282 -
UNEMPLOYMENT DURATIONS IN CLOSED AND OPEN EMPLOYMENT
The typical unemployment spell in open employment systems is
produced by a voluntary quit for search. In closed
employment systems, the spell should result from a temporary
layoff. In open systems, the spell is ended by the
acceptance of a wage offer that matches the reservation wage
chosen by the individual. In closed systems the spell in
ended by the recall. This simple scenario, of course,
assumes several things. First, that quits from open employ-
ment and layoff from closed employment are the only modes of
job separation. Second, that people during unemployment
spells do not move from one sector to another.
There is a third mode of job separation in both systems:
dismissal. Theoretically, such separations should be
frequent in open employment. They should be the exception
by definition in closed employment. Empirically they are
rare. The standard CPS question about this suggests that
only a very small proportion of those unemployed have been
dismissed. This may, of course, reflect response bias
resulting from reluctance to admit having been fired. It
also may result from the unemployment spell being very short
after dismissal. They presumably are caused by a
discrepancy between a person's performance and the current
wage rate. Reemployment opportunities are more ample the
lower the wage rate. If the spells are very short they are
under represented in cross-sectional counts of spells,
because of length bias (Salant, 1977), caused by the
selectivity of spells sampled by the cross-sectional survey
(Sørensen, 1977b). In longitudinal data, such as those
employed here, spells caused by dismissals may be more
adequately represented. However, whether they originate in
open or closed employment, they should result in search.