Wednesday, September 3, 2014

Economic analysis is mostly like reading the entrails of dead birds

Ancient Roman astrologers used many methods to divine the future, all of which we think are pretty silly - the direction and formation of flocks of birds, observing the entrails of sacrificed animals, that sort of things.  We scoff because while it was well organized and regulated, it wasn't scientific.  It was never compared to actual data observations.

And so with Economics:
So, say we raise interest rates, we expect GDP growth to slow. We know there're (at least) two effects going to happen: debtors have to pay more in interest so they'll cut back on other things. But obviously savers will get those higher interest payments and so will have more to spend. We think that savers will save some portion of those extra payments thus there will be a drop in overall demand.

But the point it that we don't actually know that this is true. We think it is, we've observed it a few times (and it's stunning how little empirical evidence we really have in this field. In any detail, for perhaps 30 countries over 5 or so business cycles over 5 or 6 decades - and that's about it, a very slim evidence base to construct an entire science upon) and it accords with what we think theory is.

But we're still open to being confounded on this. To change the example a little we would expect people to save less money in a higher inflation environment: might as well spend and get something rather than see your savings losing their value. Yet savings rates tend to go up in even medium (ie, 5 % and above) inflation rate circumstances as people, well, no one's quite sure what they think they're doing. We can construct a story to say that they're trying to maintain the absolute value of their savings but who knows?

Our models that we are computing this macroeconomy forecast with are based on all sorts of lightly observed linkages from our scarce evidence base and as we saw above, absolutely none of them managed to predict a recession when it was only months away. Or, if you prefer, we're just not very good at this macroeconomics stuff.

As to how I predicted the US recession at least, that was simple. I just listened to Dean Baker (a lefty but still a good economist) who entirely ignored all that macro stuff and instead pointed to a piece of micro. There's something called the “wealth effect”. When our assets, stocks, houses, whatever, go up in value we tend to spend more as we feel richer. This is why house price booms raise GDP. But crashes obviously put that into reverse and Baker pointed out that the 2006 or so real estate crash in the US destroyed some $8 trillion of household wealth: of course we're going to have a recession as households reel from such a blow.
Our taxes support an army of economists; we're not getting much of a return on that spend.  We may in fact be getting a negative return when you consider how their very poor model predictions feed public policy at the Fed.

2 comments:

Dave H said...

Our taxes support an army of meteorologists too, with similar returns on the investment.

Anonymous said...

Ever since that article in 2009 that I read that said that the recession had ended in 2008, all I do when I read anything said by "economists" is laugh.