Another peculiarity of Samuelson's argument is that he assumes EITHER
redistribution of the NEW hours resulting from increased demand when those
hours reach some unspecified threshold OR, *fallaciously*, that the
increased demand magically arrives (presumably at the aggregate level?) in
"job lumps" rather than in incremental hours. "A careful examination of
economic history" would show that firms respond initially to demand for
increased output by extending the hours of existing employees. It is only
after a indefinite period of time -- when the number of hours demanded have
passed a certain threshold and when employers are convinced that the
increase in demand will be sustained -- that these overtime hours get
redistributed into new jobs.
So not only does work-sharing _not_ imply a "fixed amount of work", in order
to have any impact on employment fiscal and monetary economic growth policy
_must_ assume a redistribution of hours.
"But from the point of view of the economy as a whole, the lump-of-labor
argument implies that there is only so much remunerative work to be done,
and this is indeed a fallacy. A careful examination of economic history in
different countries shows that an increase in labor suply can be
accommodated by higher employment, although that increase may require lower
real wages. Similarly, a decrease in the demand for a particular kind of
labor because of technological shifts in an industry can be adapted to --
lower relative wages and migration of labor and capital will eventually
provide new jobs for the displaced workers."
Tom Walker
http://www.vcn.bc.ca/timework/
