Wednesday, September 11, 2013

Bob responds on economics and the scientific method


Recall my concern a couple days ago that there are so many people who seem to accept the seen/unseen counter-factual argument just fine but will only give vague criticisms of actual empirical practice to deal with the problem (this is especially notable when you cite some empirical evidence but raise hackles about other empirical evidence).

I want to make two quick points about Bob's response:

1. If you're just talking about thinking carefully about something, that's fine - but that's philosophy, not science. There's no need for Bob to mock people for thinking that empiricism might have something to do with science. There's nothing childish about that idea at all. The step-by-step formula of the scientific method you learned as a kid is a little childish and ignores the iterative nature and the feedback inherent in real scientific work, sure. But if you're talking about taking empiricism out of the equation, you're talking about social philosophy. Just say that and stop the name calling.

2. Bob is quite sloppy here about assuming that (1.) politicians are on board with economists and (2.) that economists even think policy can work miracles in the first place. When we have a financial crisis, I think the best anyone hopes is that the Fed can ameliorate it anyway - not make everything perfect. As far as I know, at least. But even if you do think policymakers are miracle workers, only an an-cap could think that policymakers now are actually listening to economists.


  1. ...only an an-cap could think that policymakers now are actually listening to economists.

    Is Ben Bernanke:

    (a) not an economist,

    (b) not a policymaker,


    (c) ignoring himself?

    1. 1. If it were a one man Fed, policy would be very different.
      2. If monetary policy was the only thing that mattered, it would be very different.
      3. Even then I don't agree with the expectation of perfection. Who says the Fed can work miracles?

  2. Daniel,
    You wrote: "There's no need for Bob to mock people for thinking that empiricism might have something to do with science."
    I didn't think he mocked people; I thought he was making a serious argument. What is your evidence of mocking?

    1. In this post, the attribution of a fifth grade perspective on science - but most importantly in the original post where he accused the other side of being "intellectually bankrupt".

    2. I don't think either of those is an instance of mocking. Especially the latter. When he says "intellectually bankrupt," he's making a serious claim. He may be right or he may be wrong, but it's not mocking.

    3. I don't think it's a very serious claim, and what's more I think it's precisely the sort of thing that he gets up in arms about when others accuse him of it. I think he's wrong here and I wouldn't call him intellectually bankrupt - I'd just call him wrong.

      Russ, in his posts on the subject, has been much worse than Bob it's true. For one thing he's started with the "intellectually bankrupt" and he's also accused people of ideological bias and just being interested in making inroads with the powerful. This stuff occurs, of course, but if you're going to make a claim like that you need evidence for it. If you just have a methodological disagreement I'm not going to take someone seriously that leverages that into intellectual bankruptcy or ideological bias.

    4. Dictionary says "To treat with ridicule or contempt; deride".

      Saying someone has a fifth grade understanding of things seems to qualify as ridicule and derision to me. "Intellectual bankruptcy" certainly falls under contempt and derision. You seem to see a solid argument justifying this that I'm missing, and absent that the contemptuousness of the accusation certainly strikes me as mocking.

      What would your reaction be if I called Russ and Bob intellectually bankrupt? I don't think you'd say I'm making an argument for a reasonable claim.

    5. I'm obviously not dismissing Bob as not worth conversing with anymore over it. We'll keep on. This sort of thing happens on blogs - I'm sure I have many cases of it and I like other bloggers besides Bob that do this. So I don't want to overstate how much weight I'm putting on this particular accusation. But I'm not going to pretend it's a fair or above board thing either.

  3. The essential problem with the hypothetico-deductive method is that by adopting a suitable set of assumptions one can reach almost any conclusion. And that makes things tricky.

    But let's stay with a standard set of assumptions and look at (a) uncompensated demand curves and (b) labor supply curves.

    (a) With uncompensated demand curves, a reduction in price (say, of milk) raises real incomes and lowers the opportunity cost of purchasing milk. So long as milk is a normal good, these two effects affect the quantity demanded in the same direction, so we would conclude that the uncompensated demand curve is downward sloping. But what is our warrant for assuming that milk is a normal good? And, if it is not, our conclusions would be that *so long as the substitution effect is larger in absolute value than the income effect,* the demand curve is downward sloping. Hasn't this suddenly become an *empirical* question?

    (b) With labor supply, we once again have income and substitution effects. When the real wage rises, the value of time spent in market activities rises relative to the value of time spent in non-market activities, so people will tend to want to work more. But there is also an income effect--as the real wage rises, real income rises even if time working does not rise. And, *if leisure is a normal good,* people will react to the rise in real income by choosing to work less. Once again, it becomes an empirical question, doesn't it?

    In both these cases, how can we reach a conclusion about how people *will* react to a change in prices without investigating how they *do* react to a change in prices?

    Just asking...

    1. Yes and yes so far as I'm concerned.

      The way to keep it deductive, of course, is to keep it vague and not use it to explain any real phenomenon - to say something like "when you're dealing with normal goods this will happen, but otherwise it won't".

      Granted, if you don't get into specific assumptions (assumptions that CAN'T be derived axiomatically), it makes it much harder to make claims about policy effects on, say, labor supply. And yet precisely those sorts of claims are often made. If I were in a particularly nasty mood, I might call that "intellectually bankrupt".

  4. Part of the problem with Bob's argument is that he has taken the inherent difficulties of empirically modelling the macroeconomy and extrapolated that to all of economics. Getting the identification right in macro is very hard. (Although - note - not impossible.) I don't think anyone disagrees with this. However, it's considerably easier in, say, an applied I/O or micro settings. Bob seemingly wants to use the controversies in current macro as grounds for disqualifying the use of empirical methods in all other areas of economics. This is just plain silly.

    The other major gripe I have with his post is that it (implicitly) ignores the fact that all economic theory relies on deductive thinking. This isn't something unique to praxeology, despite all the Austrian hand-wringing about synthetic a priori axioms and so forth. And, yet, legitimate starting axioms and assumptions can lead us to different overarching theories that entail (possibly) contradictory policy prescriptions. How else to adjudicate between these in the absence of empirical evidence? I hold that this passage remains relevant:

    I challenge anyone to show me that ABCT follows solely and directly from the action axiom alone. The list of subsidiary axioms and assumptions becomes enormous once we reach the full scope of the theory. This idea of an immaculately conceived business cycle theory, of pure logical cogency and free of any auxiliary pillars is, to be frank, so fanciful that not even the most zealous praxeologist could believe it. More importantly, the "choice" between theory and empirics is a false dichotomy[...] All economic theory is essentially deductive in nature. You start with some primary axioms or propositions and work through to the implications and consequences. Yet, how do we arbitrate between competing theories or measure their importance? Well, the same way that we do for any scientific field; we test them using data from the real world. Rejection of empirical scrutiny, validation and testing means that we are no longer debating economics or any kind of science for that matter. We are now in the realm of religion.

    Embracing empirical methods in economics doesn't mean that you ignore intuition or the use of deductive logic to build your theory. (Quite the opposite, in fact, as anyone who has written an empirical paper could testify to.) It just means that you acknowledge the possible limitations of your theory and wish to better understand its applicability to the real world.

    1. And this is an interesting point - in the original post he did stick to macro (although I think it's even inappropriate there - since when has the problem being hard been a disqualifier for serious scientific inquiry??). It's only in this recent post that he's taken aim at all economics.

      The one difference I have - and this does not affect your broader argument - is that while all economic theory is deductive it does not claim to be axiomatic in the way that praxeology does. We do start with premises and assumptions, but we recognize that we are modeling - that those assumptions are not exact replicas of the real world. But we think they are good enough that we think we can get insights from building models up from them.

      I think Austrians are in this boat too (particularly w.r.t. ABCT, as you point out). Some of them just don't seem to be aware of it (some of them are more aware of it).

  5. Certainly, any attempt to dismiss the ultimate necessity of empirical confirmation of economic theory is wrong. However, I think one thing that is needed by every economist--particularly re: macro forecasts--is humility. Even in experimental sciences with "harder" microfoundations, like nutrition, there still is little consensus on pretty basic questions. Coffee--good or bad? Aspirin? Vitamins? Meat, high carb, low carb, paleo, vegetarian? A new study comes out every day overturning the previous results.

    Empirical econ results matter, obviously, but we're still very much reliant on economic intuition to make policy decisions. Perhaps in the future, with better agent-based simulations, this won't be the case. But for now it is.

    1. Sure - I don't think anyone is arguing against the idea that macroeconomists should be humble. That's certainly not my claim.

      The trouble is this tendency to replace "be humble about your work" with "your work is intellectually bankrupt". And that's not an exaggeration - two economists from top schools have made that claim now. So the discussion is well beyond the question of humility, which I hope there is agreement on (and my impression is there is agreement on).

  6. BTW, regarding Daniel your claim on my blog that the realism of assumptions is important and that's why you disagree with Friedman's famous essay, I went and dug up this quotation from David Romer's graduate macro text:

    [T]he purpose of a model is not to be realistic. After all, we already possess a model that is completely realistic—the world itself. The problem with that “model” is that it is too complicated to understand. A model’s purpose is to provide insights about particular features of the world. If a simplifying assumption causes a model to give incorrect answers to the questions it is being used to address, then that lack of realism may be a defect. . . . If the simplification does not cause the model to provide incorrect answers to the questions it is being used to address, however, then the lack of realism is a virtue: by isolating the effect of interest more clearly, the simplification makes it easier to understand.

    --David Romer, Advanced Macroeconomics (New York: McGraw-Hill, 1996), pp. 11–12.

  7. I am confident 90% of today's published macroeconomists would endorse Romer's statement.

    1. I would too.

      Romer also talks about how for example, models where agents behave according to expectations about the future are better because they're a more realistic presentation of how people behave.

    2. I would think/hope the figure would be much higher than 90%


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