Thursday, 5 August 2010

The usefulness of macro

There is an interesting discussion going on over at TVHE in which Matt Nolan puts up a defence of macro forecasters, see here and here. Read the comments on both posts, especially those myself, of course!, and Eric Crampton. My basic position would be that macro can tell us that things are happening but explaining those things is in many cases a micro issue. I think the best defence Matt can put up is the "market defence". If someone is willing to pay for macro forecasts then they are showing that they think they add value. If they didn't think that, forecasters would go out of business.

1 comment:

Matt Nolan said...

I'd say that the discussion is not so much on macro - but just on the subset of macroforecasters.

My main point is that the purpose of macroforecasters is being misinterpreted - which is making it sound like that don't add anything.

Again, it would be great if we could model the economy as a closed system with a range of incentives - but as GE theory told us we can't really identify what is going on here. And even if we made restrictive assumptions on the theory, the data is horrendously imperfect.

As a result, we do the best we can with imperfect tools - as the "first best tools" don't exist.

And as you say, it is unlikely people would go around hiring macroforecasters if there wasn't some associated value - the value of which I've discussed constantly over on tvhe.