The Unemployment Decrease Is An Aberration

In its Friday press release on the employment situation, the U.S. Bureau of Labor statistics reported that "The jobless rate was down by 0.9 percentage point since November 2010." The last time the unemployment rate fell by that much in three months was in 1983. So it's hardly surprising that the recent drop, from 9.8 percent in November to 8.9 percent last month, made headline music accompanied by back-patting at the White House.

A closer look at the recent decline, however, shows it to be an aberration.

Unemployment is directly influenced by the demand for and the supply of labor. Statistically, the primary variables, respectively, are employment and labor-force participation. Increases in employment that exceed the growth in the working-age population tend to lower unemployment. Cyclical declines in the labor force participation rate (percent of the population in the work force) also lower unemployment.

During economic recoveries, both labor-force participation and the employment-population rate are pro-cyclical and tend to rise together, the employment rate somewhat faster. As job opportunities improve, the backlog of jobless outside the labor force is attracted into the work force. Some of the job gains are absorbed by the labor-force entrants and some are taken by active jobseekers (the unemployed already in the labor force).

So a normal or typical decline in unemployment is accompanied by a rise in the employment-population rate and a rise in the labor force participation rate. But if unemployment during an economic recovery declines mainly because of a drop in labor force participation, as jobseekers who can't find work retreat to the sidelines, then the extent of the so-called improvement in unemployment is open to question.

In the recent November to February sharp decline in unemployment, the employment-population rate increased by 0.2 percentage point. Normally, that gain would have been accompanied by about a 0.1 point rise in the labor force participation rate. But labor force participation turned bearish and fell by a sizable 0.3 point, atypical of recovery behavior.

Fully two-thirds of the 0.9 percentage point drop in the unemployment rate was due to the decline in labor force participation. If the participation rate had behaved normally, the unemployment rate would have declined by only 0.3 point between November and February, to 9.5 percent last month.

The rise in nonfarm payroll jobs since November has averaged only 135,000 a month. The rise in total civilian employment averaged somewhat more, 221,000 a month. Allowing for growth in the working-age population, neither of the employment series was strong enough to lower the unemployment rate by more than a few tenths without help from jobseekers leaving the work force.

There was a similar occurrence between September and December 1983 when the unemployment rate also fell by 0.9 point in a three-month period. The participation rate declined in the face of a rising employment-population rate, contributing to the drop in unemployment. Officially counted unemployed became the "hidden" unemployed.

Earlier, between August and November 1958, the unemployment rate dropped sharply, from 7.4 percent to 6.2 percent, mainly due to a falloff in labor force participation that more than offset a rise in the employment rate.

Again, the unemployment rate fell by a full point in the three months ending September 1950, aided by a decline in labor force participation in the face of rising employment.

These instances of labor-force participation getting out of sync with employment and impacting unemployment are nevertheless atypical, and over the course of the business cycle these variables tend to return to a more normal relationship with each another.

As for the present employment recovery, it will likely continue to improve, lifting the depressed labor force participation rate. Consequently, in coming months the unemployment rate will probably hold at or near its current level, or possibly rise somewhat, as the jobless waiting on the sidelines enter the active work force to compete for jobs. With a little luck the housing market will also show improvement and lend greater geographic mobility to jobseekers. And fingers crossed that the structural gap between the available skills of the jobless and the skills being demanded by employers doesn't widen further.

Meanwhile, in assessing the condition of the job market, recent experience suggests it would be wise to put more weight on the jobs data than the unemployment rate.