Behind the Jobs Number: A Messy Reality

Among the many statistics that economic policymakers look to, none matters more than the "jobs number," and 2010 was the year it refused to drop. Today the national unemployment rate hovers near where it began the year, just shy of 10 percent.

For all its totemic power, the jobs number masks a messier reality. It is only an estimate, like a poll or a Nielsen rating, the product of a complex process of research and calculation. Without understanding the assumptions built into the figure, we can't fully understand what it can and cannot tell us.

The poor have always been with us, but the notion of unemployment is more recent. Through the 19th century in Europe and the U.S., there was no need conceptually to separate unemployment and poverty—for everyone except the landed gentry, not working basically meant being poor. English "poor laws" had a distinctly moralistic cast: They required individual parishes to set up almshouses for the "impotent poor" and workhouses for the "able-bodied" poor; beggars and vagrants were thrown in jail and even executed.

It was the Great Depression that changed public attitudes—the ubiquity of joblessness and the way it reached up into the middle class turned what had been seen as an individual moral failing into something else: a symptom of a sick economy. For the first time, governments tried to count the unemployed. The U.S. released its first numbers in 1940. Appropriately enough, the survey that provided the numbers was a Work Projects Administration initiative meant to employ some of the Depression's jobless millions.

Central to this calculation is the notion that to count as unemployed one not only has to want work but to be actively searching for it. The Bureau of Labor Statistics household survey from which the rate is derived defines someone as unemployed only if he or she has searched for a job in the past four weeks: by sending out a résumé, for example, or placing an ad. This makes sense—those not even trying are unlikely to get hired. As the BLS sees it, they are not even part of the labor force. Still, that means the official stat leaves out a lot of men and women who are not working because of the particularly grinding nature of this economy.

The BLS does keep track of these nonemployed workers; it just doesn't categorize them as unemployed. Instead, it describes them as "marginally attached to the labor force." These are people who do not have a job, who would like one, and who have looked in the past year, though not the past four weeks. Many report that they had to stop searching for personal reasons—school or illness or family responsibilities—while others say they are not looking because, in essence, they have given up. That subset is called, aptly, "discouraged workers." If the jobs number included all of these nonworkers, the November 2010 rate of 9.8 percent would climb to 11.3 percent.

Now consider those who are officially counted as employed but who are only working part-time because part-time work is all they can find. If these not-by-choice part-timers are added in, the percentage—no longer an unemployment rate, but something like a measure of the total labor underutilization of the economy—climbs to 17 percent. It's a significantly grimmer picture.

What even that bleak figure leaves out, though, is the dynamic nature of the labor market. No matter what the unemployment rate, workers are constantly moving into and out of the labor force. Even as one middle manager or mechanic is being fired, another somewhere else is getting hired, and while a freshly minted college graduate is just hitting the job circuit, his or her grandfather is preparing to retire. There is a cyclical, often seasonal quality to labor flows—retail companies hire to deal with the holiday rush, construction picks up in the spring and slacks off in the fall, and the beginning of summer sees the labor force expand when students start their school breaks. The official unemployment number controls for these factors and many others.

What worries economists today is not just the size of the unemployment number but how this flow has slowed and, for a large portion of the labor force, become stagnant. Today, 6.3 million Americans have been unemployed for more than 27 weeks. In December 2007, when the recession started, only 1.3 million were stuck in that category. These are workers who are not giving up, who are still answering ads and attending networking events and contacting employment agencies. For more than six months they've been unable to find a job, yet haven't quit the fight and dropped into the ranks of the discouraged and marginally attached. They've remained determinedly, even defiantly, part of the "jobs number."

    Before it's here, it's on the Bloomberg Terminal.