Why the unemployment rate won't keep droppingOctober 5, 2012: 3:11 PM ET
In September, the politically charged rate fell because of a huge jump in a survey that few economists trust.
FORTUNE -- It was either an average month for job growth, or one of the best months in nearly three decades. In the September jobs numbers, there was evidence for both.
Last month, employers said they added 114,000 workers to their payrolls. Not fast by any means, but not bad.
The unemployment rate, however, was a blockbuster. In September, the unemployment rate fell to 7.8%, down from 8.1%. That's the lowest it has been since Obama took office.
So is the recovery at stall speed, or finally taking off? The answer lies in how the unemployment rate is calculated.
The numbers come from a survey of individuals. The unemployment rate can go down because more people have jobs. Great. But it can also go down if a growing number of people say they have given up looking for work. Not so great. And for much of the past year the bulk of the reason for the drop in the unemployment rate, which peaked at 10% in late 2009, has been the shrinking workforce.
That changed last month. The workforce, for the first time in years, grew by a lot, up 418,000. What's more, the number of people in the U.S. who said they were unemployed dropped by 456,000. The result was one of the fastest drops in the unemployment rate not only in Obama's presidency, but also the past few decades.
The problem is add those two numbers up, and the economy appears to have added 873,000 jobs in September alone. That pace of job growth, though, seems impossibly high, especially when you consider other measures of the economy. For instance, GDP is growing just 1.3%. That suggests job growth of about 125,000, not seven times that fast.
What's more, the household survey jumps around a lot. Which is why, other than looking at the unemployment rate, most economists tend not to pay a lot of attention to it. And the number this month seems more questionable than usual.
September is always an odd month when it comes to the workforce. It's when high school and college students go back to school. To account for that, the Bureau of Labor Statistics makes seasonal adjustments, upping the number of teenagers and young adults in the workforce.
But something weird happened this year. Fed up with a tough job market, it appears student workers dropped out earlier than usual. The number of 16-to-24 year-olds in the workforce fell by more than 1 million in August. September, then, not surprisingly, had a much smaller drop than usual. That didn't stop the government, though, from making its seasonal adjustments. All told, the Labor Department says seasonally adjusted there were 448,000 more teenagers and young adults in the workforce in September than there were in August. At best, that's likely an overstatement. Either way, it accounts for the entire jump in the workforce and then some.
What about the other 425,000 jobs that were added in the month? Well, there's some evidence to believe that number is inflated as well. The number of people who took part-time jobs because they couldn't find full-time work rose by 582,000. So those people get counted as employed, but they certainly don't have the jobs they want.
Here's what's probably going on: Up until this month, the household survey has been showing weaker job growth than the survey of employers and other economic data. What we saw in September was probably a catch up for that. Combine this month with the past five, and you get an average job growth over the past six months of 157,000. That's about what survey of employers shows job growth to be and only slightly higher than the number implied by the GDP.
A number of conspiracy theorists have been saying that the White House doctored the unemployment rate to make sure it slipped below 8% right before the election. In fact, it appears that the unemployment rate has been inflated for some time, and that Obama should have gotten credit for getting the jobless number below 8% months ago.
But what's also true is that now that we have gotten below 8%, the unemployment rate is unlikely to drop much more anytime soon. For that, the economy would have to be producing much more than 150,000 jobs a month. Any suggestion that the recovery remains anything other than sluggish, still appears to be an economic, and numerical, mirage.