>20+ million managers does seem far-fetched, unless it includes
>everyone who has someone reporting to them in some capacity for one
>reason or another, and even then it boggles the mind. Why this huge
>discrepancy? Are there different definitions of manager employed by
>the various BLS surveys, and Epstein is relying on (and perhaps
>further stretching) the most elastic one?
No doubt - that's Epstein's style. The other week he really mangled an analysis of the NIPA data by economist Jason Benderly; I had to call Benderly to figure out what it was all about, and Epstein had gotten it all wrong.
I suspect he's using numbers like these: <http://www.bls.gov/news.release/empsit.t10.htm> and <ftp://ftp.bls.gov/pub/suppl/empsit.cpseea19.txt>, which come from the household survey. I was using the figure for top managers that comes from the establishment survey. I'm guessing his low number from manufacturing comes from the "production occupations" at the bottom of those tables. There's more detail on the makeup of these groups here: <http://www.bls.gov/cps/cenocc.pdf>.
And here's yet another set of numbers, using a different set of definitions, that seems a lot closer to reality than Gene's, showing about 6 million managers (including school principals and mailroom supervisors): <ftp://ftp.bls.gov/pub/news.release/ocwage.txt>.
Doug