Geography Reference
In-Depth Information
is not deterministic, but is characterised by trial and error, which means that the timely
recognition and elimination of mistakes has major signii cance for the explanation of the
resulting co-evolution. Or, as Murmann (2003, p. 222) puts it: 'no failures, no evolution-
ary adaption through selection'. The timing of individual events that ultimately have an
inl uence on shared evolution is important.
Many aspects brought up by Murmann are already being discussed in economic geog-
raphy. That i rms are not a product of their environment, but can shape it in order to
develop, has been discussed by Storper and Walker (1989), for example. Collective activ-
ity is essential, as an individual i rm can hardly shape its environment on its own, an issue
that has been discussed in the debate on the emergence of industrial districts (Schmitz,
1995). In the broader sense we can conclude from this that the co-evolution of a technol-
ogy/sector and institutions can lead to positive external ef ects that favour and establish
the development path of a sector in competition with other sectors in the territories in
which the institutions apply. Under the assumption that most institutional arrangements
are tied to a specii c territory (whether at a national level as in Murmann's study, or at a
regional one as in the debate on industrial districts) the co-evolution of sector and insti-
tution also explains the existence of a dif erentiated geography (Hayter, 2004). Or to put
it the other way around: co-evolution of a new sector and its reinforcing institutions may
only take place where regional or national societies are prepared to overcome traditional
rules and habits and are sui ciently ingenious to create new ones. Current research in eco-
nomic geography can take up on two aspects of this. First, a comprehensive debate has
developed on the signii cance of national and regional innovation systems (see Bathelt,
2003; Cooke, 2004; Iammarino, 2005). As mentioned before, however, these often see
the dynamics of technology, sector and institution as linked inseparably, which does not
pave the way for a specii c analysis of co-evolution. Second, the concept of the (regional)
lock-in of sectors (and their technologies) has been discussed in economic geography
for quite some time. On the one hand, this debate is concerned with a positive aspect
of lock-in in path-dependent processes (Martin and Sunley, 2006), which is emphasised
here with the co-evolution of technology and institutions. On the other hand, when and
why the simultaneity and reciprocality of co-evolving units becomes dysfunctional in
the form of negative lock-in often remains empirically unclear. The same is true for the
question of when it collapses, thereby inducing the end of both the technology/sector
and the supporting institutions. Geographers have been particularly interested in the
negative aspects of lock-in in institutional rigidities accompanying ageing technologies/
sectors, and hence the non-simultaneity of technical and institutional development in the
late phases of a technology/sector (e.g. Grabher, 1993; Hassink and Shin, 2005; Schamp,
2005).
In many respects the debate on the co-evolution of institutions is itself still at a stage
prior to that of i nal selection and retention. It is still a new debate and there are dif er-
ent hypotheses that dif er in the scope of the institutional concept, the temporality of the
processes to be analysed and the exactness of the links to the evolving technology. One
level of dif erentiation could be the period of time that is analysed, and thus the choice of
methodology. The co-evolution of the German dye industry covers a period of about 70
years (Murmann, 2003), for which the combined action of technology, sector and a few
institutions has been described in great detail. For an analysis of the very long-term proc-
esses of the three industrial revolutions, Von Tunzelmann (2003) suggests a very broad
