But if it is true that in subjects of great complexity we must rely to a large extent on such mere explanations of the principle, we must not overlook some disadvantages connected with this technique. Because such theories are difficult to disprove, the elimination of inferior rival theories will be a slow affair, bound up closely with the argumentative skill and persuasiveness of those who employ them. There can be no crucial experiments which decide between them. There will be opportunities for grave abuses: possibilities for pretentious, over-elaborate theories which no simple test but only the good sense of those equally competent in the field can refute. There will be no safeguards even against sheer quackery. Constant awareness of these dangers is probably the only effective precaution. But it does not help to hold up against this the example of other sciences where the situation is different. It is not because of a failure to follow better counsel, but because of the refractory nature of certain subjects that these difficulties arise. There is no basis for the contention that they are due to the immaturity of the sciences concerned. It would be a complete misunderstanding of the argument of this essay to think that it deals with a provisional and transitory state of the progress of those sciences which they are bound to overcome sooner or later.
Friedrich Hayek, 1955.
Let me make a confession. Back when I was 20 I could perceive the great progress that was being made in econometric methods. . . . I expected that the new econometrics would enable us to narrow down the uncertainties of our economic theories. . . . My confession is that this expectation has not worked out. . . . 1t seems . . . that there does not accumulate a convergent body of econometric findings convergent on a testable truth.
Paul Samuelson, 1986
- I[↩]
Are you reading ‘Hayek’s Challenge’ by Bruce Caldwell?
Concise Hayek: What a shame only a few people like me are smart enough to see through the sophistries of the Keynesians and socialists; what’s worse, those theories will still be as seductive fifty years hence as they are now, whatever the apparent progress of economic science in the meantime, because — as I was saying — it takes someone of my calibre to spot the flaws in them, and I won’t live forever.
But was he right James ?
Sure was. A right windbag.
Don – yes.
James – give me a break. Hayek was right about this stuff IMO – and though I’ve not read it in situ, I think he’s having a go at the positivism of neoclassical economics – not Keynes. His concerns with Keynesianism – as far as I’ve read – were quite different to this – and much less persuasive. I think Keynes would agree with him on this stuff and indeed on quite a bit else. Keynes described Tinbergen’s early model building as ‘black magic’.
But Keynes obviously wouldn’t have agreed with Hayek on macro-economics. I’ve not read Hayek on macro-but what I do know doesn’t sound too promising. Hayek’s intellectual ally in many things Milton Friedman regarded Hayek as a bit of a crank in macro.
the promise of certainty that econometrics and ‘precise’ modelling holds out is very difficult for young students to resist. Like Samuelson above, I fell for it as a prospective student. And the more mathematically gifted students I know still fall for it. Some eventually come round to Hayek’s view, but what percentage and at what age?
I reckon it’s broadly true, in that “the elimination of inferior rival theories will be a slow affair” for the reasons given, and progress in economics is therefore slow and limited. But Hayek and Samuelson overstate things a bit – that progess is slow but not nonexistent. Economists often lack appropriate humility but should not be posted to the Faculty of Inconsequential Studies.
Econometrics has moved on – methodologically as well as technically – quite a bit since Samuelson wrote that (inter alia, we’ve all taken the Lucas critique on board. And there are indeed theories that are widely held to have been empirically disproved – no-one is either a monetarist or a traditional Keynesian these days (and I might add that few are strict Hayekians either).
All that modelling and mathematical rigour has played a role here, though I agree its exponents are constantly tempted to underestimate just how overdetermined and stochastic (ie “messy”) the real world is. It’s not just youthful enthusiasm – the incentives of the profession reward certainty and punish diffidence.
Paul Krugman is worth reading on this, at http://www.pkarchive.org/theory/dishpan1.html (it starts off as a paper on development economics but then moves on to the methodological questions we’re focusing on here. also its also a good reminder of how much better Krugman is in his more meditative mode rather than in opinionated op-eds).
Reflecting Paul’s F’s comments above on the predilections of talented quants, I find Krugman’s stuff on methodology some of his most sloppy. You will find in it lots of implicit suggestions that formal is better than informal. That’s nonsense.
Each method has its strengths and weaknesses. The paper you site DD has some of this thought it’s a bit better than his earlier stuff on new trade theory which basically put forward what I call the ‘mountain climbing’ rationale for economic work. Why formallise new models of trade in the presence of imperfect competition. Because the techniques are there. Well fine. Go ahead, but please do so at the same time as considering, at the time, as well as in hindsight the extent to which it is likely to get you anywhere.
New trade theory never did that and so ended up with a disappointing harvest. Perhaps it was worth trying, but it is remarkable how little the problem was ever discussed as they embarked for the summit. Find me a single new trade theory paper in the 1980s or even 90s that considered these methodological issues rather than just headed off to see what could be done. Short answer – not much!
Oh, I don’t think Krugman implicitly says that formal’s better than informal – he’s perfectly explicit about it. But by the same token he wouldn’t disagree that each approach has its strength and weaknesses – he’s just saying formalise wherever possible, because that sorts out logical weaknesses and above all generates testable predictions.
So long as we add ‘informalise’ wherever possible, because that helps people make connections, make analogies which might be fruitful, helps develop intuitions about what the hell is going on inside the models and how one might try to think in models but not have one’s imagination limited by them etc etc.
A few random comments, on request from Nicholas.
First of all, Hayek is not always a joy to read, in fact The Road to Serfdom (for example) is distinctly stodgy although the argument is ok.
Scanning the link to Krugman, I am surprised at the assumption that a new field called “development economics” is required when Peter Bauer got on quite well with more or less orthodox theory when he did his stuff on Africa and Asia. Lal is travelling the same road as Bauer. Provide law and order, light taxation, and let people in other countries buy their stuff.
(Thanks to Nicholas for another heads up some time ago) – Rodrik is good on the kind of situational analysis that is required on a country by country basis to find the policy mix that is most appropriate. He is also good on the critique of regression studies. That probably makes him an exemplar of “informal” analysis.
I could say more but would be accused of stirring and it is too late to get into a big argument because I need my beauty sleep.
By the way, in case anyone thinks I have been idle this year, a running list of my Catallaxy posts.