Sunteți pe pagina 1din 18

Research Policy 34 (2005) 11731190

Knowledge bases and regional innovation systems:


Comparing Nordic clusters
Bjrn T. Asheim
a,b,c,
, Lars Coenen
a,b
a
Department of Social and Economic Geography, Lund University, Sweden
b
Centre for Innovation, Research and Competence in the Learning Economy (CIRCLE), Lund University, Sweden
c
Centre for Technology, Innovation and Culture (TIK), University of Oslo, Norway
Received 3 October 2004; received in revised form 26 January 2005; accepted 7 March 2005
Available online 27 June 2005
Abstract
The analysis of the importance of different types of regional innovation systems must take place within a context of the actual
knowledge base of various industries in the economy, as the innovation processes of rms are strongly shaped by their specic
knowledge base. In this paper, we shall distinguish between two types of knowledge base: analytical and synthetic. These types
indicate different mixes of tacit and codied knowledge, codication possibilities and limits, qualications and skills, required
organisations and institutions involved, as well as specic competitive challenges from a globalising economy, which have
different implications for different sectors of industry, and, thus, for the kind of innovation support needed. The traditional
constellation of industrial clusters surrounded by innovation supporting organisations, constituting a regional innovation system,
is nearly always to be found in contexts of industries with a synthetic knowledge base (e.g. engineering-based industries), while
the existence of regional innovation systems as an integral part of a cluster will normally be the case of industries-based on an
analytical knowledge base (e.g. science-based industries, such as IT and bio-tech). In the discussion of different types of regional
innovation systems ve empirical illustrations from a Nordic comparative project on SMEs and regional innovation systems
will be used: the furniture industry in Salling, Denmark; the wireless communication industry in North Jutland, Denmark;
the functional food industry in Scania, Sweden; the food industry in Rogaland, Norway and the electronics industry in Horten,
Norway. We argue that in terms of innovation policy the regional level often provides a grounded approach embedded in networks
of actors acknowledging the importance of the knowledge base of an industry.
2005 Elsevier B.V. All rights reserved.
Keywords: Regional innovation systems; Knowledge bases; Nordic countries

Corresponding author.
E-mail addresses: Bjorn.Asheim@keg.lu.se (B.T. Asheim),
Lars.Coenen@keg.lu.se (L. Coenen).
1. Introduction
Over the past two decades social scientist and pol-
icy makers have been paying more and more atten-
tion to regions as designated sites of innovation and
0048-7333/$ see front matter 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.respol.2005.03.013
1174 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
competitiveness in the globalising economy. The pop-
ularity of this argument can be traced back to various
empirical studies of regional success stories, such as the
rapid economic growth of networked SMEs in indus-
trial districts in the Third Italy (Asheim, 2000), the
exemplar industrial systemof Silicon Valley (Saxenian,
1994) as well as other examples of successful regional
clustering in most developed as well as developing
economies (Porter, 1990). These studies all drawon the
common rationale that territorial agglomeration pro-
vides the best context for an innovation-based global-
ising economy because of localised learning processes
and sticky knowledge grounded in social interaction
(Asheim, 2002; Asheim and Isaksen, 2002; Gertler,
2004). They have emphasised the signicance of the
regional level in economic development in addition
toand sometimes overthe national level.
Two concepts belonging to the territorial innovation
theory family (Moulaert and Sekia, 2003) have demon-
strated particular resonance in academic and policy
circles: regional innovation systems (RIS) and clusters
(Cooke et al., 2004; Porter, 2000). Even though both
concepts are closely related, they should not be con-
ated. Isaksen and Hauge (2002, p. 14) dene the latter
as a concentration of inter-dependent rms within
the same or adjacent industrial sectors in a small geo-
graphic area. A RIS, on the other hand, is dened
as interacting knowledge generation and exploitation
subsystems linkedtoglobal, national andother regional
systems (Cooke, 2004, p. 3). In principle it stretches
across several sectors in the regional economy, given
that rms and knowledge organisations interact sys-
tematically (i.e. consistently). From this follows that
clusters and RIS can (and often do) co-exist in the same
territory. In a policy context it is nonetheless crucial to
acknowledge the sector specicity of clusters and the
more generic sector orientation of RIS.
This paper takes up the issue of regionalising inno-
vation policy by looking from a bottom-up perspective
at the linkage between regional innovation systems
and clusters. Arguing strongly against any universally
valid, one-size-ts-all model, we contend that in line
with its sector specicity a differentiation needs to be
made on the basis of the clusters knowledge base. For
clusters with a synthetic (engineering-based) knowl-
edge base, the logic behind the regional innovation
system (as well as regional innovation policy) is to
support and strengthen localised learning of an exist-
ing industrial specialisation, i.e. to promote historical
technological trajectories-based on sticky knowledge.
We call this the ex-post approach. In the case of an ana-
lytical (science-based) knowledge-based cluster, it is a
questionof promotingneweconomic activity, requiring
close and systemic industryuniversity co-operation
and interaction in the context of, e.g. science parks and
incubator centres. We call this the ex-ante approach.
Based on this distinction we compare ve Nordic
clusters across a range of differing industries and
draw conclusions for the regionalisation of innovation
policy.
Section 2 introduces the notion of the learning econ-
omy as well as the main differentiation track: industrial
knowledge bases. Section 3 introduces the concept
of regional innovation system. Section 4 presents an
overview of the varieties of regional innovation sys-
tems, while Section 5 provides the empirical illustra-
tions from a Nordic comparative project on SMEs and
regional innovation systems. Finally, conclusions and
policy implications are given in Section 6.
2. Providing context: the learning economy and
industrial knowledge bases
Both the knowledge-based as well as learning econ-
omy rationale argue that in the globalising economy
knowledge is the most strategic resource and learn-
ing the most fundamental activity for competitiveness
(Lundvall, 1992; OECD, 1996). However, in academic
as well as policy oriented discourses these two concepts
have fromtime totime takenondifferent meanings with
potential importance for the theoretical understanding
of the contemporary economy as well as for policy
implications. Lundvall has always preferred to talk
about the contemporary global economy as a learn-
ing economy, while the OECD (at least the economic
sections), being inuenced by the US, has instead more
often used the knowledge-based economy. This dif-
ference can basically be traced back to the taxonomic
differentiation between high-, medium- and low-tech
industries as suggested and endorsed by the OECD
(1986). Though the initial discussion was carefully
launched, offering many necessary qualications, it
seems that the high-tech fascination has taken on a life
of its own, limitingknowledge-intensive andinnovative
activities exclusively to high-tech industries, such as
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1175
ICTequipment or pharmaceuticals. Under this assump-
tion R&D intensity is often equated with innovation
across-the-board (Hirsch-Kreinsen et al., 2003).
The key question in this context remains how inno-
vation should be understood and treated, as both de-
nitions basically refer to the importance of innovation
as a source of competitive advantage in a globalis-
ing economy. We follow Edquist (1997) in his broad
denition of innovation as new creations of economic
signicance. This encompasses radically newcreations
as well as new combinations of existing elements,
includes novelty of either a material or an intangi-
ble kind and refers to what is produced as well as
how goods and services are produced. Moreover, we
prefer to argue in terms of the learning economy
rather than the knowledge-based economy because of
its more inclusive and dynamic notion of innovation.
In a learning economy innovation is understood as
an interactive learning process, which is socially and
territorially embedded and culturally and institution-
ally contextualized (Lundvall, 1992). It emphasizes a
dynamic approach to innovation rather than the more
static approach adopted in the knowledge-based econ-
omy that emphasizes access to a stock of specialised
knowledge (Lundvall and Archibugi, 2001). In addi-
tion, this conceptualisation means an extension of the
range of branches, rm-sizes and regions that can
be viewed as innovative, also to include traditional,
non R&D intensive branches (e.g. the importance of
design in making furniture manufactures competitive
and moving them up the value-added chain). Further-
more, knowledge ows within a distributed knowledge
base (Smith, 2000)which more and more substitutes
intra-rm(or intramural) knowledge bases (which con-
stitutes the basis for the OECD taxonomy of R&D
intensity)taking place between industries with very
different degrees of R&D intensity, further weaken the
distinction between high-tech and low-tech industries.
(e.g. when food and beverage rms produce functional
food based on inputs from bio-tech rms).
Nonetheless, this position calls for a certain degree
of modication. The learning economy approach was
developed in a national context of small-sized indus-
tries relying on incremental, non R&D based product
innovations (e.g. Denmark). A problematic aspect of
learning organisations as well as the learning economy
in general has been its focus on catching up learning
(i.e. learning by doing and using) based on incremental
innovations, and not on radical innovations requiring
the creation of new knowledge. It is, of course, impor-
tant to underline the tremendous importance of incre-
mental innovation, learning by doing, by using and by
interacting in the process of technical change and dif-
fusion of innovations (Freeman, 1993, pp. 910). This
is typically what co-ordinated market economies, such
as Denmark and the other Nordic countries are very
efcient at through knowledge diffusing (and learn-
ing) organisations, with a strong developed absorptive
capacity, but being weaker when it comes to knowl-
edge creation as input for radical innovations (Hall and
Soskice, 2001). In a long-term perspective, however,
it will be increasingly difcult for the reproduction
and growth of a learning economy to primarily rely
on incremental improvements of products and pro-
cesses, for example, in the formof imitation, and not on
basically new products (i.e. radical innovations). This
would be even more problematic if it is based on exoge-
nous learning. According to Nonaka and Reinm oller,
no matter howgreat the efciency and speed of exoge-
nous learning, it will not substitute for the endoge-
nous creation of knowledge. The faster knowledge is
absorbed, the greater the dependence on the sources of
knowledge becomes (Nonaka and Reinm oller, 1998,
pp. 425426). In a dynamic and rapidly changing con-
temporary globalising economy it is, thus, necessary to
pay attention to knowledge creation as a process that
is of equal importance to the processes of learning and
competence building.
This modication acknowledges Cooke (in press)
critique of the learning economy. He argues that the
semantic emphasis on learning has become increas-
ingly myoptic with regard to the actual content of
the learning processes at hand. We, therefore, argue
that it is important to keep in mind that both learning
and knowledge remain means to achieve competitive-
ness (through innovation) rather than intrinsic goals
per se. Through processes of globalisation, i.e. increas-
ing inter-connectedness between different parts of the
world, competition has become more and more intense.
In a globalising economy competitive advantage is
based on exploitation of unique competencies and
resources (which thus excludes outdated and second-
hand knowledge). A rm or a region competes on the
basis of what they have which is unique in relation to
their competitors. A strategic perspective in the con-
temporary global economy is, thus, how to develop
1176 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
such unique competencies and resources in order to
foster competitiveness (Porter, 1990).
Maskell et al. (1998) argue that the only economic
production factors not subjected to ubiquitication, i.e.
the process of conventional production factors (e.g.
labor, capital, equipment, raw materials) becoming
more available across the world for the same price, are
localised, sticky knowledge including collective tacit
as well as disembodied codied knowledge (Asheim,
1999). Generally, territorial agglomerations play an
important role whenit comes tothe regional embedding
of knowledge and learning processes. In the same way,
as knowledge is nationally embedded due to sectoral
specialisations as well as political and cultural organ-
isations and institutions, knowledge is also regionally
embedded as a result of a historically produced territo-
rial division of labour. The exploitation of localisation
economies in territorial agglomerations accounts for an
important factor when regional specialisation should be
explained. There are numerous examples of once suc-
cessful agglomerations that were protable and com-
petitive throughout history. However, these once suc-
cessful regional clusters can also run into serious prob-
lems as a result of pathdependencyresultinginnegative
lock-in tendencies, i.e. that the dominating techno-
logical trajectory is not modied or changed before
the industry is out competed, due to lack of innovation.
Many industrial districts in the third Italy could be used
as illustrating cases. Most of them have only relied on
incremental innovations-based on tacit or in a minor-
ity of cases also synthetic knowledge, which normally
do not have the capacity of changing technological tra-
jectories. Against this background, regional policy has
clearly moved away from a re-active notion of redis-
tributing welfare to a pro-active position at the heart
of economic development. The development of the
endogenous capacity of regions to innovate in order to
create competitive advantage is today often referred to
as regional constructed advantage in which the estab-
lishment and formation of regional innovation systems
plays a strategic role (Cooke and Leydesdorff, in press).
There is no single optimal strategy in this respect.
We argue that the innovation process of rms and
industries is particularly dependent on their specic
knowledge base (Asheim and Gertler, 2005). Here we
will distinguish between two types of knowledge base:
analytical and synthetic (Laestadius, 1998), which
are summarized in Table 1. In its philosophical mean-
ing, analytical refers to the way of reasoning by which
the truth of a proposition is established independent
of fact or experience involving inference from gen-
eral principle. Synthetic, on the other hand, pertains to
knowledge having a truth value determined by obser-
vation or facts.
Here, an analytical knowledge base refers to indus-
trial settings, where scientic knowledge is highly
important, andwhere knowledge creationis oftenbased
on cognitive and rational processes, or on formal mod-
els. Examples are genetics, biotechnology and gen-
eral information technology. Both basic and applied
research as well as systematic development of prod-
ucts and processes, are relevant activities. Compa-
nies typically have their own R&D departments but
they rely also on the research results of universities
and other research organisations in their innovation
process. Universityindustry links and respective net-
works, thus, are important andmore frequent thaninthe
other type of knowledge base. Knowledge inputs and
outputs are in this type of knowledge base more often
codied than in the other type. This does not imply that
tacit knowledge is irrelevant, since there are always
both kinds of knowledge involved and needed in the
process of knowledge creation and innovation (Nonaka
et al., 2000; Johnson and Lundvall, 2001). The fact that
codication is more frequent is due to several reasons:
knowledge inputs are oftenbasedonreviews of existing
studies, knowledge generation is based on the applica-
tion of scientic principles and methods, knowledge
processes are more formally organised (e.g. in R&D
departments) and outcomes tend to be documented in
reports, electronic les or patent descriptions. Knowl-
edge application is in the form of new products or
processes, and there are more radical innovations than
in the other knowledge type. An important route of
knowledge application is new rms and spin-off com-
panies which are occasionally formed on the basis of
radically new inventions or products.
A synthetic knowledge base refers to industrial set-
tings, where the innovation takes place mainly through
the application of existing knowledge or through
new combinations of knowledge. Often this occurs in
response to the need to solve specic problems coming
up in the interaction with clients and suppliers. Indus-
try examples include plant engineering, specialised
advanced industrial machinery, and shipbuilding. Prod-
ucts are often one-off or produced in small series.
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1177
Table 1
Synthetic vs. analytic knowledge base
Synthetic Analytic
Innovation by application or novel combination of existing
knowledge
Innovation by creation of new knowledge
Importance of applied, problem related knowledge (engineering)
often through inductive processes
Importance of scientic knowledge often based on deductive
processes and formal models
Interactive learning with clients and suppliers Research collaboration between rms (R&D department) and
research organisations
Dominance of tacit knowledge due to more concrete know-how,
craft and practical skill
Dominance of codied knowledge due to documentation in
patents and publications
Mainly incremental innovation More radical innovation
Source: Asheim and Gertler (2005).
R&D is in general less important than in the rst type.
If so, it takes the form of applied research, but more
often it is in the form of product or process develop-
ment. Universityindustry links are relevant, but they
are clearly more in the eld of applied research and
development than in basic research. Knowledge is cre-
ated less in a deductive process or through abstraction,
but more ofteninaninductive process of testing, experi-
mentation, computer-based simulation or through prac-
tical work. Knowledge embodiedinthe respective tech-
nical solution or engineering work is at least partially
codied. However, tacit knowledge seems to be more
important than in the rst type, in particular due to
the fact that knowledge often results from experience
gained at the workplace, and through learning by doing,
using and interacting. Compared to the rst knowl-
edge type, there is more concrete know-how, craft and
practical skill required in the knowledge production
and circulation process. These are often provided by
professional and polytechnic schools, or by on-the-
job training. The innovation process is often oriented
towards the efciency and reliability of new solutions,
or the practical utility and user friendliness of prod-
ucts from the perspective of the customers. Overall,
this leads to a rather incremental way of innovation,
dominated by the modication of existing products and
processes. Since these types of innovation are less dis-
ruptive to existing routines and organisations, most of
themtake place in existing rms, whereas spin-offs are
relatively less frequent
1
Table 1.
1
Pavitt (1984, pp. 35365) offers a three-way taxonomy of indus-
tries based on the predominant nature and sources of technical
change. Supplier-dominated industries include agriculture and tra-
3. Regional innovation systems and localised
learning
The regional innovation system can be thought of
as the institutional infrastructure supporting innovation
within the production structure of a region. Thus, in
case the followingtwosubsystems of actors are system-
atically engaged in interactive learning (Cooke et al.,
1998) it can be argued that a regional innovation sys-
tem is in place. (1) The regional production structure
or knowledge exploitation subsystem which consists
mainly of rms, often displaying clustering tendencies.
(2) The regional supportive infrastructure or knowl-
edge generation subsystem which consists of public
and private research laboratories, universities and col-
leges, technology transfer agencies, vocational training
organisations, etc. Furthermore, Cooke et al. (1998)
emphasise the mainly informal institutional context
(i.e. norms, trust and routines) in which such interac-
tive learning takes place. This dynamic and complex
interaction constitutes what is commonly labelled sys-
tems of innovation (Edquist, 1997), i.e. systems under-
stood as interaction networks (Kaufmann and T odtling,
2001).
ditional manufacturing sectors such as textiles. Production-intensive
industries can be further subdivided into scale-intensive sectors such
as steel, consumer durables and automobiles, and specialized sup-
plier sectors such as machinery and instruments. Science-based
industries include electronics and chemicals (including pharma-
ceuticals). Our observations about industries with synthetic knowl-
edge bases correspond closely to those sectors encompassed by the
rst two of Pavitts categories (supplier-dominated and production-
intensive). Similarly, the analytical category corresponds directly to
Pavitts science-based industries.
1178 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
Taking each element of the term in turn (Asheim
and Cooke, 1999), the concept of region highlights an
important level of governance of economic processes
between the national level and the level above the local
or municipal level (Cooke and Leydesdorff, in press).
Regions are important bases of economic co-ordination
at the meso-level, although the level of regional admin-
istration can differ quite a lot across various countries.
In varying degrees, regional governance is expressed
in both private representative organisations, such as
branches of industry associations and chambers of
commerce, and public organisations, such as regional
agencies with powers devolved from the national (or,
within the European Union, supra-national) level to
promote enterprise and innovation support (Asheim
et al., 2003a; Cooke et al., 2000).
The systemic dimension of the RIS derives in part
from the associational character of innovation net-
works (Cooke and Morgan, 1998). Such relationships,
to be systemic, must involve some degree of inter-
dependence, though to varying degrees. Likewise, not
all such systemic relations need to be regionally con-
tained, but many are.
2
As the interactive mode of inno-
vation grows in importance, these relations are more
likely to become regionally contained, for example,
in the case of specialised suppliers with a specic
technology or knowledge base. Such suppliers often
depend on tacit knowledge, face-to-face interaction
and trust-based relations and, thus, benet from co-
operation with customers in regional clusters, while
capacity subcontractors are increasingly sourced glob-
ally. Further reinforcing the systemic character of the
RISis the prevalence of a set of attitudes, values, norms,
routines and expectationsdescribed by some as a
distinctive regional culturethat inuences the prac-
tices of rms in the region. It is this common regional
cultureitself the product of commonly experienced
institutional forcesthat shapes the way that rms
interact with one another in the regional economy.
Thus, we strongly disagree with Bathelt (2003), who
argues that it seems questionable that region-specic
innovation and production processes are typically asso-
ciated with the existence of regional innovation sys-
tems. To assume that such small-scale systems exist
2
In a recent study, Carlsson (2004) shows that the majority of
theoretical as well as empirical analyses of innovation systems have
a regional focus.
bears the risk of underestimating the importance of
those institutions which are negotiated and dened at
the level of the nation state. In reality, however, regional
and national innovation contexts are fundamentally dif-
ferent. Regional production congurations are often
dependent on structures and developments which are
shaped and take place outside the region (Bathelt,
2003, p. 797).
3
The key to the disagreement lies in the application
by Bathelt of social systems theory, which replaces
the element/relation dichotomy of the innovation sys-
tems approach with a system/environment dichotomy
(Kaufmann and T odtling, 2001). This leads Bathelt to
believe that one of the core problems of the regional
innovation system is that it portrays the region as
an entity which hosts a large part of an economic
value chain and has a governance structure of its own,
independent from its environment (Bathelt, 2003, p.
796). Aside from the formal systems theoretical argu-
ments, there is no substantial theory to corroborate this
statement. Empirically it may be shown that regions
can in fact contain large parts of a value chain (e.g.
Italian industrial districts) as well as having a rela-
tive autonomous government structure (e.g. regions in
federal countries, such as BadenW urttemberg in Ger-
many and Catalonia and the Basque country in Spain).
Furthermore, in a globalising economy characterised
by vertical disintegration and distributed knowledge
bases, the important perspective ought to be the inter-
dependences between regions and nations, where the
deciding criteria must be the location of core activi-
ties (and not the whole value chain as such) and the
relative importance of their connections to regional
knowledge infrastructures. With the possible exception
of the US, the argument that production congura-
tions are often dependent on structures and develop-
ments which are shaped and take place outside of
the actual territory could as easily apply to most small
and medium-sized countries as to regions, especially
if being members of supra-national organisations such
3
In a more recent article Bathelt (2005) once more underlines his
position by arguing that at this geographical level (i.e. a region, my
addition), it is unlikely that a self-referential system can develop
because regions are strongly dependent on national institutions
and other external inuences and lack important political decision-
making competencies . . .. Regional production congurations hardly
have the potential to gain and retain structural independence and
reproduce their basic structure.
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1179
as the EU, in which region formation has and con-
tinues to be evolved apace (Cooke, in press). Cooke
(in press) refers to Bathelts position as Listian writ-
ing that only nations (presumably meaning states since
there are many nations without full economic gover-
nance powers) have specicity and that they may also
be closed systems.
Yet it is essential to recognise the, by inference,
inter-locked character of a region in a wider geograph-
ical context (Howells, 1999). But even if research has
revealed that the regional level is neither always nor
even normally sufcient for rms to stay innovative
and competitive (Isaksen, 1999), and that the learn-
ing process becomes increasingly inserted into var-
ious forms of networks and innovation systems (at
regional, national and inter-national levels), the con-
tinuous importance of the regional level is conrmed
by results from a European comparative cluster sur-
vey (Isaksen, 2005). This study shows that regional
resources and collaboration are of major importance
in stimulating economic activity in the clusters. How-
ever, the survey found an increased presence of MNCs
in many clusters, and also that rms in the clus-
ters increasingly source major components and per-
form assembly manufacturing outside of the clusters
(Isaksen, 2005). Also T odtling and Trippl (2005) found
empirical support for clustering because of the impor-
tance of social interaction, trust and local institutions.
Yet they also note that both local and distant networks
are often needed for successful co-operative projects,
in particular for projects of innovation and product
development when it is usually necessary to combine
both local and non-local skills and competences in
order to go beyond the limits of the region (see also
Asheim and Herstad, 2003; Bathelt et al., 2004; Cooke
et al., 2000).
4. Varieties of regional innovation systems
The innovation system concept can be understood
in both a narrowas well as a broad sense. Anarrowdef-
inition of the innovation system primarily incorporates
the R&D functions of universities, public and private
research institutes and corporations, reecting a top-
down model of innovation as exemplied by the triple
helix approach (Etzkowitz and Leydesdorff, 2000). A
broader conception of the innovation systems includes
all parts and aspects of the economic structure and the
institutional set-up affecting learning as well as search-
ing and exploring (Lundvall, 1992, p. 12). This broad
denition incorporates the elements of a bottom-up,
interactive innovation model which alternatively could
be called learning regions (Asheim, 2001).
In order to reect the conceptual variety and empir-
ical richness of the relationships linking the produc-
tion structure to the institutional set-up in a region,
Asheim (1998) distinguishes between three types of
RISs (see also Cooke, 1998; Asheim and Isaksen,
2002). The rst type may be denoted as territorially
embedded regional innovation systems, where rms
(primarily those employing synthetic knowledge) base
their innovation activity mainly on localised, inter-
rm learning processes stimulated by the conjunc-
tion of geographical and relational proximity with-
out much direct interaction with knowledge generat-
ing organisations (i.e. R&D institutes and universi-
ties). This type is similar to what Cooke (1998) calls
grassroots RIS, and implies the broader denition
of innovation systems described by Lundvall (1992)
above.
The best examples of territorially embedded
regional innovation systems are networks of SMEs in
industrial districts. Thus, in Italys Emilia-Romagna,
for example, the innovation system can be described
as territorially embedded within that particular region
(Granovetter, 1985). These territorially embedded
systems provide bottom-up, network-based support
through, for example, technology centres, innovation
networks, or centres for real service providing mar-
ket research and intelligence services, to promote
the adaptive technological and organisational learn-
ing in territorial context (Storper and Scott, 1995,
p. 513).
Another type of RIS is the regionally networked
innovation system. The rms and organisations are
also embedded in a specic region and characterised
by localised, interactive learning. However, through
the intentional strengthening of the regions institu-
tional infrastructure, for example, through a stronger,
more developed role for regionally based R&D insti-
tutes, vocational training organisations and other local
organisations involved in rms innovation processes
these systems have a more planned character involv-
ing publicprivate co-operation. The networked sys-
tem is commonly regarded as the ideal-type of RIS: a
1180 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
regional cluster of rms surrounded by a regional sup-
porting institutional infrastructure. Cooke (1998) also
calls this type network RIS. The network approach
is most typical of Germany, Austria and the Nordic
countries.
The regionally networked innovation system is a
result of policy intervention to increase innovation
capacity and collaboration. SMEs, for example, may
have to supplement their informal knowledge (charac-
terised by a high tacit component) with competence
arising from more systematic research and develop-
ment in order to carry out more radical innovations.
In the long run, most rms cannot rely exclusively on
informal localised learning, but must also gain access to
wider pools of both analytical and synthetic knowledge
on a national and global basis. The creation of region-
ally networked innovation systems through increased
co-operation with local universities and R&D insti-
tutes, or through the establishment of technology trans-
fer agencies, may provide access to knowledge and
competence that supplements rms locally derived
competence. This not only increases their collective
innovative capacity, but may also serve to counteract
technological lock-in (the inability to deviate from
an established but outmoded technological trajectory)
within regional clusters of rms.
The third main type of RIS, the regionalised
national innovation system, differs from the two pre-
ceding types in several ways. First, parts of industry
and the institutional infrastructure are more function-
ally integrated into national or international innovation
systems, i.e. innovation activity takes place primarily
in co-operation with actors outside the region. Thus,
this represents a development model in which exoge-
nous actors and relationships play a larger role. Cooke
(1998) describes this type as dirigiste RIS, reecting
a narrower denition of an innovation system incor-
porating mainly the R&D functions of universities,
research institutes and corporations. Second, the col-
laboration between organisations within this type of
RIS conforms more closely to the linear model, as
the co-operation primarily involves specic projects
to develop more radical innovations-based on formal
analytical-scientic knowledge. Within such systems,
co-operation is most likely to arise between people with
the same occupational or educational background (e.g.
among scientists). This functional similarity facilitates
the circulation and sharing of knowledge through epis-
temic communities, whose membership may cross
inter-regional and even international boundaries (Amin
and Cohendet, 2003; Coenen et al., 2004).
One special example of a regionalised national
innovation systemis the clustering of R&Dlaboratories
of large rms and/or governmental research institutes
in planned science parks and technopoles, normally
located in close proximity to universities and technical
colleges, but, according to evidence, typically having
limited linkages to local industry (Asheim, 1995). Sci-
ence parks are, thus, anexample of a plannedinnovative
milieu comprised of rms with a high level of internal
resources and competence, situated within weak local
co-operative environments. These parks have generally
failed to develop innovative networks based on inter-
rm co-operation and interactive learning within the
science parks themselves (Asheim and Cooke, 1998;
Henry et al., 1995). Technopoles, as developed in
countries, such as France, Japan and Taiwan, are also
characterised by a limited degree of innovative inter-
action between rms within the pole, and by vertical
subcontracting relationships with non-local external
rms. In those rare cases where local innovative net-
works arise, they have normally been orchestrated by
deliberate public sector intervention at the national
level. These characteristics imply a lack of local and
regional embeddedness, and lead us to question the
capability of science parks and technopoles to pro-
mote innovativeness and competitiveness more widely
within local industries (especially SMEs) as a prereq-
uisite for endogenous regional development (Asheim
and Cooke, 1998; Longhi and Qu ere, 1993).
To summarize, the main argument that this paper
puts forward is that there are different logics behind
constructing regional innovation systems contingent on
the knowledge base of the industry it addresses as well
as on the regional knowledge infrastructure which is
accessible. In a territorially embedded regional inno-
vation system, the emphasis lies on the localised, path-
dependent inter-rmlearningprocesses ofteninvolving
innovation-based on synthetic knowledge. The role
of the regional knowledge infrastructure is therefore
mainly directed to industry-specic, hands-on services
and concrete, short-term problem solving, i.e. ex-post
support to the cluster. In a regionalised national innova-
tion system, R&D and scientic research take a much
more prominent position. Innovation builds primar-
ily on analytic knowledge. Linkages between existing
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1181
local industry and the knowledge infrastructure are,
however, weakly developed. Instead it holds the poten-
tial to promote newindustries at the start of their indus-
trial and technological life cycle. In this, the role of
the regional(ized) knowledge infrastructure is a very
central one as it provides the cornerstone for clus-
ter development (through the precarious task of com-
mercialising science) and can thus be called ex-ante
cluster support. The regionally networked innovation
system can be considered as an ideal type RIS. Simi-
lar to the regionalised national innovation system, the
knowledge infrastructure plays an indispensible role.
But in contrast to it, the cluster is not science-driven
but market-driven. In comparison to the territorially
embedded regional innovation system, the networked
RIS often involves more advanced technologies com-
bining analytic and synthetic knowledge. While ter-
ritorially embedded RIS are often found in mature
industries and regionalised national innovation systems
found in emergent industries, networked regional inno-
vation systems are typically found in the growth phase
of an industry. Firms and knowledge infrastructure
form a dynamic ensemble, combining ex-post support
for incremental problemsolving and ex-ante support to
counter technological and cognitive lock-ins. We shall
now draw our attention to ve case studies to illus-
trate these different logics in the formation of regional
innovation systems.
5. Comparison of Nordic clusters
Belowthe difference betweenanex-ante andex-post
approach is illustrated on the basis of a comparative
analysis of ve clusters in three Nordic countries: Den-
mark, Sweden and Norway
4
:
The furniture cluster of Sallingandthe wireless com-
munication cluster of North Jutland in Denmark.
The functional food cluster of Scania in Sweden.
The Rogaland food cluster and the Horten electron-
ics cluster in Norway.
On the issue of method Cooke (1998, p. 12) argues
that one of the distinct advantages of the RIS approach
4
These studies as suchhave beencarriedout byvarious researchers
through a joint research project called Nordic SMEs and Regional
Innovation Systems.
is that it allows for a systematic comparison of inno-
vation activities across various regions. Conducting
such comparable studies can lead to identication of
some functional equivalents for specic as well as
generic problems within the innovation process. How-
ever, various other researchers remaincritical andargue
that the rise of the Silicon Valley fever (Benneworth
and Hardy, 2003) has conned much work to text-book
cases in high-tech sectors (Doloreux, 2002). It is argued
that more attention should be paid to applying the
approach on other regions than the stereotypical happy
few and, more importantly, theory must be informed
by the lessons drawn from such ordinary regions. This
critique resembles the arguably more narrow and rel-
atively static understanding of knowledge intensive
industries in the knowledge-based economy as opposed
to the broader scope applied in the learning economy
perspective. Providing an example of the latter per-
spective, Kaufmann and T odtling (2000) comparative
study of old traditional industrial regions shows that
the concept allows for utilisation in ordinary regions
and mature industries as well. The purpose of the case
studies in this paper is not to provide detailed and com-
prehensive accounts of the innovation systems in the
ve regions. For this we refer to the original studies
(Asheim et al., 2003b). Instead we seek to substantiate
the arguments we made previously about the impor-
tance of the industrial knowledge base for the linkages
and support given in the context of the regional innova-
tion system by means of a comparative case analysis.
As such we choose to compare a set of clusters
5
that are
typically based on synthetic knowledge (Salling and
Rogaland) with clusters that have a stronger analytic
knowledge base component (North Jutland, Scania
and Horten). First a short introduction to the individ-
ual cases will be given followed by a more focused
discussion on the knowledge base of the respective
industries.
5.1. Introduction to the cases
In his study on the regional furniture industry in
and around the town of Salling in the North-West
5
Our selection of the cases is of course also predicated by their
inclusion in our joint research project Nordic SMEs and Regional
Innovation Systems which applied a distinct case study oriented
approach to the different clusters that were studied.
1182 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
of Jutland, Lorenzen (2003) reports on an extraordi-
nary economic growth over the past decades despite
high factors costs. Between 1972 and 1992 employ-
ment in the cluster tripled while overall employment
in Denmark decreased. The number of rms grew with
approximately 80%. In 1996, 2388 were employed in
54 rms, the majority of which are small and medium
sized. This performance is ascribed to the ability of
the cluster to collectively penetrate newmarkets, brand
products and develop newdesigns. This low-tech inno-
vativeness, which would run the risk of being ignored
in a knowledge-based economy perspective, is in turn
underpinned by a combination of stable and at the same
time exible inter-rm relationships held together by
a high level of trust and shared norms and conventions
embedded in the local community of rm managers
and workers. The other Danish cluster located in North
Jutland is specialised in wireless communication and
located around the city of Aalborg. At present it con-
sists of roughly 35 rms employing around 3220 peo-
ple. In terms of rmsize the cluster is composed of both
SMEs as well as establishments of major multination-
als. In 1997, the companies, Aalborg University and
the science park NOVI established the cluster associa-
tion NorCOM to formalize their co-operation (Dalum
et al., 2002).
It would go too far to consider the case of func-
tional foods in Scania an actual cluster as it consists
of only a handful of dedicated functional food rms
in and around the cities of Lund and Malm o. Yet it
could be suggested that a knowledge intensive embry-
onic cluster is taking shape under the wings of the
traditional food industry (Holmberg, 2003). Scania, in
the South of Sweden, has been a historically impor-
tant national centre for agricultural production (Nilsson
et al., 2002) and hosts some of the countrys largest
food processing industries. Functional food is dened
as articially developed food with added ingredients
that demonstrate scientic evidence of positive health-
related effects. It is regarded as a novel technology with
high growth and innovation potentials in a traditionally
not very innovative industry. As such it provides a case
in point of the notion of distributed knowledge bases
by which the distinction between high-tech and low-
techbecomes increasinglyblurred. Several small, R&D
intensive companies dedicated to functional food have
emerged around the University of Lund. These compa-
nies collaborate intensively with the traditional large
food companies for the actual production and market-
ing of functional foods as well as with regional research
groups and organisations for scientic research.
Rogaland, located in the South-West of Norway, is
one of the nations leading production areas in meat,
dairy foodstuffs, feed, shellsh and certain vegeta-
bles. In 1995, the region counted 161 rms manu-
facturing food and beverages (Onsager, 1999). These
employed approximately 4000 people. Several large
international companies are located here yet most rms
are small or medium sized. Onsager and Aasen (2003)
distinguish three partly differentiated, partly integrated
subsystems: agrofood production, seafood production
and life-stock production. Even though each subsys-
tem exploits separate raw materials, production tech-
nologies and end markets, functional connections and
inter-relations across the subsystemare in place in con-
nection with subcontracting, common customers and
support organisations (R&D, training and professional
forums). Due to this agglomerated sector environment
a high degree of local collaboration underpins the
innovative efforts of the rms (Onsager, 1999). The
small electronics cluster in Horten in the South-East
of Norway, hosts around 25 rms and 1900 employees
(Isaksen, 2003; Asheimand Isaksen, 2002). The cluster
contains a few large enterprises but is otherwise dom-
inated by SMEs. The motive powers in the local elec-
tronics industry are the large system houses and orig-
inal equipment manufacturers (OEM)-suppliers which
originate mainly from the Norwegian national system
of innovation. At present, these rms still collaborate
mainly with national and international research organ-
isations, universities and customers in their innovative
activities. The cluster is held together by personal net-
works of individuals within different rms and impor-
tant knowledge transfer takes place through labour
mobility. In terms of industrial relationships the system
houses and OEM-suppliers have pronounced regional
linkages only with the local subcontractors in Horten.
As specialised producers of components and software
they play a signicant role in innovation processes in
connection with transferring prototypes into effective
industrial production as well as for joint problem solv-
ing.
In the following section, we analyse the industrial
knowledge base of the respective clusters and their spe-
cic link with the region. Table 2 summarizes the result
of our analysis.
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1183
T
a
b
l
e
2
O
v
e
r
v
i
e
w
o
f
N
o
r
d
i
c
c
l
u
s
t
e
r
s
N
u
m
b
e
r
o
f

r
m
s
E
m
p
l
o
y
e
d
K
n
o
w
l
e
d
g
e
b
a
s
e
R
e
g
i
o
n
a
l
k
n
o
w
l
e
d
g
e
i
n
f
r
a
s
t
r
u
c
t
u
r
e
S
p
e
c
i

c
r
e
g
i
o
n
a
l
a
s
s
e
t
s
F
u
r
n
i
t
u
r
e
,
S
a
l
l
i
n
g
5
4
2
4
0
0
S
y
n
t
h
e
t
i
c
T
e
c
h
n
i
c
a
l
s
c
h
o
o
l
a
n
d
c
a
b
i
n
e
t
m
a
k
e
r
s
g
u
i
l
d
I
n
t
e
r
-

r
m
r
e
l
a
t
i
o
n
s
b
a
s
e
d
o
n
p
e
r
s
o
n
a
l
n
e
t
w
o
r
k
s
a
n
d
l
o
c
a
l
c
u
l
t
u
r
e
W
i
r
e
l
e
s
s
C
o
m
m
u
n
i
c
a
t
i
o
n
,
N
o
r
t
h
J
u
t
l
a
n
d
3
5
3
2
2
0
A
n
a
l
y
t
i
c
A
a
l
b
o
r
g
U
n
i
v
e
r
s
i
t
y
a
n
d
N
O
V
I
s
c
i
e
n
c
e
p
a
r
k
C
l
o
s
e
l
i
n
k
b
e
t
w
e
e
n

r
m
s
a
n
d
u
n
i
v
e
r
s
i
t
y
,
h
i
g
h
l
y
s
k
i
l
l
e
d
l
o
c
a
l
l
a
b
o
u
r
f
o
r
c
e
F
u
n
c
t
i
o
n
a
l
f
o
o
d
,
S
c
a
n
i
a
3
1
3
0
A
n
a
l
y
t
i
c
L
u
n
d
U
n
i
v
e
r
s
i
t
y
S
p
e
c
i
a
l
i
s
e
d
c
o
m
p
e
t
e
n
c
e
a
t
L
u
n
d
U
n
i
v
e
r
s
i
t
y
(
F
u
n
c
t
i
o
n
a
l
F
o
o
d
s
S
c
i
e
n
c
e
C
e
n
t
r
e
)
F
o
o
d
,
R
o
g
a
l
a
n
d
1
6
1
4
0
0
0
S
y
n
t
h
e
t
i
c
(
R
e
g
i
o
n
a
l
i
s
e
d
)
a
p
p
l
i
e
d
R
&
D
o
r
g
a
n
i
s
a
t
i
o
n
s
a
n
d
r
e
g
i
o
n
a
l
d
e
v
e
l
o
p
m
e
n
t
o
r
g
a
n
i
s
a
t
i
o
n
s
R
a
w
m
a
t
e
r
i
a
l
a
n
d
r
e
g
i
o
n
a
l
n
e
t
w
o
r
k
E
l
e
c
t
r
o
n
i
c
s
,
H
o
r
t
e
n
2
5
1
9
0
0
A
n
a
l
y
t
i
c
a
n
d
s
y
n
t
h
e
t
i
c
O
n
l
y
r
e
c
e
n
t
l
y
,
r
e
g
i
o
n
a
l
u
n
i
v
e
r
s
i
t
y
c
o
l
l
e
g
e
P
e
r
s
o
n
a
l
n
e
t
w
o
r
k
a
n
d
s
u
b
c
o
n
t
r
a
c
t
o
r
s
S
o
u
r
c
e
s
:
D
a
l
u
m
e
t
a
l
.
(
2
0
0
2
)
,
L
o
r
e
n
z
e
n
(
2
0
0
3
)
,
H
o
l
m
b
e
r
g
(
2
0
0
3
)
,
O
n
s
a
g
e
r
a
n
d
A
a
s
e
n
(
2
0
0
3
)
a
n
d
I
s
a
k
s
e
n
(
2
0
0
3
)
.
5.2. Furniture cluster in Salling
The study by Lorenzen (2003) shows that in terms
of product innovations the furniture producing rms
in Salling mainly design varieties in for example style,
materials and colors based on the existing product-line.
Completely new products types are typically intro-
duced only once a year. Process innovations necessar-
ily follow such new product designs. The shift from
hardwood to other materials, notably plywood, is con-
sidered as the most dramatic shift that the cluster
witnessed. Experimentation at the factory oor and
product revision based upon employees ideas are key
mechanisms for the rms to innovate internally, illus-
trating very well the tacit and synthetic nature of the
knowledge base in the furniture industry. The primary
source of innovation are, however, interactive innova-
tion activities which take place through vertical net-
works between producers and their suppliers (in collab-
oration with existing suppliers or by reshufing inputs
from other suppliers) and through horizontal networks
(e.g. matching product designs in order to offer fuller
product lines). Lorenzen (2003) highlights the impor-
tance of shared values and common norms among
managers within the rms to sustain these relation-
ships. Characteristic local conventions are craftsman-
ship, entrepreneurship, sense of belonging to an eco-
nomic community and local solidarity. It appears in fact
that in comparison with the localised cluster dynamics
the Salling rms have hardly any systematic learning
relationship with players outside the cluster. In addition
to the furniture rms themselves, two regional organi-
sations play an important role in sustaining the patterns
of localised inter-rm learning. Firstly, the workforce
often derives its initial training and necessary arti-
san skills from their education at the local technical
school. The school is considered as Denmarks most
specialised educational institute for furniture produc-
tion and collaborates closely with the furniture rms
to align its curriculum with their needs. Secondly, the
local cabinetmakers guild provides a crucial venue to
regularly exchange information and co-ordinate inter-
rm relationships and to (re-)produce the conventions
embedding the rms to the Salling cluster. As such, the
rms in the cluster have become rooted in the region
through their mutual inter-dependence. Over the past
20 years each of the rms has developed its own ded-
icated niche through specialisation in specic parts
1184 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
of the value chain in combination with an extensive
local network of stable yet exible inter-rm relation-
ships allowing for economies of scope. The centrality
of localised learning based on synthetic knowledge in
this regional innovation system, the embeddedness of
the rms in a local economic community and the lim-
ited role of knowledge generating organisations make
a clear case for a territorially embedded RIS. Inno-
vation policy comes predominately in the form of
local hands-on support for the existing structure of
the cluster and, thus, illustrates an ex-post approach
to constructing RIS (e.g. the local economic devel-
opment ofces help in attracting more trainees as
cabinetmakers).
5.3. Food cluster in Rogaland
For its competitiveness, the Rogaland cluster also
draws on a synthetic knowledge base mainly apply-
ing industry-specic technical knowledge. Typically
innovation takes place through constant improvements
of pre-existing product standards, packaging/design,
labelling, etc. (Onsager and Aasen, 2003). But differ-
ently from the Salling cluster, the region hosts impor-
tant food-related R&Dbodies that are actively involved
in innovation activities with the companies. An exam-
ple is Norconserv (the Norwegian Institute for Fish
Processing and Preservation Technology) which serves
as an important centre of expertise in adjustment and
development processes of production structures. Illus-
trating its systemic relationship with the local industry,
the Norconserv institute is renowned for its hands-on
research actively tailored to the sh-producing indus-
try. Also the Norwegian School of Hotel Management
at the recently established University in Stavanger,
which is the only university-based education within
this area in the Nordic countries, has R&Dprograms in
place on economic and organisational issues relevant
to the industry. Many of the R&D bodies in Rogaland
are regionalised divisions of wider national organisa-
tions (e.g. the Norwegian Crop Research Institute) that
have been located in Rogaland because of the particular
agricultural and natural conditions that prevail in the
area. National R&D organisations outside Rogaland
(e.g. the Norwegian Institute for Fishery and Aquacul-
ture Research in Troms) are also of major importance
for the rms. As we will also see in the Horten case, the
traditional dominance of the national systemof innova-
tion is characteristic for the Norwegian situation where
public R&D agencies (e.g. the Norwegian Research
Council) have a long tradition of designing R&D pro-
grammes at the national level. Nonetheless, recent
policy measures are moving towards stronger regional-
isation tendencies providing more targeted innovation
support that is more closely aligned to the needs of
the industry in place. The regional development pro-
gram Arena for Regional Commercial Development
and Entrepreneurship (ARNE) in which the private
sector, municipal and regional authorities co-operate
to support the food industry (through projects on, e.g.
ecological food and micro algae) can be seen as an
example of this. An important network organisation
for the Rogaland cluster is the Fagforum for Mat
og Drikke (Professional Forum for Food and Drink)
whose primary mission is to promote knowledge shar-
ing and competence dissemination among local rms,
education and R&D organisations. Current efforts are
directed to open The House of Food as a regional
centre of expertise on gastronomy and food technol-
ogy. Based on broad participation of the players in
the cluster it aims to become a one stop shop for
companies wishing to initiate product development
programs. As we can see, this cluster is typically depen-
dent on both the regional and national level for its
innovation support. Because of natural conditions as
well as the historical agglomeration of agriculture and
food industry in the area, relevant R&D and educa-
tional organisations have been directed to the Roga-
land region. However, this had more of a top-down
character in which R&D programs were designed at
the national level. Current innovation support is more
directly and explicitly aligned to regional needs and can
be regarded as ex-post innovation support to the cluster.
In comparison to Salling, this cluster is more depen-
dent on the knowledge infrastructure (albeit heavily
applied) provided by universities and research insti-
tutes (on national and regional level). Given current
efforts to increasingly regionalise innovation support
(endorsed by the national government) it can thus be
regarded as a networked regional innovation system.
Similar to the Scania case, this case illustrates very well
howsystemic innovation interaction between rms and
the knowledge infrastructure is based on a distributed
knowledge base. Typical for Rogaland is, however, the
wide-spectre of innovation support for the food sec-
tor, most of the time involving synthetic knowledge
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1185
though sometimes innovation based on analytic knowl-
edge can be observed (e.g. research on functional
foods).
5.4. Wireless communication cluster in Northern
Jutland
The above cases of low-tech innovations can be
contrasted with the wireless communication cluster of
Northern Jutland and the embryonic functional food
cluster in Scania. These cases are exemplars of ana-
lytic knowledge base clusters for which the regional
knowledge infrastructure has played a crucial and fun-
damental role right fromthe start of the cluster. The his-
torical overviewof the wireless communication cluster
in North Jutland by Dalum et al. (1999, 2002) shows
clearly how the presence of Aalborg university and the
NOVI science park have been essential fundaments for
the clusters establishment and growth. A real break-
through came about through the universitys establish-
ment in 1974. The regional supply of highly skilled
engineers equipped with much needed competences in
the eld of wireless communications has been a pre-
requisite for the growth of the local industry. In combi-
nation with the universitys research orientation in this
eld (basic research with a sufciently application-
oriented touch (p. 16)) this constitutes a core asset of
the region attracting major multinational companies.
Also the success of the NOVI science park demon-
strates the systemic interaction between rms and uni-
versity in and around Aalborg. The regional innovation
system presented in this case is also of the regionally
networked type. It distinguishes itself from a region-
alised national innovation system by a characteristic
regional embedding through localised inter-rm learn-
ing as well as systemic regional industryuniversity
interaction which the latter typically is lacking. In
this context, Aalborg university should be seen as a
real cornerstone for the cluster. Without it, the cluster
would not have developed its distinctive competence
and technological excellence in wireless communi-
cation on which the competitiveness of the cluster
builds. As such it illustrates the ex-ante approach to
constructing regional innovation systems as the logic
behind innovation support has been more inclined to
establishing new economic activities based on a break-
through technology rather than to support an existing
industrial specialisation. Moreover, it could be argued
that the role of a skilled workforce (consisting of
researchers and engineers) plays a more important role
as a link between the knowledge infrastructure and
the rms than in the two preceding clusters based on
a synthetic knowledge base. Even though the knowl-
edge base of the cluster is mainly analytic given the
prominence of state-of-the-art scientic achievement,
it should be noted that it has proceeded further along
its technological life-cycle compared to functional
foods.
5.5. Functional food cluster in Scamia
A fairly similar account can be given for the func-
tional foods cluster in Scania yet on a more mod-
erate scale in terms of companies (Holmberg, 2003).
In line with ndings for the Swedish biotechnology-
pharmaceutical sector (McKelvey et al., 2003) inter-
rm knowledge linkages between the small, dedicated
research rms are relatively weak despite being colo-
cated. Instead, the rms are involved in collabora-
tive research with companies and research organisa-
tions on all geographical levels. Notwithstanding this,
Lund University has been of crucial importance for
the establishment of the functional food companies
as it provided the seedbed for the original scientic
ideas underpinning these rms. But also on an on-
going basis the presence of world-class research and
education facilities in the eld of functional foods
serves as a conduit to cutting-edge science for the func-
tional food rms. This has been further reinforced by
the recent establishment of the cross-faculty research
centre Functional Foods Science Centre which is one
of the agship projects supported by the Vinnv axt
program of the Swedish public agency for innova-
tion systems, VINNOVA, to promote the formation
of regional innovation systems. In this, similar to the
wireless communication case, human capital is high-
lighted as an asset which needs to be capitalized. Also
here we see an illustration of the ex-ante approach
which takes it vantage point in the knowledge infras-
tructure because of the relative novelty of functional
foods. Nonetheless, the future of functional foods will
depend strongly on whether the traditional food sec-
tor in Scania and consumers will endorse it. There-
fore, it would be most appropriate to characterise this
case as a networked regional innovation system under
construction.
1186 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
5.6. Electronics cluster in Horten
Right from the beginning in the 1960s, the pioneer-
ing rms of the Horten electronics cluster have had
knowledge linkages mainly connecting into the Norwe-
gian national system of innovation. These rms were
in fact spin-offs from important national knowledge
organisations (e.g. the Norwegian Technical University
in Trondheim) and were established based on product
ideas that originated there. Also later on, the (radi-
cal) product development mainly drewon co-operation
with technological R&Dinstitutes and large public and
private clients in Norway and abroad, and in projects
oftenpartlynancedbythe NorwegianResearchCoun-
cil (Isaksen, 2003). Similar to the Rogaland case, this
should be conceived through the Norwegian tradition
for nation based R&D programs. In contrast to Roga-
land, the regional knowledge infrastructure was of little
value for the electronics cluster. For the technologi-
cally advanced system houses and OEM-suppliers this
is nowadays still the case. According to Isaksen (2003)
these companies have even grown out of the national
innovation system that they rose from and are increas-
ingly collaborating on an international level with rms
and R&D institutes. What ties these rms then to
Horten? This stickiness should be understood through
the build up of unique competences among key per-
sonnel attached to the locality (Asheim and Isaksen,
2002). Furthermore, the role of local subcontractors
appears to be important. These have started their busi-
ness since the beginning of the 1980s after the system
rms closed down their in-house production facilities.
While the knowledge-base of the system houses and
OEM suppliers tends to be more inclined to an analyti-
cal knowledge base, innovation activities of these local
subcontractors typically build on a synthetic knowl-
edge base. Even if the growth of a regional university
college has modied this picture somewhat in the last
years, the RIS in this region must still be characterised
as a regionalised national innovation system. Thus, at
the start of the cluster, regional innovation support was
more or less lacking. Ex-ante support, for example,
through collaborative research projects nanced by the
Norwegian Research Council, was embedded outside
the region. At present the level of technological sophis-
tication and specialisation is high which makes it hard
to construct or adjust a regional innovation system t-
ting the needs of the systemhouses and OEMsuppliers.
Instead current policy efforts are mainly targeted at
local subcontractors in order to keep this cluster tied to
the region through local network building. An example
of this ex-post approach is the exchange of production
workers between local rms and training of managers
supported through the RIS program REGINN of the
Norwegian Research Council.
6. Conclusions
In this paper, we have made the argument that in a
learning economy clusters and RIS need to be treated
as two different yet strongly inter-related concepts. The
cluster concept is substantially narrower than the RIS
concept because of the strong sectoral connotation in
clusters whereas a regional innovation systemcan tran-
scend multiple sectors. Also from a policy perspective
it is important to keep this distinction in mind due to the
difference in sector specicity versus genericness. Fur-
thermore we analysed the relationship between indus-
tries and RIS from an industrial knowledge base per-
spective on the basis of a comparison of Nordic clusters
where we differentiated between an ex-ante and ex-
post approach to constructing regional innovation sys-
tems. For industries drawing on a analytic knowledge
base, a sophisticated and aligned knowledge infras-
tructure is indispensable to embed the industry in the
region. Key mechanisms in this are the local supply of
high-skilled labour and access to scientic excellence.
Therefore, RIS policy should address the regional pro-
vision of such a knowledge infrastructure as well as
systemic regional universityindustry interaction. If
such a knowledge infrastructure is in place at the
start of the clusters life-cycle (i.e. ex-ante approach)
it facilitates co-ordination within the system thereby
increasing the potential for a networked regional inno-
vation system rather than a regionalised national one.
For industries drawing on a synthetic knowledge base,
inter-rm localised learning plays an important role.
It is essential that innovation policy recognises this. It
also implies that support from a regional knowledge
infrastructure needs to be aligned to the already exist-
ing industrial specialisation (i.e. ex-post approach). In
other words, it should be demand-led. The presence of
such a tuned-in, hands-on support infrastructure makes
the difference between a territorially embedded and a
networked regional innovation system.
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1187
This brings us to the nal point in this paper which
is that of the regionalisation of innovation policy and
the rationale from a policy point of view to apply the
RIS concept. A set of variations on the systems of
innovation approach have been developed over time,
either taking territories as their point of departure
(national and regional) or specic sectors or technolo-
gies (Fagerberg et al., 2004). The regional innovation
systemconcept was inspired by the national innovation
system concept, and it is based on a similar rationale
that emphasizes territorially based innovation systems.
It can, however, be argued that the national innovation
systems perspective is still faced with difculties to
draw the boundaries of the system and to dene its ele-
ments (Edquist, 1997). With regard to these aspects,
Lundvall (1999) admits that, for example, studies on
national business systems (Whitley, 1999) have been
more explicit and consistent in their selection of the
systems organisations and institutions. Therefore, it is
important to emphasize the connotation of innovation
networks (Kaufmann and T odtling, 2001) rather than
systems in a Luhmanian social systemapproach as they
tend to over-abstract the substantial and material con-
tent of innovative interaction (Miettinen, 2002).
Characteristic for a systems approach to innova-
tion is, thus, the acknowledgement that innovations are
carried out through a network of various actors under-
pinned by an institutional framework. On the national
level, this system is in most cases very large, complex
and involving a plethora of linkages. Because of the
high level of aggregation it remains difcult to pro-
vide a level of empirical groundedness instrumental
for innovation policy development. Miettinen (2002)
argues that principles of historicity, industry specicity
and region specicity need to be taken into consider-
ation to arrive at less abstract and more disaggregated
reduced-forminnovation systems. Also Lundvall and
Borras (1997) hint towards this when they argue that
the region is increasingly the level at which innova-
tion is produced through regional networks of innova-
tors, local clusters and the cross-fertilising effects of
research institutions (p. 39).
The case studies presented in the previous section
illustrate how studies of regional innovation systems
provide concrete, policy-relevant analysis making vis-
ible the interplay between various actors, the contri-
butions of different institutions and the relevance of
regional and national policy measures. Regional inno-
vation policy should, however, not be formulated based
on off-the-shelf, best-practice solutions drawn from
the experience of successful regions or some expert
manual (Amin, 1999, p. 371). But through the region-
alisation of innovation policy more accurate consider-
ation can be paid to the regions specic context and
circumstances in terms of the industrial structure, insti-
tutional set-up and knowledge base. Thus, it contains
the potential for innovation policy to be more focused
by providing support that is needed given the demands
generated by industrial specicities. In this the distinc-
tion between analytical and synthetic knowledge and
its important consequences for innovation policy are
an example of such sharper focus that can be catered
for at the regional level. Notwithstanding this, region-
alisation should not be understood as regionalism by
neglecting the embeddedness of regions in a national
and trans-national framework.
Acknowledgements
An earlier version of this paper has beneted from
being presented at the Regionalization of Innovation
Policy conference, organised by the German Institute
for Economic Research (DIW Berlin) on June 45,
2004, in Berlin and at the 2004 DRUID Summer con-
ference Industrial Dynamics, Innovation and Devel-
opment on June 1416 in Helsing or, Denmark. The
authors thank the participants in the project Nordic
SMEs and Regional Innovation Systems for their valu-
able contributions. Special thanks go to the guest edi-
tors Michael Fritsch and Andreas Stephan as well as
ve anonymous referees. Acknowledgement is given
to the Nordic Innovation Centre and KUNI, the Nor-
wegian Research Council, for their nancial support.
References
Amin, A., 1999. An institutionalist perspective on regional eco-
nomic development. International Journal of Urban and Regional
Research 23 (2), 365378.
Amin, A., Cohendet, P., 2003. Architectures of Knowledge. Oxford
University Press, Oxford.
Asheim, B.T., 1995. Regionale Innovasjonssystemen Sosialt
og Territorielt Forankret Teknologipolitikk. Nordisk
Samh allsgeogrask Tidskrift 20, 1734.
1188 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
Asheim, B.T., 1998. Territoriality and economics: on the substantial
contribution of economic geography. In: Jonsson, O., Olander,
L.-O. (Eds.), Economic Geography in Transition, vol. 74. The
Swedish Geographical Yearbook, Lund, pp. 98109.
Asheim, B.T., 1999. Interactive learning and localised knowledge
in globalising learning economies. GeoJournal 49 (4), 345
352.
Asheim, B.T., 2000. Industrial districts: the contributions of mar-
shall and beyond. In: Clark, G.L., Feldman, M.P., Gertler, M.S.
(Eds.), The Oxford Handbook of Economic Geography. Oxford
University Press, Oxford, pp. 413431.
Asheim, B.T., 2001. Learning regions as development coalitions:
partnership as governance in european workfare states? Concepts
and Transformations: International Journal of Action Research
and Organizational Renewal 6 (1), 73101.
Asheim, B.T., 2002. Temporary organisations and spatial embedded-
ness of learning and knowledge creation. Geograska Annaler:
Series B Human Geography 84, 111124.
Asheim, B.T., Cooke, P., 1998. Localised Innovation Networks in
a Global Economy: A Comparative Analysis of Endogenous
and Exogenous Regional Development Approaches. Compar-
ative Social Research 17, JAI Press, Stamford, CT, pp. 199
240.
Asheim, B.T., Cooke, P., 1999. Local learning and interactive
innovation networks in a global economy. In: Malecki, E.,
Oinas, P. (Eds.), Making Connections: Technological Learning
and Regional Economic Change. Ashgate, Aldershot, pp. 145
178.
Asheim, B.T., Gertler, M.S., 2005. The geography of innovation:
regional innovation systems. In: Fagerberg, J., Mowery, D., Nel-
son, R. (Eds.), The Oxford Handbook of Innovation. Oxford
University Press, Oxford, pp. 291317.
Asheim, B.T., Herstad, S.J., 2003. Regional innovation systems, vari-
eties of capitalism and non-local relations: challenges from the
globalising economy. In: Asheim, B.T., Mariussen,

A. (Eds.),
Innovations, Regions and Projects, vol. 3. Nordregio, Stockholm,
pp. 241274.
Asheim, B.T., Isaksen, A., 2002. Regional innovation systems: the
integration of local sticky and global ubiquitous knowledge.
Journal of Technology Transfer 27, 7786.
Asheim, B.T., et al. (Eds.), 2003a. Regional Innovation Policy for
Small-Medium Enterprises. Edward Elgar, Cheltenham.
Asheim, B.T., Coenen, L., Svensson-Henning, M., 2003b. Nordic
SMEs and Regional Innovation SystemsFinal Report. Nordic
Industrial Fund, Oslo (http://www.nordicinnovation.net).
Bathelt, H., 2003. Geographies of production: growth regimes in spa-
tial perspectives 1innovation, institutions and social systems.
Progress in Human Geography 27 (6), 789804.
Bathelt, H., 2005.Geographies of production: growth regimes in spa-
tial perspectives (II) knowledge creation and growth in clusters.
Progress in Human Geography 29 (2), 204216.
Bathelt, H., Malmberg, A., Maskell, P., 2004. Clusters and knowl-
edge: local buzz, global pipelines and the process of knowledge
creation. Progress in Human Geography 28 (1), 3156.
Benneworth, P., Hardy, S., 2003. In Celebration of the Ordinary
Region. In: Introduction to paper session at the RGS-IBG Inter-
national Annual Conference, September 35, London.
Carlsson, B., 2004. Innovation systems: a survey of the literature
from a schumpeterian perspective. In: Pycka, A. (Ed.), The
Companion to Neo-Schumpeterian Economics. Edward Elgar,
Cheltenham.
Coenen, L., Moodysson, J., Asheim, B.T., 2004. Nodes, networks
and proximities: on the knowledge dynamics of the medicon val-
ley biotech cluster. European Planning Studies 12 (7), 1003
1018.
Cooke, P., 1998. Introduction: origins of the concept. In: Braczyk, H.,
Cooke, P., Heidenreich, M. (Eds.), Regional Innovation Systems,
rst ed. UCL Press, London, pp. 225.
Cooke, P., 2004. Evolution of regional innovation systems
emergence, theory, challenge for action. In: Cooke, P., et al.
(Eds.), Regional Innovation Systems, second ed. Routledge, Lon-
don, pp. 118.
Cooke, P. Regional innovation systems, asymmetric knowledge and
the legacies of learning. In: Rutten, et al. (Ed.), The Learning
Region: Foundations, State of the Art, Future. Edward Elgar,
Cheltenham, in press.
Cooke, P., Leydesdorff, L. Regional development in the knowledge-
based economy: the construction of advantage. Journal of Tech-
nology Transfer, in press.
Cooke, P., Morgan, K., 1998. The Associational Economy: Firms,
Regions and Innovation. Oxford University Press, Oxford.
Cooke, P., Boekholt, P., T odtling, F., 2000. The governance of inno-
vation in europe. In: Regional Perspectives on Global Competi-
tiveness. Pinter, London.
Cooke, P., Heidenreich, M., Braczyk, H.-J. (Eds.), 2004. Regional
Innovation Systems, second ed. Routledge, London.
Cooke, P., Uranga, M.G., Etxebarria, G., 1998. Regional systems of
innovation: an evolutionary perspective. Environment and Plan-
ning A 30, 15631584.
Dalum, B., et al., 1999. The formation of knowledge-based clusters
in north jutland and western sweden. In: Paper presented at the
DRUID Conference on National Innovation Systems, Industrial
Dynamics and Innovation Policy, June 912, Rebild.
Dalum, B., Pedersen C., Villumsen, G., 2002. Technological life
cycles: regional clusters facing disruption. DRUID Working
Paper 02-10. Aalborg University, Aalborg.
Doloreux, D., 2002. What we should know about regional systems
of innovation. Technology in Society 24, 243263.
Edquist, C., 1997. Introduction. In: Edquist, C. (Ed.), Systems of
Innovation: Technologies, Institutions and Organisations. Pinter,
London.
Etzkowitz, H., Leydesdorff, L., 2000. The dynamics of innova-
tion: from national systems and Mode 2 to a Triple Helix of
Universityindustrygovernment relations. Research Policy 29,
109123.
Fagerberg, J., Mowery, D., Nelson, R. (Eds.), 2004. The
Oxford Handbook of Innovation. Oxford University Press,
Oxford.
Freeman, C., October 1993. The political economy of the long
wave. In: Paper Presented at EAPE 1993 Conference on The
Economy of the Future: Ecology, Technology, Institutions,
Barcelona.
Gertler, M., 2004. Manufacturing Culture: The Institutional Geogra-
phy of Industrial Practice. Oxford University Press, Oxford.
B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190 1189
Granovetter, M., 1985. Economic action and social structure: the
problem of embeddedness. American Journal of Sociology 91,
481510.
Hall, P., Soskice, D., 2001. Varieties of Capitalism: The Institu-
tional Foundations of Comparative Advantage. Oxford Univer-
sity Press, Oxford.
Henry, N., et al., 1995. Along the road: R&D, society and space.
Research Policy 24, 707726.
Hirsch-Kreinsen, H., Jacobson, D., Laestadius, S., Smith, K.,
2003. Low-Tech Industries and the Knowledge Economy: State
of the Art and Research Challenges. STEP Report 16-2003,
Oslo.
Holmberg, G., 2003. Functional foods in Scania, Sweden. In:
Asheim, B.T., Coenen, L., Svensson-Henning, M. (Eds.), Nordic
SMEs and Regional Innovation SystemsFinal Report. Nordic
Industrial Fund, Oslo (http://www.nordicinnovation.net).
Howells, J., 1999. Regional systems of innovation? In: Archibugi, D.,
et al. (Eds.), Innovation Policy in a Global Economy. Cambridge
University Press, Cambridge.
Isaksen, A. (Ed.), 1999. Regionale Innovasjonssystemer: Innovasjon
ogLringi 10Regionale Nringsmiljer. STEPReport 02-1999,
Oslo.
Isaksen, A., 2003. Learning, globalization, and the electronics cluster
in horten: discussing the local buzzglobal pipeline argument.
In: Asheim, B.T., Coenen, L., Svensson-Henning, M. (Eds.),
Nordic SMEs and Regional Innovation SystemsFinal Report.
Nordic Industrial Fund, Oslo (http://www.nordicinnovation.net).
Isaksen, A., 2005. Regional clusters building on local and non-
local relationships: a European comparison. In: Lagendijk, A.,
Oinas, P. (Eds.), Proximity, Distance and Diversity: Issues on
Economic Interaction and Local Development. Ashgate, Alder-
shot, pp. 129152.
Isaksen, A., Hauge, E., 2002. Regional Clusters in Europe. Observa-
tory of European SMEs report 2002 No. 3, European Communi-
ties, Luxembourg.
Johnson, B., Lundvall, B.-

A., 2001. Why all this fuss about


codied and tacit knowledge? In: Paper presented at the
DRUID ACADEMY Winter Conference, Korsr, January 18
20.
Kaufmann, A., T odtling, F., 2000. Systems of innovation in tra-
ditional industrial regions: the case of styria in a comparative
perspective. Regional Studies 34 (1), 2940.
Kaufmann, A., T odtling, F., 2001. Scienceindustry interaction in
the process of innovation: the importance of boundary-crossing
between systems. Research Policy 30, 791804.
Laestadius, S., 1998. Technology level, knowledge formation and
industrial competence in paper manufacturing. In: Eliasson, G.,
et al. (Eds.), Micro Foundations of Economic Growth. The Uni-
versity of Michigan Press, Ann Arbour, pp. 212226.
Longhi, C., Qu ere, M., 1993. Innovative networks and the technop-
olis phenomenon: the case of sophie-antipolis. Environment and
Planning C: Goverment and Policy 11, 317330.
Lorenzen, M., 2003. Low-Tech localized learning: the regional inno-
vation system of salling, denmark. In: Asheim, B.T., Coenen,
L., Svensson-Henning, M. (Eds.), Nordic SMEs and Regional
Innovation SystemsFinal Report. Nordic Industrial Fund, Oslo
(http://www.nordicinnovation.net).
Lundvall, B.-

A. (Ed.), 1992. National Systems of Innovation:


Towards a Theory of Innovation and Interactive Learning. Pinter,
London.
Lundvall, B.-

A., 1999. National business systems and national sys-


tems of innovation. International Studies of Management &
Organisation 29 (2), 6077.
Lundvall, B-

A., Archibugi, D. (Eds.), 2001. The Globalizing Learn-


ing Economy. Oxford University Press, Oxford.
Lundvall, B.-

A., Borras, S., 1997. The Globalising Learning Econ-


omy: Implications for Innovation Policy. European Communi-
ties, Luxembourg.
Maskell, P., Eskelinen, H., Hannibalsson, L., Malmberg, A., Vatne,
E., 1998. Competitiveness, Localised Learning and Regional
Development. Routeledge, London.
McKelvey, M., Alm, H., Riccaboni, A., 2003. Does co-location
matter for formal knowledge collaboration in the swedish
biotechnology-pharmaceutical sector? Research Policy 32 (3),
483501.
Miettinen, R., 2002. National Innovation System: Scientic Concept
or Political Rhetoric. Edita, Helsinki.
Moulaert, F., Sekia, F., 2003. Territorial Innovation models: a critical
survey. Regional Studies 37 (3), 289302.
Nonaka, I., Reinm oller, P., 1998. The legacy of learning: toward
endogenous knowledge creation for Asian economic develop-
ment. WZB Jahrbuch 1998, 401433.
Nonaka, et al., 2000. SECI, Ba and leadership: a unied model
of dynamic knowledge creation. Long Range Planning 33, 5
34.
Nilsson, M., Svensson-Henning, M., Wilkenson, O., 2002. Sk anska
Kluster och Prolomr adenEn Kritisk Granskning. Region
Sk ane, Malm o.
OECD, 1986. OECD Science and Technology Indicators, R&D,
Innovation and Competitiveness. OECD, Paris.
OECD, 1996. The Knowledge-Based Economy. OECD, Paris.
Onsager, K., 1999. Matindustrien i Innovative Nettverk i Roga-
land. In: Isaksen, A. (Ed.), Regionale Innovasjonssystemer: Inno-
vasjon og Lring i 10 Regionale Nringsmiljer. Step Rapport
2-1999, Oslo.
Onsager, K., Aasen, B., 2003. The Case of Rogaland Regional
Food Cluster. In: Asheim, B.T., Coenen, L., Svensson-
Henning, M. (Eds.), Nordic SMEs and Regional Innova-
tion SystemsFinal Report. Nordic Industrial Fund, Oslo
(http://www.nordicinnovation.net).
Pavitt, K., 1984. Sectoral Patterns of technical change: towards a
taxonomy and a theory. Research policy 13, 343373.
Porter, M.E., 1990. The Competitive Advantage of Nations. Billing
and Sons Ltd., Worchester.
Porter, M.E., 2000. Locations, clusters and company strategy. In:
Clark, G.L., Feldman, M.P., Gertler, M.S. (Eds.), The Oxford
Handbook of Economic Geography. Oxford University Press,
Oxford, pp. 253274.
Saxenian, A., 1994. Regional Advantage: Culture and Competition
in Silicon Valley and Route 128. Harvard University Press, Cam-
bridge, Massachusetts.
Smith, K., 2000. What is The Knowledge Economy? Knowledge-
intensive Industries and Distributed Knowledge Bases. Paper
presented at the DRUID Summer Conference on The Learn-
1190 B.T. Asheim, L. Coenen / Research Policy 34 (2005) 11731190
ing EconomyFirms, Regions and Nation Specic Institutions,
Aalborg, Denmark, June 2000.
Storper, M., Scott, A., 1995. The Wealth of Regions. Futures 27 (5),
505526.
T odtling, F., Trippl, M., 2005. Networking and project organisation
in the Styrian automotive industry. In: Lagendijk, A., Oinas,
P. (Eds.), Proximity, Distance and Diversity: Issues on Eco-
nomic Interaction and Local Development, Ashgate, Aldershot,
pp. 89110.
Whitley, R., 1999. Divergent Capitalismsthe Social Structuring
and Change of Business Systems. Oxford University Press,
Oxford.

S-ar putea să vă placă și