Geography Reference
In-Depth Information
ture and characteristics of clusters and regional networks are explored in a number of
recent studies (e.g. Cantner and Graf, 2006; Cantner et al., 2008a; Graf and Henning,
2008; Ter Wal, 2008). Apparently, when applied to a local or regional network, there is a
trade- of as the information is gained at the cost of other valuable aspects of the regional
innovation system. In the case of patent data, the observed relations will only include
actors that produce patentable innovations, services are clearly underrepresented and
important actors for regional development, such as venture capitalists or local govern-
ment, are not included.
A dif erent approach in terms of data collection is taken in a variety of studies on clus-
ters and industrial districts (e.g. Boschma and Ter Wal, 2007; Giuliani, 2007; Giuliani
and Bell, 2005). Here, smaller groups of relevant actors are asked about dif erent types
of relations to co-located actors, but because of the high ef orts related to this research
approach, these studies are - with few exceptions (e.g. Giuliani and Bell, 2008) - static
analyses.
Nevertheless, these studies provide us with insights that are to be seen as complemen-
tary to existing studies on regional innovation systems and move empirical research
in closer harmony with the requirements of an evolutionary economic geography (e.g.
Boschma and Martin, 2007; Glückler, 2007).
In his conceptual paper, Glückler (2007) opens a wide i eld for empirical research by
raising important questions about mechanisms of selection, retention, and variation of ties
in geographical networks and discussing their consequences. From the number of ques-
tions, we attempt to answer those that are related to selection mechanisms at the node and
tie level and the consequences for properties of the local network. Our central concern is
about how and through which processes the innovation system is af ected by the network
of interorganizational and interpersonal relationships. But it is not just the relations
between existing actors that shape the system, rather, we propose that a great share of the
dynamics in such a system stems from the entry and exit of actors and how they relate to
each other and integrate themselves into the existing network. What we can observe then
is a changing composition of the network that is related to the formation and cutting of
ties. Another important aspect of a regional network lies in its capacity to absorb external
knowledge (e.g. Bathelt et al., 2004; Graf, 2007). Connecting questions about the impor-
tance of internal density and the need for extra-local linkages, we search for the prevalence
of both types of relations, how they change, and which actors drive this development.
We attempt to answer these questions by providing case study-based insights into
the dynamic, developmental or evolutionary pattern of regional innovation systems.
For that, we conduct an analysis of the development of the innovation network of Jena
between 1995 and 2001. While at this point we cannot provide a coherent theory of the
evolution of networks, we still try to base our analysis on some theoretical building
blocks that we consider central for such a theory in the future in section 2. Section 3
introduces the case of Jena and our database, while in section 4 we present our empiri-
cal results. The i nal section 5 puts those results into the perspective of an evolutionary
approach towards the development of a regional innovation system.
2. Theoretical building blocks
Innovator networks are the result of interaction between various agents who transfer
and exchange knowledge, expertise and ideas among each other in order to create new
