Figure 3: Wage Dispersion and Workplace Practices Adoption - NES 1993
Hurv IhtpttMiu and So Itttt I ⅜p∙
ID! H urkplm*ι (ι∣ufr<4t >
Uugr HitparMun ι≡nd SpUna Tj∣∙
(IV- Irnmtir ( «anlrult I
ManutactarittK Imtatlrii** Maituforlur«ц ⅛at⅛ι ∣ trio
'Sar< DitparMttn und Sp Uua I pa
(111- U*ark∣iucι C un4rt⅛*)
IlR SyalBM
Sna MaaatacIurinEtaduvtntt
∖⅜ SgK IN>∣B*r*M∙t Mttl Spteni T*∣ar
(IV- 11 UtttittK I,∙ulr∙l⅜ >
Sen MaunIurIurtnE InttatirMt
no system and the one under the full system with values equal to 2.20 and 2.24 (in specification IV and
V respectively) which implies that equality is rejected at the 87% confidence level. These results are
weaker than the one found in the manufacturing case but recall that the sample size is much smaller
which reduces the precision with which the coefficients are estimated.
Finally, specification IV of table 5 shows the interactions between training for production workers 7
and the system dummies. It is interesting to see that firms that complement the employee involvement
system (either the partial or the full system) with training for production workers have lower wage
dispersion relative to firms who don’t complement the system with training. The difference is strongly
significant in the non manufacturing sector where the reduction in the managerial wage premia for
establishments under the full system is estimated to be of 0.234 percentage point. As a result the
estimated wage premia for firms under the system which also adopted complementary training comes
down to about 27%.
The results from the NES 1993 dataset are summarized in figure 3 and table 6 for both sectors. From
figure 3 one can see that the variations in the estimated managerial wage premia are similar to those
in the NES 1996. In both manufacturing and non manufacturing sectors, the estimated managerial
premium for establishments that did not adopt a system is smaller than the premium at establishments
7The training relates specifically to problem-solving meetings and teamwork.
16