Table 8. Explanatory model of regional patenting of medical equipment by manufacturing firms
(PTMEH) and by firms, individuals and institutions (PI2)______________
|
______PTMEH______ |
____________PI2____________ | |
|
Constant |
□ 126.5 ___________(□3.17)___________ |
□318.8 _____________(3.31)_____________ |
|
NETWORKING |
22.6** _____________(2.94)_____________ |
59.8*** _____________(3.23)_____________ |
|
BEDS |
0.008* _____________(2.12)_____________ |
0.002** |
|
SIZE |
0.08 |
0.24 |
|
AGGLOMERATION |
0.15 |
0.29 |
|
R = 0.847 R2 = 0.718 |
R = 0.859 R2 = 0.739 |
Notes: PTMEH □ Number of regional patents and utility models granted to Spanish firms divided by
million of inhabitants. PI2 □ Number of regional patents and utility models granted to firms, individuals
and institutions by million of inhabitants. NETWORKING □ Expenditures in computer and software by
unhabitant. BEDS □ Number of beds in clinics and hospitals. SIZE □ Average number of employees by
medical equipment manufacturing firm. AGGLOMERATION □ Percentage of region's population
living in cities of over 500,000 unhabitants, t□values between brackets *p<0.1 **p<0.05 ***p<0.01
Source: Own production
The variable NETWORKING is statistically significant at 99% and the variable BEDS
at 95%. The variables SIZE and AGGLOMERATION behave as expected but they are
not statistically significant. The bivariate correlations among the independent variables
(Table 7) show that AGGLOMERATION and NETWORKING are strongly correlated
at 99%. Therefore we have repeated the regression analysis but excluding the variable
AGGLOMERATION using an stepwise method. The results of the new model are
shown at Table 9.
The results of the model evidence that NETWORKING is the most explanatory
variable and that AGGLOMERATION is almost not significant. This is really
important because the main input of a ‘hiDtech’ activity is the market and technological
information, which it should not be restricted to any specific region. Many ‘hiDtech’
companies have their own "informal networking" of contacts with companies and R&D
centers in other regions around the world. A few studies have evidenced the feasibility
that a ‘hiDtech’ company may locate in regions with no labour or raw materials
agglomeration, but as long as the already existing infrastructure Duniversities,
communications, etcD does not prevent that the new ‘hiDtech’ company survives after
its initial location (Oakey and Cooper 1989; Martinez 1992). Some studies (i.e.
Karlsson and Olsson 1998) even evidence that peripheral regions are able to provide an
13