AUG#: +130,000 jobs.

Unemployment up at 3.7%...AUG jobs under Trump HERE

Friday, October 12, 2012

More About Those Jobs Numbers...

October Jobs Numbers Have Been Released:
October 2012 Jobs Reports Summary

They are still complaining about "cooking the books" in terms of jobs numbers.

The monthly Bureau of Labor Statistics jobs numbers and unemployment report was released a WEEK ago, but we still have miscellaneous pundits out there seeming to imply that someone somehow was "cooking the books" or otherwise miscounting jobs.

If you have just come across this post and this blog, you may need a little background.  I would suggest that you start with my posts on "873,000 More People Employed" and 
"Is the BLS Cooking the Books?"  Hopefully those two posts will explain the controversy.

July and August:  Number of jobs up but number of people working down.

In July and August 2012, the number of workers (or people employed) reported on the Current Population Survey (the survey that reports the employment numbers and the unemployment rate) went down even though the number of jobs reported on the Establishment (jobs) Survey went up.  Miscellaneous pundits back then were complaining that the jobs numbers were too high, as the number of  "people employed" went down.

September:  Number of people working up (a lot), but number of jobs not up as much.

This month the inverse has happened:  Though both numbers, jobs and people employed, went up, the number of people employed went up far more than the number of jobs.  The same chorus of complainers erupted again this month and added the charge of BLS book cooking as the increase in the number of people employed brought the unemployment rate down to 7.8%.    

To add some light to the subject, I plotted three absolute measures (not increases or decreases) of numbers of private employees (seasonally adjusted) since January 2011:  

  • The BLS private sector number of jobs as reported by employers on the "Establishment Survey" conducted by the BLS (with benchmark revisions).  These numbers are represented by the RED line on the graph below.
  • The private sector number of jobs as reported monthly by payroll processor ADP.  These numbers are represented by the BLACK line on the graph below.  ADP is NOT a government entity, unlike either the BLS Establishment report or the BLS Current Population Survey.
  • The number of people who say they are employed by a private, incorporated business on the "household" Current Population Survey report.  (This represents about 77% of the total of "people employed".  The other 23% of workers or "people employed" are self-employed workers, agricultural workers, government workers, household workers and a few other smaller groups.)  These numbers are represented by the BLUE line on the graph below.  This was the survey, the survey of "people employed" that seemed to be so over the top.
Here's the graph (though some of the specific numbers are hard to read):

A few things to note:  
  • The ADP and BLS Establishment private jobs numbers are fairly even straight lines.  The CPS "private workers/people employed by the private sector" number jumps around quite a bit more.  The CPS uses a smaller survey sample than ADP or BLS and this accounts for its volatility.  It is also self-reported by people answering a set of questions, but that may also increase the volatility.
  • Though all three numbers are different, they tend to be rising at about the same rate over time.
  • An adjustment (an increase) was made to the CPS ("people employed" in the private sector) numbers in January 2012 to account for the results of the 2010 Census.  Some of the large bulge in the January 2012 time frame is the result of this addition of about 216,000 employed people.

  • The September 2011 to September 2012 difference in numbers are, despite the much larger jump in "people employed" in September 2012, fairly similar for all three measures:  BLS private sector jobs:  +2,082,000.  CPS people employed in the private sector:  +2,192,000.  ADP private jobs:  +2,172,000.
  • The January 2011 to September 2012 differences in numbers are also fairly similar among the three measures:  BLS private sector jobs:  +3,745,000.  CPS "people employed":  +3,373,000.  ADP private jobs:  +3,514,000.   That's a difference of about 17,000 employees a month from the low (CPS number) to the high (BLS number).  I did incorporate the recent BLS jobs benchmark revisions, which boosted the BLS jobs numbers by 453,000 jobs from January 2011 to September 2012.  
  • Looking at the chart and the numbers, what is probably true is NOT that the September job estimates are so off...  but that there was unexpected volatility in the "people employed" numbers in the summer that was corrected by September's survey.
  • Most of this whole hoo haa about jobs numbers is just plain silly.  These numbers aren't perfect because they are based on survey samples, and they were never meant to be regarded as the holy grail of the economy.  They were meant to be an "indicator", but, alas, they have become so much more, and that's just plain-- absurd.          

1 comment:

  1. That is so perfect, well stated, and easy for someone like me, who is not a numbers guy to understand. You really ought to market yourself, as the quality of your work far outstrips what is generally done by amateurs and wannabees. This should be on "Newsweek" or someplace else of professional caliber. Im dead serious!


I appreciate intelligent comments and questions, including those that are at odds with anything posted here. I have elected not to screen comments before they are published; however, any comments that are in any way insulting, caustic, or intentionally inflammatory will be deleted without notice. Spam will also be immediately deleted.