BLS Employment Data Is Too Bad to Be True

As a devout skeptic of Bureau of Labor Statistics (BLS) methodology, the January 2011 jobs report filled my wildest expectations for nonsense.  In a series of moves over the last two months, the BLS has dropped the unemployment rate from 9.8% to 9.0%, while adding only 157,000 jobs.  This is an increase in jobs of 0.12% at the same time that the unemployment rate has dropped by 8% (0.8% drop/9.8%).

The unemployment rate fell by 0.4 percentage point to 9.0 percent in January, while nonfarm payroll employment changed little (+36,000), the U.S. Bureau of Labor Statistics reported today. Employment rose in manufacturing and in retail trade but was down in construction and in transportation and warehousing.  Employment in most other major industries changed little over the month.

What happened?  Here are the words of BLS.

Changes to The Employment Situation news release tables are being introduced with this release.  In addition, establishment survey data have been revised as a result of the annual benchmarking process and the updating of seasonal adjustment factors. Also, household survey data for January 2011 reflect updated population estimates.

Why the big drop in unemployment?  It is caused by the mathematical consequence of decreasing the civilian workforce by almost 800,000 over the last two months while decreasing the unemployed count by almost 1,200,000.

Unemployment rate is calculated by dividing the unemployed (13,863,000) by the civilian work force (153,186,000).  Econintersect estimates the headline unemployment rate is much higher based on falling population to employment ratios.

This drop in the unemployment rate should only work for those who believe the civilian population is dropping while 157,000 of the previously unemployed 1.2 million people have found jobs (the rest evaporated or have found employment in another country).

Econintersect did warn last week that the bad weather will have a negative influence on the reporting – but this movement was caused by more then the weather.  We need to wait until February reporting to see.

There were backward revisions throughout the data.  This is the annual benchmark adjustment that the BLS makes each year for the prior year.  See Jeff Miller discussion.  It will be finalized with the next report at the beginning of March, but further changes are likely to be small, if any.  This year’s revisions (-215,000) are much smaller than the massive 800,000+ reductions applied to the 2009 data in early 2010.

A month ago, Econintersect forecast that the ADP and BLS data would correct towards each other.  One set of data was too optimistic, the other too pessimistic.  This is the second month of correction, and ADP will adjust further in next months data towards BLS (analysis here).  Now the two data sets are starting to converge nicely.

Further highlights of the BLS report:

  • After accounting for the annual adjustment to the population controls, the employment-population ratio (58.4 percent) rose in January, and the labor force participation rate (64.2 percent) was unchanged.
  • The U-6 all in unemployment rate fell from 16.7% to 16.1% (it was 17.0% two months ago).
  • The number of persons employed part time for economic reasons (sometimes referred to as involuntary part-time workers) declined from 8.9 to 8.4 million in January. These individuals were working part time because their hours had been cut back or because they were unable to find a full-time job.
  • A measure of capacity, weekly work hours for private nonfarm employees – declined slightly form 34.3 hours to 34.2. (likely weather related)
  • Within professional and business services, employment in temporary help services was little changed in January (-11,000). Temporary help had added an average of 25,000 jobs per month over the prior 12 months.

Overall, this report was much weaker then anticipated, and includes so much revision that it will take Econintersect a few days to parse the data for a more detailed analysis.  This is terrible data for an economy in recovery – and is contrary to a variety of surveys on January 2011 employment growth including ISM (analysis here) and Federal Reserve (news here).  It flies in the face of ADP and Challenger data (analysis here).

But the National Federation of Independent Business (NFIB) seems to agree with the with the BLS.  The statement from William C. Dunkelberg, chief economist:

“GDP regained its 2007 peak in the fourth quarter, but it did so without re-employing the 8 million workers who had a job before the recession started. This is in part explained by the composition of the recovery, which continues to exclude two of the major employment sectors, construction and retail, which continue to struggle. Reported productivity numbers support this, a mathematical necessity, but also reflect structural change—firms did learn to do with less. It appears that more firms are terminating operations than new firms are being formed. In the long run, it is the net formation of new firms that produces job growth. A static population doesn’t produce net new jobs, just different jobs, as technology and spending patterns change. Look for a steady unemployment rate and not enough new jobs to keep up with population growth as the most likely outcome for January.

“Reports of net job creation continued to oscillate around the ‘0’ line. Asked about changes in total employment over the last three months, 11 percent of owners reported increasing employment at their firms by an average of 2.8 workers while 15 percent reported reducing total employment, an average of 2.9 workers per firm. Clearly, there has been no surge in hiring in the last few months. This produced a reduction of -0.15 workers per firm. In normal times, this measure would be in the range of 0.1 to 0.2 workers per firm and in a strong recovery, much larger, as was the case in 1983.

“The percent of owners reporting hard to fill job openings was unchanged from December, at 13 percent, the best reading in 24 months. This indicates that the unemployment rate should be steady or even improve a bit in the months ahead. Plans to create jobs lost 3 points, falling to a net 3 percent of all owners, an unwelcome development, but still positive. The weather has been hostile to spending on Main Street, further depressing job creation. The small business sector continues to underperform on job creation in this recovery compared to other recovery periods.”

Overall, this report seems too bad to be true.  This data is telling us that employment is below the levels needed to absorb the young entering the workforce (which needs to exceed 150,000 per month).

Please stay tuned.

Related Articles

Is It Possible the Jobs Cavalry will Ride to the Rescue? by Steven Hansen

Officially Out of Recovery and into Expansion – NOT by John Lounsbury

ADP and Challenger:  Strong Employment Gains by Steven Hansen