Geography Reference
In-Depth Information
of these knowledge-intensive industries may then be understood from evolutionary proc-
esses through which the industry gradually emerged and developed. This is not, however,
to deny that location-specii c characteristics may matter (see, for example, Bathelt and
Boggs, 2003; Boschma and Lambooy, 1999; Brenner, 2004; Glaeser, 2005; Martin and
Sunley, 2006). On the contrary, the pre-existing localised presence of knowledge, skills
and creativity may still be a good reason why knowledge-intensive industries cluster in
certain areas, and not in others, as the example of the spatial clustering of the British
automobile industry demonstrated. However, instead of saying that locations matter in a
deterministic manner (as propagated by neoclassical thinking), an evolutionary approach
to spatial clustering accounts for dynamic processes through which routines are created
and dif used (e.g. through the spinof process, the formation of a specialised local labour
market, or the establishment of institutions), and pre-existing structures in regions are
expected to condition, but not determine their spatial outcome. This emphasis on con-
tingency rel ects the idea that locations may inl uence spatial clustering of industries to
some extent, and it is up to empirical research to determine whether and to what extent
this locational inl uence varies from industry to industry (Boschma, 2007; Boschma and
Frenken, 2003). In order to determine the ef ect of pre-existing regional structures, it is
crucial to assess the extent to which individual features of i rms (i.e. their routines) matter
for their survival. While many cluster studies tend to assume that clusters have a posi-
tive ef ect (why did the industry cluster otherwise?), this can only be concluded when one
controls for the ef ect of i rm-specii c features. This is crucial, as Erik Stam observes in
Chapter 6: entrepreneurship studies often show that personal attributes of entrepreneurs
(like age, experience) provide a better explanation of entrepreneurial success than do
regional features (see also e.g. Sternberg and Arndt, 2001).
Adopting a focus on individuals and i rms could also contribute to a better under-
standing of the notion of routines, and how learning shapes and modii es routines
(Hodgson, 2009). This could be accomplished by looking at how organisational routines,
being a collective property of a i rm (Nelson and Winter, 1982), are af ected by entrepre-
neurial activities ('intrapreneurship') and the recruitment of new employees. Udo Staber
explores in Chapter 10 how knowledge may be transmitted between i rms, and how that
af ects the routines of i rms. Staber proposes that we regard ideas in the form of mean-
ings as the basic unit of transmission. Ideas are not isolated phenomena, but evolve as
bundles. An evolutionary account could focus on how these bundles of ideas remain
intact (like the set of skills in routines), but also how new ideas (through labour mobility
or spinof activity) may cause tension and incoherence in the existing set of ideas. In a
similar context, Wenting (2008) has made an attempt to assess the ef ect of post-entry
labour mobility on i rm performance in the fashion industry. Boschma, Eriksson and
Lindgren (2009) found that the hiring of new employees with skills the plant already had
inhouse had a negative impact on plant performance. This may be a result of the fact
that the new skills could be quickly integrated into the routines of the plants, but they
did not enhance overall productivity, and might even have caused conl ict and rivalry
with employees with identical skills. By contrast, the inl ow of new employees with
related (not similar) skills contributed positively to plant performance, possibly because
it added new skills to the existing set of skills in the plant, and it formed no direct threat
to employees with identical competences. The next step is to link this ef ect of post-entry
labour mobility to the industrial life-cycle approach. This concerns questions like: what
