Great Jobs Numbers? U-7 Says It Ain’t Necessarily So

Applicants meet potential employers at the Diversity Job Fair on December 6, 2012 in Manhattan, New York City. Hundreds of minorities turned out to seek jobs, a day ahead of the release of monthly national unemployment figures in Washington DC. Photo by John Moore via Getty Images.

The MSM (mainstream media) have looked at this morning’s employment data and declared them good. Better than good, in fact.

But our own U-7 unemployment statistic, designed to be all-inclusive, edged down very modestly, and the total number of Americans who reported themselves “employed” — 12 million — didn’t budge at all. Nor did the other MSM — the mainstream markets — react much either, at least not in the first few hours after the data were released.

Online, however, every one of the MSM Big Four trumpeted the new numbers in their leads:

Financial Times: “US payrolls rose by 146,000 in November, well ahead of forecasts, as an expected disruption from Hurricane Sandy did not show up in the figures.”

The Wall St. Journal: “U.S. job growth picked up in November and the unemployment rate fell as the labor market shrugged off superstorm Sandy, the latest sign of a steady economic recovery.”

Bloomberg.com: “Payrolls rose more than anticipated and the jobless rate fell to an almost four-year low in November, indicating superstorm Sandy’s effect on the U.S. labor market was limited.”

New York Times: “…The strength of the numbers was all the more notable because many experts expected a muted jobs performance because of the impact of Hurricane Sandy, which hit the East Coast on Oct. 31.”

So what’s going on?

You might call it a case of BLSMPD: Bureau of Labor Statistics Multiple Personality Disorder. That is, there are two very different BLS employment surveys administered every month — one of employers, the other of employees — and as we’ve long warned, they sometimes contradict each other. This morning, they did so once again, though subtly.

The so-called “payroll” or “establishment” survey of employers provided the positive headline: 146,000 new jobs, almost double the low consensus estimate, itself a function of Sandy shock.

And at first glance, the “household” survey of actual people made a similarly upbeat proclamation: unemployment dropping to 7.7 percent.

But if the month-to-month payroll number means little, headline unemployment fluctuations may mean even less. Reported as “U-3” by the BLS, the headline unemployment percentage excludes anyone who didn’t look for a job in the past 4 weeks. Counted as employed, therefore, is anyone who worked at least one hour in the past week, even if they said they were “part-time” for economic reasons and wanted, desperately even, full-time work.

That’s why we tally U-7, which totals everyone who says they want a job — the BLS asks this very question each month — and adds everyone who is working part-time but tells the survey taker that s/he wants to work full-time. U-7 dipped from 16.69 percent to 16.60 percent in November, hardly the headline drop from 7.9 percent to 7.7 percent. Without numbing you with the arithmetic, the U-3 drop is five times larger than what happened to U-7. So what explains the difference?

The answer lies deeper in the data. The US “civilian noninstitutional population” grew by a couple of hundred thousand but the labor force shrank by more than a quarter of a million. That’s a whole lot of people who didn’t die, didn’t emigrate. So where did they go? Given the BLS definition of “labor force,” they were no longer laborers — presumably because they hadn’t looked for work in the past four weeks.

This would jibe with another downbeat statistic. A full hundred thousand fewer Americans were reported as “employed” (even if only super-part-time, remember), though this could be Sandy related. And the shrunken labor force also corresponds to the “fact” that some 200,000 people were subtracted from last month’s “unemployed” total. They are simply no longer “unemployed” by U-3 standards. But they almost surely consider themselves to be if you take seriously the “persons who want a job” number, which rose dramatically in November.

Bottom line, today’s “healthy” drop in U-3, it would appear, is nothing to write home about. As for the payroll survey surge, relative to low projections in the wake of Sandy, I don’t know what to make of it except to proceed with caution. I can’t count the number of times, over the years I’ve been looking at the BLS data, that the payroll and household surveys have contradicted one another. As economist Dean Baker wisely counseled me long ago, look at trends, not the latest monthly blips.

In that spirit, I should point out, as the BLS did this morning, that the much-ballyhooed payroll jobs numbers of the past two months have now been revised downward by almost 50,000. Who knows what will happen to November’s?

In The Wall St. Journal: this morning, Jeffrey Sparshott and Eric Morath sum up nicely:

“While November’s rise in employment was well ahead of expectations, it followed sharp downward revisions to the prior two months. October’s nonfarm payrolls rose 138,000, versus the initially reported 171,000, and September was up 132,000, not 148,000. The lower unemployment rate, meanwhile, largely reflects people leaving the work force.”

This entry is cross-posted on the Making Sen$e page, where correspondent Paul Solman answers your economic and business questions

Follow @paulsolman