Geography Reference
In-Depth Information
interested in the description of shared histories when both sectors were co-localised in
so-called local production systems. Old industrialised areas were the preferred subject of
research in economic geography, such as the Ruhr region or textile producing regions.
Insofar as current research in economic geography is concerned with dif erent sectors
in local production systems, it is mainly interested in the mechanisms by which they
function. Therefore supply linkages between vertically linked sectors are usually exam-
ined, followed by a description of linkages of knowledge. This research is in itself useful
and may also produce valuable information for subsequent evolutionary analyses.
Nevertheless, it must be remarked that the analysis of the processes of co-evolution of
interlinked sectors in local production is just beginning.
Recent studies have i rst tried to address the issue of how a population of i rms evolves
in the 'core sector' of a local production system or 'cluster', a concept that has become
very popular in the last few years. Beginning with earlier studies, for example on the
emergence of Silicon Valley (Saxenian, 1985) or the so-called Cambridge phenomenon
(Garnsey and Hef ernan, 2007; Segal Quince & Partners, 1985), and continuing, for
example, with the analysis of the development of a cluster in the German packaging
industry (Mossig, 2005), responsibility for the development of clusters was attributed
to spin-of s and spin-outs. That spin-of s have greater chances of survival compared
to new foundations and that clusters can only emerge in this way have recently been
demonstrated by Boschma and Wenting (2007). They explain these i ndings with the
selective adoption of successful routines from incubator i rms as well as in i rms using
related technologies (e.g. automobile from bicycle production). Therefore, two processes
contribute to the evolution of a sector: i rst selective replication of successful routines,
and second variation of these routines leading to a diversii cation in the sector. In these
studies co-location remained somewhat self-evident because of an increased likelihood
that spin-of s deliberately choose geographical proximity to their incubator instead of
looking for 'arbitrary' locations elsewhere. There is no lack of plausible explanations
from traditional locational studies of small and medium-size enterprises (SMEs). A low
level of information on alternative locations and attachment to the founder's existing
lifeworld are two such explanations (Schamp, 2000) used to account for the emergence
of co-localisation within a sector. But what explains the co-localisation of co-evolving
sectors?
The reason for the co-evolution of supporting sectors in a cluster seemed to be just as
self-evident as co-location. Economic geographers, historians and economists have all
failed to provide an explanation based on evolutionary theory to show why supporting
sectors emerge as new populations of i rms. The evolution of a new sector always seems
to be associated initially with a marked vertical integration of its i rms. This can be
accounted for by the fact that incumbent sectors have dii culty in meeting the technical
requirements of the emerging sector. In consequence, the question arises as to when and
why the integrated i rms of an emerging sector start to buy products instead of produc-
ing them themselves. This is the classical question addressed by transaction cost theory.
In its more or less static form - given that transaction costs determine whether market,
network or hierarchy is the best way to coordinate economic activities - it cannot con-
tribute an answer, as the factors that determine the size of transaction costs remain exog-
enous. Jacobides and Winter (2005) argue in favour of a 'systemic' perspective on the
correlation between i rm-internal capabilities and endogenously altered transaction costs
