Geography Reference
In-Depth Information
resources. Accordingly, the founding of both i rms can be characterised as chance events
driven by sociological factors rather than economic choices of optimal locations.
These new i rms had an ef ect on regional industry structure, but initially it was a small
ef ect. It did not start the path-dependent agglomerative processes that deterministically
lead to a cluster. B&O was successful and experienced a large growth in employment.
However, this success did not create a cluster. B&O managed to contain its radio and
TV activity within one i rm. During the history of the i rm, it produced a few spinof s,
but these decided to locate outside Western Jutland, except for two in the 1980s. S.P.
Radio existed for many years before it diversii ed into maritime radio communications.
However, a growing number of i rms in a cluster are by dei nition the most important
process in its emergence. The second i rm entered as a spinof from S.P. Radio and in the
early 1980s the 'cluster' consisted of six i rms of which three were spinof s. In the early
1990s, there were 16 i rms working with wireless communications or related technologies,
and spinof s still accounted for half of the population. The strong employment growth
in North Jutland occurred in the 1980s and 1990s, while the strong employment growth
of B&O was already envisaged in the interwar period. However, the spinof process
pushed the early growth phase of the cluster in North Jutland, while Struer became a
one-company town. Thus spinof s are a key process in the emergence of clusters and it
cannot be explained by interaction between parts of Porter's diamond that are present
in a location because of historical circumstances. However, emergence of a cluster might
suddenly stop at various stages, even though the necessary conditions seem to be present
and spinof s emerge. It is not a deterministic process.
In the early growth phase of the wireless communications cluster the role of geogra-
phy was mainly coni ned to the limited mobility of start-ups. It was very hard to attract
labour to the region in the early phase. An engineer who actually moved from Storno, a
wireless communications i rm in Copenhagen, and later founded a spinof , coni rms this.
He claims that it was not attractive to move to North Jutland to get a job at Dancall in
the early 1980s, because there were only a few i rms to work for. However, as the market
grew and additional i rms entered the industry in the region and the pool of skilled labour
increased, the role of geography changed. The increased market opportunities especially
in mobile communications and the growing number of i rms and labour also increased
the breeding ground for spinof s. It also opened the possibility for labour mobility
between i rms and subsequently the creation of social networks between employees
working in various i rms. In the 1980s and 1990s, the university had an ef ect through its
growing supply of qualii ed labour and basic research in the wireless technologies. Later,
i rms were attracted to the region and entry by parent spinof s. Thus, traditional cluster
ef ects, such as knowledge spillovers, labour market pooling, and attraction of i rms, can
be found in the cluster at a later stage.
The importance of entrepreneurial spinof s is known from other studies of industry
evolution. These studies focus on the role of entrepreneurial spinof s in the evolution of
particular industries that also created well-known clusters, such as semiconductors in
Silicon Valley (Moore and Davis, 2001), wireless communications in San Diego (Simard
and West, 2003), automobiles in Detroit (Klepper, 2002a), and tyres in Akron (Buenstorf
and Klepper, 2005).
Many studies of evolution of the semiconductor industry in Silicon Valley have
shown that entrepreneurial spinof s and employee mobility have been important factors.
