[FoRK] Jobs picture redux, or, when the ugly truth slips out

rst at ai.mit.edu rst at ai.mit.edu
Mon Mar 15 20:18:03 PST 2004

J. Andrew Rogers writes:
 > The job situation and economy isn't dire.  Pretty boring actually, in a
 > vaguely uptrending sort of way.  The number of people who removed
 > themselves from the employment pool because they couldn't find work has
 > stayed constant over the last year , and to clear up a matter, these
 > people are included as "unemployed" persons long after their
 > government/unemployment benefits run out regardless of whether or not
 > they are looking for a job.  

Half right.  The official summary, at bls.gov, the site to
which you are officiously referring people, at


briefly explains that:

  In February, about 1.7 million persons were marginally attached to
  the labor force, about the same as a year earlier.  (Data are not
  seasonally adjusted.)  These individuals wanted and were available
  to work and had looked for a job sometime in the prior 12 months.

That agrees with you so far.  However:

  They were not counted as unemployed, however, because they did not
  actively search for work in the 4 weeks preceding the survey. 

Counting them as unemployed would raise the reported unemployment rate
quite a bit, to 6.7%; note that this still does not count, for
example, the 1.3 million people who are working part time because they
want full time work and can't find it.

 > People looking for work are actually
 > finding it now.  
Well, then, many fewer people must be looking for work.  Total nonfarm
payroll employment during Dubya's term in office has famously dropped
by about a couple million jobs -- something that hasn't happened
during any Presidential term since Hoover.  And a significant increase
would have been required simply to keep pace with population growth.
(Data via http://www.bls.gov/news.release/empsit.toc.htm -- the
complete set, including custom mix-and-match historical charts.  Hours
of fun).

 > If you want to know about the various sources and
 > statistical models that are used, read the reports.  A perverse waste of
 > time, but vaguely interesting.  

Oh, I don't know about that.  It's good to do a check on the people
handing out secondhand reports about what the surveys say.  *All* of

 > The unemployment figure that is normally
 > published comes from surveying individuals, but the same data for most
 > of the figures are also derived from other indirect sources in parallel.

There you go again.  The most widely cited unemployment figure is from
a survey of employers.  There is a separate survey of households,
which has been widely flogged lately by right-wing partisans because
the picture it presents is not quite as bleak as the numbers coming
out of the payroll survey -- but Alan Greenspan, of all people, has
publicly called them to task for it, as that survey is intrinsically
less reliable.  See


which has an excerpt from a New York Times article, with quotes
directly from Greenspan.

 > You can make a lot of hay by choosing which statistical source you want
 > to use for which particular statistical data point.
 > So I am to take your third-hand regurgitated media propaganda as gospel
 > truth, but when I point out that there are significant discrepancies
 > between that position and the underlying documents of authority upon
 > which your dubious source is nominally based, that's just handwaving or
 > something?

In the future, you might want to point out the discrepancies
explicitly, with direct quotes from both sources.  It's not that hard,
and it's a lot more informative for your readers.

Just a thought.


More information about the FoRK mailing list