Two days before that report appears, however, in conjunction with Moody's Analytics, Automated Data Processing, a "human capital management" company, releases its own seasonally adjusted survey of jobs added or shed in the private sector. And even though it fails time and again to come up with a count that meshes closely with those of the BLS, economists and other analysts eagerly await the results each month. Today's ADP seasonally adjusted count for July? 200,000 new private jobs. That compares with the revised count of 198,000 private jobs added in June, originally reported as 188,000.
One valuable element of the ADP report is that it breaks down jobs gains or losses by the size of the company.
• Small businesses
° 1-19 employees: 51,000
° 20-49 employees: 31,000
• Medium businesses
° 50-499 employees: 60,000
• Large businesses
° 500-999 employees: 7,000
° 1,000+ employees: 50,000
• Construction: 22,000
• Manufacturing: -5,000
• Trade/transportation/utilities: 45,000
• Financial activities: 4,000
• Professional/business services: 49,000
Unlike the ADP report, the BLS survey covers both private- and public-sector jobs. The consensus of a small group of experts surveyed by Bloomberg is that the BLS will report Friday that, seasonally adjusted, 175,000 new jobs were added July. But for the past four-plus years, the consensus has been wrong about as often as the ADP report has.
For some time now, the BLS report itself has come under considerable criticism, including from the former chief of the bureau from 2008-2012, Keith Hall. He first made his criticisms publicly known last October.
More analysis of the job report can be read below the fold.
It's not a conspiracy, as former GM CEO Jack Welch would have it, nor BLS statisticians distorting the numbers themselves. Rather it is, as I—and other critics with far better credentials—have said for several years, a question of emphasis, of formula and of the bureau's presentation. All that encourages media outlets to focus on two numbers, the headline numbers, pretty much to the exclusion of the wealth of other data included in the BLS report. Those two are the number of payroll jobs businesses report they have created (again, seasonally adjusted by formula) and the unemployment rate.
But as we critics have repeatedly noted, and Keith Hall now concedes, the numbers are tallied in such a way that it's too easy for one to be considered employed and too hard to be considered unemployed. For instance, if you report to the BLS surveyors that you worked one hour a week, you are counted as employed. If you still want a job but you've been turned down so many times you've given up looking for more than a year, you aren't counted at all. You're considered out of the labor force altogether. And that skews the headline numbers.
Simply stated, if there are 1,000 people in the labor force and 100 of them aren't working, the unemployment rate is 10 percent (100 ÷ 1,000 = 0.10). But if 50 of those 1,000 job seekers stop looking for work out of frustration, they are considered no longer in the labor force, which now clocks in at 950. The unemployment rate thus falls to 5.3 percent (50 ÷ 950 = 0.53).
A rate that falls from 10 percent to 5.3 percent makes for a good headline. But it doesn't make for a good economy if those 50 drop-outs still want a job but are too discouraged by rejections to keep looking for one.
And that, as Keith Hall said in an interview with the New York Post earlier this month, is what's been happening over the past several years. Just as many critics have been saying for years. We've seen a steep drop since the Great Recession begain in the "labor-force participation rate," which is the proportion of the population that is in the labor force, and the "employment-population ratio," which is the proportion of the population that is employed.
Result: The count of the jobless is "artificially low," perhaps 3 percentage points too low, according to Hall. Since the official unemployment rate is currently 7.6 percent, Hall puts the "real" rate at 10.6 percent. Stated another way, instead of 11.8 million being unemployed, 16.5 million are.
Add in those who work part-time jobs even though they want full-time jobs, and underemployment plus unemployment ratchets up the numbers to at least 23 million and probably closer to 25 million.