Blackwell Publishing, Ltd.
 Globalizing regional development: a
global production networks perspective
Neil M Coe*, Martin Hess*, Henry Wai-chung Yeung ,
Peter Dicken* and Jeffrey Henderson!
Recent literature concerning regional development has placed significant emphasis on
local institutional structures and their capacity to  hold down the global. Conversely,
work on inter-firm networks  such as the global commodity chain approach  has
highlighted the significance of the organizational structures of global firms production
systems and their relation to industrial upgrading. In this paper, drawing upon a
global production networks perspective, we conceptualize the connections between
 globalizing processes, as embodied in the production networks of transnational
corporations, and regional development in specific territorial formations. We delimit
the  strategic coupling of the global production networks of firms and regional
economies which ultimately drives regional development through the processes of
value creation, enhancement and capture. In doing so, we stress the multi-scalarity of
the forces and processes underlying regional development, and thus do not privilege
one particular geographical scale. By way of illustration, we introduce an example
drawn from recent research into global production networks in East Asia and Europe.
The example profiles the investments of car manufacturer BMW in Eastern Bavaria,
Germany and Rayong, Thailand, and considers their implications for regional
development.
key words globalization global production networks regional development
Asia Europe
*School of Environment and Development, University of Manchester, Oxford Road, Manchester M13 9PL
email: [e-mail]
Department of Geography, National University of Singapore
! Manchester Business School, and the School of Environment and Development, University of Manchester
revised manuscript received 10 May 2004
emphasis on endogenous institutional structures
Introduction and theoretical context
and their capacity to  hold down global networks
One of the many paradoxes of the processes of (for overviews see MacLeod 2001a; Scott 1998;
 globalization is the continued significance of Storper 1997). The other strand, focusing specifically
 regions , in the sense of sub-national spaces as foci on inter-firm networks and global commodity/value
of economic activity. Systemic processes of rapid chains (GCCs/GVCs), considers the organizational
technological change, enhanced capital mobility structures of global firms production systems and
and neoliberally inspired inter-regional competition explores how particular regions  slot into these
for investment have focused attention on the need networks with varying impacts on industrial
for regional-level interventions among a broad upgrading (see Gereffi and Kaplinsky 2001; Gereffi
community of academics and policymakers. Two 1994 1996).
recent strands of work attempt to tackle the links In their early formulations, both of these literatures
between globalization dynamics and notions of could be criticized for their failure to effectively
 regional development . One strand places particular conceptualize regional economic development in
Trans Inst Br Geogr NS 29 468 484 2004
ISSN 0020-2754 © Royal Geographical Society (with The Institute of British Geographers) 2004
 Globalizing regional development 469
an era of globalization. The new regionalism liter- and markets. These relations may be with other
ature seemed overly pre-occupied with local trans- regions within the same national territory, but
actions and institutional forms at the expense of increasingly occur at the international scale. Hence,
the many extra-local connections within which our conceptualization of a region is not as a tightly
regions are embedded, while the functional con- bounded space, but as a porous territorial forma-
nections between seemingly desirable regional tion whose notional boundaries are straddled by a
institutional configurations and actual levels of broad range of network connections (Amin 2002;
economic development were open to question Allen et al. 1998).
(Amin and Thrift 1994). The GCC/GVC approaches, The paper is organized into two main sections.
in turn, operated largely at the national scale, First, we explore how the strategic coupling of
saying little about how particular sub-national global production networks and regional assets
spaces and their institutions are integrated into, may (or may not, depending on the context) facili-
and shaped by, transnational production systems tate the processes of the creation, enhancement and
(for recent critiques, see Henderson et al. 2002; capture of value upon which regional development
Smith et al. 2002; Dicken et al. 2001). ultimately depends. Second, in order to illustrate
Recent developments in these two fields, how- how our conceptual framework might be utilized
ever, have begun to address these shortcomings empirically, we present an illustrative case study of
and to move somewhat closer together. The  new the German car manufacturer BMW and its interac-
regionalism literature, for example, places increased tions with regional development processes in East-
weight on the extra-local dynamics shaping eco- ern Bavaria, Germany and Rayong, Thailand.
nomic growth within regions (both knowledge,
capital and labour flows and also the wider institu-
 Globalizing regional development:
tional structures within which regions are embedded
towards a re-conceptualization
(e.g. Amin 2002; MacKinnon et al. 2002; Bunnell
and Coe 2001; MacLeod 2001a; Lovering 1999). In developing a broad conceptual framework for
Conversely, a number of GCC and GVC studies understanding regional development we need to
explicitly explore how regional clusters and indus- pay analytical attention both to endogenous growth
trial districts are incorporated into global production factors within specific regions and also to the
systems, and consider their implications for local eco- strategic needs of trans-local actors coordinating
nomic development and industrial upgrading (e.g. global production networks (cf. Scott and Storper
Bair and Gereffi 2001; Gereffi et al. 2001; Humphrey 2003). In our framework, regional development is
2001; Sturgeon 2001; Humphrey and Schmitz 2000). conceptualized as a dynamic outcome of the complex
In this paper, we seek to make a primarily con- interaction between territorialized relational networks
ceptual contribution to these converging research and global production networks within the context of
agendas. Drawing upon a global production net- changing regional governance structures. In that sense,
works (GPN) perspective (see Henderson et al. it resonates with Amin s topological/relational
2002), and deriving insights from both the new view (Amin 2002; see also Dicken 2004). We aim
regionalist and GCC and GVC literatures, our to specify the interactive complementarity and
approach focuses on the dynamic  strategic cou- coupling effects between localized growth factors
pling of global production networks and regional and the strategic needs of trans-local actors in
assets, an interface mediated by a range of institu- propelling regional development. We argue that it
tional activities across different geographical and is these interactive effects that contribute to regional
organizational scales.1 Our contention is that development, not inherent regional advantages
regional development ultimately will depend on or rigid configurations of globalization processes.
the ability of this coupling to stimulate processes of Despite certain path-dependent trajectories, regional
value creation, enhancement and capture. development remains a highly contingent process
We regard regional development as a set of rela- that cannot be predicted a priori. This conceptual-
tional processes (see Amin 2002). It is also, by defini- ization, however, does not mean that regional
tion, an interdependent process (Massey 1984). The institutions are unimportant. On the contrary.
fortunes of regions are shaped not only by what is Often, such complementarity and coupling effects
going on within them, but also through wider sets can be enhanced and exploited through particular
of relations of control and dependency, of competition sets and practices of  regional institutions. The
470 Neil M Coe et al.
Figure 1 A framework for analysing regional development and global production networks
term  regional must be used with care here. We sufficient, to generate regional growth in an era in
place it in scare quotes to indicate that, in reality, which competition is increasingly global. There is
regional development is not just shaped by regionally no doubt that, for development to take place, a
specific institutions, but also by a variety of extra- region must benefit from economies of scale and
local institutions (e.g. national, supra-national) that scope derived from what Storper (1997, 26) terms
will impact on activities within a region. This  scaling the  holy trinity of technology organization
of institutional influence is critical. In short, regional territory. In Figure 1, we use the term  regional
development at any particular historical moment assets to describe this necessary precondition for
requires the necessary co-presence of three inter- regional development. In general, these assets can
related sets of conditions: produce two types of economies. First, economies
of scale can be achieved in certain regions through
1 the existence of economies of scale and scope
highly localized concentrations of specific know-
within specific regions;
ledge, skills and expertise. This concentration of
2 the possibility of localization economies within
technological advantages embodied in and per-
global production networks; and
formed by social actors located in specific regions
3 the appropriate configurations of  regional
creates economies of scale in particular technolo-
institutions to  hold down global production
gies that can be exploited through the agglomeration
networks and unleash regional potential.
of firms that in turn provide employment and
We have summarized these conditions and their generate economic outputs within similar high
interactions in Figure 1. tech industries. Second, economies of scope can
exist if these regions are able to reap the intangible
benefits of learning and the cooperative atmo-
Regional advantages, global production networks
sphere embedded in these agglomerations. These
and economies of value.
Our analytical framework starts with the premise are famously known as  spillover effects. A variety
that endogenous factors are necessary, but in- of different high value-added activities may be
 Globalizing regional development 471
Table I Local and non-local dimensions of regional development
Dimensions Local manifestations Non-local forms
Firms Indigenous SMEs Global corporations
Industrial clusters Entrepreneurial subsidiaries
Intra-regional markets Distant global markets
Venture capitalists Decentralized business and financial networks
Global production networks
Labour Skilled and unskilled workers Skilled experts and technologists
Permanent migrants Transient migrants
Transnational business elites
Technology Spillover effects Global standards and practices
Tacit knowledge Intra-firm R&D activities
Infrastructure and assets Technological licensing
Strategic alliances
Institutions Conventions and norms Labour and trade unions
Growth coalitions Business associations
Local authorities National agencies and authorities
Development agencies Inter-institutional alliances
Supranational and international organizations
located or developed in these regions because the relationships, but also integrate regional and
tendencies towards learning and cooperation national economies in ways that have enormous
facilitate a broad spectrum of production and implications for their developmental outcomes. At
entrepreneurial activities. the same time, the precise nature and articulation
We argue that economies of scale and scope of firm-centred production networks are deeply
embedded within specific regions are only advan- influenced by the concrete socio-political contexts
tageous to those regions  and bring about regional within which they are embedded. The process is
development  insofar as such region-specific especially complex because while the latter are
economies can complement the strategic needs of essentially territorially specific (primarily, though
trans-local actors situated within global production not exclusively, at the level of the nation-state and/
networks. As shown in Figure 1, when such a com- or the region), the production networks themselves
plementary effect exists between regions and global are not. Global production networks  cut through
production networks, a coupling process will take national and regional boundaries in highly differ-
place through which the relational advantages of entiated ways, influenced in part by regulatory and
regions interact with the strategic needs of actors in non-regulatory barriers and local socio-cultural
these global production networks. Regional devel- conditions, to create structures that are  discontinu-
opment thus depends on such a coupling process ously territorial (see Henderson et al. 2002; Dicken
that evolves over time in relation to the rapidly and Malmberg 2001).
changing strategic needs of global production Put in these conceptual terms, it becomes clear
networks and the rather slow transformations in that local actors in specific regions (e.g. labour and
regional economies of scale and scope. Before we the state) and non-local actors in global production
analyse such a coupling process, it is important to networks (e.g. TNCs and financial capital) are dif-
unpack what we mean by the strategic needs of ferentiated by their degree of territorial embedded-
actors in global production networks. We define ness which, in turn, will have very significant
global production networks as the globally organized implications for regional development (see Table I).
nexus of interconnected functions and operations This distinction in territorial embeddedness is
by firms and non-firm institutions through which important because it shapes how value and power
goods and services are produced and distributed. are distributed in their relational interactions, a
Such networks not only integrate firms (and parts point we develop in the next section (see also
of firms) into structures which blur traditional Hudson 2001).
organizational boundaries through the develop- As key local actors in regional development, the
ment of diverse forms of equity and non-equity organizational strength and flexibility of labour is
472 Neil M Coe et al.
critical to the alignment of the region with the stra- have strongly embedded local labour markets,
tegic needs of focal firms in global production net- we argue that focal firms in global production
works. While labour has been internationalizing networks can better exploit economies of scale
through inter-institutional alliances and interna- through technology- or expertise-specific produc-
tional organizations (see Table I), the reality tion systems (e.g. in biotechnology or cultural
remains that in most cases labour is spatially industries). In regions with more flexible labour
entrapped in local labour markets (see Herod 2001; markets, economies of scope might be better
Peck 1996). To Castree et al., workers face a particu- achieved through the co-presence of a variety of
lar kind of geographical dilemma because  what different industries that reap the benefits of
might make sense for them at one geographical  untraded interdependencies . The role of state
scale may have unfortunate consequences at other institutions is important here through their regula-
scales (2004, 119). They thus recognize that tion of labour and its organizations. In some
regions, state institutions may work with labour
it s not just that workers may be tempted to put local
organizations and labour market intermediaries to
interests first, it s also that the very nature of local
increase the skill levels of labour and the flexibility
interests varies depending on the specifics of local
of local labour markets (see Benner 2003; Peck
industry, local standards of living, local living wages
2000; Jones 1999). In other regions, the adversarial
and so on. (Castree et al. 2004, 120; original emphasis)
and confrontational relationship between the state
In short, there is a prima facie case that economies of and labour may significantly reduce the region s
scale and scope in particular regions can be reaped attractiveness to focal firms in global production
more effectively by focal firms in global production networks (see Kelly 2002).
networks through labour s spatial immobility Before we move on to unravel the complexity
and flexibility in skills. The local and the regional behind the strategic considerations of focal firms in
become the most important geographical scales global production networks, it is useful to consider
through which labour interacts with the strategic one significant category of non-local actors that
needs of key actors in global production networks. impact significantly on local and regional develop-
Their interactive effects tend to favour trans-local ment: financial capital institutions. While global
actors embedded in these global production production networks may not directly encapsulate
networks because, as pointed out some time ago by financial capital in their network configuration, it
Massey (1984) among others, these global actors is useful to distinguish three types of financial
can engage in  spatial switching much more easily capital in relation to their differential territorial
than workers themselves. embeddedness: local venture capital, national bank-
Similarly, the state and its development agencies ing institutions and globally decentralized financial
are institutions that are strongly embedded locally networks (see Table I). From the perspective of
in specific regions (see Table I again). This institu- global production networks, venture capital tends
tional dimension of regional development has been to be highly localized primarily because talents
well theorized in the new regionalism literature and expertise are often embodied in people within
(e.g. MacLeod 2001b). It is sufficient to say that the a particular region that are known to venture
increasing devolution of political and economic capitalists through interpersonal networks of rela-
authority from the nation state to local and tionships. Venture capital is important to regional
regional institutions has led not only to the rise of development both in terms of its financing of high
growth coalitions within specific regions, but also risk ventures that are more likely to be at the cut-
to a higher degree of uneven regional develop- ting edge of technological development and in
ment. The latter phenomenon occurs primarily terms of its financing of supporting industries that
because different regions have very different con- supply to global production networks. The nature
figurations of state institutions that in turn shape and organization of local venture capital, however,
how these regions are articulated into global pro- is embedded within national banking systems.
duction networks. This situational power and role In some countries, venture capital is much less
of the state (and labour) and its manifestations in active because of the close relationships between
local and regional institutions has very important banks and industries (e.g. Germany and Japan).
implications for understanding the distributional Regional developmental trajectories are highly
aspects of regional development. In regions that dependent on the direction and influence of national
 Globalizing regional development 473
banking institutions (see Dore 2000; Pauly and may be used for R&D, sourcing, production and
Reich 1997). In other countries (e.g. the US and the marketing activities. These different firms often
UK), banking institutions play much less signifi- offer learning and knowledge possibilities that are
cant roles vis-Ä…-vis globally decentralized financial not available if the function is performed by a
networks that are mediated through global finan- single firm, as in the practice of global sourcing or
cial centres (e.g. New York and London). Regional R&D. As Nohria and Ghoshal (1997) argue, many
development in these countries is much less leading global corporations are increasingly tap-
dependent on the presence of banking institutions ping into differentiated advantages among differ-
and more on the articulation of those regions into ent subsidiaries and supplier networks.
global financial networks. For example, the avail-
ability of investment and equity funds has been crit- Regional development and notions of value
ical to the continuous growth and development of
creation, enhancement and capture
Silicon Valley. Such funds emerge from a variety of How then is this complex organization of different
financial networks that are decentralized in terms actors in global production networks related to
of their origin and composition (e.g. US pension regional development? In Figure 1, this relationship
funds vs Taiwanese private capital). We argue can work through the creation, enhancement and
that the uneven access to these local and non-local capture of value. Here we use the term  value to
forms of financial capital can both enhance the refer to various forms of economic rent (Kaplinsky
strategic importance of some regional economies to 1998) that can be realized through market as well
global production networks and diminish others. as non-market transactions and exchanges. Along-
These different forms of capital also embody differ- side value creation through the labour process, for
ent territorial logics, with venture capital being instance, value can take the form of technological
mostly local in its orientation, and decentralized rents by way of access to particular product or
financial networks more global in nature (see also process technologies, or may be manifested as
Clark et al. 2002). relational rents, based on inter-organizational links
Hence, spanning national boundaries and mar- improving know-how transfer and collective
ket areas, the strategic needs of focal firms  defined learning. Other forms of rent identified by Kaplinsky
as dominant firms spearheading the global organi- may derive from organizational attributes, trade
zation of production networks through their cor- policy and branding. This conception of value as
porate and market power  in global production economic rent has two significant implications for
networks do not always and necessarily intersect analysing regional development. First, different
with regional advantages. Global integration of forms of rent can be created and captured by local
activities within these production networks, for and non-local actors in global production networks
example, may not be beneficial to some regions such that some regions might be better in creating
because of the likelihood of greater external control and retaining a particular form of rent (e.g.
of the regional economy. Indeed, many focal firms technological rents in Silicon Valley vs brand-name
in global production networks may pursue differ- rents in London). A region needs neither to create
ent organizational configurations in order to reap nor to retain all forms of rent. Instead, a region that
economies of scale and scope in these networks. is endowed with certain configurations of labour,
In general, economies of scale in global production capital and state institutions might be better off by
networks can be achieved through globally inte- specializing and being competitive in one kind of
grated R&D, sourcing, production and marketing economic rent  thus reaping economies of scale. A
activities that take place only in specific locations. region with a highly competitive labour market, an
The smaller the number of firms engaging in each active pool of venture capitalists and a pro-growth
of these functions, the greater the economies of coalition of institutions will likely be engaged in
scale will be in a particular global production the creation of value through new growth industries
network. This is because each of these firms can (e.g. biotechnology) that require rapid flows of
specialize in the designated function, e.g. R&D or knowledge embodied in the local workforce, high
assembly operations. Economies of scope in global risk-taking financing and a stable institutional
production networks, on the other hand, exist environment. On the other hand, a region burdened
through differentiation in the functional activities by a weakly organized and abundant supply of
of firms in the network such that a variety of firms labour, the virtual absence of venture and banking
474 Neil M Coe et al.
capital and an unstable institutional structure may hinder the development of alternative mode of
create value through performing highly labour- learning, say, through decentralized and distanci-
intensive work for focal firms in global production ated networks facilitated by greater mobility of
networks. Endowed with different configurations actors and a series of other technologies of contact
of assets, both regions perform very different and translation (see Coe and Bunnell 2003; Amin
roles in terms of value creation vis-Ä…-vis global 2002; Bunnell and Coe 2001).
production networks. Hence, regional assets can become an advantage
Second, it should be noted that value takes on for regional development only if they fit the strate-
different forms in this spatialized network of flows. gic needs of global production networks. The pro-
At the time when value is created in one region, it cess of  fitting regional assets with strategic needs
may take a particular form of, say, relational rent of global production networks requires the pres-
in regions embued with a Marshallian-style ence of appropriate institutional structures that
cooperative atmosphere. When this value is trans- simultaneously promote regional advantages and
ferred through global production networks to enhance the region s articulation into global pro-
other regions, it may take on other forms (e.g. duction networks (see Figure 1). Again, it is crucial
technological and/or brand-name rents). This to remember that our notion of  regional institu-
multiplicity in the forms of rent indicates that the tions includes not only regionally specific institu-
analysis of value creation and capture in regional tions, but also local arms of national/supranational
development must go beyond simply tracking mar- bodies (e.g. a trade union s  local chapters), and
ket values of goods and services produced. More extra-local institutions that affect activities within
importantly, we need to unpack the different forms the region without necessarily having a presence
of rent that these values encapsulate. (e.g. a national tax authority). These regional insti-
The fact that a region is  plugged into global tutions are important because they can provide
production networks, therefore, does not automati- the  glue that ties global capital and unleashes
cally guarantee its positive developmental outcome regional potential. Three dimensions of such insti-
because local actors in this region may be creating tutional structures are crucial to regional develop-
value that does not maximize the region s eco- ment. The first dimension involves the creation of
nomic potential. A region filled with cooperative value through the efforts of regional institutions in
atmosphere should be much more successful in attracting the location of value-added activities,
creating relational rent, although in some cases cul- e.g. training and educating the local workforce,
tural and institutional impediments may prevent promoting start-up firms and supplier networks,
such value from being created. Local actors in a facilitating venture capital formation and encour-
region also may not be able to capture much of the aging entrepreneurial activities (see also Phelps
value created in the region (cf. Amin and Thrift and Raines 2003). Although it is often unclear
1992). From the regional development perspective, whether such a process involves too much  tying
the creation and retention of value within the the region to the value activities of particular focal
region is imperative. For example, a region may firms or global production networks (e.g. Phelps et
have an advantage in the quantity of labour, but al. 1998), the efficacy of this relational coupling
much of the value created in the utilization of this between the region and the focal firm hinges on the
abundant pool of labour may be transferred out of region s capacity to enhance and capture value
the region through the repatriation of profits (real- from the process. However, in the absence of such
ized value) and eventually the relocation of the a coupling process, the question of regional devel-
production networks to other regions. At the other opment remains a moot point since no value will
end of the value-creation spectrum, nevertheless, a be created, let alone enhanced and captured.
region with substantial  relational assets (e.g. More importantly, the second and third dimen-
cooperative learning and venture capital forma- sions refer to the capacity of regional institutions in
tion) may be successful in creating value in team- value enhancement and value capture. Value enhance-
based projects that require face-to-face interaction ment essentially involves knowledge and techno-
in spatially proximate clusters. However, such a logy transfer and industrial upgrading (from design
value creation process may run out of steam when and final production of commodities). The influ-
these highly localized conventions and norms in ence of regional institutions via government agen-
learning are so binding and constraining that they cies, trade unions, employer associations and so on
 Globalizing regional development 475
can be significant here. On the one hand, regional patterns and corporate governance. Clearly, focal
institutions may promote specific  regional assets firms in global production networks have enormous
(e.g. cooperative industrial relations) that are con- corporate control of resources through their ability
ducive to high value-added production activities to collect and process information on a global
because these activities incur high costs of fixed basis. This information asymmetry may afford very
investment (i.e. sunk costs) and are difficult to be strong bargaining positions to some focal firms
relocated within a short period of time. There is when they interact with regional institutions (Dicken
thus a mutually beneficial interaction between 1994 2003). The more a region is articulated into
regional institutions and regional assets (see Figure 1). global production networks, the more likely it is
On the other hand, regional institutions can pro- able to reap the benefits of economies of scale and
mote the value enhancement activities of focal scope in these networks, but the less likely it is able
firms in global production networks. This occurs to control its own fate. Put in terms of power and
when regional institutions are prepared to invest in control, this non-local origin of regional develop-
developing the infrastructure and human resources ment happens because
required for value enhancement (e.g. highly stable
the exercise of power in particular places may well
power supply and skilled engineers for wafer
originate beyond those places, at some other location,
fabrication). Over time, more value enhancement
yet remains part of power s active presence. In other
activities within global production networks may
words, the power relations in place are affected by what
occur in these regions when focal firms are induced
happens elsewhere and the network of connections of
to bring in their core technologies and expertise.
which it is a part. (Allen 2003, 180 1)
The development of sophisticated local supplier
networks, for example, is important in enhancing This dimension of the external control or dependency
the value activities of focal firms through  reverse of regional development has long preoccupied
transfer of local knowledge and experience (see economic geographers (e.g. Massey 1978; Dicken
Chew and Yeung 2001). In short, not all regional 1976).
assets are complementary to the enhancement of But equally, regional institutions may mobilize
value by focal firms in global production networks. their region-specific assets to bargain with these
The key issue is the appropriateness and comple- focal firms such that their power relations are not
mentarity of these assets, not their mere presence. necessarily one-way in favour of the latter. The
The third dimension of regional institutions in bargaining position of these regional institutions is
promoting regional development rests with their particularly high when their region-specific assets
capacity to ensure value capture. It is one thing for are highly complementary to the strategic needs of
value to be created and enhanced in some regions, focal firms  these regional institutions become
but it may be quite another for it to be captured for really powerful through their relational interaction
the benefit of these regions. The issues of power and with focal firms in selected global production net-
control are critical in the analysis of value capture works. For example, focal firms that are under
and the distributional aspects of regional develop- severe competitive cost pressures are more likely to
ment. While the concept of power is complex in allow for some forms of value to be captured in
social thought (see Lukes 1986), we follow Allen regions that offer not only significantly cheaper
(2003), who defines power not as a capacity or a factors of production (labour rent), but also highly
repertoire of resources possessed by actors, but cooperative labour relations. In this case, these
rather as relational effects of social interaction. To focal firms may choose to invest further to upgrade
Allen, the local workforce that may better support future
regional development through capturing skills
power as an outcome cannot and should not be  read
and technological rents. Moreover, a region can
off from a resource base, regardless of its size or scope
achieve greater value capture if the reinvestment
. . . It is, as suggested, a relational effect, not a property
of retained earnings in localized subsidiaries is
of someone or some  thing . (2003, 5)
critical to a particular function of the global pro-
Conceived as such, the role of regional institutions duction network (e.g. new process technology) and
in negotiating these issues of power and control the focal firm fails to secure further investments
with focal firms in global production networks is through globally decentralized financial networks
linked to their development policies, ownership (e.g. downturns in major stock markets). Such
476 Neil M Coe et al.
retention of value through reinvestment in local substitution effect (e.g. the development of super-
subsidiaries and/or suppliers may also be enhanced conductors) or financial instability (e.g. over-capacity
through the availability of venture capital forma- and over investment). Such a  global crisis in an
tion or favourable support from national banking industry may force the region in question to seek
institutions. alternative development pathways that, if successful,
In both scenarios, those local actors involved will lead to the end of its path dependency. In this
directly in these global production networks (e.g. sense, our framework allows for a dynamic view of
workforce, suppliers, venture capitalists) are likely regional evolution without placing too much
to benefit from the enhancement and retention of emphasis on endogenous structures that inhibit
value through skill upgrading, technological inno- change and transformations.
vation and new venture formation. For example, Second, our framework is explicitly comparative
Simmie (2003) recently found that the most innova- because an analysis of the interactive complemen-
tive firms in Europe tend to concentrate in a minor- tarity and coupling effects requires us to examine
ity of key metropolitan regions, combining a strong how such effects materialize in one region but not
local knowledge capital base with high levels of another region. All too often in the new regional-
connectivity to similar regions elsewhere in the ism literature, we have been told how one region
global economy. The likelihood of value capture in develops because of its endogenous growth factors.
specific regions is therefore greatly enhanced by a What is absent in this analytical approach is how
cooperative set of state, labour and business insti- other regions with similar growth factors either fail
tutions that offer region-specific assets to focal to develop or evolve through drastically different
firms in global production networks. As such, the trajectories. It also ignores the complex interde-
capacity of regions to capture value is a dynamic pendencies between regions that will shape regional
outcome of the complex bargaining process between development within regions. An explicitly compar-
regional institutions and focal firms in global pro- ative approach to regional development helps us
duction networks. The presence of region-specific appreciate better the critical mechanisms through
assets is only relevant in this process if these which some regions gain developmental momen-
assets are complementary to the strategic needs of tum whereas other regions miss the opportunity.
trans-local actors embedded in global production
networks.
Global production networks and regional
development: an illustrative example
What is new about  globalizing regional
Clearly, global production networks connect regions
development?
The above re-conceptualization of regional develop- in a complex and highly variegated manner and, as
ment from the global production network per- a result, the developmental outcomes of these con-
spective complements existing frameworks in at nections will differ in significant ways, depending
least two ways. First, it takes a dynamic approach on the focal firms strategies, the institutional
to analysing regional development as a  moving frameworks and sectoral/technological specifics.
target . While we recognize the path dependency We will illustrate this for two regions in Europe
in the evolution of regional assets (see Figure 1), and East Asia involved in the global production
our framework does allow for regions to break out network of the German car manufacturer BMW
of this trajectory of lock-in. This possibility occurs Group (see Figure 2) in order to show the strategic
when specific regions are confronted with economic coupling of a region s assets with the strategic needs
crises that do not necessarily originate from these of trans-regional actors. Given space constraints, this
regions. For example, a region may enjoy a relative example is meant merely to be suggestive of how
advantage in a particular global industry (e.g. our framework might be mobilized empirically.2
electronics) or segments of a global industry (e.g. In many ways, the case of BMW resembles the
integrated circuits). Even if the path dependency of strategies, locational impacts and organizational
this regional advantage has been set in motion, the characteristics of all major car manufacturers.
region can still experience major problems of However, while market entry modes and strate-
development when crises occur within the entire gies, production and lean management systems,
industry on the global scale. Such crises may be and the JIT-based clustering of suppliers around
due to technological change that produces a the main assembly plants are similar to those of
 Globalizing regional development 477
Figure 2 Global locations of BMW group
other firms in the sector, BMW has some specific the takeover opportunity  was the availability of
characteristics that distinguish it from other car skilled labour and, increasingly important, a
manufacturers. As a niche producer for the upmar- flexible workforce. Since unemployment in the
ket and luxury markets, the BMW Group has rela- region was high, the recruitment of people at
tively low production volumes and thus the global competitive wages was much easier than through
production network strategy is to some extent dif- the job market in the prospering Munich area.
ferent from that of the mass manufacturers. Also, Also, there was a willingness on the part of the
the company has a very strong base in Bavaria and workforce to accept flexible working hours and
is approximately 47 per cent owned by members of shift structures that allowed BMW to enhance
the Quandt family. This shareholding structure has capacity  for example, extending the machine time
a decisive impact on inter-firm relationships within to 99 hours a week  and thus reduce unit costs.
BMW s global production network. This was supported by cooperative workers
councils and regional labour unions, as well as
BMW in Eastern Bavaria, Germany government aid (through regional development
Economic development in eastern Bavaria, for a programmes) to boost the weak regional economy.
long time a peripheral and economically weak In the wake of BMW s 1982 decision to build
region, has been transformed  not least by the another assembly plant in Eastern Bavaria near
arrival of BMW s production plants  since the late Regensburg, intensive and continuous negotiations
1960s. Headquartered in Munich, the company was between BMW and the national labour union IG
looking for new manufacturing sites to expand Metall took place, illustrating the regional impact
production and in 1967 took over the small car of decisionmaking by, and relational power of, actors
manufacturer Hans GLAS with its production inside and outside the region. Against the opposition
facilities at Landshut and Dingolfing. The crucial of the IG Metall head office in Frankfurt, the regional
regional asset that attracted BMW to Eastern branch of the union engaged in cooperation with
Bavaria  apart from the proximity to Munich and Munich-based BMW to implement a work-shift
478 Neil M Coe et al.
model that has become a model for the German car significant buying power, BMW persuaded global
industry. A recent example of this process is the first tier suppliers like Lear Corp. and Modine
negotiations between Volkswagen AG and IG (both from the US) to establish plants next to each
Metall over a new tariff contract (cf. Pries 2002). other in the innovation park. This not only guaran-
So far, BMW has invested some 7 billion Euro teed the functioning of the production network,
and now operates three plants and one supplier but created an innovative context, where suppliers
park in the area, directly employing about 35 000 (some of them competitors on the world markets)
people, whilst an estimated further 20 000 jobs share tacit knowledge and continuously improve
have been created by local first-tier suppliers. This products and processes (cf. Hess 2001). That way,
represents more than one tenth of all manufactur- new innovative structures have been created
ing employment in the region. The wages and through the BMW-induced arrival of foreign firms
salaries paid to the employees at the company s in the region, contributing not only to direct
regional sites exceed 2 billion Euro per year and employment but also providing the environment
constitute a considerable source of consumer pur- for spillover effects that benefit the regional econ-
chasing power within the region. omy. Some of these global suppliers branch plants
The development of BMW s supplier and inno- have now become leading plants within their par-
vation park in Wackersdorf near Regensburg in the ent companies, setting benchmarks for other plants
1990s has increased the integration of its operations within the production network. The globalization
in Eastern Bavaria with global production net- of BMW s production network itself resulted in the
works. After the German government was forced establishment of a logistics centre at the same site
to abandon the building of a nuclear waste man- in Wackersdorf, from which all of BMW s interna-
agement facility at Wackersdorf, private companies tional parts and components distribution to its for-
were able to use the location as an industrial park, eign plants in the US, South Africa, Russia and East
administered by the agency formerly responsible Asia is organized, with a daily shipment of 2.5
for setting up the nuclear site. Investors were million parts carried out by a third party logistics
attracted by the low prices of land available and provider.
about 500 million Euro in compensation payments Eastern Bavaria s regional economy has, without
to the region, paid by the federal and state govern- doubt, benefited from globalizing processes linked
ments, which accelerated infrastructure develop- to the region via BMW s production network.
ment. Again, this opportunity matched BMWs Apart from value creation in the form of both
strategic needs at that time. The distribution of domestic and foreign investment, as well as direct
power between the relevant political and societal and indirect employment, the skills and technology
actors in Eastern Bavaria, however, has proven to transfer from BMW and its foreign-owned suppliers
be complex and has played out at different scales  to local companies ensures a noteworthy degree
analogous to the different scales of influence from of value enhancement and capture, which is essen-
the labour union s and workers side. The federal tial for the region s sustainable economic develop-
and state governments have been under pressure ment. The strategic coupling process between the
to find alternatives to the planned nuclear facility regional assets and BMW s needs to develop its
and thus were obliged to regional political and GPN has not only led to repeated rounds of capital
civil actors, improving the bargaining power of the and technological investments, but also provided
latter. BMW could reap the benefits from this polit- opportunities for regional actors to capture differ-
ical struggle and implement their plans in the ent forms of rent, notably organizational and
region at comparatively lower costs. relational rents. The process of value creation and
The Just-In-Time system, introduced at BMW s enhancement has been mediated through ongoing
plants in Regensburg and Dingolfing, increasingly interaction and negotiation between firm, govern-
required the co-location of major suppliers, while ment and labour representatives at regional and
the modularization of production forced compo- trans-regional scales. However, as BMW is one of the
nent manufacturers to integrate and coordinate leading economic actors within a quasi-hierarchical
their business. Initially, BMW started to produce regional network, it has got considerable power
convertibles at the new Wackersdorf plant, after it vis-Ä…-vis institutions and other firms, whereby
promised to create at least 1600 jobs in the region, quite a strong dependency on its commitment to
but later changed the plans for this area. Using its the region remains.
 Globalizing regional development 479
network. Initial investments of 25 million Euro in
BMW in Rayong Province, Thailand
BMW remained, until very recently, essentially a the manufacturing facilities for the 3 series cars
German-based company. Like other major car will be topped up by an additional 15 million Euro
manufacturers, however, it has had to respond to to install a new assembly line for the production of
globalizing forces by creating a geographically 7 series cars in 2003, to be sold in the domestic and
more extensive production network (Figure 2). In regional markets. Currently, the Rayong plant of
this context, BMW s entry into East Asia is BMW employs about 250 people, assembling nearly
potentially very significant both for the company 4000 cars annually from vehicle kits imported from
itself and for the region, particularly Thailand. Germany. These rather small figures suggest a
Since the 1960s, Thailand has become the centre of negligible contribution to regional development in
Southeast Asia s automotive industry, employing Thailand s Rayong province.
about 120 000 people in the sector. Motor vehicles Indeed, while local content regulations existed
have become the third biggest export category until the year 2000, the bulk of value added parts
after computers and electrical circuits: 230 000 are not manufactured in the region, but rather
cars from a total output of 760 000 in 2003 were brought in from abroad. Since the market is not yet
exported (Bangkok Post 2003). Due to a nationally big enough, low production volumes do not allow
implemented cluster policy (cf. Lecler 2002), most for economies of scale by establishing a production
of the companies in this sector, including BMW, site, and therefore completely knocked down
are located to the south of Bangkok, in the Rayong (CKD) assembly with comparatively little regional
and Samutprakarn provinces of Thailand s Eastern value-added prevails. For a BMW car, about 40
Seaboard (see Figure 2). To date, there are almost per cent of value added is achieved through local
two-dozen car manufacturers operating in the content, but this means production and sourcing
region, surrounded by more than 700 first-tier within ASEAN countries, therefore the value
suppliers, 50 per cent of which are fully or partly added within the Thai auto cluster is lower than
foreign-owned. The Thai auto component industry the 40 per cent figure suggests. Furthermore, the
is generating an annual turnover of about US$4bn. regional assets in the South of Thailand do not nec-
In this case, the nation state plays an important essarily match all the needs of car manufacturers
role in coupling the regional assets with the like BMW, as the level of skills among the work-
strategic needs of global companies and their force and organizational sophistication have yet to
networks, not least due to the weaknesses or lack reach the required standards. The fact that most of
of regional institutions. Since BMW is one, rather the suppliers are partly or wholly owned by for-
small player among a fairly large number of eign companies reflects the problems in upgrading
companies within the Thai car industry, its power the industrial base and transferring skills and tech-
in relation to regional and national institutions is nology to local companies. There have been consid-
limited. Hence the bargaining position of these erable initiatives by the Thai government to adapt
institutions is considerable, supported by the fact to the changing strategic needs of manufacturers
that south-central Thailand has become first choice like BMW and to participate more strongly in their
location for many foreign firms and thus creates value-added networks. Through governmental and
opportunities towards the enhancement of regional quasi-governmental organizations, most notably
assets. the Board of Investment and the Thailand Automo-
Like many other car manufacturers, BMW has tive Institute, a series of attempts are being made
chosen Thailand s gulf area as its prime location to improve the vocational skills of the Thai work-
for the Southeast-Asian region, because the coun- force and to support domestic SMEs, in close
try had no national car programme and hence collaboration with foreign manufacturers (see also
this sector was fairly liberalized compared to other Lauridsen 2003; Techakanont 2003). In addition,
countries in the region, especially Malaysia and foreign assemblers train their workforce either
Indonesia (cf. Tucher 1999). In anticipation of a in-house or put them on training courses within
potentially large market after the completion of an the parent company abroad, with BMW being no
Asian Free Trade Agreement (AFTA), the rationale exception. However, know-how transfer to domes-
for production in Asia was mainly to avoid current tic suppliers is still rather limited.
tariff and non-tariff trade barriers as well as to The success of BMW s Thai venture depends,
integrate the region into BMW s global production to a great extent, on the final implementation of
480 Neil M Coe et al.
supranational economic integration under an Asian ment to create regional institutions that help to
free trade agreement. BMW has chosen the Rayong transform and enhance the regional assets, espe-
plant, its only wholly owned facility in Asia, to cially in the field of education and vocational train-
become an integrated production site once the ing, laid out in the current Automotive Industry
institutional framework (AFTA) allows. Rayong Master Plan, and progress in supra-national negoti-
will become a full production site whilst its other ations to pave the way towards an integrated
Southeast Asian locations will serve as BMW cen- production system, with the global forces of multi-
tres of excellence. In this way, economies of scale national companies like BMW and supra-national
will be achieved through production specialization economic policy arrangements being the main
and exchange within Southeast Asia as a whole. drivers for regional development in Rayong and
Other companies, e.g. Toyota, follow similar strate- Samutprakarn.
gies of regional complementation (cf. Yoshimatsu As this empirical example has illustrated,
2002), which provide an opportunity for the Thai regional development is strongly linked to external
auto cluster to upgrade and develop in the region. influences in the form of both non-firm institutions
Once a critical mass is reached, BMW will not only and economic actors (see Figure 3). While the acti-
be able to attract additional foreign suppliers to the vation of endogenous resources to foster sustaina-
region, but also will be more likely to invest more ble development is an important task for regional
in upgrading and developing local suppliers which economic policy, it is not sufficient in itself but
the company might use in the future. As it stands, rather has to take into account the strategic cou-
BMW is currently bridging the gap between its pling process between global production networks
strategic needs and the territorial assets in the form and regional assets. Figure 3 shows a strong intra-
of local supply companies through a  mediated regional connectivity between actors for the region
process of technology transfer and industrial of Eastern Bavaria, both in terms of material flows
upgrading. and technological/organizational cooperation, sup-
For example, in order to secure local sourcing for ported by regional institutions. However, the main
side glass for its E46 cars assembled in Thailand, drivers of development are extra-local, based on
BMW approached a supplier they used elsewhere. BMW s production strategy and investment, while
However, the negotiations failed and since produc- previous policy decisions and subsequent capital
tion volumes were too low to persuade its German flows from the Bavarian and Federal Governments
glass supplier to locate in Rayong, BMW went to helped kick-start regional development. The BMW-
the Thai subsidiary of a Japanese glass manufac- linked production network in the Rayong/Samut-
turer and arranged a technological cooperation prakarn area, on the other hand, currently shows
process between the German supply company and comparatively few regional linkages. Most of the
the Japanese/Thai manufacturer to upgrade and parts and components are imported from Germany
technically release their products and processes. via the Wackersdorf logistics centre, and invest-
That way, the German supplier became the main ment as well as technology transfer to Thai suppliers
facilitator of technology and know-how transfer, has so far been rather modest. Additionally, the
based on long-term relations and familiarity with future of a prospering automotive cluster at Thai-
BMW s technical and organizational standards, land s Eastern Seaboard depends to a large extent
without BMW having to deal with double invest- on supra-national free trade negotiations, making
ments. A similar, triangular technology transfer the international dimension of regional develop-
arrangement exists between BMW, one of their ment ever more obvious.
European suppliers and a domestic, fully Thai-
owned supplier. That way, technological rents are
Conclusion
generated within the Thai auto cluster, although to
date they are still rather confined to joint ventures In this paper we have proposed an integrated
and foreign-owned suppliers. conceptual framework for  globalizing regional
Summing up, in order to achieve the goal of development that takes external or  global forces
upgrading, a number of adjustments are needed to as well as regional assets into account. In short,
facilitate a positive strategic coupling process our framework highlights the dynamic  strategic
between global production networks and regional coupling of global production networks and regional
assets. These include efforts by the Thai govern- assets, an interface that is crucially mediated by a
 Globalizing regional development 481
Figure 3 BMW s GPN and regions in the EU and ASEAN
range of institutional activities across different the newly industrialized countries, national politics
spatial scales. Our key argument is that regional sets the dominant framework for regional develop-
development depends on the ability of this ment, with regional institutions often weakly
coupling mechanism to facilitate processes of value developed or completely missing. On the other
creation, enhancement and capture. Empirically, hand, in countries with a more decentralized
due to the constraints of space, we have only structure, regional institutions attempt to develop
been able to offer a brief illustration of how our their bargaining power vis-Ä…-vis focal firms in the
framework might be mobilized for a very simple context of nation-state governance structures and
global production network configuration consisting inter-regional competition. In every case, however,
of one firm s activities across two regions. the exercise of power is a multi-scalar process, with
However, the BMW case clearly shows that the varying combinations of actors cooperating or
developmental impact of the coupling process is playing off one against the other. Knowledge about
highly variable and contingent, and by no means these territorially specific power configurations,
automatically beneficial for the region. therefore, is elementary for regional institutions to
Clearly, regional development does not take take appropriate measures for transforming a
place on a level playing field. For the processes of region s assets and to maximize their bargaining
value creation, enhancement and capture to benefit power and impact.
economic development in particular regions, the More specifically, what policy lessons might be
balance of power between the different actors drawn from our analysis for actors at the regional
involved is a crucial variable in determining the level? Four are particularly pertinent, although
potential for value enhancement and, ultimately, they can only be mentioned here because of space
value capture. Governance structures in different constraints. Firstly, policymakers clearly need to
territorial contexts are variable and hence the accumulate considerable stocks of knowledge not
possibilities for developmental policies to impact only about the various assets contained within
on a region s assets will differ as well. In many of their region, but also about how they relate to the
482 Neil M Coe et al.
regional development processes. While we acknow-
needs of various global production networks,
ledge the term is imperfect and may be perceived as a
many of which will originate outside their terri-
rather crude structural interpretation of regional devel-
tory. Secondly, such knowledge is clearly sectorally
opment, it is used here for heuristic purposes and
specific, and effective policy interventions will
should be understood as the coupling process between
have to be designed for, and targeted at, particular
regional economies and global production networks
segments of particular industries. Such flexibility
that is mediated through specific action and practices
would appear to be a necessary precondition for
of key actors and institutions. In this sense, we hope
effective intervention in the contemporary era.
to offer a middle-ground interpretation that is both
Thirdly,  out there knowledge clearly needs to
structural and actor-centric.
extend beyond the needs of specific production 2 A brief note on methodology. This illustrative case
study is derived from the ESRC funded project
networks to incorporate a reflexive understanding
Making the Connections: Global Production Networks in
of the multi-scalar institutional configurations in
Europe and East Asia (Grant # R000238535) that ran
which all policymakers are situated. Some of these
from 2000 to 2003 and explored economic connections
institutional relations will pull against a particular
within and between Europe (including Central and
endeavour, whilst others will support it. An appre-
Eastern Europe) and East Asia in three sectors: auto
ciation of this complexity is a crucial first step in
components, retailing and telecommunications. Framed
appropriately harnessing (or resisting) extra-local
within the GPN conceptual framework (Henderson
connections. Finally, it is clear that policy interven-
et al. 2002), over 150 semi-structured corporate and
tions need to be underpinned by an awareness of
institutional interviews were conducted in 12 countries
the differences between value creation, enhance- in addition to an extensive programme of secondary
data and documentary analysis. The aim of the
ment and capture, so that strategic couplings that
research was both to  map out leading GPNs in these
prioritize the latter two processes can be identified
three sectors and to explore their developmental
and supported. All of these are policy issues that
impacts in different localities.
require substantial attention in future work on
regional development in a globalizing world.
References
Allen J 2003 Lost geographies of power Blackwell, Oxford
Acknowledgements
Allen J, Massey D and Cochrane A 1998 Rethinking the
This article draws on work conducted under the
region Routledge, London
auspices of the Economic and Social Research
Amin A 2002 Spatialities of globalisation Environment and
Council project, Making the Connections: Global Pro- Planning A 34 385 99
Amin A and Thrift N 1992 Neo-Marshallian nodes in glo-
duction Networks in Europe and East Asia (Grant # R
bal networks International Journal of Urban and Regional
000 238535). We are grateful to the ESRC for their
Research 16 571 87
support. Thanks also to Adam Tickell and three
Amin A and Thrift N 1994 Living in the global in
anonymous referees for their insightful comments
Amin A and Thrift N eds Globalisation, institutions and
on an earlier version of this paper.
regional development in Europe Oxford University Press,
Oxford 1 22
Bair J and Gereffi G 2001 Local clusters in global chains: the
Notes
causes and consequences of export dynamism in Torreon s
1 Our notion of  strategic coupling here may appear to blue jeans industry World Development 29 1885 1903
resonate quite closely with the coupling process Bangkok Post 2003 Year-end economic review Bangkok,
between the mode of social regulation and regime of Thailand
accumulation in the French regulationist theory of Benner C 2003 Labour flexibility and regional develop-
economic development (see Boyer 2000; Tickell and ment: the role of labour market intermediaries Regional
Peck 1992) and its refinement through Jessop s (1990 Studies 37 621 33
2001) strategic-relational approach to state theory. Boyer R 2000 The political in the era of globalization and
However, our use of the concept is not a direct import finance: focus on some Régulation school research Inter-
from the regulationist theory that focuses primarily on national Journal of Urban and Regional Research 24 274 322
the coupling process of two meta-structures within Bunnell T G and Coe N M 2001 Spaces and scales of
particular national economies (i.e. the national scale). innovation Progress in Human Geography 25 569 89
At the risk of developing a functionalist interpretation Castree N, Coe N M, Ward K and Samers M 2004 Spaces
of regional development, we use the term to charac- of work: global capitalism and geographies of labour Sage,
terize the very complex scalar juxtapositions that drive London
 Globalizing regional development 483
Chew Y-T and Yeung H W-C 2001 The SME advantage: Hudson R 2001 Producing places Guilford, New York
adding local touch to foreign transnational corporations Humphrey J 2001 Opportunities for SMEs in developing countries
in Singapore Regional Studies 35 429 46 to upgrade in a global economy (http://www.ids.ac.uk/
Clark G L, Mansfield D and Tickell A 2002 The German ids/global/pdfs/clusterchainsv4.pdf) Accessed 7 February
social market in the world of global finance: pension 2003
investment management and the limits of consensual Humphrey J and Schmitz H 2000 Governance and upgrad-
decision-making Transactions of the Institute of British ing: linking industrial cluster and global value-chain
Geographers 27 91 110 research Institute of Development Studies, University of
Coe N M and Bunnell T 2003  Spatializing knowledge Sussex, IDS Working Paper 120
communities: towards a conceptualization of transna- Jessop B 1990 State theory: putting capitalist states in their
tional innovation networks Global Networks 3 437 56 place Polity Press, Cambridge
Dicken P 1976 The multiplant business enterprise and Jessop B 2001 Institutional re(turns) and the strategic-
geographical space: some issues in the study of external relational approach Environment and Planning A 33
control and regional development Regional Studies 10 1213 35
401 12 Jones M 1999 New institutional spaces: TECs and the remak-
Dicken P 1994 Global local tensions: firms and states in ing of economic governance Jessica Kingsley, London
the global space economy Economic Geography 70 101 19 Kaplinsky R 1998 Globalisation, industrialisation and sus-
Dicken P 2003 Global shift 4th edn Sage, London tainable growth: the pursuit of the nth rent Institute of
Dicken P 2004 Geographers and  globalization : (yet) Development Studies, University of Sussex, IDS Discus-
another missed boat? Transactions of the Institute of Brit- sion Paper 365
ish Geographers 29 5 26 Kelly P F 2002 Spaces of labour control: comparative
Dicken P and Malmberg A 2001 Firms in territories: a perspectives from Southeast Asia Transactions of the
relational perspective Economic Geography 77 345 63 Institute of British Geographers 27 395 411
Dicken P, Kelly P F, Olds K and Yeung H W-C 2001 Lauridsen L 2003 The role of the state in linkage forma-
Chains and networks, territories and scales: towards a tion between TNCs and local Thai enterprises Paper
relational framework for analysing the global economy presented at the workshop on  Understanding FDI-assisted
Global Networks 1 89 112 Economic Development 22 25 May 2003 TIK-centre,
Dore R 2000 Stock market capitalism, welfare capitalism: University of Oslo
Japan and Germany versus the Anglo-Saxons Oxford Uni- Lecler Y 2002 The cluster role in the development of the
versity Press, Oxford Thai car industry International Journal of Urban and
Gereffi G 1994 The organization of buyer-driven global Regional Research 26 799 814
commodity chains: how US retailers shape overseas Lovering J 1999 Theory led by policy: the inadequacies of
production networks in Gereffi G and Korzeniewicz the  New Regionalism International Journal of Urban and
M eds Commodity chains and global capitalism Praeger, Regional Research 23 379 95
Westport CT 95 122 Lukes S ed 1986 Power Blackwell, Oxford
Gereffi G 1996 Global commodity chains: new forms of MacKinnon D, Cumbers A and Chapman K 2002 Learn-
co-ordination and control among nations and firms ing, innovation and regional development: a critical
in international industries Competition and Change 1 appraisal of recent debates Progress in Human Geography
427 39 26 293 311
Gereffi G and Kaplinsky R eds 2001 The value of value MacLeod G 2001a New regionalism reconsidered: globali-
chains: spreading the gains from globalization IDS Bul- zation and the remaking of political economic space
letin 32 International Journal of Urban and Regional Research 25
Gereffi G, Humphrey J, Kaplinsky R and Sturgeon T 804 29
2001 Introduction: globalisation, value chains and MacLeod G 2001b Beyond soft institutionalism: accumula-
development IDS Bulletin 32 1 8 tion, regulation, and their geographical fixes Environment
Henderson J, Dicken P, Hess M, Coe N and Yeung H W- and Planning A 33 1145 67
C 2002 Global production networks and the analysis of Massey D 1978 In what sense a regional problem?
economic development Review of International Political Regional Studies 13 233 43
Economy 9 436 64 Massey D 1984 Spatial divisions of labour Macmillan,
Herod A 2001 Labor internationalism and the contradic- London
tions of globalization: or, why the local is sometimes Nohria N and Ghoshal S 1997 The differentiated network:
still important in a global economy Antipode 33 407 26 organizing multinational corporations for value creation
Hess M 2001 Globalisierung und regionale Innovation- Jossey-Bass, San Francisco
snetzwerke: Beispiele für  Glokalisierung in der deut- Pauly L W and Reich S 1997 National structures and
schen Fahrzeugindustrie in Schätzl L and Grotz R eds multinational corporate behavior: enduring differences
Regionale Innovationsnetzwerke im internationalen Vergleich in the age of globalization International Organization 51
LIT, Münster 83 100 1 30
484 Neil M Coe et al.
Peck J 1996 Workplace: the social regulation of labor markets regional economies Progress in Human Geography 26 41
Guilford, New York 63
Peck J 2000 Workfare states Guilford, New York Storper M 1997 The regional world: territorial development in
Phelps N and Raines P eds 2003 The new competition for a global economy The Guilford Press, New York
inward investment: companies, institutions and territorial Sturgeon T 2001 How do we define value chains and pro-
development Edward Elgar, Cheltenham duction networks? IDS Bulletin 32 9 18
Phelps N, Lovering J and Morgan K 1998 Tying the firm Techakanont K 2003 Globalization strategy of assemblers and
to the region or tying the region to the firm: early changes in inter-firm technology transfer in the Thai auto-
observations on the case of LG in South Wales European mobile industry Working paper series vol 2003-23 The
Urban and Regional Studies 5 119 38 International Centre for the Study of East Asian Devel-
Pries L 2002 5000x5000: Ende gewerkschaftlicher Tarif- opment, Kitakyushu
politik oder innovativer betrieblich-tariflicher Sozial- Tickell A and Peck J 1992 Accumulation, regulation and
pakt? Industrielle Beziehungen 9 222 35 the geographies of post-Fordism: missing links in
Scott A J 1988 New industrial spaces Pion, London regulationist research Progress in Human Geography 16
Scott A J and Storper M 2003 Regions, globalization, 190 218
development Regional Studies 37 579 93 Tucher M v 1999 Die Rolle der Auslandsmontage in den
Simmie J 2003 Innovation and urban regions as national internationalen Wertschöpfungsnetzwerken der Automobil-
and international nodes for the transfer and sharing of hersteller VVF, Munich
knowledge Regional Studies 37 607 20 Yoshimatsu H 2002 Preferences, interests, and regional
Smith A, Rainnie A, Dunford M, Hardy J, Hudson R integration: the development of the ASEAN industrial
and Sadler D 2002 Networks of value, commodities cooperation agreement Review of International Political
and regions: reworking divisions of labour in macro- Economy 9 123 49