Re: people, places, things, and ideas

Robert S. Thau (rst@ai.mit.edu)
Sat, 9 Jan 1999 08:59:11 -0500 (EST)


Mike Masnick writes:
> Besides, just because economics on the whole might not be able to be
> "falsifiable", doesn't mean that certain theorems or models can't be tested
> and falsified. That's what we learn from. And the ability to look back
> and later explain why something in the past happened, is useful in creating
> future models.

Ummmm... I was reacting to your *characterization* of economics,
from several notes back, which was:

> The thing about economics is that in the long run, pretty much anything can
> be *explained* by it. However, predicting with economics is a hell of a
> lot trickier, because there are very few people (if anyone) who can
> realistically factor in *all* the relevant variables. That's why
> predictive economics is pretty much a crapshoot of who's got the best model
> (determined conclusively only after the fact) - which is, afterall, just a
> model.

I think we agree that individual economic models can be falsified; my
complaint, I guess, is that I don't see much value in "theories" so
general that they can "explain" anything post hoc. It's not that I
think all of economics is worthless, just that you seem to be holding
them to too low a standard, particularly given the enormous influence
they have over policy decisions these days.

> I think my stand on this is pretty clear. Arguing any further doesn't seem
> likely to get us anywhere, unless you honestly feel there's more to be
> added, I think I'm through...

Yup. It's dead, Jim.

rst