What the August Jobs Number Misses
The jobs numbers came out this morning – although a lot of us, especially the president, probably wish they hadn't. The official tally is 96,000 jobs added and 8.1% unemployment. As Daniel Gross puts it, "a triple-decker Meh sandwich, slathered with Blah."
As I say below, these figures still mask a lot of underlying pain in the labor market. Dig a bit into the household survey, and you see that 360,000 people dropped out of the labor force, which artificially lowered the unemployment rate. Sure, the nonfarm payroll report gave a business-end measurement of 96,000 jobs added – but that includes the double-counting I describe below.
The broader household survey, which surveys regular folks, actually saw a decline in employment of 119,000.
Below the headline numbers, the report contains all the bad news we predicted. June and July job creation was revised down a total of 41,000, average weekly earnings fell, and perhaps worst of all, the median duration of unemployment jumped to 18 weeks. As former CBO chief Doug Holtz-Eakin put it, "The August jobs report tells a tale of an exhausted and discouraged American workforce." Any "good" news is more a product of the nonfarm payroll report's methodology – specifically, its labor-force masking effect.
Read on for why and how the headline August numbers miss the real story.
On Friday, at 8:30am, all eyes will turn to one, economy-shaking, campaign-driving figure: the number of American jobs added in August. Economists are expecting anywhere from 70,000 to 150,000. Two months before Election Day, the former would be a big Obama blow, while the latter would mean champagne in Chicago.
But the headline number’s source, the nonfarm payroll report (NFP), is worth a more critical look. The report does provide a decent snapshot of month-to-month payroll change. But it also understates the weakness of the labor market. It leaves out a lot of bad news: stuff like less labor force participation and longer unemployment spells. No matter the jobs-added number, both campaigns will try to spin the report. But caveat voter: there’s a lot more data – much of it, troubling – lurking below the morning’s main announcement.
The Labor Department comes up with tomorrow’s big number by asking about 141,000 nonfarm businesses how many workers they have on payroll. They survey the same firms each month, and simply report the change. Compared to the agency’s household survey – which provides a broader picture of employment by calling people at over 60,000 households – the NFP is basically a snapshot. It’s a glance at the labor market’s speedometer, but not the road.
“I never look at [the NFP] report myself,” said Professor David Card, Berkeley labor economist and program director of the National Bureau of Economic Research’s Labor Studies program. “The report is designed to give a very short-run prediction … They have to do it every month and have a limited budget.”
Due to its short-termism, the NFP can mask underlying trends. As Yale economist Ray Fair explained, if you have two jobs, you’re counted as two workers in the survey. Moreover, the NFP doesn’t capture growing population or a shrinking labor force. Over the course of the recession and recovery, the size of the labor force – everyone who has a job or is looking for one – has plummeted, as dejected applicants give up their job search, claim disability, or go back to school.
The morning’s second headline number, the official unemployment rate – last month, 8.3% -- comes from the household survey, which in itself understates the degree of labor market frustration. Since 2008, the number of adults who have searched for work, failed, and dropped out of the labor force has skyrocketed, and is still near its recession-era height (just over 800,000 people). So has the number of adults claiming disability during the downturn. And so has the number of middle-aged workers leaving the workforce to enroll in community college. Those people don’t make it into the “official” rate.
That number, which the media and markets will pounce on, will headline the Labor Department’s household survey. But a better look at labor force woes will come from the report’s “alternative measures of labor underutilization” – most notably, its “U-6 measure,” which includes all of the official unemployed, plus those “marginally attached to the labor force,” plus those who have a part-time job, but can’t get a full-time one. That rate in July? 15 percent.
There’s more bad news that won’t show up in the headlines. The ++median duration of unemployment is now 16.7 weeks,++[ http://research.stlouisfed.org/fred2/series/UEMPMED] above recession-era highs. If people don’t return to work as population grows, potential job growth is more easily outpaced. As Card puts it, “There has to be some kind of continuous employment growth to keep up with population growth.” Longer unemployment duration also increases the chances of labor force dropouts.
The NFP understates these problems, and even the household survey’s official unemployment rate gives an incomplete picture.
To be sure, a look at payroll changes on the business end is incredibly valuable to the markets, which are already rallying on NFP optimism – as Fair explained, it makes sense for payroll figures to be more important to businesses and their investors. As Card put it, “The payroll survey is really getting a very quick pulse on the business end.”
But for the rest of us, the headline jobs number says too little. The nonfarm payroll report leaves out a lot of economic pain, and provides an incomplete picture of changes in the labor market. It’s not quite a presidential report-card. And if it is, it’s likely to be an overly rosy one. You have to look at the rest of the data – and particularly, the household survey – for a fair shake.