Geography Reference
In-Depth Information
data showing non-collaborators in clusters are small but perform better than non-
collaborators of any size outside clusters.
5. Contrasts in stylised interpretations of non-collaborator clustering: neoclassical and
evolutionary
We have identii ed a somewhat unexpected practice by a signii cant portion of small
and medium-sized enterprises in UK ICT. This is that there is sui cient attractiveness to
locate in the midst of what are known to be clusters containing specialised and related
variety i rms with whom inward locators do not seek to collaborate in any intentional
way. For the privilege, they are probably paying up to three times the land and labour
rents they would pay in a not-too-distant science park environment outside the main
clusters (e.g. M4 corridor, Cambridge or Oxford). They are mostly small, even micro-
i rms rather than even medium-sized ones. Their only expressed reasons for this loca-
tional practice are possibly to hear of sub-contracts or of skilled labour availability. To
repeat, this is not to 'piggy-back' a possible consortium engaged in innovative actions,
of which there are numerous ones in such settings, such as
Symbian
for 'Bluetooth' and
3.5/4.0 generation mobile telephony
in Cambridge (Cooke and Huggins, 2003). Rather
they are superi cially 'free-riders' hoping to benei t from moderate 'knowledge spillovers'
for which they are willing to pay up to a 300 per cent locational premium in conditions
of great uncertainty. The uncertainty comes from the strong likelihood that to receive
such spillovers they would have to have some knowledge with which to trade, which on
the face of it seems unlikely, or worse, that they might expect knowledge in a cluster to
be 'free-l owing' when in reality it may be more likely to be preserved in the 'club' atmos-
phere of locations such as Cambridge, where some such incumbents (e.g. in computer
games) were socially excluded from networks involving the rather exclusive Cambridge
Network Ltd, a i rm established in 1986 on the San Diego CONNECT model to enhance
network- based
knowledge- l ow among incumbent members (Cooke and Huggins,
2003).
We may stylise neoclassical and evolutionary interpretations of such, by no means
exceptional, i rm-practice, according to criteria listed in Table 11.4. First, with regard to
motivation; second, with regard to uncertainty; third, relating to utility; fourth, informa-
tion (resources); i fth, expectations (e.g. of cluster); sixth, price; and i nally, the verdict
as to whether such practice might be deemed economically rational or irrational - for
Table 11
.
4
Stylised neoclassical and evolutionary interpretation of non-collaborative
clustering
Criterion
Neoclassical
Evolutionary
Motivation
Externalities
Knowledge spillovers
Uncertainty
High
Learning
Utility
Proi t optimisation
Knowledge search
Information
Low
Selective
Expectations
Specialisation
Variety
Location price
Very high
Spillover cost
Rationality
Irrational
Rational
