3. Empirical results and discussion
To explain the spatial distribution of innovative activities in the medical equipment
industry we have carried out a regression analysis by OLS using regional patent and
utility model granted to Spanish residents. The dependent variables used for the study
have been two: the regional number of patents and utility models by employee granted
to Spanish firms (PTMEH), and the regional number of patents and utility models by
unhabitant granted to firms, institution and individuals in that region (PI2). Both
variables are for the 1979□1995 period and are related to patenting in industrial
companies (PTMEH) and patenting in all economic agents (PI2). The independent
variables used have been:
• SIZE: average number of employees per firm in the regional medical equipment
industry.
• BEDS: number of beds in regional hospitals and clinics. A large number of beds is an
indicator of a greater demand for medical equipment but also of a greater
specialization and qualification of physicians than at regions with smaller hospitals.
• AGGLOMERATION: percentage of region’s population living in cities of over
500,000 unhabitants.
• NETWORKING: Logaritm of regional expenditure per inhabitant in computer and
software.
Table 7 indicates the descriptive statistics and correlations of the independent variables.
The values of the independent variables are for 1996. Then, Table 8 shows the results
of the regression analysis for each dependent variable. Both are explained by the model
in more than 70%, which indicates that the approach developed in this paper is valid to
both firm’s and total’s patenting at regional level.
Table 7. Descriptive statistics and correlations
|
Mean |
SD |
1 |
2 |
3 | |
|
SIZE_______________ |
27.4 |
32.1 | |||
|
BEDS____________ |
4,476.3 |
678.0 |
□0.019 | ||
|
AGGLOMERATION |
9.8 |
18.2 |
0.239 |
0.171 | |
|
NETWORKING |
4.2 |
0.4 |
□0.130 |
0.000 |
0.636*** |
Notes: Observations = 17 ***p<0.01
Source: Own production
12