The problem with experts is that they're awful at predicting the things they're supposed to predict. Nassim Taleb makes this point in both Fooled by Randomness and The Black Swan. The problem with reporters is that they don't bother to check how well the experts have done in the past.
The CU Boulder's Leeds Business School released its Business Economic Forecast for 2009. Both the Post and the Rocky reported on it, doing little more than repeating the predictions. Now, the whole thing's available for download here.
Looking back over the 2005-2007 predictions, they've done all right, but nothing spectacular. And this year's prediction spent time on the housing market, but completely missed the financial crisis and the mortgage industry's spillover into the rest of the economy. The 2008 forecast doesn't even mention the word "mortgage" outside of the housing industry survey.
In terms of percentage change in total employment, they were off by 8% in 2005, 1% in 2006, and 17% in 2007. That's not bad, but their sector-by-sector results aren't so good. They've consistently underestimated the Mining sector's growth, as well as government growth. They're almost never under 10% relative error on these estimates. When the change is near 0, that's going to happen, but of the 33 sector predictions over the last 3 years, only 6 have been under +/-1000.
I saw the same thing when I was valuing a construction company at the brokerage. The American Institute of Architects tries to predict construction activity, but their sector predictions are almost never with 10% of the actual number.
We use these numbers because we don't think we have anything better. That doesn't mean they're actually good.
The CU Boulder's Leeds Business School released its Business Economic Forecast for 2009. Both the Post and the Rocky reported on it, doing little more than repeating the predictions. Now, the whole thing's available for download here.
Looking back over the 2005-2007 predictions, they've done all right, but nothing spectacular. And this year's prediction spent time on the housing market, but completely missed the financial crisis and the mortgage industry's spillover into the rest of the economy. The 2008 forecast doesn't even mention the word "mortgage" outside of the housing industry survey.
In terms of percentage change in total employment, they were off by 8% in 2005, 1% in 2006, and 17% in 2007. That's not bad, but their sector-by-sector results aren't so good. They've consistently underestimated the Mining sector's growth, as well as government growth. They're almost never under 10% relative error on these estimates. When the change is near 0, that's going to happen, but of the 33 sector predictions over the last 3 years, only 6 have been under +/-1000.
I saw the same thing when I was valuing a construction company at the brokerage. The American Institute of Architects tries to predict construction activity, but their sector predictions are almost never with 10% of the actual number.
We use these numbers because we don't think we have anything better. That doesn't mean they're actually good.