Four Pinocchios for the Washington Post Fact Checker

In the Washington Post Fact Checker column today, Glenn Kessler got really exercised about Bernie Sanders’ totally accurate description of a Congressional Budget Office (CBO) report on job losses that will occur if spending caps in the Budget Control Act (BCA) are not loosened in coming years. In the end, Kessler’s “fact check” is much more misleading than anything Sanders and his staff released.

The CBO provided a range of estimates of job losses that would occur in 2016 and 2017 if these spending caps are not lifted (alternatively, one could describe these as job gains that would be realized if the caps are lifted). The high end of these estimates was 800,000 jobs in 2016 and 600,000 in 2017. That’s how much higher total employment in the United States could be if the caps were lifted, relative to a counterfactual baseline where the caps stay in place.

So, what has Kessler so angry? First, that Sanders cited the high end of the CBO range—even though he and his staff are clearly identifying it as the high end. Kessler writes:

First of all, note that the Sanders’ statement says that “as many as” 1.4 million jobs would be lost. That’s a signal that a politician is using the high-end of a range.

Kessler must be a real Washington insider to know that saying “as many as” is a “signal” that the high end of a range is being used. I’m thrilled that he’s sharing this savviness with us rube readers. On the other hand, I’d argue this phrase is not a “signal” per se, but is instead is literally saying that the high end of a range is being used. And as such, this makes the Sanders description not only perfectly accurate, but perfectly transparent too. Is there any reason to cite the high end rather than averages? Of course. The Sanders position is that these spending caps are destructive policies that will harm job growth—and highlighting just how large this damage might end up being is a perfectly fine way to make this case.

Second, Kessler argues that one cannot add the 800,000 jobs in 2016 to the 600,000 jobs in 2017 to get 1.4 million jobs in total lost to spending caps. There is the tiniest grain of truth here—the proper way to express this would be 1.4 million job years. Another way to say this is that over the two years in question, the total hours of full-time employment that would be extinguished if the BCA caps are not lifted would equal 1.4 million full-time jobs. Kessler is saying that one cannot claim from the CBO numbers that total employment in 2017 would be 1.4 million jobs higher if the spending caps were lifted. But Sanders and his staff never claimed that.

The concept of the job-year is well accepted by macroeconomic analysts—see this report by the Council of Economic Advisers in assessing the impact of the Recovery Act in job-year terms. If Christina Romer—one of the preeminent macroeconomists in the world—is comfortable with this term and concept then I’m not sure what standing Glenn Kessler has to quibble with it.

Finally, Kessler is angry that Sanders doesn’t try to minimize these numbers as much as possible by scaling them to total employment in the U.S. economy. But why is this the right scalar? By this metric even the peak-to-trough losses in employment during the Great Recession don’t sound enormous. Even at the worst point of the recession, more than 93 out of every 100 employed Americans held on to their jobs…

Maybe a better measure would be to compare these estimated CBO job losses to total payroll employment growth in the most recent 12 months, which at 2.9 million new jobs was a pretty good year. If job growth over that year had been depressed as much as the CBO’s high-end estimate for how much the spending caps will slow growth in 2016, the pace of job creation would have been slowed by more than 27 percent. That doesn’t sound trivial to me.

All in all, this particular “fact-check” is bizarrely tendentious. There is nothing in the Sanders description of the CBO numbers that is misleading or inaccurate. In fact, Sanders was spot on.