A Puzzlement on the Labor Force

Where did all the people go?

The strangest number this month in the jobs report is the number for the labor force. It went down 720,000 from September. That is the largest monthly drop since December 2009.

Is that real? Almost certainly not.

So what happened? Some of it may relate to the government shutdown â€" although logically it should not have. Some may result from seasonal adjustments. But most of it seems to be related to the fact the household survey can be very volatile.

The establishment survey, which polls employers and is used to calculate how may jobs were added, was relatively strong. It added 204,000 jobs.

The household survey, which produces the unemployment rate, came up with that labor force oddity. It also reported a sharp fall in the number of people with jobs.

Some of the confusion no doubt comes from the fact the two surveys were supposed to treat furloughed federal workers differently. They had jobs in the establishment survey but were unemployed for the household survey. To make that even more confusing, the government thinks some federal workers who were on furlough were miscounted in the household survey. But it thinks that miscounting caused them to be counted as employed, which would not explain the labor force decline.

It is conceivable that a few of those workers were counted as being out of the labor force, although that should not have happened, given the way questions are phrased. Tom Nardone, the associate commissioner of the Bureau of Labor Statistics, tells me that confusion may account for no more than 20,000 or so of the labor force decrease, when other data are factored in.

The household survey each month adds new households to the survey, and drops others. But among those who were on both the September and October surveys, the largest change went from employed to out of the labor force, Mr. Nardone said. That would seem to be a reason to think the decline may have some validity.

Normally, the labor force grows in October, and so the seasonal adjustments assume an increase. That magnified the decrease in the labor force reported this month.

The household survey can be volatile, and this may just be one of those outliers. The last similar labor force decline, in December 2009, seems to have been a result of such volatility. Within three months, all of them were back.

“We have to wait for additional months of data,” said Steven Haugen, a Bureau of Labor Statistics economist.