Imagine that I'm president, and I sign a bill that says a bridge will be built over the Potomac. Imagine, further, that the bridge is never actually built.

Now, do you think that The New York Times would report that our imaginary bridge exists anyway, simply based on a government report on regional commerce that is based on earlier assumptions that the bridge would be built?

The answer should be "no," but I'm not so sure we can count on the Times to live up to our expectations in this regard. For once again, it is citing these CBO reports as proof that the stimulus created net jobs:

The American Recovery and Reinvestment Act, known as the stimulus, produced or saved at least 1.9 million jobs and as many as 4.7 million last year, according to the Congressional Budget Office.

Very, very misleading, and a disservice to your readers. If you're going to use this number, at least include what CBO says about it: It's based on assumptions made before the stimulus package passed Congress. CBO's numbers, as the CBO's own report says, do  not reflect the current economic situation, but only prior assumptions that do not seem to have played out -- at least not in the minds of the 68 percent of Americans who think the stimulus money was "mostly wasted."

CBO has never represented its figures as an actual measurement of jobs created, even if the Obama administration continues to portray them that way and the New York Times helps the effort. CBO's director, Doug Elmendorf, specifically made this point in a public appearance last year, that these reports are "repeating the same exercises we [aleady] did rather than an independent check" on the success of stimulus job creation.

If unemployment stood at zero today, CBO would have given the same estimate, under the methodology it used. If unemployment were at 50 percent, CBO would have also given the same estimate.