Friday's news release on the June employment situation was less than cheering. Employer reported nonfarm payroll jobs declined by 125,000 owing to a reduction in temporary employees working on the 2010 Census. Private sector jobs nudged up a modest 83,000 and the unemployment rate edged down to 9.5 percent from 9.7 percent in May.
Ordinarily, so small an increase in jobs would not be sufficient to prevent the unemployment rate from rising. In June, however, labor force participation dropped as more of the jobless retreated to the sidelines and were not counted as officially unemployed.
But one or two month's data, particularly when volatile and subject to revision, don't make a trend. We need to take a longer look at employer-reported payroll jobs as well as the government's more comprehensive civilian employment series collected by household survey.
Since December 2009 payroll jobs have risen by 882,000 or by an average of 147,000 a month, enough to keep up with the normal increase in the labor force due to population growth but not enough to meaningfully cut into unemployment. With the preponderance of new jobs in the private sector, employment since year-end at least looks to be on an upward path.
Total civilian employment has put on a better showing than the payroll data, rising by more than a million since year-end and pulling down the unemployment rate from 10 percent last December. The labor force participation rate last month was slightly higher than in December, so the credit for lower joblessness goes to employment creation, primarily from the private sector, and not withdrawal from the work force. The household survey employment data explain this year's reduction in unemployment. The payroll data do not.
When the comparability of the civilian employment data is improved by adjusting the series to the definition of payroll employment and smoothing out the effects of the government's annual population control revisions, the contrast between the two series is even more dramatic. (The adjusted series is published monthly on the Bureau of Labor Statistics website.)
In the past six months, civilian employment so adjusted has risen by 1.7 million, averaging a hefty 285,000 a month, and painting a picture of a solid recovery.
Question: Is the faster rebound in civilian employment simply a reflection of its relatively steeper decline during the recent recession? No. From the business cycle peak in December 2007 to December 2009, the adjusted employment series fell by 7.6 million. Payroll jobs fell by more - by 8.3 million.
In this year's job recovery to date, the increase in payrolls is equal to 10 percent of the jobs lost in the last recession. Adjusted civilian employment, by comparison, has recovered 22 percent of its recession losses. But we need to keep in mind that both figures are upward biased by the recent bulge in temporary Census hiring. If government is excluded from the data, the job recapture rates are two to three percentage points lower.
The reasons why household-measured employment is doing better than the payroll data this year are not clear. Trends in the two series have diverged from time to time in the past, but for periods of several months or longer there is no good reason to reject one series in favor of the other. For individual month-to-month changes, the payroll data are usually preferred because of their smaller sampling error.
In such a predicament, one could do worse than simply take an average of the payroll and the adjusted household employment data, in which case the recovery in employment so far this year can be characterized as moderate and better than we thought. The calculation indicates an average monthly gain in employment since year-end of about 215,000. A continuation of that trend would be sufficient to slowly but steadily reduce unemployment, allowing for some recovery in labor force participation.
A continuation of this year's comparatively smaller average increase in payroll jobs wouldn't be enough to meaningfully reduce joblessness in the months ahead. But a job recovery equal to the stronger average showing this year in the Bureau of Labor Statistics adjusted civilian employment data would have the unemployment rate breaking below nine percent in a year's time and below eight percent in two years time.
Alfred Tella is a former Georgetown University research professor of economics.
Read Full Article »