4. Adjustment costs, wage elasticities and national labour markets
In Table 3 we summarise our main findings: MNEs exhibit a speed of adjustment that in many cases is
undistinguishable from unity, whereas NEs have adjustment lags that are longer on average and with a
much higher cross-country variation. Despite rather similar short-run (impact) wage elasticities, long-
run values are considerably higher for NEs in almost all countries. Moreover, while long-run elasticities
do not appear to vary much across countries in the case of MNEs (all values are in the range —0.45, -
0.9), the elasticities of national firms seem much more affected by country-level characteristics. A
similar picture emerges when comparing short and long-run output elasticities.16
Table 3 — Short-run and long-run wage and output elasticities
|
Speed of |
short-run |
long-run |
Short-run |
Long-run | ||||||
|
NE |
MNE |
NE |
MNE |
NE |
MNE |
NE |
MNE |
NE |
MNE | |
|
Belgium |
0.09 |
0.80 |
-0.53 |
-0.45 |
-5.77 |
-0.56 |
0.20 |
0.33 |
2.13 |
0.41 |
|
Denmark |
0.32 |
1.07 |
-0.74 |
-0.43 |
-2.30 |
-0.40 |
0.61 |
0.42 |
1.91 |
0.39 |
|
Spain |
0.36 |
0.98 |
-1.06 |
-0.73 |
-2.96 |
-0.75 |
0.31 |
0.61 |
0.87 |
0.62 |
|
Finland |
0.78 |
1.03 |
-0.42 |
-0.54 |
-0.53 |
-0.53 |
0.50 |
0.39 |
0.63 |
0.38 |
|
France |
0.69 |
1.00 |
-0.91 |
-0.73 |
-1.32 |
-0.73 |
0.64 |
0.57 |
0.93 |
0.57 |
|
Germany |
0.52 |
0.92 |
-0.88 |
-0.71 |
-1.68 |
-0.77 |
0.68 |
0.56 |
1.30 |
0.61 |
|
Italy |
0.59 |
1.00 |
-0.96 |
-0.90 |
-1.63 |
-0.90 |
0.60 |
0.73 |
1.01 |
0.73 |
|
The Netherlands |
0.23 |
0.86 |
-0.58 |
-0.47 |
-2.51 |
-0.55 |
0.51 |
0.44 |
2.21 |
0.51 |
|
Norway |
0.85 |
0.97 |
-0.75 |
-0.68 |
-0.89 |
-0.70 |
0.37 |
0.45 |
0.44 |
0.46 |
|
Sweden |
0.55 |
1.01 |
-0.31 |
-0.50 |
-0.56 |
-0.50 |
0.51 |
0.65 |
0.93 |
0.65 |
|
United Kingdom |
0.13 |
0.92 |
-0.46 |
-0.43 |
-3.55 |
-0.47 |
0.52 |
0.53 |
4.01 |
0.58 |
|
Definition |
1 -Y 1 |
1-Y1 -Y6 |
Y 4 |
Y4 + Y9 |
Y4 1-Yι |
γ4 + γ9 1-Y1 -Y6 |
Y 2 |
Y2 +Y7 |
Y2 |
y2 + y7 1-Y1 -Y6 |
Source: calculations on coefficients reported in Table 2
The finding that MNEs have lower long-run wage elasticities can be rationalised in terms of the
higher skill intensity of their work force. Skilled personnel generally cover important roles in a firm, and
with time it cumulates specific human capital that is essential to the activities of the firm and that
cannot be disposed of easily. Regrettably, we cannot control for the skill mix in our sample, but the
finding that average labour cost per employee is invariably higher in MNEs, even after controlling for
size and industry is consistent with the presumption that MNEs are relatively skill intensive.
Lower adjustment costs are the dominant factor explaining MNEs' faster speed in adjusting
employment at the desired levels. Adjustment costs, in turn, are related to hiring and firing costs, but
procedural obstacles could also be relevant. Hiring costs are related to labour availability -- in terms of
both skill levels and willingness to work -- while firing costs are related to the ease of dispose of
unwanted employment (the degree of “labour market rigidity”). Procedural costs can also be due to
16 By imposing a unitary speed of adjustment to all enterprises (assumptions that work against NEs, because according to
estimated coefficients in Table 2 in the short run they are off their labour demand curves), we have also used repeated cross
section estimation as a robustness check. We find that MNEs and NEs are undistinguishable in the short run, and we take
12