a-Based on a sample of 5423 observations for 1993 and 3828 for 1996. Individual hourly
wages for managers and production workers. Base category is front-line in small manu-
facturing firms.
b-Test of joint equality of coefficients.
The idea is to see whether cross- establishment variations approximated by cross-”industry*size” variations
in the differentials is affected by the inclusion of the experience variables. To test that point, a F-test for the
joint equality of the wage differentials is performed.
Inter-establishment variations in wage differentials have been approximated first in a large way by interacting
dummies for manufacturing and non manufacturing firms and for large (more than 500 employees), medium
(between 100 and 500 employees) and small firms. The last part of the table show the results of the estimations
with a more detailed division of industries and firm size. More precisely, 6 different types of firm size are interacted
with 20 different types of industries. The results show the F-test from the regressions.
Looking at the estimated wage differentials in the first estimations with a large definition of establishment,
one can see a clear drop in the coefficients when education is added to the specification (2). The drop is much
weaker when experience is added to specification (2). Experience has an impact on the manager-production
workers wage differentials but the question is whether the effect is similar across establishments. An F-test for
the joint equality of the differentials across manufacturing and non manufacturing large, medium and small firms
show that including experience does not eliminate the variations in wage differentials (the test reject the null
of equality of the wage differentials). With a finer characterization of establishments (bottom of the table), one
arrives to the same conclusions.
In summary, manager-production-worker wage dispersion within establishments does not seem to be strongly
modified when controlling for experience. In addition, the variations across establishments are better explained
by education than experience.
33