Content uploaded by Robert Hassink
Author content
All content in this area was uploaded by Robert Hassink on Mar 10, 2018
Content may be subject to copyright.
How to Unlock Regional Economies
from Path Dependency? From Learning
Region to Learning Cluster
ROBERT HASSINK
University of Duisburg-Essen, Institute of Geography, Duisburg, Germany
ABSTRACT Since the Industrial Revolution the cyclical processes of rise and fall of regional
economies have been accelerating. Many of the specific problems of the falling part of clustering,
that is old industrial areas, are related to path dependency and lock-ins. Particularly political
lock-ins hinder the necessary restructuring processes in old industrial areas. They can be
considered as thick institutional tissues aiming at preserving existing industrial structures and
therefore unnecessarily slowing down industrial restructuring and indirectly hampering the
development of indigenous potential and creativity. Of the recently born offspring of the family of
territorial innovation models, the learning region concept seems to be most focused on
overcoming and avoiding political lock-ins in old industrial areas. Most scholars consider
learning regions as regional development concepts in which the main actors are strongly, but
flexibly, connected with each other and are open both to intraregional and interregional learning
processes. Policy-makers in learning regions are involved in learning from institutional errors
made in the past and by doing that in avoiding path-dependent development. Empirical evidence,
however, shows that the learning region is of limited importance to unlock regional economies
from path dependency, due to three weaknesses: its fuzziness, its normative character in its
squeezed position between national innovation systems and global production networks. A less
normative and more process-oriented concept is proposed in this paper, namely that of the
learning cluster.
Introduction
Since the Industrial Revolution the cyclical processes of rise and fall of regional econom-
ies have been accelerating. Modern theoretical concepts in economic geography, however,
mainly try to explain the rising part of geographical clustering of industries. Since these
concepts are heavily based on observations in a small number of exceptional regional
economies, such as Silicon Valley, the Third Italy and Baden-Wu
¨rttemberg, they have
relatively little to offer to regional policies focused on the specific problems of the
falling part of clustering, that is old industrial areas. These policies, therefore have been
mainly deriving from lessons learned from success regions and thus have been consisting
Correspondence Address: Robert Hassink, University of Duisburg-Essen, Institute of Geography, Lotharstr. 65,
47048 Duisburg, Germany. Email: hassink@uni-duisburg.de
ISSN 0965-4313 print=ISSN 1469-5944 online=05=040521– 15 #2005 Taylor & Francis Group Ltd
DOI: 10.1080=09654310500107134
European Planning Studies Vol. 13, No. 4, June 2005
of measures such as science parks and technology transfer agencies to boost small and
medium-sized enterprises.
Of the few theoretical concepts that try to explain the decline of industrial areas, evol-
utionary regional economics, in general, and path dependency and lock-ins, in particular,
are powerful ones, because they stress the importance of history and institutional context
for regional development. Grabher (1993) developed political lock-ins furthest in his
studies on the Ruhr Area in Germany. They can be considered as thick institutional
tissues aiming at preserving existing industrial structures and therefore unnecessarily
slowing down industrial restructuring and indirectly hampering the development of indi-
genous potential and creativity.
Of the recently born offspring of the family of territorial innovation models (Lagendijk,
2003; Moulaert & Sekia, 2003), the learning region concept seems to be most focused on
overcoming and avoiding political lock-ins in old industrial areas (Schamp, 2000; OECD,
2001; Boschma & Lambooy, 1999). Although there are several definitions and perspec-
tives, most scholars consider learning regions as regional development concepts in
which the main actors (politicians, policy-makers, chambers of commerce, trade unions,
higher education institutes, public research establishments and companies) are strongly,
but flexibly connected with each other and are open both to intraregional and interregional
learning processes (Morgan, 1997; Boekema et al., 2000; Butzin, 2000; Hassink, 2001;
Wink, 2003). Policy-makers in learning regions are involved in learning from institutional
errors made in the past and by doing that in avoiding path-dependent development. Illus-
trated by recent research on highly specialized declining regional economies in Germany
and South Korea, this paper, however, will reveal the learning region concept’s limitations
to unlock regional economic path dependency in old industrial areas. It will show its main
weaknesses and will propose an alternative concept, namely that of the learning cluster. In
the following section, the concept of lock-ins and path dependency in relation to the
restructuring of old industrial areas is discussed. The section after that will portrait the
learning region from a conceptual point of view, whereas the next section will verify
the application of the concept in two old industrial areas. The last section will reveal
the learning region concept’s limitations to unlock regional economic path dependency
in old industrial areas and will come up with the alternative concept of learning clusters.
Lock-ins, Path Dependency and the Delayed Restructuring
of Old Industrial Areas
The line between successful and open regions and old industrialized, insular, inward-
looking industrial districts can be very thin (Grabher, 1993; Hamm & Wienert, 1989;
Fromhold-Eisebith, 1995; Hassink, 1997; Maskell & Malmberg, 1999). As milieus tend
to change more slowly than industries, a sclerotic milieu can remain in a region even
after the industrial structure to which it belonged already has disappeared. Maskell and
Malmberg (1999) distinguish ‘good’ from ‘bad’ agglomerations by pointing at their
ability to ‘un-learn’, which involves the removal of formerly significant institutions
which now act as a hindrance to further development. There appears a great variation
in the ability of regions to ‘un-learn’, “which makes it possible in some regions but not
in others to inaugurate new institutions and simultaneously dissolve ones” (Maskell &
Malmberg, 1999, p. 179).
522 Robert Hassink
In some old industrial areas, insular, inward-looking production clusters suffer from a
combination of three negative lock-ins: functional lock-ins (inter-firm relationships),
cognitive lock-ins (a common world view that might confuse secular trends with cyclical
downturns) and political lock-ins that might come up in a production cluster (Hamm &
Wienert, 1989; Grabher, 1993; Hassink & Shin, 2005; Morgan & Nauwelaers, 1999;
Boschma, 2003). In other old industrial areas, however, political initiatives and insti-
tutional renewal have led to the successful promotion of new industrial activities and
thus restructuring (Cooke, 1995). In the unsuccessful ones, political lock-ins hinder
renewal and restructuring. They can be regarded as thick institutional tissues aiming at
preserving existing traditional industrial structures and therefore unnecessarily slowing
down industrial restructuring and indirectly hampering the development of indigenous
potential and creativity. Institutional tissues consist both of organizations (“formal struc-
tures with an explicit purpose”), such as political administrations at all spatial levels, trade
unions, large enterprises and business support agencies, and ‘things that pattern behaviour’
such as norms, rules and laws (Edquist, 1997, p. 26). With regard to the latter part there
seems to be, therefore, a strong relationship between cognitive lock-ins and political
lock-ins. Such a particular and thick institutional tissue together with the firms and
workers can form a so-called self-sustaining coalition (Grabher, 1993; Hassink & Shin,
2005). In such a situation, large companies are unwilling to sell unused sites to local auth-
orities for the attraction of inward investment, as they are afraid to lose qualified employ-
ees to competitors. Local authorities do not see the point to attract inward investment or to
promote restructuring in another way, as large tax incomes are paid by large local enter-
prises. In some regional production clusters the spirit of the Schumpeterian entrepreneur
might dwindle due to an increasing industrial concentration and the domination of large
companies. The self-sustaining coalition also lobbies for sectoral interventions often at
a national or supra-national level, which hamper the restructuring process more than
they support it, as they remove the incentives to take initiatives for entrepreneurs and
thus paralyse competition and tranquillize large industries (Hamm & Wienert, 1989).
Morgan and Nauwelaers (1999) stress that in these kind of networks status is privileged
over knowledge, power over learning and past over present. Related to the issue of
political lock-ins are discussions about network failures and anti-developmental networks
(von Tunzelmann, 2004; Iammarino, 2005). Indicators to measure political lock-ins
might be the amount of subsidies spent to support the industry, the number of lobbying
organizations and, more importantly, their impact, long-term stability of institutions
involved in supporting the industry and the weak support of new industries.
Grabher’s lock-in concept has very often been cited (see for instance Cooke & Morgan,
1998, p. 111; Schamp, 2000, p. 139), showing its importance as an explanatory concept not
only for classical old industrial areas, such as the Ruhr Area, but for the decline of a large
variety of differently structured industrial areas (see Shapira & Fuchs, 2005). There have
been much fewer thoughts and papers about regional development concepts that can
unlock old industrial economies from negative path dependency. The learning region is
potentially such a concept.
The Learning Region
In the framework of the contemporary transformation from an industrial to a knowledge-
based economy, the learning economy (Lundvall, 1996) and recently also learning regions
How to Unlock Regional Economies from Path Dependency? 523
have been propagated as future concepts for successful economic development in many
countries of Europe (Morgan, 1997; Hassink, 2001; van Geenhuizen, 1999; Butzin,
2000; Scheff, 1999; Boekema et al., 2000; OECD, 2001; Landabaso et al., 2001; Fu
¨rst,
2001; Kunzmann & Tata, 2003). The capacity of both individuals and organizations to
engage successfully in learning processes is regarded as a crucial component of economic
performance in the knowledge-based economy. Oinas and Virkkala (1997) even speak
about the 1990s as being the era of the learning economy and the learning region and
Malmberg (1997, p. 576) refers to the ‘learning turn’ in economic geography.
Reading the recently rapidly expanding amount of literature on the learning region, two
angles can be distinguished from which this concept has been launched. First, some
authors have written about the relationship between entrepreneurial learning, innovation
and spatial proximity at the micro level (theoretical, actor-related perspective) (Oinas &
Virkkala, 1997; Lorenzen, 2001; Boekema et al., 2000). Secondly, most authors have
launched the concept as a theory-led regional development concept from an action-
related perspective at the meso level. This distinction bears resemblance to a distinction
Boekema et al. (2000) made in a book on the theory and practice of learning regions,
where they distinguish between regional learning (mainly company-initiated cooperation
between actors in a region through which they learn) and the learning region (institutional
networks that develop and implement a regional innovation strategy). Since this paper
aims at analysing the potential of the learning region strategy to break through negative
path dependency, it will focus on the second, action-related perspective. In this second
perspective, the learning region is seen as a new theory-led regional development
concept which aims at achieving and/or supporting collective learning processes
(Morgan, 1997; Fu
¨rst, 2001; Butzin, 2000; OECD, 2001).
In many countries a general shift of innovation and labour market policies can be
observed from the national to regional levels of decision-making, partly supported by
supra-national organizations such as the European Union (EU) and the World Bank
(OECD, 2001). The regional level is more and more seen as the level that offers the great-
est prospect for devising governance structures to foster learning in the knowledge-based
economy (Cooke & Morgan, 1998; Lorenzen, 2001; Boekema et al., 2000; Fritsch, 2004;
Koschatzky, 2001, 2005).
The definitions of learning regions are quite vague and diverse, since seldom concrete
examples can be shown and since policy-makers, who have been eager to use the concept
as a label for their development plans, have not made efforts to define what they mean by
learning regions. The concept seems to travel easily from academic circles to policy-
makers and back without deep thoughts about its meaning (Hassink & Lagendijk,
2001). Small wonder, therefore, that Martin (2001, p. 198) considers learning regions,
together with institutional thickness and un-traded interdependencies, as fuzzy concepts,
as he calls these concepts ‘vague and impressionistic neologisms’. In a recent edited
book on learning regions, Boekema et al. (2000, p. 12) even make this situation worse,
as they want “to avoid an unproductive discussion on what is or is not a ‘learning
region’” and launch the learning region as a paradigm that does not need to be defined.
According to the Organization for Economic Cooperation and Development (OECD)
(2001, pp. 23 – 24) the learning region “constitutes a model towards which actual
regions need to progress in order to respond most effectively to the challenges posed by
the ongoing transition to a ‘learning economy’”. It is “characterised by regional insti-
tutions, which facilitate individual and organizational learning through the co-ordination
524 Robert Hassink
of flexible networks of economic and political agents” (OECD, 2001, p. 24). Regional
policies are crucial for stimulating individual and organizational learning, because
policy-makers can address path dependency that goes beyond the interest of single
agencies and firms (OECD, 2001). Both changing the industrial structure and institutional
unlearning are issues that can fruitfully be addressed by regional policy-makers. As the
learning region is a model, it is not possible to identify examples of actually existing learn-
ing regions (OECD, 2001). There are various trajectories towards the goal to become a
learning region. To affect social capital in regions is an important element of the learning
region strategy.
Morgan (1997) calls learning regions the new generation of regional policy, which,
compared to traditional regional policy, focuses on infostructure instead of infrastructure,
on opening minds instead of opening roads and branch plants and which devises policies
with small and medium sized enterprises (SMEs) instead of just policies for SMEs. Other
characteristics of this concept are: bottom-up concept, transparent, face-to-face relations,
integrated solving of problems (crossing of policy fields) and permanent organizational
learning with feedback effects. This network is open to learning, both to intraregionally
and interregionally, and willing to unlearn. These characteristics of a learning region,
however, only describe the method of working and the attitude of regional economic
policy-makers. The concrete contents of the innovation policy need to vary according
to the economic profile and demand in individual regions.
The learning region can thus be defined as a regional innovation strategy in which a
broad set of innovation-related regional actors (politicians, policy-makers, chambers of
commerce, trade unions, higher education institutes, public research establishments and
companies) are strongly, but flexibly connected with each other, and who stick to a
certain set of ‘policy principles’ (OECD, 2001). The following ‘policy principles’,
which are general in scope, leaving regional policy-makers to adapt them to specific con-
texts and demand for innovation policies in the various regions, are a crucial part of a
learning region strategy (OECD, 2001; Fu
¨rst, 2001):
.carefully coordinating supply of and demand for skilled individuals
.developing a framework for improving organizational learning, which is not only
focused on high-tech sectors, but on all sectors that have the potential to develop
high levels of innovative capacity
.carefully identifying resources in the region that could impede economic development
(lock-ins)
.positively responding to changes from outside, particularly where this involves unlearning
.developing mechanisms for coordinating both across departmental and governance
(regional, national, supranational) responsibilities
.developing strategies to foster appropriate forms of social capital and tacit knowledge
that are positive to learning and innovation
.continuously evaluating relationships between participation in individual learning,
innovation and labour market changes
.developing an educational and research infrastructure for knowledge society
.encouraging openness to impulses from outside
.fostering redundancy and variety
.ensuring the participation of large groups of society in devising and implementing
strategies.
How to Unlock Regional Economies from Path Dependency? 525
As the learning region can be considered as an eclectic concept (Fu
¨rst, 2001), it is strongly
linked to several existing theory-led development models and policy-oriented innovation
stimulation concepts, which have been coined as new regionalism (Lovering, 1999) or the
family of territorial innovation models (TIMs) (Moulaert & Sekia, 2003). Examples of
these concepts are regional innovation systems, industrial districts, innovative milieus
and regional clusters. It would go beyond the scope of this paper to discuss and explain
the concept’s exact location within this plethora of concepts, but what is important to
know is that of all these concepts the learning region concept has been most clearly con-
nected to the solving of problems of old industrial areas. By focusing on the learning
ability of regional actors, the learning region concept might, in contrast to the other con-
cepts, be able to explain why in some regions learning by interacting and collective tacit
knowledge can turn from a strength into a weakness (negative path dependence). Here the
learning region clearly adds something to existing concepts. Compared with other con-
cepts, learning regions are more involved in learning from institutional errors made in
the past and by doing that in avoiding path-dependent development (OECD, 2001). The
latter point is illustrated by the research question Gertler and Wolfe (2004, p. 93) are
putting in their study on a regional innovation system in Ontario, Canada: “how reflexive
is the [regional innovation] system as a whole in terms of monitoring its successes or fail-
ures and adopting the features associated with a learning regions elsewhere?” Learning
regions, therefore, seem to be reflective and monitoring or ‘virtuous’ (OECD, 2001,
p. 11) regional innovation systems. Furthermore, in contrast to the earlier described
theory-led development models, which are mainly based on experiences in growth
regions such as Silicon Valley, Baden-Wu
¨rttemberg and the Third Italy, the learning
region concept is not derived from experiences in any particular kind of region (although
in a recent article Benner (2003) describes Silicon Valley as a learning region). Therefore,
it can be applied to a broader range of regions than the other models, which turned out to
be difficult to transfer to structurally weak regions.
Although many observers of the learning region have criticized the lack of empirical
evidence (Fu
¨rst, 2001; Blotevogel, 1999, p. 56), recently both semi-academic empirical
work on the learning region by the OECD (2001) and numerous policy initiatives launched
under the label of learning regions (Lagendijk & Cornford, 2000) provide us with a
considerable amount of empirical information on the learning region phenomenon. The
OECD (2001) recently published the first in-depth empirical study on the concept of learn-
ing regions. With the help of an interesting mix of a quantitative correlation analysis and a
qualitative analysis on the basis of case-studies several conceptual relationships of the
learning region are investigated. Other recent empirical studies on the learning region
include work on the central part of the Ruhr Area in Germany by Pommeranz (2000)
and on the Graz Region in Austria by Scheff (1999). Moreover, the learning region
concept has arrived in one form or another on the desks of regional policy-makers in
Europe for quite some time now. A short internet and press survey carried out by
Lagendijk and Cornford (2000) in August 1998 revealed that nine regions labelled
themselves as learning regions, and this number has dramatically increased since then.
In May 2004 an internet survey by the author of this paper revealed a total of 5000
hints on learning region in the world wide web and no less than 11,000 for the German
equivalent Lernende Region (the Dutch and French equivalents (lerende regio and
re
´gion apprenante) had only 218 and 185 hints, respectively). The popularity of the
German term can be explained through the federal programme called Lernende Region,
526 Robert Hassink
which is partly funded by the European Social Fund (www.lernende-regionen.info), and
which supports 72 learning regions in Germany. A strong appeal to regional policy-
makers has been that, with the help of the buzzword learning region, they could broaden
out narrow technology policies to areas of business development, labour market policies,
skill improvement and particularly lifelong learning (Lagendijk & Cornford, 2000,
p. 216). Furthermore, partly based on the learning region concept, the EU has started a
new generation of regional policies (Landabaso et al., 2001), which aim at improving the
institutional capacity for innovation of less-favoured regions, which, in turn, should lead
to higher absorption capacity for innovation funds from the EU and national governments.
In sum, the learning region concept could serve to solve the question what distinguishes
‘good’ from ‘bad’ industrial agglomerations. Traditional theoretical concepts as well as
recent studies on regional networking and collective learning in Europe not only focus
too much on success regions, they also lack the equipment to distinguish ‘good’ from
‘bad’ industrial agglomerations. The limited learning ability of regional actors could be
the explanatory factor why the coordination of collective activities in some regions
turns from a strength into a weakness (path dependence). Since, conceptually, the learning
region is regarded as most focused on overcoming and avoiding political lock-ins in old
industrial areas, the following section will analyse whether learning region strategies
are applied in declining regional economies in order to unlock them from negative path
dependency.
Learning Region Strategies in Old Industrial Areas?
This section will verify whether learning region strategies are applied to unlock regional
economic path dependency in two highly specialized, declining old industrial areas,
Mecklenburg-Vorpommern in Germany, which has an economy dominated by the ship-
building industry, and Daegu in South Korea, which specializes in the textile industry.
It will focus on the following selection of the earlier-mentioned policy principles that
can be regarded as important in the context of negative path dependency:
.carefully identifying resources in the region that could impede economic development
(lock-ins)
.positively responding to changes from outside, particularly where this involves unlearning
.developing mechanisms for coordinating both across departmental and governance
(regional, national, supra-national) responsibilities
.encouraging openness to impulses from outside
.fostering redundancy and variety.
Shipbuilding dominates the regional production structure of Mecklenburg-Vorpommern,
Germany (Eich-Born & Hassink, 2005), one of the new La
¨nder in reunited Germany situ-
ated in the north-east (see Figure 1). Due to the transformation from central planning to
market economy de-industrialization (employment in shipbuilding dropped from 55,000
in 1989 to around 5000 in 2003) led to dramatically high unemployment rates of
around 20%. In order to save the industry from total collapse, a political consensus was
built between various interest groups on different geographical levels: yard managers,
workers councils, regional trade unions, mayors of yard cities, regional policy-
makers as well as the representatives of the German Shipbuilding and Ocean Industries
How to Unlock Regional Economies from Path Dependency? 527
Association and the Coordinator for the Maritime Economy in the Federal Ministry of
Economics and Technology. Dissolution of the state-owned enterprise and privatization
were the strategies applied by the German national and regional government. Nowadays
the yards are in the hands of mainly Scandinavian shipbuilding concerns. The moderniz-
ation of the production capacities was mainly financed by subsidies provided by the
German government, the state of Mecklenburg-Vorpommern and the EU. Over a period
of 5 years, the German government invested more than DM 6 billion (E3 billion) in the
construction of new docks, which means state support of about 1 million DM
(E500,000 per job) (Ro
¨ller & von Hirschhausen, 1996, p. 17). For each DM of state aid
only about 0.09 DM of private investment was attracted. In order to avoid a strong increase
in over-capacity, the federal government and the European Commission agreed that the
shipyards in Mecklenburg-Vorpommern were not allowed to build more ships than a
certain annual capacity limit until 2005.
Thus, regional industrial policy in Mecklenburg-Vorpommern is very much focused on
preserving the existing modern shipbuilding complex, rather than focused on developing
new products and industries. That is not only shown by the large amount of subsidies made
available for the shipbuilding industry, it is also shown by the successful lobby of the con-
sensus group of actors to release the EU capacity limitation and to sue the South Korean
government at the World Trade Organization (WTO) for supposedly illegally supporting
Korean yards. It would be unfair, however, to make the impression that the regional
industrial policy is only active in lobbying activities in order to preserve existing structures.
It does also a lot to support innovative small companies and innovation projects. However,
these projects, such as the Maritime Alliance in the framework of the federal support
programme InnoRegio, also mainly support the innovativeness of the existing cluster.
Figure 1. Map of Mecklenburg-Vorpommern, Germany, with location of shipyards
528 Robert Hassink
In a recent paper, Cho (2004) discusses the restructuring process of the textile industry
cluster of Daegu, the third largest city in South Korea (Figure 2), a process the author has
analysed himself, during field-work in 2003 in South Korea. The cluster, which started to
grow in the 1960s and which geographically consists of Daegu, Gumi and Gyeongsan, is
characterized by a specialization in the production and weaving of chemical fibres and has
been strongly focusing on export. Textile business constitutes the largest segment of man-
ufacturing in Daegu: 31.3% of total establishment, 34.7% of total employment; which
means a location quotient of 4.1, 34.6% of total production, 54.2% of total export and
30.9% of total value added in 2002 (Cho, 2004). The high rate of automation in the
1980s brought about problems of overcapacity and overproduction, which in turn led to
financial difficulties in textile business. Moreover, textile companies were faced with
increasing competition from low-cost neighbouring countries, China in particular, and a
shift of Korean producers to China in the 1990s. Forty years of path dependent evolution
Figure 2. Map of Daegu in South Korea
How to Unlock Regional Economies from Path Dependency? 529
led to a specialization in the narrow low-value added and low-tech middle stream of the
textile value chain, whereas high-value added and high-tech downstream activities were
nearly totally absent. Consequently, the employment in Daegu’s textile industry decreased
from 91,000 in 1981 to 82,000 in 1986 to 47,000 in 2000 (Cho, 2004).
In response to this situation, the central government launched an ambitious project
called the Milano Project (1998 – 2003) aiming at restructuring the present middle-
stream textile of Daegu into a high value added down-stream textile which comprises
apparel, design and fashion as a competitive edge. Milano is a symbolic target for the
high-road restructuring of textile in Daegu. In April 1998, President Kim Dae-Jung,
who came to power with regionalist full support from his home province in the south-
west, visited Daegu to mollify the south easterner’s regionalist antagonism against him
and officially promised (kongyak) a full policy support for revitalization of Daegu’s
decaying textile industry, which materialized into the Milano Project. It consists of 19 pro-
jects in four sectors, which require a total of 650 million euros for 5 years. As of 30 April
2003, the overall rate of project implementation was 75%. The main promoters of the
project, the central government and the City of Daegu, aim at promoting both new
activities (fashion and design) and projects with new actors (research institutes, univer-
sities, design schools, banks, etc.), whereas the actors with a vested interest, local textile
producers and their lobby organizations, oppose these plans. The latter argued that
Daegu’s textile cluster should maintain its competitive edge in the branch of weaving
and dyeing, whose technology, know-how and market accessibility were believed to be
at the top of the world. Therefore resistance to and conflict around the restructuring are
widely witnessed in the process of project implementation. It shows that lock-ins
oppose learning initiatives to restructure the regional economy.
The Learning Region: A Policy Concept to Unlock Regional Economies
from Path Dependency?
The earlier presented cases share with each other the problems of regional economic
decline and path dependency and to some extent also political lock-ins. Although in
most cases the broad set of innovation-related regional actors (politicians, policy-
makers, chambers of commerce, trade unions, higher education institutes, public research
establishments and companies) that are strongly, but flexibly connected with each other
can be found, none of them can be called a learning region. The main reasons for this
is that they did not stick to the ‘policy principles’ that characterize a learning region. In
neither of the regions we can observe a carefully identifying of resources in the region
that could impede economic development (lock-ins), nor did the regional actors positively
respond to changes from outside, particularly where this involves unlearning. On the
contrary, in many ways the regions are characterized by negative responses to changes
from outside. Furthermore, what is lacking in the regions is an encouraging openness to
impulses from outside and a fostering of redundancy and variety. These cases have
shown how difficult it is to set up a learning region strategy in declining, mono-structural
economies due to political lock-ins. They have also shown that in both regions the national
and supra-national institutional level increasingly affects the leeway of regional actors.
A learning region strategy, therefore, will not be successful if it ignores the impact
of national and even international innovation systems on inter-firm cooperation and
innovative behaviour.
530 Robert Hassink
Despite these problems, the learning region concept potentially certainly bears poten-
tials to become a theoretical basis for modern regional innovation policies. Learning
regions and regional innovation systems can be seen as an “intellectual basis for the devel-
opment of particular forms of sub-national intervention” (OECD, 2001, p. 25). Therefore
geographers and planners dealing with the learning region seem to belong to the “signifi-
cant numbers of economic geographers [who] have been working on policy-relevant
topics and problems, including ... local industrial clusters, local high-technology
milieux” (Martin, 2001, p. 193). This clearly can be judged positively seen against the
background of the observed decrease in public policy relevance of geography research,
in general (Martin, 2001). In comparison with other concepts in economic geography,
the theory-led learning region concept is even more focused on a direct transfer of aca-
demic insights to local and regional innovation policies. Despite this merit, three weak-
nesses prevent the learning region from becoming a fully-fledged theoretical concept in
regional studies (for another critical voice see Hudson, 1999).
First, one group of authors hardly define the concept (see for instance Boekema et al.,
2000), leaving it to be a fuzzy concept, that is a concept characterized by both lacking con-
ceptual clarity, rigour in the presentation of evidence and clear methodology and difficul-
ties to operationalize (Markusen, 2003). In this group of literature seldom concrete
examples are shown. To make thing worst, most policy-makers, who have been eager
to use the concept as a label for their development plans, do not make efforts to define
what they mean by learning regions.
Secondly, and to some extent contradictory to the first point, the learning region can be
considered a normative concept from the perspective of the OECD study on learning
regions (OECD, 2001). Concepts differ to what extent they are normative in character,
such as the learning region, or based on real situations in regions (industrial districts)
(Hassink & Lagendijk, 2001). The learning region is also in many ways a normative
concept. As shown in an earlier section it is regarded as a model and it sticks to certain
‘policy principles’. In contrast to other concepts with an empirical base, such as industrial
districts, it clearly has a normative outlook. Moreover, since learning is a process which
takes place in any economy at any time, treating the learning region or learning
economy as a model is questionable. There is much written about what the ideal learning
region is, but little is understood about how we get there. As long as we do not know
enough about the learning processes that lead to a learning region strategy, we cannot
speak about a learning region yet.
Thirdly, although in theory open to impulses from outside, learning regions are in prac-
tice squeezed between national innovation systems and global production networks.
Speaking about learning regions, one should namely not forget the role of nations and
industries. In order to stay competitive, companies must integrate locally specific compe-
tence with codified, generally available knowledge, or, in other words, they must link their
own innovation systems with national innovation systems and international knowledge
flows (Bathelt et al., 2004). According to Gertler (1996) the increasing impact of national
regulatory and innovation systems on the behaviour and strategy of individual firms
narrows the leeway for regional innovation policy. Furthermore, depending on national
political-administrative systems, the leeway for developing learning regions strategies
differs considerably. A learning region strategy, therefore, will not be successful if it
ignores the impact of national and even international innovation systems on inter-firm
cooperation and innovative behaviour. Moreover, the learning region concept does not
How to Unlock Regional Economies from Path Dependency? 531
pay much attention to industry differences and the position of firms in global production
networks. By stressing the supply architecture for learning and innovation, it tends to
neglect that “different kinds of products will ‘demand’ different kinds of innovation
systems” (Storper, 1997, pp. 107– 108). Firms in different industries need different part-
ners for technological learning (chemical industry – public research establishments; build-
ing industry– customers) at different distances. Regional learning processes, therefore,
vary to a large extent, depending on the industry and the position of firms in global pro-
duction networks. Most regional policy-makers, however, have little knowledge about the
global production networks regional firms are embedded in (Herrigel, 2004). Since the
economic well-being of firms is increasingly affected by their position in global pro-
duction networks, policy-makers have more and more difficulties to update their firm
and industry knowledge which is necessary for a tailor-made regional innovation
policy. Regional learning is therefore less and less confined to the local, whereas learning
region strategy mainly focus on supporting intraregional learning processes.
Due to these weaknesses, the author has difficulty in using the term learning region.
There are seldom regions that are characterized by only one dominant industry and a
strong regional government (Bathelt & Depner, 2003). In reality, different clusters
within one region are found, with differing learning processes, different global production
networks and different national administrative systems and therefore a different leeway for
learning region strategies (for empirical evidence on this point, see To
¨dtling & Trippl,
2004; Steiner & Hartmann, 1998). In order to develop a learning region strategy, first
there is a need to understand the different learning processes that take place in the different
clusters found in regions.
Therefore the author wants to propose a related, but slightly different concept, that of the
learning cluster. The learning cluster concept is able to bridge the gap between regional
learning, which increasingly crosses the borders of regions and nations due to the globa-
lization of production networks, and the learning region strategy, which focuses on the
regional SMEs active in a variety of different clusters with different characteristics. It is
a concept that combines the strengths of both the learning region and clusters concept
in tackling the problem of lock-ins in regional economies.
On the basis of theoretical thoughts on geographic clustering by Porter (2000) and
Enright (2003), a rapidly increasing number of policy initiatives to support clustering of
industries have been emerging in many countries of the world (see for instance Porter,
2000; Gordon & McCann, 2000; Cumbers & MacKinnon, 2004). Like learning regions,
clusters, therefore, seem to be an empirical and theoretical basis for newly oriented
regional innovation policies (Koschatzky, 2005). According to Lagendijk and Cornford
(2000, p. 217) policy actors applying the cluster concept are more concerned about defin-
ing the concept than their counterparts using the learning region concept. The latter seems
to be much more an undefined buzzword in the regional development industry.
Although clusters clearly can be criticized as being vague, ambiguous and even suppor-
tive to specialization and lock-ins (Martin & Sunley, 2003), they have some strengths that
compensate for the weaknesses learning regions have concerning the unlocking of
regional economies from path dependency. First, they are less normative and more
process-oriented in character. Secondly, they are geographically less confined to politi-
cal-administrative borders. The supporting institutions of a cluster strategy follow their
companies along the lines of the global production network. At the same time,
however, clusters have a significant weakness that could be compensated for by the
532 Robert Hassink
learning region strategy. The cluster approach lacks namely strategies on how to prevent
the emergence of negative lock-ins and path dependency in regional economies. The
learning region strategy, in contrast, does pay attention to these dangers, as it applies
lock-in avoiding principles. The main actors in such a strategy share a reflective attitude
and are prepared to change a winning team. The author believes that by developing learn-
ing cluster strategies in future research, most of the learning region’s weaknesses dis-
cussed in this paper can be alleviated, without losing its main strengths, that is its
relevance for regional innovation policy-makers and its institutional capabilities to
avoid the emergence of negative lock-ins.
Acknowledgements
The first draft version of this paper was presented at the conference Regionalization of
Innovation Policy—Options and Experiences, 4 – 5 June 2004, Berlin. During that confer-
ence the author learnt a lot from the criticism of Philip Cooke, David Charles and Franz
To
¨dtling on the content of this paper. In addition, the author is grateful to the two anon-
ymous referees for their useful comments on this paper. The usual disclaimers apply,
however, and responsibility for the paper’s content rests solely with the author.
References
Bathelt, H. & Depner, H. (2003) Innovation, Institution und Region: Zur Diskussion u
¨ber nationale und regionale
Innovationssysteme, Erdkunde, 57, pp. 126– 143.
Bathelt, H., Malmberg, A. & Maskell, P. (2004) Clusters and knowledge: Local buzz, global pipelines and the
process of knowledge creation, Progress in Human Geography, 28, pp. 31 –56.
Benner, C. (2003) Learning communities in a learning region: The soft infrastructure of cross-firm learning
networks in Silicon Valley, Environment and Planning A, 35, pp. 1809 – 1830.
Blotevogel, H. H. (1999) Zur Neubewertung der Region fu
¨r die Regionalentwicklung und Regionalpolitik, in:
ARL (Ed.) Europa
¨ische Einflu
¨sse auf die Raum- und Regionalentwicklung am Beispiel des Naturschutzes,
der Agenda 2000 und des regionalen Milieus, pp. 44– 60 (Hannover: ARL).
Boekema, F., Morgan, K., Bakkers, S. & Rutten, R. (2000) Introduction to learning regions: A new issue for
analysis, in: F. Boekema, K. Morgan, S. Bakkers & R. Rutten (Eds) Knowledge, Innovation and Economic
Growth: The Theory and Practice of Learning Regions, pp. 3 – 16 (Cheltenham/Northampton: Edward
Elgar).
Boschma, R. (2003) An evolutionary approach to the restructuring of old industrial regions in Europe. Paper pre-
sented at the International Workshop The Restructuring of Old Industrial Areas in Europe and Asia, Bonn.
Boschma, R. & Lambooy, J. (1999) The prospects of an adjustment policy based on collective learning in old
industrial regions, GeoJournal, 49, pp. 391–399.
Butzin, B. (2000) Netzwerke, Kreative Milieus und Lernende Region: Perspektiven fu
¨r die regionale Entwick-
lungsplanung, Zeitschrift fu
¨r Wirtschaftsgeographie, 44, pp. 149– 166.
Cho, M.-R. (2004) Locking out through high-road restructuring—the case of textile industry in Daegu, Korea
(Seoul: Dankook University, unpublished manuscript).
Cooke, P. (Ed.) (1995) The Rise of the Rustbelt (London: UCL Press).
Cooke, P. & Morgan, K. (1998) The Associational Economy; Firms, Regions, and Innovation (Oxford: Oxford
University Press).
Cumbers, A. & MacKinnon, D. (2004) Introduction: Clusters in urban and regional development, Urban Studies,
41, pp. 959–969.
Edquist, C. (1997) Systems of innovation approaches—their emergence and characteristics, in: C. Edquist (Ed.)
Systems of Innovation: Technologies, Institutions and Organizations, pp. 1 –35 (London/Washington, DC:
Pinter).
Eich-Born, M. & Hassink, R. (2005) On the battle between shipbuilding regions in Germany and South Korea,
Environment and Planning A, 37, pp. 635–656.
How to Unlock Regional Economies from Path Dependency? 533
Enright, M. (2003) Regional clusters: What we know and what we should know, in: J. Bro
¨cker, D. Dohse &
R. Soltwedel (Eds) Innovation Clusters and Interregional Competition, pp. 99–129 (Berlin: Springer).
Fritsch, M. (2004) Von der innovationsorientierten Regionalfo
¨rderung zur regionalisierten Innovationspolitik,
in: M. Fritsch (Ed.) Marktdynamik und Innovation, pp. 105 – 127 (Berlin: Dunker & Humblot).
Fromhold-Eisebith, M. (1995) Das ‘kreative Milieu’ als Motor regionalwirtschaftlicher Entwicklung: Forschung-
strends und Erfassungsmo
¨glichkeiten, Geographische Zeitschrift, 83, pp. 30– 47.
Fu
¨rst, D. (2001) Die ‘learning region’—Strategisches Konzept oder Artefakt?, in: H.-F. Eckey, D. Hecht,
M. Junkernheinrich, H. Karl, N. Werbeck & R. Wink (Eds) Ordnungspolitik als konstruktive Antwort auf
wirtschaftspolitische Herausforderungen, pp. 71– 89 (Stuttgart: Lucius & Lucius).
Geenhuizen, M. van (1999) De lerende regio: hype of nieuw perspectief?, Geografie, 8, pp. 12 – 16.
Gertler, M. S. (1996) Barriers to technology transfer: Culture and the limits to regional systems of innovation.
Paper presented at the EU-RESTPOR Conference ‘Global Comparison of Regional RTD and Innovation
Strategies for Development and Cohesion’, 19–21 September, Brussels.
Gertler, M. S. & Wolfe, D. A. (2004) Ontario’s regional innovation system: The evolution of knowledge-based
assets, in: P. Cooke, M. Heidenreich & H.-J. Braczyk (Eds) Regional Innovation Systems: The Role of
Governances in a Globalized World, pp. 91–124 (London/New York: Routledge).
Gordon, I. R. & McCann, P. (2000) Industrial clusters: Complexes, agglomeration and/or social networks?,
Urban Studies, 3, pp. 513–532.
Grabher, G. (1993) The weakness of strong ties; the lock-in of regional development in the Ruhr area, in:
G. Grabher (Ed.) The Embedded Firm; on the Socioeconomics of Industrial Networks, pp. 255 – 277
(London/New York: Routledge).
Hamm, R. & Wienert, H. (1989) Strukturelle Anpassung altindustrieller Regionen im internationalen Vergleich
(Essen: RWI).
Hassink, R. (1997) What distinguishes ‘good’ from ‘bad’ industrial agglomerations?, Erdkunde, 51, pp. 2–11.
Hassink, R. (2001) The learning region: A fuzzy concept or a sound theoretical basis for modern regional
innovation policies?, Zeitschrift fu
¨r Wirtschaftsgeographie, 45, pp. 219– 230.
Hassink, R. & Lagendijk, A. (2001) The dilemmas of interregional institutional learning, Environment and
Planning C: Government and Policy, 19, pp. 65–84.
Hassink, R. & Shin, D.-H. (2005) Guest editorial: The restructuring of old industrial areas in Europe an Asia,
Environment and Planning A, 37, pp. 571–580.
Herrigel, G. (2004) Space and Governance in New Old Economy Manufacturing Industries (Bonn: SECONS
Discussion Forum No. 7).
Hudson, R. (1999) The learning economy, the learning firm and the learning region: A sympathetic critique of the
limits to learning, European Urban and Regional Studies, 6, pp. 59 – 72.
Iammarino, S. (2005) An evolutionary integrated view of regional innovation systems: concepts, measures and
historical perspectives, European Planning Studies, 13(4), pp. 497–520.
Koschatzky, K. (2001) Ra
¨umliche Aspekte des Innovationsprozesses (Mu
¨nster: LIT Verlag).
Koschatzky, K. (2005) The regionalization of innovation policy: New options for regional change?, in: P. Shapira
& G. Fuchs (Eds) Rethinking Regional Innovation and Change: Path Dependency or Regional Break-
through, pp. 291–312 (New York: Springer).
Kunzmann, K. R. & Tata, L. (2003) Learning City Initiatives: A Review (Dortmund: Department of Spatial
Planning in Europe, University of Dortmund (unpublished manuscript)).
Lagendijk, A. (2003) Towards a conceptual quality in regional studies: The need for subtle critique—a response
to Markusen, Regional Studies, 37, pp. 719– 727.
Lagendijk, A. & Cornford, J. (2000) Regional institutions and knowledge—tracking new forms of regional
development policy, Geoforum, 31, pp. 209–218.
Landabaso, M., Oughton, C. & Morgan, K. (2001) Innovation networks and regional policy in Europe, in:
K. Koschatzky, M. Kulicke & A. Zenker (Eds) Innovation Networks: Concepts and Challenges in the
European Perspective, pp. 243–273 (Heidelberg: Physica Verlag).
Lorenzen, M. (2001) Localized learning and policy: Academic advice on enhancing regional competitiveness
through learning, European Planning Studies, 9, pp. 163–185.
Lovering, J. (1999) Theory led by policy? The inadequacies of the ‘new regionalism’ (illustrated from the case of
Wales), International Journal of Urban and Regional Research, 23, pp. 379 – 395.
Lundvall, B.-A
˚. (1996) The Social Dimension of the Learning Economy (Aalborg: DRUID Working Paper No.
96-1).
534 Robert Hassink
Malmberg, A. (1997) Industrial geography: Location and learning, Progress in Human Geography, 21,
pp. 573–582.
Markusen, A. (2003) Fuzzy concepts, scanty evidence and policy distance: The case for rigour and policy
relevance in critical regional studies, Regional Studies, 33, pp. 869–886.
Martin, R. (2001) Geography and public policy: The case of the missing agenda, Progress in Human Geography,
25, pp. 189–210.
Martin, R. & Sunley, P. (2003) Deconstructing clusters: Chaotic concept or policy panacea?, Journal of Economic
Geography, 3, pp. 5–36.
Maskell, P. & Malmberg, A. (1999) Localised learning and industrial competitiveness, Cambridge Journal of
Economics, 23, pp. 167–185.
Morgan, K. (1997) The learning region: Institutions, innovation and regional renewal, Regional Studies, 31,
pp. 491–503.
Morgan, K. & Nauwelaers, C. (1999) A regional perspective on innovation: From theory to strategy, in:
K. Morgan & C. Nauwelaers (Eds) Regional Innovation Strategies; The Challenge for Less-Favoured
Regions, pp. 1–18 (London: The Stationary Office and Regional Studies Association).
Moulaert, F. & Sekia, F. (2003) Territorial innovation models: A critical survey, Regional Studies, 37,
pp. 289–302.
OECD (2001) Cities and Regions in the New Learning Economy (Paris: OECD).
Oinas, P. & Virkkala, S. (1997) Learning, competitiveness and development; reflections on the contemporary
discourse on ‘learning regions’, in: H. Eskelinen (Ed.) Regional Specialisation and Local Environment—
Learning and Competitiveness, pp. 263– 277 (Stockholm: NordREFO).
Pommeranz, J. O. (2000) Lernende Region Ruhrgebiet—eine regionalpolitische Leitperspektive fu
¨r das 21.
Jahrhundert, Zeitschrift fu
¨r Wirtschaftsgeographie, 44, pp. 183– 200.
Porter, M. E. (2000) Locations, clusters and company strategy, in: G. L. Clark, M. Feldman & M. Gertler (Eds)
The Oxford Handbook of Economic Geography, pp. 253 – 274 (Oxford: Oxford University Press).
Ro
¨ller, L.-H. & Hirschhausen, C. von (1996) State Aid, Industrial Restructuring and Privatization in the New
German La
¨nder: Competition Policy with Case Studies of the Shipbuilding and Synthetic Fibres Industries,
Discussion Paper FS IV 96-13 (Berlin: WZB Social Science Research Centre).
Schamp, E. W. (2000) Vernetzte Produktion: Industriegeographie aus institutioneller Perspektive (Darmstadt:
Wissenschaftliche Buchgesellschaft).
Scheff, J. (1999) Lernende Regionen: regionale Netzwerke als Antwort auf globale Herausforderungen (Wien:
Linde).
Shapira, P. & Fuchs, G. (Eds) (2005) Rethinking Regional Innovation and Change: Path Dependency or Regional
Breakthrough (New York: Springer).
Steiner, M. & Hartmann, C. (1998) Learning with clusters: A case study from Upper Styria, in: M. Steiner (Ed.)
Clusters and Regional Specialisation, pp. 211–225 (London: Pion).
Storper, M. (1997) The Regional World: Territorial Development in a Global Economy (New York/London:
Guilford Press).
To
¨dtling, F. & Trippl, M. (2004) Like phoenix from the ashes? The renewal of clusters in old industrial areas,
Urban Studies, 41, pp. 1175–1195.
Tunzelmann, N. von (2004) Network alignment in the catching-up economies of Europe, in: F. McGowan,
S. Radosevic & N. von Tunzelmann (Eds) The Emerging Industrial Structure of the Wider Europe,
pp. 23–37 (London/New York: Routledge).
Wink, R. (2003) Transregionales Lernen—Leerformel oder regionalpolitische Zukunftsstrategie, Gesellschaft fu
¨r
Regionalforschung, Seminarbericht, 46, pp. 1– 17.
How to Unlock Regional Economies from Path Dependency? 535
