U.S. employers added a scant 51,000 jobs in September, according to the government's Payroll Survey, far below market expectations.
But the "other" employment survey, the Household Survey, which is based on households rather than businesses, indicates a much stronger September job market: 271,000 jobs added. This is the survey that reports ethnicity. Here are the month's employment gains/losses by racial group:
Since about half of Hispanics are foreign-born, we use Hispanic employment as an indicator of immigrant employment—data which the government, disgracefully, does not collect.
September marked the third month in a row in which Hispanic job growth lagged that of non-Hispanics. This is, of course, anomalous in recent history. Since George W. Bush took office (January 2001) Hispanic employment has risen by 3,438,000 positions – a gain of 21.3 percent—while 3,636,000 jobs were filled by non-Hispanics—a gain of 3.0 percent. This powerful seven-fold Hispanic job growth trend is reflected in VDARE.COM's American Worker Displacement Index [VDAWDI]. But monthly data seems to be prone to noise, as was also apparently the case in 2004.
We have long argued that the "official" Payroll employment stats were flawed—that the actual rate of job creation was much larger, closer to that reflected in the Household Survey.
The surveys tell very different stories. The Payroll Survey estimated that 135.6 million workers held jobs in September. The Household Survey counted 144.9 million—i.e. more than 9 million more.
Since Bush's inauguration in January 2001, Payroll employment first declined, then recovered. In September 2006 Payroll employment was 3,159,000 above where it was at the start of the Bush administration. But the Household Survey employment is currently 7,072,000 above its January 2001 level.
Why the gap? Some economists have argued that it's the "new economy". Self-employed workers such as part-time consultants, eBay entrepreneurs, and even real estate agents show up in the Household Survey, but not in the Payroll Survey.
We have argued that there's a better explanation: illegal aliens. They will tend not to show up in the Payroll Survey for the simple reason that employers who admit to hiring them risk stiff penalties. (Even though the Bush Administrations appears to have quietly abandoned enforcing these laws.)
By a very cautious count (for example, in this Federal Reserve Bank paper) there are 8 to 10 million illegals living in America. (It may be as high as 20 million.) About 6 million of them are working. By some estimates, illegals account for half of the overall growth in adult immigrant employment since 2000.[A Jobless Recovery? Immigrant Gains and Native Losses By Steven A. Camarota, Center For Immigration Studies]
The gap between the two employment surveys (9 million jobs) strikingly resembles the estimated number of illegal immigrant workers (6 million).
Official Washington may be coming around to our view. In a statement released with the September figures, Acting Bureau of Labor Statistics Commissioner Philip L. Rones announced an unusually large upward revision to payroll employment:
"Preliminary tabulations of employment from state unemployment insurance tax reports indicate that the estimate of total nonfarm payroll employment for March 2006 will require an upward revision of approximately 810,000, or six-tenths of one percent. The historical average for the benchmark revision over the prior 10 years has been plus or minus two-tenths of one percent. BLS currently is researching possible sources for this larger-than-normal expected revision [our italics]…" [Statement of Philip L. Rones, October 6, 2006]
Stay tuned: Illegal aliens may finally be included among the "possible sources" of our employment undercount.
VDARE.COM's recommended next step for government data moles: collect good data on immigrant displacement of Americans. Judging from that long-run Hispanic employment trend, it may be even worse than we think.