Geography Reference
In-Depth Information
Table 17
.
3
Network statistics for the networks of technological relatedness
tech
P1
tech
P2
No. of actors
139
189
Share of largest component
97.1%
94.7%
Share of isolates
1.4%
0.5%
Density
0.149
0.165
Mean degree
a
19.165
28.730
Network centralization
0.602
0.717
Overall graph clustering coei cient
1.228
1.200
Average distance (among reachable pairs)
2.003
1.879
Note:
a
Networks have been dichotomized.
Summarizing these results for the Jena region, we i nd decreasing fragmentation
combined with a higher degree of connectedness where the large, core actors within the
network increasingly focus on formal cooperation, while the smaller, surrounding or
peripheral actors rather intensify their contacts through informal, personal relations. In
the following, we want to analyse more closely the determinants of this increasing
connectedness of the Jena network of innovators that took place despite an increase
in the total number of actors. We start with the analysis of the changing potentials to
cooperate.
Potential to cooperate
As noted above, technological relatedness is one of the preconditions for actors to coop-
erate and to exchange knowledge. Any increase in this relatedness may be a cause for
personal relations to become more frequent and for the whole system to become more
interconnected. In this sense, technological relatedness or proximity can be interpreted
as the potential to cooperate.
Innovators are assumed to be close in technology space if they have patented in the
same technology classes. For the technological classes, we use a concordance between the
IPC code of the patent classii cation and 30 technologies that has been elaborated jointly
by the Fraunhofer-Institut für Systemtechnik und Innovationsforschung (FhG-ISI),
the Observatoire de Sciences et des Techniques (OST), and the Science and Technology
Research Policy Unit of the University of Sussex (SPRU). The more classes two innova-
tors have in common, the higher is their degree of technological overlap and thus their
technological relatedness. These technological relations are used to build up a network
of technological overlap that we interpret as the potential to cooperate.
Table 17.3 shows the statistics with respect to this network. Obviously the cohesion of
the network increases - according to the measures density and mean degree - with the
number of actors from P1 to P2. Thus the potential to cooperate increased over time.
Since we consider a very broad concept of technological overlap with only 30 tech-
nological classes, in both periods, the networks turns out to be highly connected. For
each we observe roughly 95 per cent of the actors being part of the largest component.
The potential network has become increasingly connected as evidenced by the measures
of density (0.149 to 0.165) and also by an increase of the mean degree from 19.165 to
