Tyler makes some thoughtful points on ZMP.  Replies:

1. There has been plenty of evidence for “labor hoarding”; oddly,
once the ZMP workers start actually being fired, the concept suddenly
becomes controversial.  The simple insight is that firms don’t hoard so
much labor any more.

Clever, but at least at Princeton, no one claimed that “hoarded labor” had zero productivity.  The idea is merely that firms pay workers more than their current productivity during downturns to avoid the costs of hiring replacements when conditions improve.  If productivity declines 10%, wages don’t fall, and employers don’t get rid of you, that’s labor hoarding, but hardly ZMP.  The fact that today’s employers are more likely to get rid of you during recessions isn’t a symptom of ZMP either.

2. The ZMP worker concept can overlap with the sticky wage
concept.  If a person is a prima donna who will sabotage production
unless paid 120k a year and given the best office, that person has a
sticky wage.  That same person also can be ZMP.  Very often the concept
is about bad morale, not literal and universal incompetence; the ZMPers
are often quite effective at sabotage!

Excellent point.  But as I told Eli, this eliminates the policy difference between Keynesianism and ZMP.  If ZMP is caused by sticky wages, inflation effectively raises productivity.  In any case, note that sabotage need not drive net productivity down to zero to cause unemployment.  As long as net productivity falls faster than the fall in the nominal wage, employers won’t want to cut wages.

3. No one thinks a worker is ZMP in all possible world-states.

True.  As Greg Clark says in A Farewell to Alms, the Industrial Revolution did disemploy a lot of horses.  I just don’t see any evidence that relatively low-skilled humans are the horses of the Information Age.

4. The high and rising premium for good managers is another lens for
viewing the phenomenon.  More workers could usefully be employed if we
had more skilled supervisors, and thus the shadow value for a skilled
supervisor is especially high.

Yes.

5. Virtually everyone believes in the concept, although opinions
differ as to how many workers it covers.  How about the people who are
classified as having given up the search for work altogether?  There’s
quite a few of them.  Put aside the blame question and the moralizing,
can’t at least a few of these people — who aren’t even looking to work — be considered ZMP?

Of course.  I just don’t think that the fraction of the healthy adult population with ZMP has risen much since the last time unemployment was 5%.