of new sectors. The environment is considered to be of minor importance at the initial stage of
development of a sector when there exists a gap between the requirements of the new industry (in terms of
knowledge, skills, etc.) and its surrounding environment. Windows of locational opportunity are open in
emerging industries. Historical accidents, creative strategies of new industries, and increasing returns, are
considered to drive the agglomeration outcomes. In principle, this may be based on purely evolutionary
mechanisms, in which routines and knowledge are reproduced between organizations by labour mobility,
supplier-buyer linkages, spin-offs, etc. However, non-evolutionary mechanisms may also stimulate this
self-reinforcing process, such as the increase of more specialized suppliers, better infrastructure and a
more diversified labour market. As such, a supportive regional environment (such as a specific knowledge
base and institutional set-up) is more likely the outcome of a long process of co-evolution rather than the
determinant of such a process. For example, the favourable institutions for Germany’s chemical industries
were constructed during the process of industry formation (van den Belt and Rip 1987). Utmost, space
plays a generic role (such as labour markets with generic skills), which only becomes specific and
supportive in those areas in which some critical mass of firms already emerged.4 This also explains why
the assumption of empty spaces in evolutionary models, which address the evolution of new industries, is
justified, while the same assumption is more questionable when dealing with economy’s as a whole.
Where institutional and evolutionary approaches to some extent diverge is that institutionalists view
territories as units with particular institutions that affect the nature of economic activity, evolutionary
economic geography views the region as an entity that is not pre-defined. Comparative analysis is only of
limited use in an evolutionary framework, because economic performance is supposed to be related more
to routines and knowledge built up in the past rather than the present state of institutions. Space matters
almost automatically in institutional economic geography, being basically an approach in which macro-
structures (i.e. institutions) determine the behaviour of agents, and the intensity and nature of relations
between actors at the micro-level. Space may matter in evolutionary economic geography, but not
necessarily, because spatial patterns can also be the outcome of a combination of chance events, human
agency and increasing returns (Boschma and Lambooy, 1999). Moreover, evolutionary economic
geography accounts for contingency, which means space may influence, but does not determine actions
and relations between agents. As such, an evolutionary approach focuses on the dynamic interplay
between macro (structure) and micro (agency) that gives specific outcomes in different places. To put it
differently, the institutionalist approach takes the region as unit of analysis from the very start (it is more
about ‘space of places’), while in an evolutionary approach, a place becomes especially relevant at the end
of the analysis as a node in a ‘space of flows’. This conceptualisation of space comes close to what Bathelt
and Gluckler (2003) have described as the main focus of ‘relational economic geography’, that is, the
evolution of relations between economic agents in space.
Evolutionary economics also suggests that comparative analysis of places, as a benchmarking
exercise, is limited. Since the knowledge and competence base tends to accumulate, and the institutional
setting is quite durable due to its systemic nature and high sunk costs, both assets are intangible forces that
are reproduced in the course of time, but are hard to copy or imitate by outsiders (Gertler, 2003). The core
4 These expectations are confirmed in a recent study of the Dutch software industry (Boschma and Weterings, 2004).
The emerging spatial pattern of this industry in the Netherlands could be best explained by a combination of dynamic
capabilities of firms (capacity to solve problems), spin-off dynamics (spin-offs inheriting successful routines of
parents in terms of background in ICT-business) and urbanisation economies (availability of labour with generic
ICT-skills).
21