Fed Chairman Bernanke's Jobs Bifurcation

Last Wednesday Federal Reserve Board chairman Ben S. Bernanke presented the Fed's Semiannual Monetary Policy Report to the Congress. In his remarks on the labor market he noted that "After two years of job losses, private payrolls expanded at an average of about 100,000 per month during the first half of this year, a pace insufficient to reduce the unemployment rate materially."

But did Mr. Bernanke tell the full "truth" as he himself has defined it?

To understand the context, we need to go back to a speech that then Governor Bernanke gave at an economics conference at Carnegie Mellon University in Pittsburgh on November 6, 2003. The speech was about the jobless recovery following the 2001 recession. In it Mr. Bernanke presented an in-depth analysis of the differences between the government's monthly employer survey of nonfarm payroll workers and its monthly household survey of civilian employment. The analysis included a discussion of survey coverage, sample size, reporting errors, benchmarking, population scaling, and immigration effects.

Mr. Bernanke pointed out that multiple job holders are counted more than once in the payroll survey, but only once in the household survey, and that the more comprehensive household survey includes such groups as the unincorporated self employed, unpaid family workers, and farm workers who are not included in the payroll survey. He rightly stated, "One can correct for these differences and use the household survey to estimate nonfarm payroll employment, the concept measured by the payroll survey."

When the household data were so adjusted, said Mr. Bernanke, the results showed that "in contrast to the payroll survey, the household survey suggests that employment has recovered somewhat since the recession trough. ...More than 600,000 jobs have been created since the recession officially ended in November 2001." The payroll data showed no increase. During the 2001 recession the decline in the adjusted household job count was found to be significantly less than the drop in jobs measured by the payroll survey.

"Which, if either, of these two surveys are we to believe?" asked Mr. Bernanke, who then discussed the factors leaning one way or the other and concluded: "...we do not fully understand the differences in employment reported by the payroll and household surveys, and the truth probably lies in between the two series. However, because of the larger sample used in the payroll survey and because of possible problems with the population estimates used to scale the household survey, somewhat greater reliance should probably be placed on the payroll survey."

Two words jump out of Mr. Bernanke's summary - "truth" and "somewhat."

If the truth lies between the two series, then why doesn't the chairman mention the household data in his testimony or in the Fed's semi-annual reports to the Congress?

Every month the Bureau of Labor Statistics (BLS) publishes on its website the same household survey data adjusted to the definition of payroll jobs that Mr. Bernanke used in his analysis. The data are seasonally adjusted and reported back to 1994. No mention is made of these data in the monthly BLS press release on the employment situation and they're ignored by the media.

One might argue that it's not possible to know exactly where the truth lies between the two series - a valid point. But a range from the two measures could be given which would certainly help fill out the jobs picture.

There's another alternative. Mr. Bernanke placed "somewhat" greater weight on the payroll survey. Somewhat is more than a little, but less than a lot. As an approximation, let's assign a two-thirds weight to the payroll data and a one-third weight to the adjusted household series in an attempt to get closer to the chairman's "truth." Based on such a series, what would the jobs picture look like in the employment turnaround so far this year?

The average monthly increase in the weighted job series in the last six months amounts to 161,000. That's not an impressive performance, but it's a lot better than the 99,000 average monthly gain in reported payroll jobs and reflects a respectable improvement in employment from the BLS household survey count adjusted to the definition of payroll jobs.

There may be a reason chairman Bernanke seems to be turning away from his earlier perspective on jobs data. Perhaps by taking a minimalist view of job growth he hopes to put pressure on congress and the administration to add more stimulus to the economy. That would take pressure off the Fed, which is running low on ammunition. The Fed could then save what bullets it has left in case the economy and the job situation take a turn for the worse.

Let's not bury useful job market information. In assessing employment, let's shoot for Mr. Bernanke's "truth."