The Associated Press's Jeannine Aversa started off her Friday evening report on the day's economic news showing, as she and her AP colleagues have for several months, that they either don't understand very basic concepts relating to the information they're attempting to digest and convey or are deliberately reporting it inaccurately:
Pink slips piled up and jobs disappeared into thin air in May as the nation's unemployment rate zoomed to 5.5 percent in the biggest one-month jump in decades. Wall Street swooned, and the White House said President Bush was considering new proposals to revive the economy.
..... Help-wanted signs are vanishing along with jobs, so the unemployment rate is likely to keep climbing, a government report indicated .....
Make no mistake, the news was bad. On a seasonally adjusted basis, the economy lost 49,000 jobs in May, and the seasonally adjusted unemployment rate rose by more than it has in any single month since the mid-1980s.
But that doesn't change the fact that Aversa either was deliberately inaccurate when she wrote that "pink slips piled up," or that she doesn't comprehend the subject matter she is supposed to be covering.
Here is the chart from the Bureau of Labor Statistics (BLS), updated Friday, showing how many jobs were actually added (yes, added) in May:
(To replicate, go to this BLS link and select the top report in the "Not seasonally adjusted" column.)
The key, of course, is that the chart is not "seasonally adjusted." That is, it represents BLS's best estimate that 648,000 more Americans were actually working in May than were working in April.
The government, using commonly accepted statistical techniques, adjusts monthly data for "seasonality." Seasonally adjusting the raw data smooths reported results, and represents an attempt to give appropriate context to what happened during the previous month relative to previous years. To make a much longer story very short, the fact that the increase in the number of people actually working in May 2008 is less than was added in the previous three years seen above, plus many years previous years not seen, goes a long way towards explaining why the seasonally adjusted jobs reduction in May was 49,000.
May's news, as noted, was clearly not good, and I'm certainly not pretending otherwise. But I'm not going to sit by and watch reporters from AP and other outlets continue month after month to conjure up, out of "thin air," false images of tens of thousands of people thrown out onto the streets and employers en masse slamming their doors shut on new hires, without lodging a reality-based objection. Those things emphatically, and obviously, are not happening. As you can see above, BLS estimates that the economy has actually added 2,481,000 jobs during the past four months -- a number that, while impressive in isolation, is significantly, unfortunately, and unacceptably lower than the numbers seen in the previous three years, and most relevant years before that.
While not good enough, the February through May results do not represent "pink slips piled up," and they sure as heck don't show "jobs disappearing into thin air." Correctly written up, Aversa would have told us that May hiring was disappointingly low for the fourth straight month, leading to a net loss in seasonally adjusted jobs.
But that wouldn't have fit the sky-is-falling, ever more recession-obsessed narrative Aversa and other AP reporters have been pushing for years, even though last two quarters of GDP growth has been tepid but positive, and even though the widely-followed Institute for Supply Management Manufacturing and Non Manufacturing indices have, on a weighted average basis, showed economic expansion for the past two months.
It would also appear to be beyond AP to have any curiosity whatsoever as to why an alternative report on jobs issued by payroll and outsourcing giant ADP (full report is at this PDF file) is showing a year-to-date 554,000-job difference from BLS in seasonally adjusted private nonfarm payroll jobs.
As you can see, BLS says that 357,000 seasonally adjusted jobs have been lost so far this year, while ADP is showing 197,000 gained:
It just may be that ADP, which starts with client employee paychecks issued and projects what they see onto the rest of the economy, is picking up on hiring shifts at small and medium-sized businesses -- clearly favorable this year, especially in the Service-providing sector, according to the company -- sooner than Uncle Sam's folks at BLS, who rely on state-provided jobs and unemployment data plus various surveying techniques.
Evidence that there may be something to this premise is that in two of the past three years, BLS had to make what it calls "Comprehensive Annual Revisions" to the jobs numbers. The two revisions involved had the effect of adding at least 1.3 million jobs (Feb. 2007 - over 900,000 added; Feb 2006 - 400,000 added, per this New York Times article) that were not part of the information routinely reported in BLS's Employment Situation report (and thus went relatively ignored by the business press).
Aversa's repeated misrepresentations over "pink slips," plus other worse-than-rookie mistakes noted earlier this week in an unbylined AP report (see closing paragraphs at this NewsBusters link; at this BizzyBlog link) justify concern as to whether the folks at the self-described "Essential Global News Network" even comprehend the previous three paragraphs. But assuming for the moment that they do, I suspect that their interest in following up on the BLS v. ADP issue before about mid-2009 would be low. That's because what they find in comparing the two, and in dissecting their relative accuracy, might disturb the wire service's almost invariably shrill, too-often ignorant, and sometimes falsehood-based narrative.
So don't expect groundbreaking investigative reporting about the reliability of the monthly jobs numbers anytime soon from AP. There's a recession to keep shouting about, and if they shout long enough, they may help give us a real one -- just in time for November.
Cross-posted at BizzyBlog.com.