AP's Aversa Continues Job Reporting Malpractice

July 10th, 2008 4:58 PM

The Associated Press's disgraceful coverage of last week's Employment Situation Report from Uncle Sam's Bureau of Labor Statistics (BLS) got left behind in the holiday weekend hubbub, but calls out for comment nonetheless.

The AP's Jeannine Aversa reached into her Thesaurus as she began her report with what has become the wire service's standard monthly error of treating reported seasonally adjusted job reductions as reflecting real people thrown out on the streets by mean old employers (as you will see after the jump, reality, as usual, differed):

The nation lost jobs for a sixth month in a row in June, a storm of pink slips drenching this year's July Fourth holiday for more than 60,000 Americans and leaving thousands more worried about the future.

Weighed down by energy prices and the housing crisis, employers laid off workers in stores, factories and forsaken building sites.

..... In June alone, employers got rid of 62,000 jobs, bringing total losses so far this year close to a staggering half-million — 438,000, according to the Labor Department's report released Thursday.

What happened on a not seasonally adjusted basis (i.e., in the real word) differed (go to this link at BLS to access tables referred to at this post):

BLSjobsAdded0608

As you can see, 241,000 jobs were added on a not seasonally adjusted basis (translation: This is Uncle Sam's best estimate thus far of what actually happened in in June). As was the case in the preceding four months, fewer jobs were added than in preceding years, which goes a long way towards explaining why the trend of reported seasonally adjusted job losses continued.

To be clear: The job market's performance during the past six months or so has been very sub-par and unacceptable. But that doesn't change the fact that for the fifth consecutive month, a period during which 2,712,000 jobs have been added (again, an unacceptable number), employers have NOT "gotten rid of" hundreds of thousands of employees, and hundreds of thousands of individuals have NOT been handed a "storm of pink slips."

When you dig a bit deeper (gee, I thought this is what reporters are supposed to do), this month's standard malpractice by Aversa was worse than usual:

BLSpvtGovtNotSAjobsAdded0608

As you can see, the overall real-world result has two distinct components. Private employers added 640,000 jobs in June, while the government sector reduced headcount by almost 400,000. I'd be tempted to say that mean, heartless governments issued a "shower of pink slips," but as you can see from previous years, a reduction of this magnitude is typical for June. The drop likely has a lot to do with the end of the K-12 school year.

The number of private jobs added still trails previous years, and is very troubling. But it remains the case that this later paragraph in Aversa's report is demonstrably false:

Heavy job losses were reported in construction, manufacturing and financial services — the worst casualties of the housing, credit and financial debacles. Cutbacks also came in retailing, temporary help, trucking, publishing and elsewhere. That more than swamped job gains in other places including health care, education, hotels, bars and restaurants and the government.

The not seasonally adjusted facts that contradict Aversa's claims of job gains or losses in the previous paragraph are these:
- Construction, +128,000 (-43,000 seasonally adjusted)
- Manufacturing, +68,000 (-33,000 seasonally adjusted)
- Financial services (finance and insurance, per BLS), +14,400 (-10,000 seasonally adjusted)
- Retailing, +58,000 (-7,500 seasonally adjusted)
- Temporary help, +2,700 (-30,400 seasonally adjusted)
- Trucking (truck transportation, per BLS), +10,900 (-7,400 seasonally adjusted)
- Publishing (publishing industries except Internet, per BLS), +2,200 (-2,300 seasonally adjusted)
- Government, -397,000 (+29,000 seasonally adjusted)

The question remains open as to whether Aversa, or her AP editors, even understand any of this.

Next month will be an interesting one, as July is a month when employment typically goes down. If the overall employment situation picks up a bit, leading to seasonally adjusted job gains, will the AP reporter assigned to the task suddenly discover the not seasonally adjusted statistics?

Cross-posted at BizzyBlog.com.