The CBO knows its basic forecasts are wrong
The authors, Kevin Kliesen and Daniel Thornton, show that CBO’s budget forecasts are less predictive than a so-called “random walk” model that assumes next year’s deficit will be the same as last year’s deficit. That makes the CBO look pretty bad. Or, it does until you realize the CBO agrees with Kliesen and Thornton, and their budget forecasts account for the huge errors in their basic model.
According to Kliesen and Thornton, CBO’s problem isn’t that it can’t predict the path of the economy. It’s that it can’t predict the decisions made by policymakers. CBO has to yoke its basic forecasts to “current law.” But current law is often clearly wrong. Right now, for instance, it says that the Bush tax cuts will expire in 2013. But the Democrats want to extend 80 percent of the Bush tax cuts and the Republicans want to extend all of them. Under those conditions, the Bush tax cuts are very unlikely to fully expire. And the CBO knows it.
That’s why they release an “alternative-fiscal scenario” in which the implausible things that current law says will happen — huge tax increases, huge Medicare cuts — don’t happen. So then you see the difference between current law — which, at this point is implausibly austere — and a scenario in which Congress simply refuses to make any hard decisions. Right now, that difference is about 4 percent of GDP. Current law will drive deficits down to 1.5 percent of GDP and the alternative scenario will leave them at 5.4 percent of GDP. Current law is absurd. The alternative scenario is probably closer to right, albeit a little pessimistic. And the CBO knows that, which is why they provide both, so people can make an informed judgment based on the range of plausible options.