Why did the national unemployment rate rise while 244,000 new jobs were created in April? The difference can be blamed on the two surveying methods that the Bureau of Labor Statistics uses. They disagreed last month. This has actually been fairly common during the labor market recovery that began in 2010.

As a refresher, the household survey provides the national unemployment rate, while establishment survey tells you how many jobs were added or lost in various industries. First, I'll just show you the chart, then I'll explain it:

tale of two bls 2011-04.png

Back in February, I explained some of the reasons why these surveys differ. They boil down to four, according to the BLS:

  1. The household survey includes people on unpaid leave among the employed. The establishment survey does not.
  2. The household survey is limited to workers 16 years of age and older. The establishment survey is not limited by age.
  3. The household survey has no duplication of individuals, because individuals are counted only once, even if they hold more than one job. In the establishment survey, employees working at more than one job and thus appearing on more than one payroll are counted separately for each appearance.
  4. The household survey includes agricultural workers, the self-employed, unpaid family workers, and private household workers among the employed. These groups are excluded from the establishment survey.

The giant 434,000-job difference for April cannot be easily described by either of the first two -- those are all likely to create minor differences here. The third we cannot account for based on any of the BLS statistics. But the final difference we can actually correct for: BLS breaks out these workers, which you can subtract from the household survey total.* This is how I calculated the "adjusted household survey" shown in the chart.

It provides a closer approximation of how these two surveys should compare. Really, the only potentially significant difference we aren't approximating for is the possibility of people working multiple jobs. It's pretty hard to believe that this difference could be responsible for the vast disparities that remain in certain months, however. These variances instead likely differ mostly because different surveying methods provide different results. That's a frustrating outcome when trying to understand how much the labor market is improving.

* It turns out that the BLS does not provide seasonally adjusted data for unpaid family workers or private household workers. In practice, however, these are pretty small groups. So I just used the seasonally unadjusted data for these two stats.