Whoops: Economic Forecast Wrong Within Weeks

05/07/2009 05:12 am ET | Updated May 25, 2011

The business of making long-term projections is generally a safe one: If you wind up on the wrong end of a forecast, enough time has often passed that nobody remembers the blunder.

But not if you're wrong within weeks of making the projection. Such has been the fate of the Congressional Budget Office.

In early February, we reported that CBO forecasts were no better than wild guesses because the models they used only accounted for post-World War II recessions and couldn't factor in a simultaneous financial-sector collapse.

The GOP latched on to the optimistic CBO projections that ignored the financial-system collapse. Presuming a shallow dip and a quick turnaround, the CBO projected that the stimulus bill could actually slow growth in the long term. "As CBO has stated, this bill will not work. In fact, this bill will hurt our economy," Sen. Tom Coburn (R-Okla.) said in at the time in a statement typical of the genre.

The CBO, however, has already been shown to be wrong. It's January projection for the unemployment rate for all of 2009 was 8.3 percent. (Click on economic projections.) By March, unemployment was already 8.5 percent. The CBO has since revised its unemployment rate forecast and now says it'll be 8.8 percent by the end of 2009. (Notably, they left their 2010 projection at the same 9.0 percent, meaning that even as they acknowledge the economy's falling faster than they thought it would, at the same time they think it'll turn around at an even faster rate.)

They stand to be proven wrong again -- and soon.

"It's virtually a done deal that when we get the April numbers it'll be 8.9 or 9 [percent]. It could even be 9.1," says Dean Baker, an economist with the liberal-leaning Center for Economic Policy and Research, who argued in February that the CBO estimates weren't to be believed.

Economist James Galbraith also called it accurately. "We are working with a set of economic projections which assume, for mechanical reasons, that the economy is going to start turning around at the end of this year. There is no analytical foundation for that," Galbraith said in February. "Those models that are based on the period after 1945 aren't going to work. They just aren't going to be right."

Galbraith says that as long as the CBO and other economists continue to rely on bogus assumptions, they'll continue to be wrong.

"CBO assumes a natural rate of unemployment, to which the economy naturally reverts. That's the major long-term problem," says Galbraith. "The major short-term problem is an assumption that all recessions last about the same amount of time, and so can be expected to trough a given number of months after they began."

Baker says that the CBO and other economists who use similar models -- which is almost all of them -- err because they refuse to recognize bubbles. So when a bubble pops, the economists expect that prices will again return to pre-pop levels when the economy turns around. As Pets.com shareholders are all too aware, that's not how it works in reality.

"The basic principles here are very, very simple and there's a real effort just to obscure the fact that the guys in charge, who are paid lots and lots of money and given a lot of respect, really blew it," says Baker.

And, notes Baker, they're still in charge, still making lots of money, still respected, still using the same models and still dispensing advice.

The CBO did not return calls requesting comment.

Suggest a correction