Geography Reference
In-Depth Information
Table 11.1 indicates non-collaborators enter clusters to access knowledge spillovers. To
begin with Table 11.3 shows some strong support for swifter knowledge exchange, but
also shows less importance, as noted, placed on the core neoclassical presumption of
such entry 'reducing uncertainty'. Finally, it is noteworthy, if unsurprising, that despite
all the enthusiasm and practical action taken by collaborators and non-collaborators to
access communicative connectivity in a cluster, far fewer actually get into contractual
innovation cooperations.
Puzzlingly, 20 per cent of avowed non-collaborators also engage in contractual
cooperation. From interviews conducted subsequent to the analysis of results with a
representative sub-sample of the respondent population, two explanations arise. First,
they perceive arm's length contractual relations as 'cooperation' and second, they actu-
ally do get into contractual innovation and other forms of cooperations as dei ned in
this research to some extent, since they are truly 'opportunistic'. Cognitively dissonant
or rationally utility-maximising? Perhaps i rm-species mutation through learning to
search for and select opportunity from the cluster ecosystem would be the most plausible
explanation. But it also has to be entertained that as active non-collaborators they i nd
themselves more excluded than they expected given their main expressed knowledge
spillovers interest in recruiting talent or contractual opportunities. Alternatively what is
clearly 'in the air' in the cluster is information about patenting and R&D that is not of
as much interest to them or is beyond their absorptive capacity. Probably a combination
of both lies close to the heart of the explanation, perhaps a 'diseconomies of scale' i rm-
cluster problem for non-collaborators is being experienced compared to the reverse for
collaborators (also Table 11.1)
Hence we i nd convincing evidence from these results relative to the following three
key dimensions. First, i rms that collaborate perform better on nearly all performance
indicators than i rms that do not. Collaboration thus gives to i rms in these industries
an added competitive advantage. Second, collaborating i rms in clusters perform better
than collaborators not in clusters. Thus collaboration is good for business but geo-
graphical proximity is best. This means the cluster begins to take on the characteristics
of what we wish to call 'constructed advantage'. This is a dynamically derived form of
advantage constructed on the static qualities of agglomeration, which is transformed
into a cluster by interactivity. Given these i rms are in an ICT platform consisting of
computing and communication hardware, software and services businesses, they derive
evolutionary 'energy' from
related variety
.
Finally, the cluster of ers an unexpectedly
large portion of even non-collaborating (56 per cent) ICT i rms' constructed advantage.
This arises from their conscious aspiration to access knowledge spillovers from the inter-
action ef ects and knowledge 'free-riding' opportunities available to i rms within earshot
of other incumbents with whom they have no intention of collaborating. The possibility
that recruitment of talent is an element of knowledge spillover advantages being sought
by such i rms must also be taken into account. However, Table 11.3 shows they are
frequently disappointed and some evolve to i t the cluster ecosystem by becoming col-
laborators against their avowed intent. Recall the signii cantly smaller employment size
of these i rms, which suggests that though this may not have been their primary interest,
needs must. In general, the constructed advantage of the knowledgeable cluster thus
derives from its local linkages and conveys degrees of competitive advantage directly
and indirectly to its collaborators and non-collaborators alike. This is underlined by
