The Bureau of Labor Statistics (BLS) released data on February US unemployment.

The report showed initial claims remained at 9.7% (seasonally adjusted). This is known as the seasonally adjusted U3 rate. The BLS makes adjustments to the data every month to compensate for known annual pattern distortions in the population employment data. They do this to account for such predictable annual things as when students look for jobs every summer. It is helpful to have both the nominal (non-adjusted) as well as adjusted data.

The data:

Not Seasonally Adjusted:

  • U3 rate: 10.4%. that is 16M people.
  • U6 rate: 17.9%. basically all unemployed + underemployed + those that have give up but still need work. That is 27M Americans.
  • 41% have been unemployed for 27 weeks or more. Normally, roughly 20% of unemployed are out of work for 27 weeks or more.

Seasonally Adjusted:

  • U3 rate: 9.7%.
  • U6 rate: 16.8%.

Thank goodness for seasonal adjustments. 16M initial unemployed is much better than 15M.

It is interesting to observe a pattern with the White House doing something of a head-fake prior to a government report on economic data. Several times in the past six months, someone from the White House has delivered a statement or interview that claims or hints economic data to be released shortly may appear weak. The data is subsequently released and is less bad than the lowered expectations. We saw this happen earlier this week when Larry Summers indicated in an interview that the jobs data to be presented on Friday would be weak, but the winter storms were partly to blame. Darn you, Frosty the Snow Man!

By virtue of the fact that such a large portion of unemployed fall into the U6 dataset and not the U3 dataset, I suggest that the U3 dataset is less valuable as an indicator. Let’s be honest, how can we ignore 9M Americans that happen to fall into the more tragic U6 bucket – and just as important – what does this say about the strength of our economy ?

Lastly, I want to point out that snow storms happen in the US in February. Granted, they were more devastating this year. But the BLS data released today is based on who gets paid, not who did not show up for work (perhaps because of a snow storm). So I refute Larry Summers position and indeed suggest that snow storms may have increased the number of paychecks in February, not subtracted.