Is It Time to Put Unemployment Out to Pasture?

A workhorse of economics, journalism, and politics over the past century has been the unemployment rate. It's a deceptively simple statistic: the number of people who don't have work but are looking for it divided by the entire labor force.

But what is the labor force? Ah, there's the rub. Last month's unemployment figures (via Drudge) exhibit the vagaries of the labor force:

The unemployment rate soared from 5 percent in April to 5.5 percent in May. That was the biggest one-month jump in the rate since February 1986. The increase left the jobless rate at its highest since October 2004.

Yikes! With a labor force of about 170 million, a half-percentage point rise means that nearly a million people lost their jobs. That's a huge economic crash! But wait, the numbers don't quite line up:

...nervous employers cut 49,000 jobs.

In an economy where millions of jobs are cut and created every month ("churning", we call it), a paltry 49,000 were lost, on net? That's 0.028% of all jobs. So where does a change 18 times that size come from?

The government said the number of unemployed people grew by 861,000 in May... The over-the-month jump in unemployment reflected more workers losing their jobs as well as an increase in those coming into the job market -- especially younger people -- to look for work, the Bureau of Labor Statistics said. [emphasis added]

The headline should read: Labor Force Grows Sharply. Nearly a million non-workers looking for jobs is big news - but it's much different news than nearly a million workers losing jobs. Economists have long known that people move in and out of the labor force freely. If a 20-year-old college student starts looking for a part-time job, he's in the labor force. When he decides it's not worth it, he's out. The young, the recently retired, and married women are the most often switchers - and their participation can be read more than one way.

If young people keep joining the labor force, and jobs never materialize, that's a sign of a weak economy, and the increase in unemployment correctly implies a downturn. However, if wages go up (as they did in May) and more non-workers decide it's a good time to look for a job, it's harder to interpret the increase in unemployment. The new job seekers did not, after all, find jobs, so we're not seeing evidence of a boom. But they did decide that looking for work is better than not looking, so they can't be completely pessimistic about their chances.

The bottom line is that unemployment rate as it's currently measured - like some statistics in baseball and elsewhere - is not a great metric of what it claims to measure. We might want to put this old horse out to pasture and let labor economists fight over a better summary statistic of labor force dynamics.