Geography Reference
In-Depth Information
readily be acquired. So the second dimension, foreshadowed above, is the
associative
governance
of the cluster. Conceptually, this involves a major shift from state-regulation
of economic af airs to a degree of self-regulation by responsible groups in economy and
society, but not strictly 'liberal market' governance. In Hirst's (1994) formulation this
means ceding some aspects of economic governance to associations at large capable
of managing certain aspects of communal provision (such as vocational training or
technology transfer), supported by appropriate i nancial mechanisms. It also implies
decentralised, transparent and consultative governance. Institutional learning is a crucial
part of an associative approach. It presumes no fount and origin of all wisdom; rather
it assumes the processes of economic development and especially innovation are inter-
active ones in which institutions on the user side (e.g. customers) may be as important
as producers (e.g. scientists) of the innovations in question. Localised cluster evolution
can be signii cantly assisted by associative institutions and organisations such as these.
They in turn may occasionally act collectively where a regional agency has capabilities
to supply further, more strategic, support. In the next section the chapter moves into a
report of research i ndings into the question of whether i rms in clusters perform better
than their ICT sectoral equivalents if they exist in clusters and collaborate on research
and innovation activities.
4. Measuring the ef ect of proximity and collaboration on i rm performance
The research to be reported on here administered postal questionnaires to structured
samples of UK healthcare biotechnology (not covered in this chapter; for methodo-
logical detail of the project, see Cooke et al., 2007) and ICT companies (264 hardware,
software and services respondents) enquiring about comparative i rm performance of
collaborator and non-collaborator i rms in and outside clusters. Clusters were dei ned
as being located among sectoral neighbours and actively collaborating with them in a
general, possibly informal way or cooperating specii cally and contractually on some
topic such as R&D, knowledge transfer or marketing, as well as networking distantly in
the same respects. Non-collaborators dei ned themselves by reporting they did no such
partnering, formal or informal, with i rms or organisations inside or outside the cluster.
They claimed they only engaged in market exchanges with customers or suppliers. This
addresses a matter of key practical importance to this contribution, namely whether i rm
performance is af ected by business 'clustering'.
Given what has already been written above regarding the
specialisation
emphasis in
the work of neoclassicals and that perspective's neglect of non-market exchange inter-
actions, it would be unlikely to hypothesise collaboration though it might hypothesise
contractual cooperation. The richer evolutionary approach hypothesises both, and as it
turns out, most fruitfully (on related issues, see Boschma and Frenken, 2006 for further
discussion). In fact, from the evolutionary economic geography perspective, the really
pressing question is whether it has a better explanation for clustering than the neoclas-
sical approach, and if so why. To tackle this, the data i rst report more general interac-
tive characteristics of i rms in clusters. It is important to establish the extent to which
i rms consider themselves 'collaborative', whether in geographical proximity or not.
In particular it is interesting and important to separate collaborator performance from
general performance, and key indicator data for ICT are presented in Table 11.1, which
compares key performance indicators for i rms that collaborate or do not collaborate
