Tuesday, 7 December 2010

Why do new ideas fail?

Paul Krugman in "Bourbon Economics" (and his commenter Peter von zur Muehlen) complain that we've had new ideas for decades in macroeconomics, but they don't take hold.
By 1988, it was already obvious that equilibrium business cycle theory had failed. Shiller had already circulated his devastating demonstration that asset prices were much too volatile to be explained by fundamentals...nothing happened. Real business cycle theory continued to prosper, developing an increasing stranglehold over the professional journals. Behavioral finance stayed on the margins. The equilibrium guys had learned nothing and forgotten nothing...
Our problem, in short, isn’t lack of nifty new ideas; it’s the refusal of too many economists to face up to the fact that some of their preferred theories don’t work
I sympathise - as an adherent and practitioner of behavioural finance, I could hardly not. But it's too easy to blame this on the establishment for not listening. And really, if Krugman and Shiller aren't part of the economic establishment, who is?

von Muehlen, in the comments, points out that the alternative theories aren't really very good either:
Everything you say is correct---sadly. But this criticism holds equally for all extant competing macroeconomic theories.
...Caballero notes:
"What does concern me about my discipline, however, is that...the dynamic stochastic general equilibrium approach — has...begun to confuse the precision it has achieved about its own world with the precision that it has about the real one."
This indictment affects all of macroeconomics. Yours, mine, theirs.
Those, including me, who want to promote new theories, have to face up to this: it's our responsibility to persuade people our ideas are true. If we have failed so far, then don't sit bitterly on the sidelines, bitching about how the adults won't listen to us. Engage with them, understand the debate in its own terms and participate in it...and if we do it right, and if our theories are true, we'll move the battle to our own territory eventually.

In the case of behavioural finance, I believe this means developing whole-economy models - like the Arrow-Debreu equilibrium model, like the DSGE macro models - which incorporate insights from behavioural and cognitive research. Even a simpler model would do: if we can start with the undisputed results of cognitive experiments and derive clear rules about how a single market will behave, that would be a great step forward.

David Laibson is doing a bit of this; some people like Roger Farmer, though a little more in the rational camp, are doing work that we could learn from too. There's no point complaining - let's improve our proofs and persuade people that way instead.

1 comment:

PunditusMaximus said...

That assumes that the correctness of the theories is relevant to their adoption. This is, of course, heroic; the only thing that gets economic theories adopted is the combination of their utility to moneyed classes and their attractiveness to the autistic mindset of libertarians.

Since we live in a time period where looting is more valuable than production, then theories which allow looting will be accepted and ones which engender widespread prosperity will be denigrated. That is, the thing which will absolutely doom any economic theory is correctness.