Geography Reference
In-Depth Information
fact that the spatial evolution of network structures may, in turn, af ect the degree of the
dif erent forms of proximity. That would really contribute to our understanding of the
spatial evolution of networks as a truly endogenous process.
3. Firm dynamics, industrial dynamics and spatial clustering
In Part 2 of the topic, the focus shifts from broad conceptual issues to the specii c case
of i rm dynamics and industrial dynamics in space. Here, we start from the micro level,
focusing on the locational behaviour of i rms, and how i rms compete and learn on the
basis of their routines in time and space. This leads to i rm dynamics in the economic
landscape: new i rms will enter the market, some i rms will do well and increase their
market share in an industry, while other i rms will stagnate and exit the market. Moving
to the meso level, one can investigate how these i rm dynamics lead industries to evolve
through dif erent stages of development in time and space. The contributions in Part 2
of the topic explore how such an evolutionary framework can be fruitfully applied to
topics in economic geography at the micro level of i rms (like the geography of entre-
preneurship) and the meso level of industries (like spatial clustering). More particularly,
the contributions examine how i rms behave, compete and learn in space, whether i rm
dynamics at the industry level lead to spatial clustering, and whether spatial cluster-
ing brings positive or negative externalities to cluster i rms along the life-cycle of an
industry.
According to Boschma and Frenken (2006), evolutionary economic geography exam-
ines how the spatial structure of the economy emerges from the micro-behaviour of indi-
viduals and i rms.
4
Instead of describing the behaviour of individuals and i rms as if they
optimise, they follow Simon's (1955) concept of bounded rationality to claim that i rms
are subject to cognitive constraints (Dosi et al., 1988; Nelson and Winter, 1982). In order
to reduce uncertainty, i rm behaviour is guided and constrained by routines. Because of
their tacit and cumulative nature, routines are not easy to change, and are very dii cult
to imitate for other i rms (Heiner, 1983). Evolutionary theory predicts that most i rms
innovate incrementally, exploiting the knowledge they have built up in the past. Nelson
and Winter (1982) have described this as a 'local search process'. And when i rms diver-
sify and grow, they tend to expand into related products, that is, into those products that
are technologically related to their current products (Penrose, 1959).
Taking such a micro-perspective, an evolutionary approach to economic geography
can describe the evolution of the economic landscape as changes in the time-space dis-
tribution of routines over time (Boschma and Frenken, 2003), that is, how new routines
come into existence, and how they dif use in time and space. The economic landscape (as
it manifests itself in the spatial clustering of industries, for instance) is then the result of an
evolutionary sequence in which some variations of routines have been selected because,
for some reason, they are better adapted than others. As mentioned above, selection
occurs at the micro level of the i rm (through its routines), but also at the macro level of
markets and institutions. Market competition acts on variety as a selection device, which
determines which (old and new) routines survive and prosper, and which ones decline
and go out of business (Ormerod, 2005). In a dynamic economy, i tter routines become
more dominant over time through selection, enabling more ei cient i rms with i tter
routines to expand their production capacity and market shares at the expense of less
ei cient i rms. The selection environment not only includes markets but also institutions,
