Tuesday, August 2, 2011

Gary on Paradigms and Economics

In a comment thread Gary writes: "Most scientific disciplines have some sort of over-arching paradigm to them (or series of paradigms which are some times linked, sometimes not) - in biology you have evolution, cell theory, etc. Nothing like that exists in economics - obviously economics has gone through a variety of paradigmatic shifts (e.g., Smith, then Ricardo, then marginalism, then Keynes, etc.), but the paradigms are never really complete or uncontested and thus they tend to be fairly weak. And yes, there are debates about evolution by evolutionary scientists, but those debates regard specific mechanisms and so forth, no one doubts the basic story about the evolution of species. Instead of something like the Big Bang or wave particle duality or what have you, you've got in economics something more a kin to one of the undemonstrated UFTs. Most of the grander economic theories that are supposed to explain a lot look a lot more like M theory or loop quantum gravity theory; they don't seem to be able to convince people like the Big Bang does or evolution or quantum theory in chemistry."

I think he's deeply misunderstanding economics. First we do have at least two paradigms that are universally accepted:

(1.) the gains from trade/the law of supply and demand (the "theory" is supply and demand but of course the accepted understanding of the gains from trade is what motivates the behavior in question). The law of supply and demand is like natural selection for economics. It supports everything and is universally accepted.

(2.) marginalism/subjectivism

I think the paradigm shifts we go through are also fairly thorough and happen at about the rate that they do in any other science. In my mind the first real thorough paradigm we had was Ricardianism. Then there was marginalism (a micro paradigm), then Keynesianism (a macro paradigm). And I'm not sure we've ever come out of the Keynesian paradigm.

Moreover, Keynesianism as a paradigm is relatively uncontested. This is what Arnold Kling refers to as the AS/AD mentality. I think Kling unfairly caricatures the perspective, but that's another discussion entirely. We don't use the word "Keynesian" to refer to the whole paradigm, of course. When we use it in every day language we're really refering to a more specific class of models and perhaps even some policy persuasions. So we oppose "Keynesian" to "Monetarist" and "New Classical", etc. when really monetarists and new classical economics are very firmly in the Keynesian paradigm - they just tweak a few assumptions in their models to try to solve puzzles that they feel earlier work didn't solve. The paradigm, though, is fairly stable. To borrow Gary's words from the comment, the big debates you see between economists "regard specific mechanisms". I'd probably agree with Gary that a lot of models are comparable to UFT work - but then nobody ever claimed that we've broken a new paradigm in economics and nobody in physics is quite ready to claim that for UFTs yet. So? This is where the paradigm leads us. Do seemingly intractable puzzles bother Gary? Kuhn says these are solved in a couple ways: first, with new empirical technology, or with a paradigm shift (and often the former facilitates the latter). We're getting new empirical technology all the time. We have longitudinal employer-employee household datasets now. We didn't have that in the 1990s. Now we can test the new advanced in theory in the 1980s that was just awarded a Nobel Prize last year. Who knows where these advances are going to eventually lead us? I don't.

But Gary's simply wrong on the paradigms. We have a few big ones and we definitely have laws that are universally accepted.


People get really weird when they talk about the scientific status of economics, despite it's very long and very firm status as such. The reason, it seems to me, is that people don't like to be pinned down and studied. That's what we do to inanimate objects and that's what we do to animals but we like to think there's something "special" about humans. We don't like to be reminded that we're just a highly evolved primate and that we are as amenable to scientific study as other primates.


  1. I'm sorry, but no. Even the law of supply and demand is highly contested in your field; see Paul Samuelson for an example.

    "Moreover, Keynesianism as a paradigm is relatively uncontested."

    Evolution is not "relatively uncontested." Neither is quantum theory in chemistry.

    When you have to throw in modifiers like that I don't take your claims seriously.

  2. Also, from what I can tell, marginalism isn't (a) universally accepted, (b) when it is it is often combined with some other concept to make it useful (von Neumann) and (c) for a couple of decades it fell completely out of favor.

  3. Ok, I have a busy day. My last word for the day is this:

    the more I look at economics the more it looks like history to me, and the less it looks like physics or chemistry or biology.

  4. Wait - who isn't a marginalist?!?!?!

    And what doesn't Samuelson accept about supply and demand?

  5. As for your last comment on people not liking to be studied; well, that doesn't apply to me, otherwise I wouldn't like to read stuff like this:


    Personally, I love epigraphy (there are volumes and volumes of published epigraphy out there if you are interested - one of the great contributions of positivist history).

    Marxists clearly don't accept marginalism, and it seems to me that some sort of "objective" measure of value is always creeping into one or another analyses of marginalism.

  6. re: "Marxists clearly don't accept marginalism"

    And creationists don't accept natural selection. So?

    (I'll have to take your word that this is true - I don't know much Marxism)

  7. Dk wrote:

    We don't like to be reminded that we're just a highly evolved primate...

    I think we had this argument before, Daniel, but what do you mean by the word "just" here? Couldn't a physicist recommend doing economics using Newton's laws (or quantum mechanics if you prefer) and saying, "People are just collections of atoms, as amenable to our study as particles in an accelerator." ?

  8. And I believe the last time we took the time to have a think on the implications of my use of the word "just" I was a little curious why you thought a collection of atoms was so impoverished. In other words - I'm not sure I'm smart enough to justify decisively divorcing matter from meaning.

    Whatever the implications of materialism, I certainly don't intend "just a highly evolved primate" to imply any lack of meaningfulness or spirit or whatever else on the part of humans.

  9. Paul Samuelson said, "I side with Sraffians" ("A Modern Post-Mortem on Böhm's Capital Theory: Its Vital Normative Flaw Shared By Pre-Sraffian Mainstream Capital", Journal of the History of Economic Thought, V. 23, N. 3 (2001)).

    Paul Samuelson also said, "My vantage point in the discussion was not neoclassical. It was Sraffian!" ("Samuelson's 'Reply on Marxian Matters'" Journal of Economic Literature, V. 11, N 1 (March 1973)).

    Paul Samuelson also said, "In this age of Leontief and Sraffa there is no excuse for mystery or partisan polemics in dealing with the purely logical aspects of the problem." ("Understanding the Marxian Notion of Exploitation: A Summary of the So-Called Transformation Problem Between Marxian Values and Competitive Prices", Journal of Economic Literature, V. 9, N. 2 (June 1971)).

    John von Neumann's exploration of prices of production in his growth model is also not marginalist and not subjectivist.

  10. The full-cost pricing controversy was a challenge to marginalism in the middle of the 20th century.

  11. "Keynesianism as a paradigm is relatively uncontested"

    is this true? when I open a macroeconomics journal I don't see IS/LM or AD/AS models, I see variants of Kydland and Prescott. Even the New Keynesian stuff is just DSGE with Keynesian frictions built in.

    but getting back to the point, economics for sure has a dominant paradigm. Agents in economic models are constrained maximizers. The rest is just details. This approach is ubiquitous in economics, and is pretty much what differentiates us from the other social scientists.

  12. anonymous -
    Well a lot of the old Keynesian relations still pop out of DSGE models.

    I think it's important to differentiate models from paradigms. Certainly models change. When I think about a Keynesian paradigm I'm thinking about it more in the sense that Friedman did when he said "we're all Keynesians now".

  13. At least quote Friedman correctly:

    "In one sense, we are all Keynesians now; in another, nobody is any longer a Keynesian."


  14. "The full-cost pricing controversy was a challenge to marginalism in the middle of the 20th century."

    I like Fred Lee on the full-cost controversy. Dan Hausman and Philippe Mongin have an interesting 1998 article on how economists ignore empirical evidence.

    To me and looking at it from the perspective of the sociology of economics, economics is, in Kuhnian terms, in a pre-paradigm state. I don't think our host will acknowledge that. As far as I can see, mainstream economics is a matter of the willfully ignorant indoctrinating the willfully ignorant.

  15. Gary - I did quote him correctly, I just didn't say the whole thing. The whole quote of course supports what I'm saying as well - and Friedman's later clarification when everyone went nuts over it supports what I'm saying.

    We're all "Keynesian" in terms of how we think about and talk about the macroeconomy scientifically. None of us are Keynesians (not even me) if you're simply thinking of Keynes as a model.

  16. RV -
    You're right I don't agree with that. Kuhn noted several indicators of passing out of pre-paradigm stage, and economics seems to me to pass all of them.

    The Ricardian period is squishy (sociology always is, isn't it?). But I would say micro passed out of it over the period from 1870 to 1890 - from the marginalists to Marshall's publication. And macro passed out of it somewhere between 1936 and 1947.

    Perhaps I'll post on his or write on this in the future when I have Kuhn sitting next to me (I'm on a site visit right now), but he's pretty specific about what makes a discipline move out of the pre-paradigm stage. If you think that's not relevant, you're talking about a different theory of scientific progress.

  17. Daniel,

    I'm sorry, but no - he clearly said that one uses "Keynesian language" (and that is what he meant by his claim); he did not claim that Keynes brought some sort of scientific revolution to the field (Keynes would have to have been a scientist for that to be the case anyway).

    What does Kuhn mean by paradigms? In simplest terms he means an achievement of such significance that it is (for a time) acclaimed and used universally. No economic idea has met that criteria; they are all pre-paradigm.

  18. Gary,
    yet at the same time, virtually every first year PhD student receives the exact same training.

    economics is done with constrained maximization. this isn't about individual models, this is about how we build models and the fact that we build models at all. this is universal. so I don't know what you're using as a definition of paradigm, but I think it is very clear how economists interpret data, organize ideas, build new theories, and test theories against each other (at least to academic economists), and it is hard to do it any other way and get published in good journals (again public information to economists).

    btw Daniel, Keynesian economics is not the uncontested paradigm of macroeconomics. Friedman said his thing along time ago, before modern macroeconomics came to be. sure some ideas cross over, but as a framework for studying the macroeconomy, things like the Lucas critique are much closer to an organizing principle. how exactly is modern macro organized by the Keynesian paradigm?

  19. Anonymous,

    Theodosius Dobzhansky (from whom much of what we think of as the evolutionary synthesis derives - it is from him that we get the concept of gene mutation) stated the following (here I quote from memory - the phrase being a title of a paper of his):

    "Nothing in biology makes sense, except in the light of evolution."

    I struggle to come up with a similar claim regarding economics.

  20. Gary -
    1. What language we use to talk about things is a huge deal in Kuhnian understandings of paradigms.

    2. What about "nothing in economics makes sense, except in light of supply and demand" or "nothing in economics makes sense, except in light of constrained optimization"???

    What exactly do you think economics consists of - reason.com blog posts about why government is bad and Daily Kos blog posts about why government is good?

    Anonymous -
    How is the Lucas critique opposed to the Keynesian paradigm? I'm a little confused by your critique. You rightfully tell Gary that economics isn't about individual models, but then you seem to turn to me and make a big fuss about the evolution of individual models in macroeconomics!

    I think Gary was on to something with his original point about UFT and modern macro. This - I think - is true. And this accounts for the menagerie of macro models we have. Macroeconomics emerged with a paradigm shift away from general equilibrium approaches to whole economies - however, after that break there have still been various attempts to provide microfoundations consistent with marginalism for those macro models. And that's produced a ton of stuff and a great big question mark for economists. But I don't see every shift in the favorability of these models as a new paradigm shift - at least in any Kuhnian sense (which is of course the sense in which I'm using it).


All anonymous comments will be deleted. Consistent pseudonyms are fine.