When the press is not inquisitive

Tony Fratto: The White House "Jobs-Saved" Deception


After nearly twenty years in Washington I thought I've seen every trick ever conceived, but the White House claims of "jobs saved" attributed to the stimulus bill is unrivaled. What causes the jaw to drop is not just the breathtaking deception of the claim, but the gullibility of the Washington press corps to continue reporting it.
News stories from President Obama's event last week hailing the 100-day mark since the stimulus was passed typically repeated the assertion that the stimulus has already "created or saved 150,00 jobs." ("And that's just the beginning," the President crowed.)
Here's an important note to my friends in the news media: the White House has absolutely no earthly clue how many job losses have been prevented because of the stimulus bill. None. Not Christina Romer. Not Jared Bernstein. Not Austen Goolsbee.
Each of these distinguished economists would have failed Statistics 101 for making such a laughable claim. But we see them now repeating these assertions to reporters who have seemingly abandoned all skepticism.
Forget that only a trickle of stimulus spending has yet made its way into the real economy. Set aside your views on whether or not the stimulus has any job-saving or -creating impact. And leave for another day the White House's failing to account for changing macroeconomic conditions and seasonal adjustments.
There is only one necessary data point to make the "jobs-saved" claim: an accurate measure of expected employment levels in the future. That baseline data is critical to measure what the employment level would be in the absence of the stimulus. Unfortunately for the White House, they cannot possibly know that measurement within any degree of confidence -- and they know it.
To understand just how unknowable this data point is, it's not necessary to be an economist, a mathematician or a statistician.
You only need to know this: the Bureau of Labor Statistics (BLS) - thousands of the most professional and rigorous counters and analyzers of labor data in the history of mankind - makes TWO revisions of employment data for their ESTIMATE of the PREVIOUS month! And even then the reports are mere estimates - an annual benchmark survey is required to reset the nation's payroll baseline.
That is, the best employment statisticians the world has ever known, people whose lives are dedicated to employment data, conducting labor surveys and research, constantly refining their complex models, have a difficult time telling you how many jobs were created in the PAST!
In fact, monthly BLS revisions of past job creation estimates are routinely off by tens of thousands of jobs, and on occasion by more than a hundred thousand jobs. The annual benchmark surveys always reset employment levels by hundreds of thousands of jobs.
And we're supposed to believe that the Council of Economic advisors have acquired the clairvoyant ability to estimate payrolls in the future? Please.
Romer, Goolsbee and Bernstein are smart people, and yet they haven't learned from even their recent misadventures with payroll data projections -- having already experienced the folly of attempting to project the range of possible jobs levels if the stimulus were passed. Projecting job creation with any degree of accuracy was always inherantly impossible, and should never have been taken seriously.
If I -- or even my predecessors in the Clinton Administration -- had tried to pull off this ridiculous gimmick we would have been run out of town. I don't even believe it's possible to look back and accurately measure the "job-saving" impact of Bush or Clinton Administration policies, let alone to measure in real time, or project into the future.
On Friday the BLS will release its estimate of May job losses. They will also report their revisions for March and April. And White House officials will once again gear up the spin machine on how many jobs have been "saved".
A self-respecting press corps would vigorously question the White House on their claims. We'll see if we have one.