Economic forecasters aren’t all that bad

Economists get a lot of criticism for failing to polish their crystal balls thoroughly enough before reading the economy. So how good are people forecasting other complex systems? Physicist types. Real Scientists. Let’s ask a weather forecaster how accurate their forecasts are a couple of days out. Days, not months, quarters, or years.

…if our forecasts are about 60 percent accurate or higher, then we consider that to be a good estimate.

Well, that’s slightly better than flipping a coin.

3 replies
  1. VMC
    VMC says:

    I am thinking you posted this to provoke a little debate, so here goes. There are many aspects to a weather forecast such as will it rain? if it will rain, how much?, what will the max, or min, temperature be? etc etc. And then there are follow up questions such as how close is close enough to regarded as “accurate”? And then there are also questions about how far out are we interested in, a day or two, or maybe a week or two. I note that the UK met service claims to be: 94% accurate with forecasting max temperatures a day ahead, and 90% 2 days ahead; 84% accurate at forecasting minimum temperatures; and 94% accurate at forecasting temperature. They publish their accuracy forecasts and have them audited. (Sounds like a good idea to me). But I suspect that the main issue for most ordinary people is whether the weather forecaster predicts abnormal events that might cause real mischief if left un-forecast (such as snow for farmers, etc). My sense is that NZ forecasters do pretty well here, even if they dont always predict the exact amount of rain we will get they generally let us know in advance that a flood is likely. And I also think that its here that economic forecaster fall short. While some people did forecast disaster ahead of the GFC, they tended not to be mainstream economists. Not sure if you have seen “Inside Job” (its a great documentary, BTW) – but it would suggest that a number of academic economists were actually complicit in what happened. And its here that I think weather forecasters have a natural advantage over economic forecasters – an economic forecast might actually affect the economy while a weather forecast never can!

Comments are closed.