I’ve always believed meteorology has been a good way to think about economic forecasting models. Simply, economists can, via statistical analysis, have some visibility on what might happen next, but the system (be it ecological or economical) is so vast and complex we just don’t have the models to know with precision what will happen, even in the immediate future. So, after the Irene panic over the weekend, we instead get headlines like ‘People assume we can predict everything’…; NYT: Experts Misjudged Structure and Next Move…; IRENE: A PERFECT STORM OF HYPE…, and so on. It’s not that the meteorologists did a bad job—they’re just limited in what they can do, and in a situation where lives are on the line they will err on the side of caution.
This is basically—almost precisely—how to think about economic forecasts. Are we headed for another recession? Maybe. My sense is probably not. But recent weak economic data don’t guarantee anything either way. Quarterly and monthly data especially is lumpy, and never moves in a straight line. So, you get a headline like this that surprised a lot of folks this week:
Statistical economic analysis in forecasting, even just a month ahead, is at this point a lot like meteorology—the system is too complex to predict with perfect accuracy.