Illustration of an Asteroid Hitting the Earth

A few weeks ago, I posted that CBO’s estimate that the stimulus created saved 1.5 million jobs was not based on any actual examination of the post-stimulus economy. Instead, CBO essentially re-released their initial prediction that the stimulus would work, and presented that as proof that it did work.

This is like a weather forecaster saying that the high yesterday was 65 degrees, because that is what had been predicted — even though it actually never topped 50 degrees.

Now, CBO director Doug Elmendorf has finally conceded that they never actual examined this stimulus bills’ affect on the economy. Responding to a questioner following a recent speech, he admitted that the CBO’s jobs count was “essentially repeating the same exercise” as their initial projections. When asked if this means their jobs projections would have ignored any failures of stimulus spending to perform as CBO predicted, Mr. Elmendorf responded “that’s right.” (Exchange begins at 38:20.)

CBO never actually counted the jobs. Nor did their analysis take into account the rising unemployment rate. Or the economic figures. Or how effectively the money was spent. They merely assumed this government spending “must have” saved 1.5 million jobs.

An asteroid could have destroyed the entire Unites States outside of Washington, D.C., and (as long as the money was spent), the CBO’s model would have still claimed the stimulus saved 1.5 million jobs

More specifically, CBO first programmed their economic model to automatically assume that stimulus spending creates/saves millions of jobs. And then (surprise!), their model concluded that the stimulus created/saved millions of jobs. This is a classic case of the “begging the question” fallacy, also known as assuming what one is trying to prove.

Calculating the stimulus’ impact on jobs is never easy or precise. Yet any jobs count that ignores the actual post-stimulus economy, and instead uses a computer model that is pre-programmed to guarantee a specific jobs figure should not be taken seriously.

More here.

Cross-posted at The Corner.