Geography Reference
In-Depth Information
Table 11
.
2
Collaboration and performance of UK ICT i rms - correlations
Collaboration
ef ect on
performance
Respondent
perception of
collaboration
Perceived
innovation
ef ect
Capacity to introduce new products/
services
0.059
0.213**
0.024
Market share improvement
0.273**
0.181
0.112
Patent announcements
0.158*
0.100
0.075
Turnover increase
0.172*
0.236*
0.199*
Employment change
0.013
0.182*
0.145
R&D change
0.038
0.107
0.086
Improvements of company's best
products
0.102
−0.096
0.323**
New products for company but not
0.029
0.117
0.142
new for market
New products new for company and
0.230**
0.152*
0.203**
new for market
Number of new products/services
0.102
0.011
0.157*
Number of changes in products/
services
0.081
−0.003
0.096
**Correlation is signii cant at the .01 level (2-tailed); *correlation is signii cant at the .05 level (2-tailed).
Source:
CASS ICT Collective Learning Survey.
i gures between 2000 and 2003. Thus collaboration clearly pays in most dimensions of
measurable i rm performance. Hence we continue to believe on the basis of our evidence
that collaboration provides a competitive advantage. Of course, high performance may
attract collaboration, which would be an excellent evolutionary inference. Accordingly,
it is not unreasonable to propose that ICT i rms engaging in collaborative activity with
others are more capable on the R&D and patenting input side of the innovation relation
and they benei t on the output side with greater market share. This is also coni rmed in
Table 11.2 which shows how collaboration is signii cantly and positively associated with
key performance and innovation indicators.
A further key question is the extent to which ICT collaborators - not forgetting non-
collaborators - are found consciously locating in clusters. The answer provided below
shows a picture where geographical (cluster) proximity for UK ICT i rms is important,
as shown by the number of non-collaborating i rms that consciously decide to co-locate
in a cluster (56 per cent), but whether this is proven to be benei cial for i rms' perform-
ance is somewhat unclear. The operation of knowledge spillovers seems to be important,
where substantial numbers of non-collaborators are found in clusters indicating that
there is a 'knowledge spillovers' attraction ef ect even for those who envisage non-
collaborative relations with their neighbouring i rms. These may be assumed to be those
seeking to exploit knowledge that is 'in the air'. Interestingly, collaborators in clusters
perform only marginally better, but consistently so, than those not in clusters, except on
R&D expenditure increase 2000-03 where collaborators increased most. This could be a
