Dec. 5, 2009: The November 2009 Employment Report: Green Shoots At Last?
On Dec. 4, the Bureau of Labor Statistics released the employment report for November 2009. The headline numbers were 11,000 jobs lost but the unemployment rate fell from 10.2% in October to 10.0% in November. You might ask: how can the unemployment rate fall when the economy is still losing jobs? The answer is that the two headline numbers come from two separate surveys: the jobs number from the establishment survey—a survey of employers—and the unemployment number from the household survey—a survey of households. The establishment survey is larger and has a smaller sampling error, but cannot provide information on unemployment. Hence, the BLS also surveys households to get information on unemployment.
In November, the household survey showed that employment actually increased by 227,000, after showing losses of about 1.5 million during September and October. This number may be the first real “green shoot,” indicating that the long recession is finally ending. There also was a downward blip back in July, when the unemployment rate fell from 9.5% in June to 9.4% in July, but that was driven by a 637,000 person decline in the size of the labor force rather than by a rise in the number of employed. During November, for the first time since April, employment actually rose. On the downside, the number of workers leaving the workforce because of discouragement and other reasons continued, with 291,000 members of the labor force leaving during November.
Evidence from more comprehensive measures of unemployment also was positive. The U-6 measure, which includes discouraged workers and part-time workers who want to work full time, fell from 17.5% in October to 17.3% in November. The only real negative news was from those unemployed for 15 weeks or longer, which rose from 5.7% to 5.9% of the labor force.
From the establishment survey, there was more good news, as the job losses reported for September and October both were revised downward by a total of 169,000. In addition, the average workweek ticked up from 33.0 hours to 33.2 hours. Employers will first utilize existing workers for longer workweeks before hiring new workers, so this increase provides additional evidence that the labor market is finally on the mend.
It is important to keep in mind that these results are from samples with large margins of error. We will need a couple of more months of positive data to confirm that the November good news is not simply the result of statistical noise or faulty seasonal adjustments by BLS staff. Most of the month-to-month changes in unemployment are the result of seasonal trends, which BLS staff “removes” using statistical models.
That said, we all should welcome this nascent "green shoot."
