true and non-trivial social scientific propositions

A tangent from Teppo’s post…  A well-known, true story goes like this.  Nobel-winning economist Paul Samuelson, arguably the single biggest influence behind the mathematization of economics in the 20th century, was once challenged by the mathematician Stanislaw Ulam, who was skeptical of the value of social science, to name a single social scientific proposition which is both true and non-trivial.  Samuelson couldn’t think of anything on the spot, but years later he realized what, in his mind, is the correct response:  comparative advantage (see wikipedia).

Regarding comparative advantage, Samuelson said:

That it is logically true need not be argued before a mathematician; that it is not trivial is attested by the thousands of important and intelligent men who have never been able to grasp the doctrine for themselves or to believe it after it was explained to them.

I think Samuelson’s choice of comparative advantage is an excellent one and probably the best choice from economics.

But what would be other good answers to Ulam’s question?  Is there an obvious answer from sociology?  From political science?

About these ads

Written by mikemcbride

June 19, 2008 at 12:19 pm

21 Responses

Subscribe to comments with RSS.

  1. Answers require agreement regarding triviality, but perhaps:
    * cognitive dissonance
    * fundamental attribution error
    * loss aversion / prospect theory (garnering an econ nobel also)
    These are largely micro concepts. As for true and non-trivial macro concepts, I think we’re early in that process, but some candidates may be:
    * constructivism (irony alert, in that constructivism would problematize Ulam’s concept of “true”)
    * influence via structural equivalence
    * decoupling


    June 19, 2008 at 1:00 pm

  2. I’d always heard it as ‘true, *surprising*, and non-trivial.’

    Michel’s Iron Law of Oligarchy seems to qualify. At least, apparently-intelligent people keep getting surprised by it. Perhaps also the rising-expectations (vs. immiseration) finding about when revolts and resistance happen. And I suspect Timur Kuran’s model of preference falsification/ preference cascades also fits the bill.

    Jacob T. Levy

    June 19, 2008 at 1:28 pm

  3. Feynman tells a story somewhere about getting shit from the math grad students for working in such a heavily applied field such as quantum mechanics, and. If I remember right, their line was that if some proposition couldn’t be rigorously proven then it wasn’t really true. Feynman’s response was to pick up on a tendency of the math people to discount lots of results as trivial, so he would insist that if they could rigorously prove something then ipso facto is was a trivial result.


    June 19, 2008 at 1:29 pm

  4. - Brokerage (a la Ron Burt)
    - Path dependence via increasing returns

    Arguably true but, if so, non-trivial:

    - Religious economies/rational choice models of religion
    - Laitin/Fearon models of ethnic conflict


    June 19, 2008 at 1:31 pm

  5. In economics, the existence of equilibria in any finite game. Wasn’t obvious to me, and it’s pretty darn hard to prove.


    June 19, 2008 at 2:31 pm

  6. Back in econ, I gather that the put-call parity theorem qualifies.

    Jacob T. Levy

    June 19, 2008 at 3:29 pm

  7. From sociology. It’s not a logical theorem, but a landmark finding is the global uniformity of occupational prestige. No matter which country you go to, jobs are ranked in remarkably similar ways.


    June 19, 2008 at 3:46 pm

  8. The single most counterintuitive, true idea in economics is the notion that economic growth depends on increasing productivity. Comparative advantage is merely a corollary to that notion and is subject to as many qualifications. Solow’s exogenous growth theory and Romer’s endogenization of it seem to meet both criteria, as do the Coase theorem, opportunity cost, and the law of one price (no-arbitrage condition – mentioned by Jacob Levy). The entire body of theory subsumed under the rubric of portfolio analysis, including various versions of the efficient market hypothesis, seems to be equally true and equally counterintuitive, based upon the popularity of notions like momentum investing. Frankly, in my experience, even the law of demand and marginal cost pricing meet a lot of cognitive resistance from my students. To be sure, knowing a little math greatly increases the likelihood that many of these notions will seem intuitive, which may explain why Samuelson found the question so difficult. Rational expectations meets the counterintuitive test; I am not sure about the truth part. And so it goes. These notions also meet the pragmatic test – they can be useful as well pretty. On the other hand, prospect theory seems like a singularly bad example of a theory that is counterintuitive as well as true. Rather, what it says is that statistical decision theory, especially Bayesian theory, is frequently and predictably counter intuitive.

    I have no training in sociology, but from what I have seen the idea that human artifacts – tools, collectivities of all sorts, money, indeed the entire financial economy, etc. – are socially constructed realities seems to meet the test, as does the folk theorem, although perhaps only to an economist. There are probably a lot of other examples, although the iron-law of oligarchy doesn’t seem counterintuitive, besides Aristotle said it first. Maybe, Ziph’s law.

    Once upon a time, I would have wholly denied that PS had produced anything meeting both tests, but in Aaron Wildavsky’s hands Mary Douglas’s cultural theory of risk appears to me to do so. He came close earlier when he applied incrementalism to the politics of the budgetary process.

    Fred Thompson

    June 19, 2008 at 5:08 pm

  9. Fabio, I not familiar with any literature on the uniformity of occupational prestige. Are you saying that nobody has attempted to derive this from some form of model, right? Sounds like low-hanging fruit. Or with the result being well-known, does nobody feel it necessary to derive it formally? I am also surprised by your mention of the existence of equilibrium. I guess I wasn’t thinking of it as a social scientific proposition, more of a mathematical result that has implications when applied in social science. But don’t get me wrong, I’m glad you said it.

    Cognitive dissonance is a pretty good one, brubineau.

    Jacob, You win the “I wise I had said that” award for mentioning the Iron Law of Oligarchy.

    Trey, You win the “And I didn’t even pay you” award for mentioning rat choice models of religion.

    Back to economics…
    – I think Keynes’s notion of aggregate demand is pretty significant, both theoretically and for policy.
    – Olson’s discussions of latent groups and the importance of selective incentives in groups are huge. They dramatically changed the way political scientists thought.
    – Schelling’s focal points for solving coordination problems.
    – I think Arrow’s Impossibility Theorem is highly subtle and significant, though I wonder if it is too abstract for this list.

    From cognitive science, I think inattention blindness and change blindness have very deep significances even if magicians have known and taken advantage of them for centuries.


    June 19, 2008 at 5:10 pm

  10. Fred, I would almost include Duverger’s Law as an example from polisci, but there are some glaring counterexamples which suggest that the its simple formulation needs more detail. I guess someone could also claim that my Olson and Arrow examples would fit under polisci even though they came from economists.


    June 19, 2008 at 5:22 pm

  11. Mike, occupational prestige is not my specialty, but I think it has so far been empirical. Job prestige (and status more generally) is strongly linked to education and income. There’s very little R-squared left over after that. Some strat researchers even suggest that the main prestige measure ought to be a weighted sum of income and education, which is how the NORC/GSS computes prestige. When this came out in the 50s-70s, people gave some rough arguments about prestige (e.g., Kingsley Davis argued that it was just usefulness to society). If there’s a more theoretical and modern approach, I am not aware, but I’m not a strat expert.


    June 19, 2008 at 6:04 pm

  12. Mike McBride: I’ll grant you Arrow’s impossibility theorem, but Olson’s interest group theory is merely an application of Samuelson’s 1954 theory of public goods. If you know Samuelson, Olson is trivial, although I grant that I didn’t see it until I read Olson & Zeckhauser on alliances. But is it political science merely because it is in the domain of collective choice? In that case, Coase, Arrow, and Williamson were doing sociology when they worked in the domain of organizations. I think Riker’s theory of coalitions or his heresthetics is a more apt example, although neither are obviously true (and the former can viewed as a gloss on Olson and the latter on the impossibility theorem).

    Fred Thompson

    June 19, 2008 at 6:46 pm

  13. An obsevation: If economics were really as performative as it has become popular to believe in the STS corner of econ soc, it would be hard to think of an economic theory that did *not* meet these criteria.

    My favorite candidate for this is Feld and Grofman’s class size paradox. See especially, Feld’s fun 1991 AJS paper, Why Your Friends Have More Friends Than You Do. (Note that this is so non-obvious that John Jost and I show [in our 2001 SPQ paper, 'What Makes You Think You're So Popular?'] that most people– well, at least, most Chicago undergraduates, and we know how representative they are!– get it backwards).

    Ezra Zuckerman

    June 19, 2008 at 7:06 pm

  14. If economics were really as performative as it has become popular to believe in the STS corner of econ soc, it would be hard to think of an economic theory that did *not* meet these criteria.

    If the performativity thing were strongly happening, then as time went by the subsequently enacted findings would all seem trivial, because obvious, inevitable truths about human nature or society. As it happens, you sometimes see this argument (under the guise of reflexivity) put forward in defense of Sociology: as sociology’s counterintuitive research findings become incorporated into the stock of public knowledge about society, the discipline seems to lack counterintuitive findings.


    June 19, 2008 at 7:16 pm

  15. “as sociology’s counterintuitive research findings become incorporated into the stock of public knowledge about society, the discipline seems to lack counterintuitive findings.”

    So, Kieran, are you saying that sociology is a self-structuring structure? … oh, I’m too good!


    June 19, 2008 at 7:20 pm

  16. Kieran: Funny irony, that. Though it’s not clear to me (and I’m pretty sure, not clear to many performativity enthusiasts either) whether the performativity of the theory is supposed to lie in its being widely *accepted* as a dominant model or in its being widely *understood*. You seem to assume the latter. Perhaps. But if the former, you could have a theory that is widely enacted without the enacters really understanding why it should work. For instance, people could liberalize their economies because western economists tell them to do so without ever understanding those economists’ theories. (You might even see those theories as cockamamie, but see that it’s easy to justify them and that you might get rich by doing so).

    Anyway, to clarify my cryptic observation, my point was that if performativity were strong, even silly theories would become true; but since silly, would be non-obvious.

    A couple of quick notes on Fabio and Mike’s dialogue (though I don’t consider myself a strat expert either):

    * The universality of the occupational prestige hierarchy has been challenged– I think most famously and effectively, by Coxon and Jones, who argued that it was partly an artifact of forcing respondents to do a ranking. I think the aggregate rankings were less consistent when subjects were asked to use their, more open-ended “pile-sort” technique.

    * There is definitely no consenus that occupational prestige is a function of education and income. For instance, Abbott’s 1981 AJS article ‘Status and Status Strain in the Professions’ argues convincingly, mainly through counterexample, that income and education are not the basis for occupational prestige. Following Shils, Abbott basically argues that the prestige of the professions (relative to lower-status occupations) has more of a cultural foundation. Essentially, prestigious occupations are those that are recognized as solving fundamental problems of disorder (e.g., life and death, law and order, etc.) I think this area of research has mostly been dormant lately, with the exception of Zhou’s 2005 AJS article, which also built on Shils to make a somewhat different culturalist argument.

    (FYI, I’m going dark now)

    Ezra Zuckerman

    June 19, 2008 at 7:44 pm

  17. Fabio and Ezra, Thanks for the extra info on organizational prestige. It strikes me that prestige in a given culture or community would be closely related to the scarcity of the goods being provided by those occupations being ranked. Thus, you could have differences in prestige across cultures based on peculiar local scarcities but also some commonalities due to scarcities common across cultures.

    Fred, I agree that Sameulson’s 1954 paper is highly significant, but I don’t think Samuelson anticiapted many of the things of which Olson wrote. I suppose there’s always room for disagreement when alotting credit for ideas. My own sense was that Samuelson didn’t really think of it as an organizational problem, which is where I see Olson really making his contribution.

    Ezra, “…it would be hard to think of an economic theory that did *not* meet these criteria.” Well, this is great! It means I’ve only published true and non-trivial social scientific propositions! Where are those retirement papers…


    June 19, 2008 at 8:50 pm

  18. Ezra (in the dark): thanks for the up date. My strat knowledge is about 10 years old now!


    June 19, 2008 at 11:38 pm

  19. I think you can pick at the occupational prestige thing on methodological grounds, and I think it does not qualify to be on this list because it is just a giant empirical generalization waiting for a theory (although this theory would be able to explain it) not some sort of “proposition” but my sense is that the major people in stratification (the ones that deal with the nitty gritty numbers that is) still think that the fact that all occupations get ranked the same everywhere you look (I think Hodge and Treiman [1966] get the credit for this, so if this was Physics we would be talking about the “Hodge-Treiman effect”) it’s probably the single most important and surprising empirical finding in the field (for instance Hout and Diprete 2006 rank it at number one in their list of accomplishments in the stratification field) so Fabio’s hunch in my view was correct in this respect, even with 10-year-old strat knowledge.


    June 21, 2008 at 1:48 am

  20. I find it difficult to believe that anyone would find the law of demand counterintuitive. I have seen children as young as 8 understand it easily.


    October 8, 2009 at 8:33 pm

  21. With time, I believe the mathematical “problem solving” orientation of twentieth-century economics will be viewed as piece-meal and insufficient, especially as we look at the viability of the market-mechanism itself in the wake of the financial crisis. Relatedly, Samuelson’s “mathematization” of economics treats the discipline as though it were a science–ignoring the inherent limits to prediction of a human system, whether social, political or economic. For more, pls see


    December 14, 2009 at 7:36 pm

Comments are closed.


Get every new post delivered to your Inbox.

Join 971 other followers

%d bloggers like this: