Saturday, December 1, 2007

Estimates May Have Overstated Job Growth

The NY Times reports:

The American economy appears to have created far fewer jobs this spring than has been reported so far, a new government report indicated yesterday. That could provide further impetus for the Federal Reserve to lower interest rates when it meets Dec. 11.

The report included a sharp downward revision of the government’s estimate of personal income growth for the second quarter. Because the changes were made as soon as better employment figures were available, the revisions made it seem likely that figures on job creation are also likely to be revised downward in coming months.

The new report concluded that personal income from wages and salaries grew at an annual rate of 1.6 percent in the second quarter, far below the 4.5 percent that had previously been estimated.
The government did not explain why the revision was made, and it is possible that some of it came from reducing estimates of wages or of the profits that employees received from exercising stock options. But it was most likely, said Robert J. Barbera, the chief economist of ITG, that the largest part of the revision came from a change in employment estimates.

If so, he said, he expected the government would revise its estimate of the number of jobs created in the quarter, to as little as 50,000 a month from 126,000 a month. That would indicate that the economy was much weaker than had been thought.

The Commerce Department had estimated that the overall economy grew at a 3.8 percent rate in the spring quarter, surging to a 4.9 percent pace this summer. But those figures are still subject to revision.

The government’s first estimates of personal income are based in part on the Labor Department’s monthly survey of employers. The revision of the personal income statistics yesterday came after the statisticians first saw figures from unemployment tax collections during the quarter. Those figures, based on the number of employees actually on payrolls, are viewed as more accurate and will later be used to revise the job growth figures.

The employment situation this year has been confused by a sharp difference in the two surveys the government releases each month. Through October, the payroll survey estimated that 1.2 million jobs were added to the American economy.

The household survey, which is based on talking to a sample of workers, has shown a decline of 72,000 jobs so far this year.

“This tells us the payroll number is truly flawed,” Mr. Barbera said. “The number the Fed has to key on is the household time series.”

The payroll survey in the past has been considered more reliable and less volatile. But because it is based in part on assumptions about how many jobs are being created by new employers, it “is likely to have some difficulty producing reliable estimates at economic turning points or during periods when there are sudden changes in trend,” the Labor Department stated on its Web site. A large part of the job growth reported this year has come from those estimates, rather than from the survey itself.

The payroll survey underestimated job growth in 2003, when the household survey showed a better employment picture. In the end, the payroll figures were revised higher.

Widget