guess the distinction between rigorous logic and rigorous mathematics is lost on the Financial Times writer. Are there any equations in Knight’s work?

UPDATE 2: Paul Krugman apparently thinks that the caliber of criticism of Samuelson’s work is very low. This is what he culled from comments on his eulogy. He should read ThinkMarkets.

Addendum: I previously posted on a very recent article by Paul Samuelson.

34 thoughts on “The Death of Paul Samuelson and Selection Bias

  1. A very good summing up, Mario. There is a long article, maybe even a book, waiting to be written about PAS’s influence on (I deliberately didn’t say “contributions to”) economics. Or maybe the book has been written, by Philip Mirowski, “More Heat than Light: Economics as Social Physics, Physics as Nature’s Economics,” although it’s about more than PAS’s influence on the discipline.

    What about his influence on the social milieu as expressed in the growth of the State–and statist thinking–in the post-WW II era? I assume he’d deserve more than a footnote (or maybe not) in any reckoning of that.
    My introduction to economics was his textbook, which I found a bit offputting, for reasons I couldn’t fathom at the time.
    I guess it struck me as a bizarre combination of cutseyness and faux formalism, mixed with a certain arrogance–his self-ragrding reference to the Nobel Prize in economics wasn’t matched in any physics text I ever perused. How was it possibly relevant to learning the subject? I hadn’t become a libertarian or Austrian yet, but I distinctly remember the teacher singing his praises to the sky, and then bashing business (in general) in the next breath.
    And I must add that, whenever I read his articles (not more than half a dozen or so), I always felt that I was in the grip of someone with a frankly totalitarian mindset. This even though he said he wouldn’t want to work for the government, because then he (supposedly) couldn’t write or say what he thought–as if JFK was going to tell him what to think!

    The NYT obituary mentions his so-called contributions to a number of topics, including the idea of randomly fluctuating stock prices being normally distributed. This was taken as gospel until Benoit Mandelbrot studied actual markets in the early 1980s, and discovered, lo and behold, that these distributions were fat tailed, not normally distributed. He published a revisionist theory in The {Mis)Behavior of Markets, an eye opening book.
    But we know from Samuelson that it was the Austrians, with their a priori proclivities, who were out to lunch. That’s a lunch I’d like to sit down to–without PAS.

  2. Certainly the advent of Samuelson’s methods did not result in a Pareto improvement for all social scientists — nor could any possible innovation deliver such.

    As you have pointed out, some schools of thought found their ideas subject to an increased skepticism — whether deserved or not. A subset of those managed, after some reflection, to create stronger ideas from the criticisms and survive today; others didn’t and have fallen into the proverbial dustbin of history.

    It is to the Austrians’ credit that the movement has stood up to (and, to some extent, adapted to) the onslaught of mathematical analysis. I expect, however, that the consumers of economic analysis of all stripes have been enriched by Samuelson’s revolution, whether from the analytical component itself or from the competition-driven improvements in non-mathematical side of the discipline that followed.

  3. To merely second Mario’s insightful and right on target remarks, Samuelson (to use Jevon’s phrase about Ricardo) helped to shunt economics off on the wrong track.

    We should recall the enthusiasm with which he helped to herald the triumph of Keynesian Economics (see his contribution to Seymour Harris’ edited 1947 volume, “The New Economics,” which is a “worshipful” essay in which Samuelson was so happy to be under 30 years of age when “The General Theory” appeared, and therefore mentally open and receptive to the new revelation.)

    His “Economics” textbook indoctrinated generations of students into the “failures” of the market that could only be corrected by discretionary fiscal and monetary policy guided by wise and competent economics “doctors” like himself.

    He arrogantly insisted on the mathematical formalizing of economics. And though he studied with and personally liked Schumepeter during his student days at Harvard, he spoke in contemptuous terms of poor Schumpeter’s amateurish attempts to learn and do math.

    He was one of the intellectual godfathers of those who tried to introduce the social engineering mindset during the Kennedy “New Frontier”days — that generation of the “best and the brightest” that gave us, finally, both the Vietnam War, the Great Society, and Stagflation.

    When I took my first economics undergraduate class, I was assigned the 7th edition of Samuelson’s “Economics.” It had that, now, infamous chapter on comparative economic systems in which he extrapolated GNP growth rates for the U.S. and the U.S.S.R. until the the end of the 20th century. And where he projected the Soviets exceeding U.S. GNP by the 1980’s perhaps, but certainly by 2000.

    (Not only was his prediction about Soviet GNP wrong, he failed to predict that there would be no Soviet Union by 2000!)

    Samuelson epitomized all that went wrong with the mainstream economics profession from the 1930s to the current time.

    But as Mario correctly observes, his passing will be used to rationalize the formalistic bias and the interventionist policy views of many in that mainstream of the economics profession.

    If there is ever an economics Nuremberg Trials, Samuelson should be posthumously found guilty of intellectual “war crimes” against freedom and prosperity.

    Richard Ebeling

  4. I think we need to remember that Samuelson was influential because he made persuasive arguments. Whatever error rate we may wish to attribute to Samuelson, he said things worth listening to and thinking about 100% of the time or nearly so. We are so infused with Samuelsonian economics that we may not recognize the ways we have all been influenced by him. The thinking of even his critics have been shaped by him. I suppose Le Chatelier principles might be the most obvious example of that. His 1965 “Proof that Properly Anticipated Prices Fluctuate Randomly” is another. Those results, whether learned directly or through textbooks or later contributions, are just part of our thinking. We do not have to agree with his political economy or macroeconomics to recognize his greatness. And of course, as Mario suggests up front, we should mourn the passing of a great mind and a fellow human.

  5. I am simply taking this opportunity when people feel the need for praise to add some necessary balance.

    Consider the New York Times making grandiose statements: “Paul A. Samuelson, the first American Nobel laureate in economics and the foremost academic economist of the 20th century, died Sunday at his home in Belmont, Mass.” Emphasis added.

    So let’s see the other side.

  6. In chap. 7 (“The Ironies of Physics Envy”) of his book _More Heat than Light_, Philip Mirowski says that Le Chatelier’s Principle was invoked “as a metaphor for the reactions of a social system to outside stresses [by Robert Lindsay], and imported as a metaphor for monetary stability [by Francis Divisia]” (p. 380). This was not mentioned by Samuelson; nor is it mentioned in the Wikipedia article on Le Chatelier’s Principle.

    Discussing Samuelson’s 1941 article, Mirowski continues:
    “As in the previous bit of physics mongering, the actual mathematical model in the thermodynamics had little relationship to Samuelson’s mathematics, and therefore Samuelson was not invoking an exact metaphor. Both the central issues of the relevant conservation principles and the irreversability of phenomena dictated by the second law of thermodynamics were ignored, and therefore the Samuelson/Le Chatelier principle had little connection to the content of the thermodynamics. As one physicist has commented, Samuelson’s discussion ‘is entirely devoid of dynamical considerations’ (p. 381; references omitted).”

    Perhaps Samuelson was doing the economics equivalent of the discredited cold fusion.
    In the social sciences someone could get away with this, at least until an economist with actual training in physics, and the guts and integrity to call a spade a spade, exposed the Emperor for what he was.

  7. To Roger: True, Samuelson was persuasive to many, but that’s simply to repeat what we already know. It says nothing about the quality of his work.

    The persuasiveness of an argument is quite different from its validity or usefulness. In policy matters, he repeatedly, blatantly committed the Nirvana fallacy. I think his persuasiveness tended to be a kind of preaching to the anti-market choir.

    His methodological and theoretical contributions are persuasive to the extent one sees economics as mechanics; to me, it’s scientism.

    Whenever I pick up Conquest’s “Harvest of Sorrow” I’m reminded of Samuelson’s difficult question: “Is the superior economic performance of a Soviet system worth the cost in freedom?” What kind of a science is it where the great minds could be so utterly confused on matters of theory, empirical fact, and ethics?

  8. We should noted that Samuelson acknowledged the empirical failure of his Operationalist / Machian conception of economic science (taken mostly from Schumpeter) which had presumed a confirmation process whereby the functionalist quantitative relationships of math econ. would be verified by econometrics. Samuelson seems to have held on to his Operationalist / Machian conception of “good science” — and to have abandoned his faith in the presumed verification role of econometrics.

    So Samuelson at least acknowledged things were not working out as all as he had planned and assumed in his first great economic treatise.

    And Samuelson in the 1980s finally acknowledge that Hayek was right and Schumpeter was wrong about socialism and the economics of socialism, even in Samuelson never really understood what Hayek was saying in any non-superficial sense.

    Samuelson’s behavior was worse when it came to the failure of his scientistic “Revealed Preference” construct, or the Cambridge Capital debates, or his smearing of Hayek in his textbook visa vie The Road to Serfdom.

    As to “Austrian” economics, it’s clear that Samuelson never understood in the least Hayek’s explanatory strategy or Hayek’s epistemic approach or Hayek’s philosophy of science. Samuelson has a very crude conception of Hayek as an “apriorist” or “scholastic” — and least he pretended to in order to diminish Hayek as an economist.

    Who knows what Samuelson really believe or knew.

  9. I agree with Charles Steele. Everyone knows that Samuelson was not Milton Friedman in terms of values or ideology. However, that is not my point. My point in this regard is that his political economy was poor analysis. I do not criticize him because he did not share classical liberal values but, among other things, because his analysis of state action was simplistic.

  10. […] Mario Rizzo, an economist at NYU, would “like to strike a discordant note,” lamenting the quantification of economics that Samuelson’s influence brought about. I believe that economics would be far better with more philosophy and less mathematics. I believe that the profession has, because of the formalistic direction in which it has traveled, more than its share of idiot savants.    The interesting thing for me is that when a man so influences the course of economics as Samuelson did those who enter the discipline after him are largely (though not entirely) those who like the kind of intellectual constructs, methods, attitudes he promoted – those who have a comparative advantage in this sort of thing. So in a way the whole phenomenon is self-congratulatory. “That Samuelson was great. He made economics into something I am really good at.” Selection bias at work. […]

  11. My friend Barkely Rosser comments on Samuelson’s passing.

    I agree with him, of course, that “His influence is so great in so many areas that the key papers by him that lie behind the standard textbook accounts in many areas do not even bother to cite them.” I was getting at the same point above. Barkley makes an interesting point that goes to Mario impatience with idiot savants. Barkley says Samuelson “himself was generally personally aware of the flaws and limits of many of his own ideas” and “the greater sins have been by simplifying followers of his, the “sons of Samuelson,” who have been more responsible for codifying and spreading and enforcing the more simplistic versions of what he posited. The man himself was more complicated and self-aware than many gave him credit for.”

  12. Samuelson’s greatest achievement was to turn economics into a product which attracted lecture halls full of engineering students looking for an easy grade, and which could be easily taught and graded by TA’s.

    Some of these would be engineers recognized how easy this math game was, and decided to take over the profession.

  13. While it is true that Samuelson did make fun of Austrians on more than one occasion as Mario Says, it is also true that he expressed his appreciation for the work of Gottfried Haberler, especially in the field of international trade, and its explicit Austrian roots.

  14. Great post, Mario.

    Charles Steele also had a great point: “True, Samuelson was persuasive to many, but that’s simply to repeat what we already know. It says nothing about the quality of his work. … The persuasiveness of an argument is quite different from its validity or usefulness.”

    Far too many libertarians (maybe Austrians too, certainly conventional economists) forget this insight and equate truth with that-which-sells or persuades.

  15. I think Stephan and Charles err in saying that persuasiveness tells us “nothing about the quality of [a scholar’s] work.” It is no “proof” to be sure. I recall Keynes great quote puzzling over how Das Capital could have “carried fire and sword half way round the world.” One does not abandon one’s own judgment to others, even where there are many others who disagree! Nevertheless, academic success, especially on Samuelson’s scale, is a strong sign of quality. If Samuelson erred, surely his work had merit; it was high quality stuff. And if it had merit, perhaps we should ask ourselves why?

    Keynes confessed that he could not understand why Das Capital were so successful. But his open puzzlement reveals that he saw the problem. The right presumption is that the it is a great book. I don’t have to think the concept of surplus value is right in order to recognize that it addresses a real and serious question in economics, namely, where does the interest payment come from. Schumpeter later puzzled over the same problem and decided it was a tax on profits. I don’t agree, but I hope I see why that analysis has quality.

    Anyone who has given Samuelson the back of his hand should take pause and consider why other competent serious and informed minds have come to a very different judgment. The result of that inquiry might be to change one’s mind and decide Samuelson was right after all. But it need not be. However deeply I dive into, say, Foucault, I probably won’t come back to the surface a postmodernist. I might come back wiser, however, about the limits and weaknesses of my own system of thought. And I would be correspondingly better able to push that system forward and make progress with others (especially younger others) who sincerely believe Foucault to have the better end of the argument. Same goes for Samuelson. Coming after Mises, for example, he might have stumbled across one or two points missed by Mises and it can only improve one’s understanding to recognize the merits of his work.

    My comments do little more than remind us that John Stuart Mill made a fine argument for freedom of speech in “On Liberty,” and we should not neglect the considerations he raised in that great essay.

  16. Koppl: “I think Stephan and Charles err in saying that persuasiveness tells us “nothing about the quality of [a scholar’s] work.””

    I didn’t say that, Roger. I suppose it could say something about the quality of the work. But what I value is not the “quality of the work” but rather its correctness. And no, I don’t think persuasiveness is an indicia of truth. This to my mind is the intellectual equivalent of might makes right.

  17. I am glad that Rizzo gives an honest appraisal of Samuelson upon his death. It is rather silly for otherwise critical people to be flattering when someone dies. It isn’t some kind of virtue to die, so doing such doesn’t require any particular praise on the part of those still living.

    To further praise Samuelson upon his death only furthers an already-existing injustice: that he and his followers are praised and acclaimed instead of economists who deserved such praise (Mises, Rothbard, Hayek).

  18. Um, Stephan, er, you quote Charles Steele favorably saying precisely that. But I see that you didn’t really mean to deny the (imperfect) correlation between persuasiveness and quality. And I do not identify either quality or persuasiveness with truth. So any differences between us on these particular points might not be so big.

  19. “If Samuelson erred, surely his work had merit; it was high quality stuff. And if it had merit, perhaps we should ask ourselves why?”

    At least _part_ of the answer? His work created more publications and dissertation topics for others — and it was damn easy to teach.

    Follow the incentives.

    Isn’t that what economics teaches us?

    I understand this is only part of the story, but I’d submit it is the most important part — especially given the shown failure of central parts of Samuelson’s research project, and the self-evident failure of his research project as “science”.

  20. “academic success, especially on Samuelson’s scale, is a strong sign of quality.”

    There are various kinds of “academic success”. Samuelson’s peers were impressed with his work, but far less impressed than the young who didn’t know much, and now most economists a less impressed than economists of, say 40 years ago.

    Science advances one death at a time, just like perceptions of taste and fashion.

  21. Roger: You’ve smuggled an evaluation into your argument, I think w/o realizing it. You say we should “consider why other competent serious and informed minds” find merit in Samuelson.

    “Seriousness” isn’t in question. But whether one thinks Samuelson’s ideas are correct or not is precisely what “competent” means. (I’m embarrassed to cite my dictionary, but it confirms my belief that “competent” means having “suitable or appropriate knowledge.”) You’ve assumed that the mainstream is competent simply because it is the mainstream. But what if the mainstream (i.e. the majority) happens to be utterly wrong? In that case, their “competency” has to be questioned. (“Competent” is a somewhat loaded word; “correct” is better.)

    “Informed?” Paul Krugman wrote a brilliant essay in the Medema & Samuels methodology book in which he argues that before you critique something you must understand it deeply. But Samuelson and his followers were happy to dismiss Mises and Hayek as “hocus-pocus” w/o any serious understanding of what they were dismissing. Do you doubt this? Then where are their arguments that show the error of the Austrian approach?

    Samuelson had many ideas, he developed them in a very impressive form that persuaded many, and these ideas were wrong. No one here doubts that Samuelson wrote his ideas beautifully and persuasively. So what? The fact that most economists found them persuasive says zero about their validity.

    The “market test” is not in any way a test of truth or goodness — it is a test of whether people like something enough to buy it. Samuelson’s buyers liked Samuelson and there were many such buyers. That isn’t a test of the truth or validity or usefulness of his work. It is a test of the importance of his work, simply because of who the buyers are. Similar things can be said for Marx’s terribly flawed work. And yes, it can be interesting reading. But that’s of zero importance in assessing its value.

    Kinsella is quite right right that popularity of something cannot be used to determine a work’s value — if that value is something distinct from popularity.

    Finally, to other readers, I have to say that for the many years I’ve known Roger I’ve noted and greatly appreciated the great lengths to which he goes to be fair, charitable, and deal honestly with people. I think he’s utterly mistaken about Samuelson’s value, but not he’s not mistaken in his insistence that we take the ideas of everyone seriously and treat them fairly.

  22. Were people who were impressed by Samuelson’s arguments actually competent and informed, or is their being impressed evidence that they are not competent (at least in economic analysis)?

    Socialism is quite popular among those alleged to be intelligent.

  23. Ah, gheez, Charles. The generous remarks at the end of your post completely deflate my bluster balloon. Talk about fighting dirty! 🙂

    You seem to be imputing views to me that I have not expressed. When did I say popular = true? I did say popular (among relevant academics) correlates with *quality,* but I don’t think that’s what you meant by “value” is it?

    I’m not sure a remark I made over a The Austrian Economists applies to our exchange, but I liked it and I’m gonna repeat it here:

    I don’t agree with the general thrust of his policy preferences. I rank the Mises-Hayek trade cycle higher than he seems to have. I don’t agree with his early methodological writings, although he never practiced what he preached as Machlup showed. (Which suggests *error* not any moral failing, BTW.) And so on, But the guy sure knew his stuff. He sure knew how to construct a solid argument. The puzzle in my mind is why anyone would question that obvious point.

  24. Sorry for piling on comments one after the other, which I generally dislike. But I want to cut and paste a statement by Barkley Rosser from comments to a post at Econospeak.

    Samuelson’s influence, for better or worse, and not all of it has been good by any means in many ways, is so deep that one does not even notice it half the time. In very many fields he is like the structure behind the walls that one does not even refer to because it has been in place for so long and is so widely accepted. From the Principles level to much higher (although not in game theory), he is like the building itself.

    I think that’s right, which makes it hard IMHO to dismiss Samuelson as someone whose work had no value.

  25. Excuse me for making grand statements but Mario’s post has me wondering what is state of modern philosophy in general education? Many economists appreciate Samuelson because he makes his arguments clear, at least superficially. And clearly ideology was muddling the discourse of economics in Samuelson’s time. In my opinion, making ideas clear is the job of a philosopher; so how come mathematics was able to subvert philosophy in such a task? I think part of the answer has been mentioned in Mario’s post and the ensuing comments but it does make me wonder why philosophy didn’t have a better run (purely from the perspective of the consumer (academic economists).

  26. Roger: A valuable theory is a true theory, or at least one that helps us get closer to truth.

    An “unvaluable” theory is one that confuses or takes us away from the truth.

    Samuelson, like Keynes, produced a great deal of “unvaluable” work. It’s cleverly written, often framed in very sophisticated rhetoric and mathematical modeling, and since it is popular with people who believe it, and those people have been in powerful positions in academia, it is, in your definition, “quality.”

    I think none of us contest Samuelson’s persuasiveness (and that really does seem to be your standard of quality). But his ideas are wrong.

    We’re now threatened by all sorts of corrections of “market failures,” such as Copenhagen climate fixes and health care reform. We have Samuelson’s very bad positive economics to thank for this. If his economics were true true, his policies would follow. It’s his erroneous positive economics to which I object.

  27. I am no longer certain the place you’re getting your information, however good topic. I must spend some time studying much more or working out more. Thank you for great info I was on the lookout for this info for my mission.

  28. Greetings! I’ve been following your website for some time now and finally got the courage to go ahead and give you a shout out from Kingwood Texas! Just wanted to say keep up the excellent job!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s