The Unemployment Decline: Good or Bad News?

The official unemployment rate fell by an impressive (and statistically significant) 0.4 last month, from 9.0 percent in October to 8.6 percent in November.  8.6 percent is the lowest rate since March 2009.

"Signs of Hope," cheered the New York Times. The Washington Post said that economists "liked what they saw." (But not all economists.)

Was the unemployment decline really good news, or was it bad news disguised as good?

Let's start with jobs. Payroll employment rose only modestly in November, by 120,000. The payroll data, collected from employers, have a monthly sampling error of about 100,000, so the November gain passed the test of statistical significance. Nevertheless, it was barely enough to absorb the normal increase in the labor force due to growth in the working-age population. There was nothing left over to lower unemployment.

And keep in mind that the payroll data are a measure of jobs, not a head count of employed persons. To the extent that new jobs go to people already working in another job, the increase in the number of persons employed is less than the recorded rise in jobs.

By comparison, the number of civilians employed as measured by the government's household survey showed a 278,000 rise last month. However, the change was well within the large 436,000 monthly sampling error range for that survey, so it failed the test of statistical significance. Economists have traditionally placed greater reliability on the payroll employment data than on household-measured employment.

What did mainly lower unemployment?

A look at the labor force participation rate tells the story. In November it fell by a statistically significant 0.2, from 64.2 percent to 64.0 percent, as more of the unemployed in the face of a weak job market gave up looking for work and withdrew from the labor force. Job search is not costless.

Had the participation rate not declined last month, with employment as it was, the November unemployment rate would have been 8.9 percent - a rate not statistically different from October's 9.0 rate. (It takes a monthly change of 0.2 in the unemployment rate to be statistically significant.)

So what we really had was mostly bad news. Officially measured unemployment declined mainly because more jobseekers lost hope and joined the already swollen ranks of the "hidden" unemployed. More jobseekers are also likely dropping out of the work force as their unemployment insurance expires. It's not known how many dropouts are working off-the-books.

This is not the story of just one month, but of the last four years. The labor market is still in deep recession, and more and more of the jobless have disappeared from the officially measured unemployment data. Both the employment-population rate and the labor force participation rate continue to bump along their cyclical low points. Short-term unemployed have become long-term unemployed, and more of the long-term jobless have become structurally unemployed - and potentially unemployable.

Since the pre-recession peak in December 2007, the work force participation rate has fallen by two full points, from 66.0 percent to 64.0 percent last month. Some of the decline reflects compositional changes in the labor force. Nevertheless, much of it is cyclical reflecting labor force withdrawal and non-entrance - a rise in hidden unemployment due to job shortage.

By my estimate, if the hidden unemployment rate is added to the officially reported unemployment rate (the government's U-3 series), the adjusted total rate is currently over ten percent - well above the official 8.6 percent rate. And that doesn't include the understatement in unemployment due to response errors in the data collection process or the full-time equivalent of partial unemployment, i.e., the involuntary loss of working hours among the part-time employed.

We need to be skeptical of the "good news" about unemployment headlined in the press. Do your own homework. Misinterpretations can sway votes. If the unemployment rate continues to decline because of labor force shrinkage, it could secure President Obama's re-election.