In a post about Scott Sumner, "The Economics of Unobservable Variables", Arnold Kling makes a good point about Mr. Sumner's use of NGDP expectations as a means of describing recessions as predictions of slow NDGP growth:
Unfortunately, this puts a huge emphasis on something that is unobservable, namely “expected NGDP 1, 2, 5 and 10 years out in the future.” Any time you have a theory that relies on such an unobservable, you drift further away from the realm of science and nearer to the realm of circularity.
In Sumner’s defense, he wants expectations for nominal GDP to be observable, by having tradable NGDP futures contracts. Even so, it is rare for any such contracts to go out more than a year ahead.
I don't think recessions are a prediction of economic diaster as most economists would understand it, that is as a change for the worse in the path of aggregate demand.
After all, if people are profit maximizers, as most economics would also supposes, they are maximizing their profit which means that they are doing to make bets about the particular sectors which affect them. It's their own particular profits that they care about, not the economy as a whole. They are in effect bets about particular demands, not aggregate demand.
With these two considerations, we find that a heavy emphasis on aggregate demand is neither scientific nor does it entirely mesh with what entrepreneurs are trying to do in the markets. There is never really a drop in aggregate demand so much as there are drops in particular markets just as entrepreneurs don't really care about the path of aggregate demand so much as they care about profits in the specific sector that concern them.
Comments