producer service industries have on average a higher knowledge intensity than knowledge
intensive manufacturing industries, according to this classification.
In Table 4.3, the results from a 3SLS estimation of the system of equations in
Equation (13a) and (13b) over the total employment in advanced (knowledge intensive)
manufacturing and the total employment in knowledge intensive producer services are
presented. In terms of which estimates that are significant, the results mirror those
obtained in Table 4.2. The only difference is that the dummy for urban regions is
significant in the equation for the producer service sector. However, that the dummy is
significantly negative should not be interpreted as that those municipalities have a lower
employment in producer services overall. Instead, the dummy is negative due to the
steeper relationship for these municipalities, which presses down the intercept. A striking
result, though, is that the size of the estimate for the interaction variable in the producer
service equation is much larger in this case compared to the estimate based on the
aggregate data. This implies that the knowledge intensive industries within the producer
service sector in urban regions are more dependent on the accessibility to knowledge
intensive manufacturing compared to the overall pattern in the same type of regions. This
amounts to say that the hypothesis of a stronger input coefficient between these types of
industries within the two sectors seems to be correct, but only for urban regions.
Table 4.3. 3SLS estimations of Equations 13a and 13b, knowledge intensive manufacturing and
producer services.
|
Variable |
Parameter |
Estimates |
Estimates (producer services)__________ |
|
Intercept,_____________________________ |
a, δ________ |
-1.20 (-1.50)___________ |
-2.26 (-2.90)*____________ |
|
Acc. producer services |
Φ1___________ |
0.02 (2.26)* |
- |
|
Acc. manufacturing |
/1__________ |
- |
0.001 (0.07) |
|
Wage-sum |
Φ 2__________ |
0.00002 (1.85)** |
- |
|
Knowledge intensity |
/ 2_________ |
- |
25.60 (3.66)* |
|
Dummy urban regions |
Φ 3, /3 |
0.23 (0.51) |
-1.76 (-2.13)* |
|
Interaction variable (D*Pa ) |
Φ 4__________ |
0.16 (4.11)* |
- |
|
Interaction variable (D*Ma ) |
/ 4 |
- |
0.30 (4.51)* |
|
adj. R2__________________________________ |
- |
0.50______________ |
________0.42________________ |
|
No. of observations__________________ |
- |
81________________ |
_________81_________________ |
|
Hausman Specification test__________ |
- |
__________________33.95 (9.49)__________________________ | |
*)denotes significance at the 0.05 level.
**)denotes significance at the 0.1 level.
***) t-values are presented within brackets. For the Hausman specification test, the figure within
brackets is the critical value at the 0.05 level.
Turning to the estimation of the dependency between the non-knowledge intensive
industries within the two sectors, the results are presented in Table 4.4. As before, the
estimates have the expected signs. It can immediately be seen that the major difference is
that the estimate for the accessibility to manufacturing employment in the producer
service equation is now significant. In other words, there is a significant relationship
16