Geography Reference
In-Depth Information
the co-evolution of knowledge sources which are internal and external to
the firm, with the external sources being at least partially internal to the
industry (Griliches 1986; Pavitt and Patel 1991; Cefis 1996). Indeed, the
empirical evidence in industrial economics shows that there are significant
differences across sectors in the patterns of industrial competition (Dosi
et al. 1995, 1997; Marsili 2001). These differences exist across various
dimensions in terms of both static and dynamic measures of competition,
such as the degree of market concentration, firm size distributions, rates
of entry and exit, growth rates, the persistence in firms' attributes and
performance, the volatility of market shares of incumbent firms, and the
degree of diversification. For example, entry rates are supposed to depend
on the economic incentives for entry into an industry relating to the trade-
off between profits and entry barriers. However, when looking at the
industry-specific factors that might explain such differences, the empiri-
cal evidence suggests that those factors which are related to the nature of
economic incentives, such as profitability, capital intensity, demand size
and growth, have little explanatory power with regard to the observed
variety of sectoral patterns. In order to explain this variety, it is necessary
to take into account the nature of technology in different industrial sectors
(Marsili 2001).
Some stylized facts on the evolution of industrial structures are pointed
to by evolutionary economists within the so-called Schumpeterian sce-
nario of 'institutionalizing innovation' (Nelson and Winter 1982; Dosi et
al. 1988, 1995; Audretsch 1997, 1998). Firstly, industrial demographics in
a particular sector are shaped by the rate and direction of technological
change in that sector. These in turn depend on the nature of technol-
ogy, such as whether it favours entrants over incumbents or the reverse.
Secondly, the sources and nature of knowledge influence the decisions
to enter the industry in the first place, while technological learning and
market selection determine the growth and survival of firms.
8
Hence, if
cumulative learning occurs mostly through sources inside the sector, such
as via in-house R&D and knowledge which is specific to industrial appli-
cations, a pattern of industrial dynamics characterized by low entry and
high concentration is likely to prevail. If instead new opportunities come
from sources outside the sector, such as academic research and generic and
non-systemic knowledge, high entry and low concentration are likely to be
dominant. Thirdly, the overall process of industrial evolution is character-
ized by structural change, with the emergence of new industries and the
decline of old ones. Radical innovation will often lead to the creation of
new economic activities, while conversely, incremental innovation acts as
the growth engine of existing activities in a specific institutional context
(Verspagen 2000). Consequently, the evolution of firms in an industry is
