Abstract This article presents further opportunities to develop the Global Production Network (GPN) approach by reopening the ‘black box’ of the multinational enterprise (MNE) through a structuration perspective. It emphasises three aspects to a renewed focus on the agency of MNEs, namely: the importance of the variety of relationships within MNEs between parent and subsidiaries; the importance of dynamic capabilities in underpinning corporate change; and, the micropolitics of MNEs and subsidiaries which impact on firm-institutional change within regional economies. The agency exercised by MNEs in these ways influences the ‘selection’ of investment locations, ‘coupling’ processes, and the depth and pace of host territorial institutional change. In conclusion, this article argues that future research needs to place greater emphasis on the contribution of dynamics internal to the MNE to understand evolution in regional economies and GPNs. 1. Introduction In recent years, the Global Production Networks (GPNs) perspective has provided important insights into the geographies of contemporary capitalism (Yeung, 2016). This approach argues that firms work through networks which are a ‘nexus of interconnected functions, operations and transactions through which a specific product or service is produced, distributed and consumed’, encompassing both vertical and horizontal relations and embeddedness (Coe et al., 2008, 274). Curiously, the multinational enterprise (MNE), and its internal dynamics, have not been at the forefront of analysis. Indeed, there remain opportunities to further focus on the role of firm behaviour within GPNs, even in the recently presented GPN version 2.0 (Coe and Yeung, 2015). In this article, we argue for a need to look within the ‘black box’ of the firm and thus at the agency and practices of the MNE within GPNs, and particularly processes of coupling with ‘host’ regions where both lead and supplier (e.g. ‘strategic partners’) firms can be MNEs. In this sense, our multidimensional framework is concerned not just with lead corporations, but also incorporated firms with various roles within GPNs, and recognising they often fulfil diverse roles and strategies (Pickles et al., 2006). Renewed understanding of the MNE can be an important element in a structurationist perspective vital to understanding the (spatially configured) relations between regional economies and GPNs. Our purpose, therefore, is to outline important concepts which can contribute to understanding the ‘agency’ contained within the ‘black box’ of corporate practices in the structuration of economies by way of: (i) the corporate selection of regions for particular forms of investment; (ii) the potential for strategic coupling/decoupling; and (iii), following MacKinnon (2011), the mechanisms of co-evolution between MNEs and institutions in the processes of coupling. More specifically, our framework builds upon GPN 2.0 by recognising the importance of ‘competitive dynamics’ and ‘managing risk’ (as advocated in GPN 2.0), but focusing on the practices constituting and influencing firm strategies and cost-capability ratios, and therefore the ‘how’ of organising and governing GPNs. In the next section, we examine the conceptual insights of GPN 1.0 and 2.0 in understanding the configuration of the contemporary global economy, but identify where a greater appreciation of the agency of MNEs would further contribute to the GPN approach. The proceeding section then examines the three important elements outlined above that contribute further to GPN 2.0’s conception of firm strategies through a structuration perspective. In conclusion, we underline the value of reopening the black box of the MNE to the GPN approach through a greater appreciation of the intricacies of this particular actor. 2. GPNs 2.1. The basis of the GPN approach The GPN approach argues that there has been a disaggregation of vertically integrated production chains into individual firm specialisation within externally networked value chains, as lead MNEs have sought to develop transnational ‘technological’ and ‘organisational’ ‘fixes’ for value creation (Henderson et al., 2002; Yeung, 2009a). Value is created by efforts to increase rent, coming about through various realms of production, including technological, organisational and networked production processes (Henderson et al., 2002), as well as recognising the ‘ongoingness of value’ (Herod et al., 2014, 423). Efforts towards enhancement are based on the desire to produce greater value and thus acquire increasing returns from particular production activities (Werner, 2016). The ability to capture rents by firms and territorial regional bodies where segments of GPNs are located, relies upon the dynamics of power and embeddedness of such activities in the regional economy (Coe et al., 2004). Yet, critical within such conceptions is the need for sensitivity towards the organisational motives, decision-making arrangements and practices of MNEs as actors within GPNs, not least because it is their desire to capture rent which drives participation and power relations. GPN accounts emphasise the contingent, uneven and constantly emergent nature of such networks, which are produced through geographical, temporal and organisational arrangements (Hess and Yeung, 2006). This represents a movement away from the analysis of macro social structures to an understanding of actors and relations constituting the production of structural network arrangements within contingent spatial-temporal settings, and recognises the dynamic co-constitution of structure and agency (Bathelt and Glückler, 2005). Through such thinking the GPN approach argues that relations within networks encompass both horizontal and vertical arrangements between firm and non-firm actors (e.g. trade unions) and institutions, with networks tied together through actors and relations (Levy, 2008). These networks are understood to vary in size and complexity as arrangements in which (governing) structure, practice and agency are produced through power-laden and contested relations (Blažek, 2016). The expedient actions of actors in deliberation with others in production networks is critical, traditionally viewed in the GPN perspective by way of bargaining and the power ascribed in the possession of assets and competences (see Henderson et al., 2002). In such thinking those possessing important but scarce assets and competences have greater bargaining power, underpinning recent concerns with a ‘dark side’ of GPNs (Coe and Hess, 2011; Werner, 2016). Recent GPN accounts understand power in more pragmatist terms, recognising the experienced, expedient and experimental nature of power (see, e.g., Murphy, 2012). It is enacted by actors through particular actions within (spatially-configured) production networks and has to be experienced in order to be realised (i.e. to change the actual behaviour and intentions of other actors in accordance to those actors seeking to enact change). Rather than deriving from possession of particular resources per se, it also depends on the nature of actors, their position in networks and resources available, and how they are mobilised (Hess, 2008; Allen, 2016). Central to the GPN perspective is the understanding that there are considerable power relations arising from the ‘inter-firm division of labour’ within individual network supplier relations (Yeung and Coe, 2015). Both collaboration and various forms of power relation, such as subordination, characterise this division of labour. Such processes have a considerable influence on network relations, strategic coupling and thus regional development (Coe and Hess, 2011). For instance, those at the lower end of the inter-firm division of labour, typically located within developing or emergent regions with lenient regulatory regimes and subject to cost-based decisions within competitive markets, experience more stringent power relations through contracting, the regions concerned being vulnerable to closures or disinvestment. Value and power are related to different types of embeddedness, referring to how GPN actors and production network activities are (heterogeneously) embedded within particular territories. Three such forms of embeddedness are elucidated, including, first, societal embeddedness relating to the position and constitution of GPN actors through broader historically-contingent institutional and cultural arrangements, such as the importance of home country institutions/culture in guiding firms (Hess, 2004). Secondly, network embeddedness includes the relations, position and connectivity of actors within GPNs (Henderson et al., 2002). Finally, territorial embeddedness understands that GPN segments and actors have degrees of dependence and connectivity in particular territorialised production spaces, with such embeddedness dynamic and subject to change as firms seek greater value creation and capture (Hess, 2004). Recognising the underdevelopment of a causal framework with which to explain the complex historical, dynamic and variable organisational configuration of GPNs, Coe and Yeung (2015) have proposed a GPN 2.0 approach. The purpose of GPN 2.0 is to elucidate the causal factors contributing to the strategies and organisational configurations of production networks, and which subsequently impact on regional development, through greater explanatory analysis of actor-specific strategic responses to ‘competitive dynamics’. The latter encompass four key processes and elements in which the capital accumulation process occurs, building upon various theoretical concepts of regional development. ‘Cost-capability ratios’, considered one of the key internal dynamics of the firm in relation to the environment, is constituted by labour, capital, technology and knowledge. The purpose of the ratio is to understand the decision-making processes by which actors capture value, not simply cost-based factors (such as outsourcing), but also the capability to organise, manage and utilise resources. It therefore facilitates an understanding of the rationale behind production decisions within GPNs. The ratio is dynamic in nature, encompassing responses by firms to global competitive markets, and their efforts to continually optimise cost-capabilities. GPN 2.0 further advances our understanding of the increasingly complicated nature of markets, moving beyond a concern with the buying and selling strategies of large firms in production networks. Instead, the concept of sustainable ‘market development’ eludes to the role of producers and consumers in creating markets through processes of dynamic negotiation, but tied to their particular motives (such as in consumer demand for higher quality yet lower cost goods) (see, e.g., Smith et al. (2014) on the importance of market dynamics in influencing the Slovak clothing industry). Firm-specific strategies and GPNs are increasingly influenced by considerations of finance, encapsulated within the concept of ‘financial discipline’. Financialisation of the economy has a major influence on the strategies and activities of firms by providing opportunities and constraints, not least in terms of the greater onus on production networks as a means in which to perform well in the stock market, further reinforcing the causal influence of finance. Finally, the increase in risks and greater awareness of such risks, has led to the need for firms to ‘manage risks’ so as to reduce uncertainties, including that relating to the disaggregation of production through outsourcing and increasing technological changes. Such risks are extensive, occurring within the realm of the economic, product (e.g. damage to a brand), regulatory (e.g. new international standards), labour (e.g. industrial action) and environmental (e.g. natural hazards). Together, it is the ‘unique’ combination of these competitive dynamics, with their specific historical-geographical configurations, which influence strategies and constitute the organisation and coordination of particular GPNs and their socio-spatial relations (Coe and Yeung, 2015). They achieve this by shaping four particular strategies undertaken by actors that configure GPNs and impact on territorial development. Strategies around ‘intrafirm coordination’ relate to how production is internally organised by GPN firms in response to competitive market dynamics (e.g. cost control). Alternatively, ‘interfirm control’ occurs where a firm outsources production responsibilities, largely because of high internal costs and often in relation to low value-added activities, but where they still retain relative control over suppliers and contractors. In contrast, ‘interfirm partnership’ encompasses more horizontal and collaborative relations between lead firms, strategic partners and specialised suppliers, with the purpose of taking advantage of differential cost-capabilities, high market need, financial discipline and managing risks. Finally, strategies of ‘extrafirm bargaining’ include relations between firms and non-firms, encompassing the nexus between market and non-market (e.g. societal, political) values and priorities. These strategies are the causal instruments by which specific GPN configurations are produced, including particular types of strategic coupling. 2.1.1. Strategic coupling Emphasis within the GPN perspective has focused on the concept of ‘strategic coupling’ between production networks and ‘territorialised’ regional assets and actors, with regional development contingent on the articulation of the latter into GPNs (Coe et al., 2004; Yeung, 2009a). The GPN approach defines strategic coupling as a dynamic ‘mutually dependent and constitutive [deliberative] process involving shared interests and cooperation between two or more groups of [regional and transnational] actors who otherwise might not act in tandem for a common strategic objective’ (Yeung, 2009a, 212). Within this context, coupling is characterised by varying degrees of congruence between regional institutions and corporate production networks, including the former seeking to produce the regional assets required by the latter for ‘value capture’ purposes, and driving a measure of territorial embedding of GPNs within particular places (Yang et al., 2009). These elements come to influence the nature and degree of articulation of regions with GPNs (i.e. coupling), which varies along a spectrum from the very strong forms of coupling characterised by highly developed regional economies, to regions characterised by cost-based foreign direct investment (FDI) that is susceptible to disinvestment and rationalisation (Coe and Hess, 2011). From this basis, Yeung (2009b) and Coe and Yeung (2015) identify three particular forms of strategic coupling: First, ‘indigenous coupling’ characterised by the endogenous development of GPN segments, often as part of a broader industrial development strategy. Secondly, ‘functional coupling’ involves opportunities for value capture by regions as firms form a particular role in GPNs, either as part of vertical specialisation or through more partnership-based means. Finally, ‘structural coupling’ refers to limited, often cost-based outsourcing and subcontracting responsibilities for firms and regional economies, characterised by greater external dependence for regions. There is hence a close association between the responsibilities of the MNE subsidiary and regional assets, which can encompass region-specific intangible assets, and ubiquitous assets where regional advantages are typically cost-based (Horner, 2014). For MacKinnon (2011), GPN studies have tended to understate the power relations between territories and GPN lead firms, as well as paying little attention to the historical constitution and evolution of relations between networks, regional assets and institutions (see, also, Horner, 2014). MacKinnon (2011) subsequently relates the concept of strategic coupling to broader evolutionary thinking around processes of institutional change through social agency advocated by Martin (2010). For MacKinnon (2011), corporate investment selection decisions are able to produce decoupling (disinvestment) and recoupling (reinvestment), based on the extent to which regional assets and regional market conditions are congruent with the strategic aims of the MNE. When selection produces new rounds of investment there is ‘layering’ of additional new organisational practice and competences, involving gradual institutional change. Where existing rules and practices are reconfigured in terms of form and function there is institutional ‘conversion’. Processes of ‘recombination’ involve the linking of regional assets with corporate investment in subsidiaries through new institutional rules and practices, and where reconfigured forms of coupling occur through the creation of new institutional combinations, including processes of institutional de-locking. MacKinnon (2011) thus extends understanding of coupling processes through the identification of certain institutional and corporate elements (e.g. mode of entry) that have causal impacts, though conceivably the agency exerted by MNEs remains underplayed. 2.2. Advancing the GPN approach While Coe and Yeung (2015) appreciate the critical role of firm resources and capabilities, there remain conceptual challenges in GPN 2.0 which we examine in this section. The basis of the argument we present is that GPN 2.0 relies heavily upon ‘firm strategies’ as a means in which to elucidate the causality of competitive dynamics on GPNs and their geographical relations. While four such firm strategies are presented, their actual explication remains abstract, leaving room for further conceptualisation of the actual practices of actors within and of MNEs in producing, organising and governing GPNs. 2.2.1. Intra-MNE processes Cost-capability ratio decisions and ‘managing risk’ are constituted not just by cost but also the capability to manage, which builds on the resource-based view of the firm as a ‘managerial device’ for organising resources (Coe and Yeung, 2015). There is a danger of adopting a ‘classical’ interpretation of strategic management as purely within the realm of rationalistic decision-making in response to transparent market dynamics, and where strategy brings about cohesive order (Whittington, 2001). Firm strategies in GPN 2.0 are presented in a broad manner, potentially downplaying the complexities of strategic and operational decision-making and coordination, which are characterised by power relations and politics, both within and beyond the MNE in production networks (Whitley, 2009). Furthermore, MNEs are heterogeneous organisational entities constituted by various values, strategies, systems, procedures and practices (see Geppert and Dörrenbächer, 2013). These are the internal dynamics, comparable to competitive dynamics, which determine and mediate strategies and the management of capabilities and resources (Balogun et al., 2011). Secondly, while the MacKinnon (2011) framework presents an important advance by emphasising the relationship between GPNs and forms of institutional change in processes of coupling, there remains a dearth of explanation for the causal internal corporate practices influencing ‘selection’, coupling/de-coupling and institutional change as part of firm strategies (see Dawley, 2011), and as evidenced in empirical studies of coupling (e.g. Horner, 2014; Kleibert, 2016). This relates more broadly to Murphy’s (2012) concern that GPN analysis lacks appreciation of the role of agency and practices, and the intricacies of power relations, by which firms interact with regions. For instance, the ‘selection’ perspective of MacKinnon (2011) brings the role of agency to the fore, but there can usefully be a deeper analysis of this process, taking account of processes of selection rather than viewing them as an end result (see Kleibert, 2016). 2.2.2. MNE agency and change Capability development, so critical to positional power within GPNs, relies upon the efforts of firms to advance their competitive position in networks and the marketplace, and change their ‘value capture trajectories’ (Coe and Yeung, 2015). Transition to higher capabilities or the reinforcement of capabilities and ‘market development’, involves change through Schumpeterian entrepreneurship and innovation (Yeung, 2009a). Yet, while a focus on cost-capability ratios advances our understanding of the decision-making criteria involved in ‘organisational fixes’ (Yeung, 2009b) around the four strategic approaches, the actual practices employed by actors, and particularly the ‘why’ and ‘how’, which come to constitute the dynamism and performativity of the ratio and practices of firm strategies, are not fully developed in GPN 2.0. Indeed, value creation within GPNs is premised on the reconfiguration of the cost-capability ratio with markets understood to be in constant flux, in part because of the limited prospects of reducing costs and the greater requirement to innovate and change. This suggests far greater sensitivity towards the practices of reconfiguring resources within MNEs and production networks when producing capabilities than is the case within GPN 2.0 and FDI accounts (e.g. Pavlınek and Zızalova (2016) account of supplier networks in the Czech automotive industry). 2.2.3. Extrafirm bargaining and micropolitics of MNEs GPN 1.0 has been accused of insufficient conceptualisation of the state and institutions, since it is more concerned with the role and activities of the state, rather than deliberations and politics (Murphy, 2012; Smith et al., 2014). Similarly, there has been criticism of the approach for downplaying negative aspects of coupling relationships, including exploitation and subordination (MacKinnon, 2011; Dawley, 2011) apparent in special jurisdictions or enclaves (Phelps et al., 2015). In response, GPN 2.0 places particular prominence on ‘extrafirm bargaining’ between GPN firms and state and non-state actors that have a major impact on GPNs, and the institutional underpinnings of strategic coupling (Coe and Yeung, 2015). Critical in such practices is the need to understand how firms go about achieving strategic aims, as the important element for GPN 2.0 is the relationship between the interests and ‘pressures’ of firms, and the capacity and priorities of non-economic actors. However, as yet, the GPN 2.0 approach is focused more on explicating the overarching outcome (e.g. societal legitimacy), rather than actual practices. By taking an actor-centred approach one may overlook the processes of deliberation and governance in production networks advanced in GPN 1.0. What is therefore important is the further elucidation of a framework in which to conceptualise the micropolitical strategic motivations and deliberative actions of MNEs as they seek to influence regional actors, and the geographical relations by which such practices occur. Furthermore, with the emphasis on bargaining in GPN 2.0, there is the need to guard against treating power relations as expressive of rationalist actors and a zero-sum game in which there are only winners and losers. Economic actors are irrational, socialised entities subject to and encompassing alternative values and motives, and living through heterogeneous and often contradictory societal conventions and formal and informal institutional landscapes (Williams and Lee, 2011). Such conditions characterise MNEs that are themselves heterogeneous, constituted by diverse aims and rationalities. One must treat bargaining therefore as a socialised, emergent and irrational affair subject to various social processes (Geppert, 2003). An emphasis of this kind is critical given that MNEs, with all their resources and capabilities, remain susceptible to the ‘experimental’ nature of power because of the deliberative manner of action (Morgan, 2009). This stems from a pragmatist belief in the anti-foundational nature of knowledge and power which is created through social transactions rather than pre-existing (Allen, 2008). Power is exercised and experienced in an experimental manner with no guarantee that actors will achieve their desired outcomes. This is because ‘power in its various guises is mediated relationally through space and time’, since it relies on the deliberate actions of actors as they seek to enact power (e.g. organising resources), their social interactions with other actors who can have different conceptions of truth, and disparate environments and situations (Allen, 2008, 30). Understanding of the ‘micro-politics’ of actors in mediating such landscapes is therefore critical. 3. A structurationist perspective of MNE agency within GPNs As outlined, there are possibilities in which to expand GPN 2.0 by exploring further aspects of the ‘black box’ of the MNE, but understanding they fulfil roles throughout GPNs (e.g. generic supplier). Giddens’ (1984) scheme for understanding structuration in terms of institutional domains, modes of structuration and agency is utilised as a conceptual framework in which to examine the role of the MNE within GPNs and regional economies, and how these factors influence the four types of firm strategy and their implementation (Figure 1). Figure 1 View largeDownload slide Corporate agency and the structuration of territorial economies and GPNs. Source: Adapted from Giddens (1986, 29). Figure 1 View largeDownload slide Corporate agency and the structuration of territorial economies and GPNs. Source: Adapted from Giddens (1986, 29). 3.1. Differential sources of agency within the MNE A structuralist perspective can focus on processes internal to GPN MNEs, which reflect a view of these agents not as a monolithic entity but as networked and ‘federative’ in nature. Here, the subsidiary emerges more clearly as an actor mediating between the MNE parent, production networks and particular host territories (Phelps and Fuller, 2000; Fuller, 2005). Also, the MNE is opened up as a trans-territorial and networked actor whose strategic goals are produced and contested via the ‘interpretive schemes’ of both parent and subsidiary entities. Drawing directly from Giddens (1984), interpretative schemes (as ‘modalities’) are recognised as the forms of knowledge that actors utilise to assign meaning to their own actions and those of others and, as such, they underpin structures of meaning, or signification. They are both enabling, as they inform and guide action, and constraining, in that they can reinforce existing norms, values and forms of knowledge (Giddens, 1984). Interpretative schemes involve transmission through social interaction or regularised institutionalised routines (Stones, 2005). In the case of the MNE, the corporate centre seeks to control by constructing particular structures of signification that are transmitted through everyday micro-social interaction, institutional routines (as rules seeking to guide behaviours) and firm strategies. These are interpreted and acted upon by subsidiary management through their own interpretative schemes, and have implications for production networks (Geppert and Dörrenbächer, 2013). Power relations are intrinsic to intra-corporate relations and interpretative schemes. There are different modalities of power, including the more formal (coercion, domination and authority) and informal (seduction, manipulation and inducement), that are intrinsic to intra-corporate relations (Allen, 2016). In a recent intervention, Allen (2016) argues that the exercise of power is increasingly topological in nature, based on the ability of actors to ‘stretch’ into spaces to exercise particular modalities of power, or ‘fold’ socio-spatial relations so that the distant becomes proximate, demonstrating the ability to act at a distance, but where we must recognise that power is experimental with no assurance of successfully achieving your aims. Any consideration of structuration needs to take account of the practices of actors as they mobilise power, the spatiality that such mobilisation of power entails, and the influence of spatial relations on the deployment and impact of power relations (Faulconbridge and Hall, 2009). As highlighted above, there is a need for greater appreciation of the complexities of MNEs in different segments of GPNs, and how these influence and constitute firm strategies. In essence, these corporate dynamics, evident in MNEs fulfilling various GPN roles, are causal and mediating factors in firm strategies, such as in complicating interfirm partnerships, or disrupting interfirm coordination. Within International Business (IB) studies such thinking has underpinned an understanding of MNEs as ‘heterarchies’ with dispersed resources, responsibilities and power relations (Ghoshal and Bartlett, 1990), while working through and being influenced by disparate institutions, conventions and rationalities (Geppert and Dörrenbächer, 2013). MNEs are considered not just ‘differentiated networks’ but complex ‘federations’, in which the ‘position’ of an entity within the corporate value chain does not necessarily equate to a particular power relation (Andersson et al., 2007). The broader geographical and organisational implications are ones in which the MNE works through many cost-capability ratios and processes of ‘selection’. Critical to such federative arrangements is the role of knowledge. For Ciabuschi et al. (2011), knowledge is intrinsically related to the territorialised context in which a subsidiary’s knowledge production activities function, as well as the relational topological production networks in which they operate. As knowledge is territorialised and network-specific, and involves working through particular networked production relations, then corporate HQs can never possess perfect knowledge of subsidiary activities (Fuller and Phelps, 2004; Ciabuschi et al., 2011). Such knowledge is critical in underpinning the capabilities central to the cost-capability ratio in GPN 2.0, since it produces competitive advantages relating to the ability to organise and deploy resources through capabilities, and is particularly important in ‘interfirm partnership’ arrangements where capabilities held by certain MNEs and value enhancement are critical, as is the need to ensure financial discipline and market responsiveness (Teece, 2007). The rise of more federative corporate configurations has therefore been associated with greater opportunities for subsidiaries to lobby and influence parent companies through particular actions and interpretative schemes (such as ‘seducing’ corporate HQs with the promise of new capabilities); develop strategies, knowledge and networks that are beyond the ‘topological’ control and observation of corporate HQs; and produce further rounds of selection, acquiring funding and/or legitimacy and authority from the corporate centre in which to pursue new investments, or develop capabilities (Birkinshaw, 1997; Fuller and Phelps, 2004; Mudambi, 2008). Within these processes subsidiaries can utilise a power relation, based on the ability to circumvent and influence corporate HQs through specialist knowledge possession, and thus utilising particular forms of signification (i.e. interpretative schemes) that rely upon territorialised and spatially relational knowledge (Andersson et al., 2007; Awate et al., 2015). This includes subsidiary managers negotiating, manipulating and resisting corporate strategies that negatively impact upon subsidiary development, often through informal topological political relations with intra- and extra-corporate entities (Clark and Geppert, 2011). Where particular subsidiaries possess specific capabilities they can convey greater bargaining power (e.g. through ‘inducement’) and deviant autonomous behaviour that disrupts ‘intrafirm coordination’ and thus actual value enhancement intentions (see Phelps and Fuller, 2000), and this can occur in MNEs fulfilling various roles in a GPN. There is also the possibility of subsidiaries influencing the power relations of production networks, such as in subsidiaries providing the capabilities in which to subordinate suppliers in the case of ‘lead’ firm strategies around ‘interfirm control’, or disrupting such control by way of collaborating in ‘interfirm partnerships’ where this requires a supplier subsidiary to develop critical capabilities beyond its mandate (see Clark and Geppert, 2011; Phelps and Fuller, 2016). These processes of subsidiary knowledge possession and deployment of interpretative schemes are thus the causality underpinning corporate selection of particular GPN suppliers/contractors and/or regions. Rather than conceptualising ‘selection’ as occurring through market mechanisms, as is the case with evolutionary thinking, what is important here is a need to understand the heterogeneous intra-corporate actors constituting MNEs. From this basis selection is far more complex, not undertaken by a homogenous MNE, but by particular actors within the MNE, such as subsidiaries pursuing autonomous activities to develop particular capabilities as part of ‘interfirm partnership’ strategies, HQs selecting subsidiaries within regions in the case of repeat investment (such as in the case of ‘intrafirm coordination’ strategies where the emphasis is on cost reduction), or selecting regions (with particular assets) through greenfield investment or other means (e.g. joint ventures). The growing relative importance of subsidiaries within federative MNEs does not deny the importance of the corporate HQ. It remains a key actor with the ability to guide and subordinate subsidiaries, and production networks, through strategic decision-making powers and deployment of resources (Bouquet and Birkinshaw, 2008). Within a GPN 2.0 framework, it is most evident in ‘intrafirm coordination’ strategies as firms seek to ‘internalise’ and ‘consolidate’ production, with the purpose of controlling costs and retaining proprietary capabilities. Strong HQ control has also been extensively examined in FDI studies in economic geography which argue that the operational positon of subsidiaries within MNEs has an impact on their regional development impacts and embeddedness (e.g. Phelps et al., 1998). More recently, Kleibert (2016) argues that elements of the ‘branch plant syndrome’ within vertically orientated offshoring remain important, typically because they lack decision-making powers. The corporate HQ remains powerful by way of the types of topological relations outlined by Allen (2016), with power not so much (spatially) concentrated, but based more on the ability to stretch across space in real-time to bring the distant within close proximity. Indeed, for Buckley (2009) and Yamin and Sinkovics (2009), more advanced forms of IT permit the corporate centre to impose real-time control over subsidiaries that increasingly have smaller ‘fine-sliced’ responsibilities, and more dispersed value chains, labelled as ‘global factories’. For Johnston (2005) and Andersson et al. (2007), hierarchical forms of corporate decision-making and organisation arises when HQs have significant knowledge of subsidiary responsibilities and their production networks. This suggests limits on the autonomy of subsidiaries, which arises from HQs being able to contest and subordinate efforts of subsidiaries through their own signification processes. Such processes can characterise ‘intrafirm control’ strategies where forms of subordination are strong as HQs seek to increase value enhancement, and ‘inter-firm control’ as lead firm HQs seek to lower cost-capability ratios through externalisation and control of suppliers to reduce costs, such as in the automotive sector (see Pavlínek and Ženka, 2010). In such hierarchical arrangements, investment ‘selection’ decisions will be exogenous to the subsidiary and region, with subsequent types and degrees of coupling/decoupling and institutional change significantly influenced by the corporate HQ in the case of lead and strategic partner firms, but also where the decisions of high capability firms (e.g. specialist supplier) are important in terms of coupling. Various studies also suggest this is evident in subsidiaries with production mandates lacking significant value creation in a GPN (i.e. specialist or generic suppliers, such as in interfirm control firm strategies), and isolated from intra-corporate and GPN knowledge networks (Monteiro et al., 2008). In such instances, coupling relates to what Coe and Yeung (2015) term ‘structural coupling’, where relations with state bodies and other regional organisations are based on financial incentives and limited efforts to develop incremental capabilities and institutional layering where possible (Phelps and Fuller, 2016), but which are vulnerable to decoupling by decisions taken at the corporate HQ. To summarise, a range of corporate processes influence firm strategies and their cost-capability ratios, corporate selection of regional assets, forms of coupling and institutional change. As such, future GPN 2.0 orientated studies need to take account of the complex intricacies of these corporate processes and interpretative schemes, since they constitute, influence and mediate firm strategies, particularly as competitive dynamics are interpreted through these internal firm dynamics. 3.2. Agency and change: the role of ‘dynamic capabilities’ GPN 2.0 could usefully incorporate greater analysis of the role of actors and practices in (re)configuring capabilities to change ‘value capture trajectories’ (recognising the dynamic nature of cost-capability ratios and other strategic decisions), as well as understanding the motives for participation in GPNs through firm strategies (see Murphy, 2012). In Figure 1, the ‘facilities’ that embodied material and non-material resources in Giddens’ (1984) original scheme, and which permit the mobilisation of power to achieve particular aims through the exercise of agency, have been replaced by ‘dynamic capabilities’ as a mode of structuration. As we discuss below, the ‘capabilities’ based view of the firm strongly implies change within the MNE, at individual subsidiaries and in GPNs by way of firm strategies. As highlighted by Teece (2007), and following Giddens’ (1984) view of configured resources as the media of power, dynamic capabilities represent the enactment of agency (and thus power) through the mobilisation of resources by actors within the MNE and through GPNs, be they subsidiaries seeking to challenge or reconfigure corporate arrangements and production networks, or corporate HQs reproducing or developing new forms of (behavioural) control. Such dynamics then impact upon relations, processes and configurations of GPNs. Dynamic capabilities therefore provide a window into the heterogeneous nature of the MNE (as a site of agency and power relations) and resulting forms of selection, coupling and institutional change. Pitelis and Teece (2010) argue that MNEs comprise heterogeneous, accumulated portfolios of tangible and intangible capabilities which are difficult to transfer to other firms and can be regarded as assemblages of ‘high-order’ routines (i.e. ‘dynamic capabilities’) (Winter, 2003). The latter are broadly defined as ‘the organizational and strategic routines by which firms achieve new resource configurations as markets emerge, collide, split, evolve and die’ (Eisenhardt and Martin, 2000, 1107). Dynamic capabilities, as higher order capabilities, arise from learning and are embedded within the firm as part of a set of ‘systematic methods’ for reconfiguring strategies, production and routes (Zollo and Winter, 2002). They arise from the co-evolution and interaction of processes of experience-based knowledge (i.e. ‘trial and error’), articulation of knowledge through different organisational functions and forms (e.g. R&D departments), and codification (Gavetti and Levinthal, 2000). Dynamic capabilities encompass the reconfiguration or displacement of existing firm routines and capabilities, and underpin efforts to develop new products and production processes in response to market pressures, while at the same time they maintain or reorientate external resources associated with the business environment (Kay, 2010). For Pitelis and Teece (2010) dynamic capabilities are critical to the objectives and nature of what can be considered lead and strategic partner MNEs, and through which competitive dynamics and firm strategies impact on other GPN actors. They involve MNEs combining co-specialised and complementary assets which are typically dispersed throughout GPNs. Markets are subsequently co-created and captured with various agents in the host ‘eco-system’ through processes of combination-based innovation and entrepreneurship, which is often international in scope, since it involves cross-border exchange (Somaya et al., 2011). Where MNEs cannot find relevant assets within markets, or where value capture from co-specialisation is likely to be greater from the MNE undertaking such tasks, they will produce internal cross-border integration and induce foreign direct investment rather than market transactions. Alternatively, they will undertake co-specialisation through value chains where assets are beyond the MNE, and the costs of producing specialisation internally outweigh potential value creation (Defee and Fugate, 2010). The nature of firm capabilities will have a considerable impact on the types of strategy deployed by lead firms and suppliers in GPNs. For instance, in the case of the MNE possessing capabilities and unable to find assets in the marketplace (as well as there being other conducive competitive dynamics, such as a low market imperative), ‘intrafirm coordination’ will be of importance—with lead corporate units producing greater forms of coupling where they are located (Phelps and Fuller, 2000). In contrast, where they do not possess capabilities or assets, MNEs will seek market exchange through ‘interfirm control’ (when seeking to reduce costs through externalisation in a mature market), or ‘interfirm partnership’ (as lead MNEs seek to reduce costs and/or develop or acquire assess to capabilities within GPNs) as a means of value capture. There is thus the potential for certain types of power relations in production networks, coupling and institutional arrangements to develop, depending on the importance of complementary assets held beyond the MNE (Almeida and Phene, 2004) (see, e.g., Sturgeon and Kawakami (2010) on the electronics sector). The dispersed yet territorially embedded nature of regional assets requires MNEs, with various roles in a GPN, to develop dynamic capabilities for fulfilling various production roles, which will in turn create and capture value (Dunning and Lundan, 2010; Pitelis and Teece, 2010). A critical element is the unique ‘orchestration’ capabilities, as modalities, of MNEs in creating and configuring ‘difficult to replicate’ dispersed assets in the process of knowledge creation and learning (Katkalo et al., 2010). Orchestration refers to the everyday ‘routine’ and ‘exceptional’ decision-making processes of the agency of individual firm managers (see Sirmon et al., 2011) as they assemble and establish relevant combinations of resources within the firm (producing co-specialised assets), and through relations with external actors and resources by way of particular firm strategies (e.g. interfirm partnerships) (Pitelis and Teece, 2010). This includes, for example, the case of strategic partners working with lead firms to create particular capabilities and products (e.g. see Ernst (2009) on collaborations in the electronics sector, including that between Huawei and Siemens). It is also comparable to ‘novelty’ in evolutionary economic geography since it is concerned with the enactment of various forms of (organisational and spatial) creative change (e.g. transformation) to firm/sectoral practices, products, institutions and production networks through innovation and recombination. However, the concept of orchestration places greater emphasis on conceptualising the actual interactive practices of actors as they seek to change and/or enact completely new routines through recombining and/or introducing new ‘heterogeneous elements’ (see Stam, 2010), and it therefore appreciates the agency of, and within, MNEs as they influence geographies of capitalism. Following our structuration perspective (Figure 1), dynamic capabilities therefore represent a ‘modality’ in which firm managers seek to influence overarching corporate arrangements (e.g. acquire new responsibilities) and firm roles within particular GPNs, including the nature and power relations of coupling. As a ‘mobilisation’ of power in network relations, this occurs through the key orchestration practices of dynamic capabilities: sensing, seizing and transforming (Teece, 2007). Sensing relates to the ability to perceive opportunities and threats in the environment through intelligence and R&D, and act upon these by ‘shaping’, and is intrinsic to many GPN studies as they recognise the (re)configuration of production networks in response to cost and capabilities-based forms of competition (e.g. Horner, 2014; Dörry, 2015). Seizing opportunities relies upon managerial capabilities in devising new rules and processes of resource (re)configurations when acting upon ‘sensing’ opportunities, including building loyalty among partners (Teece, 2007). One particularly important example of this is Horner’s (2014) examination of the growth of Indian MNCs as they seized the opportunity to produce pharmaceutical products for the local market, and which arose from central government legislation to restrict foreign MNCs in India from the 1970s and characterised as decoupling. Finally, ‘transformation’ encompasses the reconfiguration of resources and organisational and partnership structures in the actual implementation phase, and in essence characterises collaborations between GPN partners as they develop dynamic capabilities. Indian pharmaceutical MNCs, for example, reconfigured production towards indigenous bulk production, deploying ‘chemistry skills’ based on the imitation of foreign MNC products (Horner, 2014). Similarly, Dörry’s (2015) account of the international financial centre status of Luxembourg demonstrates the critical role of restructuring in advanced business services. These have transformed in response to market-driven cost pressures, where profit margins are low, resulting in standardisation, automation and outsourcing as firms seek to remain competitive. Such actor-led transformations have particular geographies, characterised by outsourcing through production networks to cheaper locations for low value-added activities, while increasing the nodal importance of Luxembourg. In this we see orchestration as the actions of corporate managers to bring about changes to corporate strategies and practices, and GPNs; such as investing in new technologies or deploying alternative working practices that seek to ultimately lead to the redesignation of subsidiary responsibilities, or which support a subsidiary in fulfilling a more value-added role in a GPN. More broadly, for Teece (2007), the increasing pace and knowledge-based nature of capitalism requires transformation towards decentralisation and co-specialisation through production networks, but at the same time as allowing the firm to coordinate these spatially diverse elements, largely through network relations. The implications for GPN in particular are critical: namely, that this performative element, involving the reconfiguration of resources outlined in Figure 1, is key to unlocking the constant enactment of capabilities and, thus, cost-capability ratios, as well as firm responses to other competitive dynamics. These practices of dynamic capabilities represent the mobilisation of power, both within the firm and through production networks, as change is initiated and different forms of power are experienced. For instance, Felin and Powell (2016) argue that seizing involves power relations of persuasion as partners need to be convinced of the need to pursue new opportunities. Where such forms of power operate through GPNs it is likely to occur in less coercive ‘interfirm partnerships’ involving strategic partners and specialist suppliers. Such processes are evident in the ‘northern migration’ of groups of Taiwanese electronics ‘strategic partner’ MNEs from the Pearl River Delta to the Yangtze River Delta in China (see Yang, 2008). ‘Transformation’ can encompass new forms of influence, such as ‘seduction’ as a form of power relation, whereby lead firms are able to influence and ‘capture’ regional state bodies by making their proposed future actions attractive, typically by suggesting new development benefits (see Allen, 2016). Alternatively, there can be a threat of disinvestment by lead firms over regional state bodies, where new incentives are sought (as part of selection) or completely new practices are required (e.g. workforce skills), representing recombinant institutional change. In each instance, we can witness the reconfiguration of broader structuring power relations through various modes, manifest in changes to the nature of GPN relations, coupling and differing forms of institutional change. It is through such a framework that it is possible to identify two geographical elements of dynamic capabilities. First, it is difficult and costly to codify territorialised dynamic capabilities into tangible knowledge and generic organisational routines for intra-corporate and inter-firm GPN transfer across borders (Katkalo et al., 2010). Such situations are likely to be related to strong forms of territorialised coupling and co-evolutionary institutional tendencies, such as ‘indigenous’ or strong forms of ‘functional’ coupling and recombinant change during periods of subsidiary transition. Secondly, elements of dynamic capabilities can be clarified and codified into routine transferable knowledge over time, and therefore lose their salience for subsidiaries within MNEs and GPNs (Dunning and Lundan, 2010). Such clarification and codification increases potential decoupling in the future, accentuating the need for the subsidiary to produce further dynamic capabilities which can reinforce their role within GPNs, and foster new rounds of selection, recoupling and institutional change. Ultimately, a concern with dynamic capabilities can add greatly to GPN 2.0 by providing a conceptual framework in which to elucidate the causal actions of actors as they produce and constitute firms strategies and associated geographical relations. 3.3. The micropolitics of MNE-host institution interactions For Levy (2008), Phelps and Wood (2006) and Hess and Yeung (2006), MNE actions take place through and are mediated by historically and geographically evolving institutional landscapes, as well as being able to influence the latter. We therefore consider it critical to understand the emergent and causal nature of formal and informal institutions, and their interaction with the ‘micro-political’ actions of MNEs and their subsidiaries in promoting institutional change in regional economies. In response to the view that GPN takes an economistic perspective on power relations, Levy (2008) advocates a Gramscian and institutionalist perspective on the way GPNs constitute and influence broader discursive ideologies and institutions, which act as forces of consent and coercion, and with the possibility of resistance and change by particular actors (e.g. ‘institutional entrepreneurs’). What Levy (2008) does not do, however, is conceptualise the actual practices constituting power relations, including institutional isomorphism, and how actors seek to resist and reconfigure ‘organisational fields’. For this we turn to the concept of ‘legitimacy’ as a means in which to understand how MNE subsidiaries, fulfilling various roles in GPNs and as agents operating in particular foreign sites, seek to conform, resist or reconfigure regional institutional arrangements, but involving both discursive and material actions since these are interwoven (Morgan, 2009). Embedded within an institutionalist perspective, micropolitics and shared norms of behaviour are the modes by which MNE subsidiaries seek to legitimate their activities, representing social structures of ‘legitimation’ within Giddens’ (1984) structuration framework; and is a feature of the relationship, or degree of coupling, between firms and territorialised regional institutions found in the GPN perspective (Coe and Hess, 2011). Through such an approach we recognise that MNE subsidiaries seek to work both through existing geographically and historically-configured (e.g. territorialised) norms and value systems so as not to be subject to sanctions (and in effect confirming to discursive institutional formations), as well as attempting to change them when mobilising extensive sociospatial relations as emphasised by Levy (2008) (see also Kostova et al., 2008). By taking such a perspective we recognise that MNEs cannot be defined as institutions per se, but as ‘organisations’ influenced and partly constituted by institutions (Morgan, 2009). These are understood to be both formal (e.g. national legal system) and informal (e.g. norms, values and habits), and involve various sociospatial relations such as those concentrated within particular territories or mobilised across networks (Storper, 2010). The picture that emerges from the IB literature, as well as accounts of MNEs in economic geography (e.g. Phelps et al., 1998; Dawley, 2007), is one of the subsidiary entangled in a highly politicised and geographically contingent set of interactions and decision-making, in which the (territorialised and relationally configured) societal institutions of host regions have a potentially important role in the construction of capabilities (Dunning and Lundan, 2010). The existence of such institutions does not necessarily mean that coordination and control is easily achieved, not least because (dynamic) capabilities and everyday organisational routines are typically embedded within corporate units and the geographical spaces through which they operate (Cantwell et al., 2010). Of course, the GPN 2.0 perspective does advance our understanding of political relations but this is conceptualised in terms of the different realms of bargaining, such as ‘social and political legitimacy’, which are based on an end state, and the actual desired outcomes of lead firms. Despite recognition of the critical role of ‘strategies’, the actual tactics and practices of the MNE as it exerts (bargaining) agency are not extensively conceptualised in GPN 2.0 or, and as suggested above, in Levy’s (2008) account, and for this we turn to the institutional concept of ‘legitimacy’. The search for legitimacy by way of a micropolitics of MNE-host regional institutional interactions emerges as an important ingredient in a structurationist account of corporate agency under the GPN approach (as depicted in Figure 1). Kostova et al. (2008) argue that MNEs, working through networked geographical relations, do not function within ‘organisational fields’ of cohesive organisational structures and actions. Instead they operate through ‘fragmented, nested, or often conflicting institutional environments’ in which they are less inclined to seek legitimacy in host localities beyond territorial-based legal and regulatory requirements, largely because it is impossible to conform to all these multifaceted institutional influences (Kostova et al., 2008, 998). This confers with many geographical accounts of the asymmetrical relations between MNEs (engaging in cost-based investment decisions) and regions, where the former can exploit competition between regions for investment, and corporate HQs in home countries possess the ability to close such foreign subsidiaries (e.g. Phelps and Waley, 2004). These disparate sociospatial conditions restrict the extent to which the production of shared and legitimate organisational structures, strategies and actions are possible. For Kostova et al. (2008), this provides MNEs and subsidiaries with considerable scope within which to engage institutions and adopt organisational practices which they deem relevant from their territorialised host site, to acquire the legitimacy required to operate, while at the same time being able to circumvent the adoption of other practices. This has been evident in many FDI accounts of regional development, where MNEs will negotiate locating in a region with the purpose of acquiring financial and resource assistance on the basis of legitimately confirming to particular regional development priorities, but at the same time potentially obscuring the actual qualitative impact they will potentially make (Fuller and Phelps, 2004; Phelps and Wood, 2006). Isomorphism and legitimacy are therefore not necessarily critical, particularly as MNEs are viewed as important change agents for localities in that they introduce and foster entrepreneurship, knowledge production and innovation which is critical economic development. Legitimacy in this sense is an integral element of the GPN’s conceptual approach to examining embeddedness, since efforts by MNEs to acquire a particular form and level of legitimacy within host territories and societies, irrespective of their position within the GPN, is a causal force in influencing the form and level of subsidiary embeddedness. The concept of ‘legitimacy’ thus contributes to explanations of the degree and nature of embeddedness by focusing on the (dynamic and heterogeneous) social agency of MNEs (and incorporating the ‘genetic code’ of their home countries in guiding action), but understanding that their intent and motivation is connected with the institutionalised territorial conditions (e.g. business networks) constituting host localities, and with both therefore forming part of the constitution of embeddedness (Hess, 2004). Following Levy (2008), this legitimacy therefore represents a variable degree of adherence to societal norms, values and beliefs embedded within (nationally and regionally-configured) informal institutions and which, conversely, guides regional actors in and around GPNs, forming part of the nature and depth of strategic coupling. This would include, for instance, regional development agencies framing MNEs as legitimate on the basis that they will provide multiplier effects (e.g. local supplier linkages). Where such positive impacts are not achieved (see Phelps et al. (1998) on LG in South Wales), these agencies will question the legitimacy of such MNEs, based on societal conceptions that obviously vary between countries and regions (see Morgan and Kristensen (2006) on the differences between the pro-market UK and regulatory state of Germany). What is important in the Kostova et al. (2008) account is recognition of the critical role of the ‘social agency’ and micropolitics of MNEs in institutional change and embeddedness. In processes of legitimacy creation, MNEs will typically negotiate status through political processes and communication in which they will have a proactive influence on institutions and their creation, since institutions comprise shared understandings and outcomes that arise through social interaction. For Kostova et al. (2008), social environments, as constituted through various sociospatial relations of proximity (i.e. territorialised institutions) and distance (i.e. topologically networked MNEs), are therefore ‘products of a continuous process of sense-making, enactment and negotiation of political interactions’ as MNEs seek both legitimacy and distinctiveness, but where shared meanings are produced (1002). This is congruent with Murphy’s (2012) conceptualisation of the importance of legitimacy in social practices of ‘relational proximity’ between regions, elements of GPNs and lead firms. The Bolivian wood products sector encompasses ‘conventional GPNs’, characterised by global buyers ensuring their suppliers meet international and national standards, and in relation to issues such as the sustainability of forest operations and labour force conditions (Murphy, 2012). Based on this understanding that MNEs seek legitimacy and distinctiveness, Dunning and Lundan (2010) identify three types of micropolitical MNE behaviour that can provide an important framework in conceptualising the causal processes underlying the coupling tendencies in which MNEs and their subsidiaries are implicated. First, there is ‘institutional experimentation’ by corporate subsidiary managers as they seek to create institutions conducive to their aims, particularly in developing place-specific and ‘relational’ knowledge production (Dunning and Lundan, 2010). This implies substantive qualitative change and would underlie processes of strategic coupling through recombination, involving significant sense-making, negotiations and reconfiguration of territorial norms and values by subsidiary managers. For instance, within a GPN framework, Dörry’s (2015) analysis of advanced business services in Luxembourg demonstrates that these firms have developed ‘strategic couplings’ based on the important strategic position of Luxembourg as a financial centre. The state seeks to be sensitive to the requirements of global financial services, but where the latter actively lobby state institutions during periods of new regulations, representing efforts at fostering shared values and norms but which adhere to MNE aims. Such processes are also evident in foreign apparel MNEs aligning their interests with the neoliberal reforms of the Romanian government, which facilitated the growth of cost-based assembly production (Plank and Staritz, 2015). Secondly, where ‘institutional adaptation’ occurs it takes place through MNEs as they adjust their capabilities and practices to be more congruent with their environment, the purpose of which is to seek legitimacy and influence with largely territorialised political institutions. This is particularly evident in the case of MNEs investing in China where ‘lead’ and supplier firms in the automotive sector have to strictly adhere to national and regional institutions, largely because of the power of the state to determine access to the Chinese market and workforce (see Liu and Dicken, 2006). While for Perkmann (2006), the competitiveness of MNEs in South Tyrol, Italy, was partly based on engaging labour market institutions. Similarly, MNE ‘independent suppliers’ can engage extra-firm actors, such as state institutions and universities, as a means in which to develop capabilities and upgrade their roles within GPNs (see Ernst, 2009; Coe and Yeung, 2015). Finally, ‘institutional avoidance’ occurs where MNEs do not have substantive engagement with the institutions of host countries, making it possible for disinvestment and closure with ease, or qualitatively limited forms of layering based on cost motivations (see Morgan and Kristensen (2006) on UK market-based institutions). This is particularly evident in Dawley’s (2007) account of the closure of a semiconductor plant in northeast England. While Siemens had acquired substantial development agency support, it was able to close the plant with ease despite high levels of productivity. Relative to plants in the home country, the MNE had not actively been embedded within the formal and informal institutions of the northeast. In other instances, retail MNEs enacting new market-seeking FDI have failed to fully engage with host institutions, resulting in exit after a short period (see Bianchi and Arnold (2004) on Home Depot’s disinvestment in Chile following a failure to present a suitable merchandise range and store atmosphere). What the micropolitical approach presented here therefore provides GPN 2.0, in contrast to previous accounts, is a conceptual means in which to examine the practices and motives constituting extra-firm strategies, by way of it engagement with issues of institutional legitimacy. 4. Conclusion We have outlined three elements of the agency exercised by MNEs that remain underplayed in the GPN approach. The sole purpose of the article is to emphasise the need to understand the role of these three elements in constituting, influencing and mediating firm strategies and practices by MNEs fulfilling various roles in the GPN, since this is the organisational spaces that co-constitutes competitive dynamics and risk environments. To reiterate, first, the subsidiary, as an actor fulfilling various GPN roles, emerges as a key nexus within such relations and all that this implies for the prospects for economic development in some regions. In terms of future research, this suggests greater appreciation of the internal arrangements and practices of MNEs, and their structuration through particular sociospatial GPN roles, (power) relations and firm strategies. Second, an emphasis on the agency of actors within and of MNEs—and in particular a focus on dynamic capabilities rather than resources—can assist in the recognition of continuity and change in production networks and regional economies by way of firm strategies. For GPN analysis, this would involve greater emphasis on the actors and social practices implied in the ‘orchestration’ of dynamic capabilities, and their structuration through various geographical relations. Third, the interaction between MNEs and institutional environments is critical to the understanding of coupling processes, but is a process where greater emphasis could usefully be placed on the causal agencies of MNEs. Processes internal to the MNE are implicated in the micropolitics they pursue, particularly when trying to reconfigure external institutions in ways that support their strategic aims. Following Geppert and Dörrenbächer (2013), such an analysis necessitates examination of actual discursive practices, including sense-making and framing, as well as negotiation by MNEs with various organisations and institutions. Finally, the thrust of our article has been to emphasise the importance of focusing on the ‘black box’ of the MNE, to understand one form of ‘social agency’ and its enactment through firm strategies. Of course, such an approach does run the risk of reifying the causal properties of MNEs and their behaviours when they operate in structural circumstances not of their own choosing (Walker, 1989). What we have proposed here is to make explicit the MNE as a key agent in the structuration of territorial economies and ‘relational’ networks of production implicit in the GPN approach. A fuller appreciation of corporate agency has important implications for future research in the GPN approach, and its connections to perspectives deriving from the IB literature. GPN 2.0 can add to the IB literature where, geographically speaking, the onus has been on ‘local’ business networks with very limited consideration of the role of the (diverse) geographical relations constituting MNEs. GPN provides IB studies with a set of conceptual tools for understanding those heterogeneous (geographical) production relations constituting governance networks that stretch beyond the MNE, in a way that is more sophisticated than found, for example, in the ‘global factories’ (Buckley, 2009) perspective and, as such, IB could significantly advance by engaging the GPN perspective. Funding There is no funding award associated with this paper. References Allen J. ( 2008) Pragmatism and power, or the power to make a difference in a radically contingent world. Geoforum , 39: 1613– 1624. Google Scholar CrossRef Search ADS Allen J. ( 2016) Typologies of Power . Oxford: Routledge. Almeida P., Phene A. ( 2004) Subsidiaries and knowledge creation: the influence of the MNC and host country on innovation. Strategic Management Journal , 25: 847– 864. Google Scholar CrossRef Search ADS Andersson U., Forsgren M., Holm U. ( 2007) Balancing subsidiary influence in the federative MNC: a business network view. Journal of International Business Studies , 38: 802– 819. Google Scholar CrossRef Search ADS Awate S., Larsen M. M., Mudambi R. ( 2015) Accessing vs. sourcing knowledge: a comparative study of R&D internationalization between emerging and advanced economy firms. Journal of International Business Studies , 46: 63– 86. Google Scholar CrossRef Search ADS Balogun J., Jarzabkowski P., Vaara E. ( 2011) Selling, resistance and reconciliation: a critical discursive approach to subsidiary role evolution in MNEs. Journal of International Business Studies , 42: 765– 786. Google Scholar CrossRef Search ADS Bathelt H., Glückler J. ( 2005) From substantive concepts towards a relational perspective. Environment and Planning A , 37: 1545– 1563. Google Scholar CrossRef Search ADS Bianchi C., Arnold S. J. ( 2004) An institutional perspective on retail internationalization success: home depot in Chile. The International Review of Retail Distribution and Consumer Research , 14: 149– 169. Google Scholar CrossRef Search ADS Birkinshaw J. ( 1997) How multinational subsidiary mandates are gained and lost. Journal of International Business Studies , 27: 467– 496. Google Scholar CrossRef Search ADS Blažek J. ( 2016) Towards a typology of repositioning strategies of GVC/GPN suppliers: the case of functional upgrading and downgrading. Journal of Economic Geography 16: 849– 869. Google Scholar CrossRef Search ADS Bouquet C., Birkinshaw J. M. ( 2008) Managing power in the multinational corporation: how low-power actors gain influence. Journal of Management , 34: 477– 508. Google Scholar CrossRef Search ADS Buckley P. ( 2009) Internalisation thinking: from the multinational enterprise to the global factory. International Business Review , 18: 224– 235. Google Scholar CrossRef Search ADS Cantwell J., Dunning J., Lundan S. ( 2010) An evolutionary approach to understanding international business activity: the co-evolution of MNEs and the institutional environment. Journal of International Business Studies , 41, 567– 586. Google Scholar CrossRef Search ADS Ciabuschi F., Dellestrand H., Martín O. ( 2011) Internal embeddedness, headquarters involvement, and innovation importance in multinational enterprises. Journal of Management Studies , 48: 1612– 1639. Google Scholar CrossRef Search ADS Clark E., Geppert M. ( 2011) Subsidiary integration as identity construction and institution building: a political sensemaking approach. Journal of Management Studies , 48: 395– 416. Coe N. M., Hess M., Yeung H., Dicken P., Henderson J. ( 2004) ‘Globalizing’ regional development: a Global Production Networks perspective. Transactions of the Institute of British Geographers , 29: 468– 484. Google Scholar CrossRef Search ADS Coe N., Dicken P., Hess M. ( 2008) Global Production Networks: realizing the potential. Journal of Economic Geography 8: 271– 295. Google Scholar CrossRef Search ADS Coe N. M., Hess M. ( 2011) Local and regional development: a Global Production Network approach. In Pike A., Rodriguez-Pose A., Tomaney J. (eds) The Handbook of Local and Regional Development . London: Routledge. Coe N., Yeung H. ( 2015) Global Production Networks: Theorizing Economic Development in an Interconnected World . Oxford: Oxford University Press. Google Scholar CrossRef Search ADS Dawley S. ( 2007) Fluctuating rounds of inward investment in peripheral regions: semiconductors in the north east of England. Economic Geography , 83: 51– 73. Google Scholar CrossRef Search ADS Dawley S. ( 2011) Transnational corporations and local and regional development. In Pike A, Rodriguez-Pose A., Tomaney J. (eds) Handbook of Local and Regional Development . London: Routledge. Defee C. C., Fugate B. S. ( 2010) Changing perspective of capabilities in the dynamic supply chain era. The International Journal of Logistics Management , 21: 180– 206. Google Scholar CrossRef Search ADS Dörry S. ( 2015) Strategic nodes in investment fund Global Production Networks: the example of the financial centre Luxembourg. Journal of Economic Geography , 15: 797– 814. Google Scholar CrossRef Search ADS Dunning J., Lundan S. ( 2010) The institutional origins of dynamic capabilities in multinational enterprise. Industrial and Corporate Change , 19: 1225– 1246. Google Scholar CrossRef Search ADS Eisenhardt K. M., Martin J. A. ( 2000) Dynamic capabilities: what are they? Strategic Management Journal , 21: 1105– 1121. Google Scholar CrossRef Search ADS Ernst D. ( 2009) A New Geography of Knowledge in Electronics Industry? Asia's Role in Global Innovation Networks . Singapore: Institute of Southeast Asian Studies. Felin T., Powell T. ( 2016) Designing organizations for dynamic capabilities. California Management Review , 58: 78– 96. Google Scholar CrossRef Search ADS Faulconbridge J., Hall S. ( 2009) Organisational geographies of power: introduction to special issue. Geoforum , 40: 785– 789. Google Scholar CrossRef Search ADS Fuller C. ( 2005) Corporate repeat investment and regional institutional capacity: the case of after-care services in Wales. European Urban and Regional Development Studies , 12: 365– 473. Fuller C., Phelps N. ( 2004) Multinational enterprises, repeat investment and the role of aftercare services in Wales and Ireland. Regional Studies , 38: 783– 801. Google Scholar CrossRef Search ADS Gavetti G., Levinthal D. ( 2000) Looking forward and looking backward: cognitive and experiential search. Administrative Science Quarterly , 45: 113– 137. Google Scholar CrossRef Search ADS Geppert M. ( 2003) Sensemaking and politics in MNCs: a comparative analysis of vocabularies within the global manufacturing discourse in one industrial sector. Journal of Management Inquiry , 12: 312– 329. Google Scholar CrossRef Search ADS Geppert M., Dörrenbächer C. ( 2013) Politics and power within multinational corporations: mainstream studies, emerging critical approaches and suggestions for future research. International Journal of Management Review , 16: 226– 244. Google Scholar CrossRef Search ADS Ghoshal S., Bartlett C. A., ( 1990) The multinational corporation as an interorganizational network. Academy of Management Review , 15: 603– 625. Giddens A. ( 1984) The Constitution of Society: Introduction of The theory of Structuration . Berkeley, CA: University of California Press. Henderson J., Dicken P., Hess M., Coe N., Yeung H. W. C. ( 2002) Global Production Networks and the analysis of economic development. Review of International Political Economy , 9: 4436– 4464. Google Scholar CrossRef Search ADS Herod A., Pickren G., Rainnie A., McGrath Champ S. ( 2014) Global destruction networks, labour and waste. Journal of Economic Geography , 14: 421– 441. Google Scholar CrossRef Search ADS Hess M. ( 2004) Spatial relationships? Towards a reconceptualization of embeddedness. Progress in Human Geography , 28: 165– 186. Google Scholar CrossRef Search ADS Hess M. ( 2008) Governance, value chains and networks: an afterword. Economy and Society , 37: 452– 459. Google Scholar CrossRef Search ADS Hess M., Yeung H. W.C. ( 2006) Whither Global Production Networks in economic geography? Past, present and future. Environment and Planning A , 38: 1193– 1204. Google Scholar CrossRef Search ADS Horner R. ( 2014) Strategic decoupling, recoupling and Global Production Networks: India’s pharmaceutical industry. Journal of Economic Geography , 14: 1117– 1140. Google Scholar CrossRef Search ADS Johnston S. ( 2005) Headquarters and Subsidiaries in Multinational Corporations . New York: Palgrave Macmillan. Google Scholar CrossRef Search ADS Katkalo V., Pitelis C., Teece D. ( 2010) Introduction: on the nature and scope of dynamic capabilities. Industrial and Corporate Change , 19: 1175– 1186. Google Scholar CrossRef Search ADS Kay N. ( 2010) Dynamic capabilities as context: the role of decision, system and structure. Industrial and Corporate Change , 19: 1205– 1223. Google Scholar CrossRef Search ADS Kleibert K. M. ( 2016) Global Production Networks, offshore services and the branch-plant syndrome. Regional Studies , 50: 1995– 2009. Google Scholar CrossRef Search ADS Kostova T., Roth K., Dacin M. ( 2008) Institutional theory in the study of MNCs: a critique and new directions. Academy of Management Review , 33: 994– 1007. Google Scholar CrossRef Search ADS Levy D. L. ( 2008) Political contestation in Global Production Networks. Academy of Management Review , 33: 943– 962. Google Scholar CrossRef Search ADS Liu W., Dicken P. ( 2006) Transnational corporations and ‘obligated embeddedness’: foreign direct investment in China’s automobile industry. Environment and Planning A , 38: 1229– 1247. Google Scholar CrossRef Search ADS MacKinnon D. ( 2011) Beyond strategic coupling: reassessing the firm-region nexus in Global Production Networks. Journal of Economic Geography , 12: 227– 245. Google Scholar CrossRef Search ADS Martin R. ( 2010) The Roepke Lecture in economic geography: rethinking regional path dependence: beyond lock-in to evolution. Economic Geography , 86: 1– 27. Google Scholar CrossRef Search ADS Monteiro L., Arvidssonl N., Birkinshaw J. ( 2008) Knowledge flows within multinational corporations: explaining subsidiary isolation and its performance implications. Organization Science , 19: 90– 107. Google Scholar CrossRef Search ADS Morgan G. ( 2009) Globalization, multinationals and institutional diversity. Economy and Society , 38: 580– 605. Google Scholar CrossRef Search ADS Morgan G., Kristensen P. ( 2006) The contested space of multinationals: varieties of institutionalism, varieties of capitalism. Human Relations , 59: 1467– 1490. Google Scholar CrossRef Search ADS Mudambi R. ( 2008) Location, control and innovation in knowledge-intensive industries. Journal of Economic Geography , 8: 699– 725. Google Scholar CrossRef Search ADS Murphy J. T. ( 2012) Global production networks, relational proximity, and the sociospatial dynamics of market internationalization in Bolivia's Wood Products Sector. Annals of the Association of American Geographers , 102: 208– 233. Google Scholar CrossRef Search ADS Pavlınek P., Zızalova P. ( 2016) Linkages and spillovers in Global Production Networks: firm-level analysis of the Czech automotive industry. Journal of Economic Geography , 16: 331– 363. Google Scholar CrossRef Search ADS Perkmann M. ( 2006) Extraregional linkages and the territorial embeddedness of multinational branch plants: evidence from the south Tyrol Region in northeast Italy. Economic Geography , 82: 421– 441. Google Scholar CrossRef Search ADS Phelps N. A., Lovering J., Morgan K. ( 1998) Tying the firm to the region or tying the region to the firm? Early observations on the case of LG in South Wales. European Urban and Regional Studies , 5: 119– 137. Google Scholar CrossRef Search ADS Phelps N. A., Fuller C. ( 2000) Multinationals corporations, intra-corporate competition and regional development. Economic Geography , 76: 224– 243. Google Scholar CrossRef Search ADS Phelps N. A., Waley P. ( 2004) Capital versus the districts: a tale of one multinational company's attempt to disembed itself. Economic Geography , 80: 191– 215. Google Scholar CrossRef Search ADS Phelps N. A., Wood A. ( 2006) Lost in translation? Local interests, global actors and inward investment regimes. Journal of Economic Geography , 6: 493– 595. Google Scholar CrossRef Search ADS Phelps N. A., Atienza M., Arias M. ( 2015) Encore for the enclave: the changing nature of the industry enclave with illustrations from the mining industry in Chile. Economic Geography , 91: 119– 146 Google Scholar CrossRef Search ADS Phelps N. A., Fuller C. ( 2016) Inertia and change in multinational enterprise subsidiary capabilities: an evolutionary economic geography framework. Journal of Economic Geography , 16: 109– 130. Google Scholar CrossRef Search ADS Pickles J., Smith A., Bucěk M., Roukova P., Begg R. ( 2006) Upgrading, changing competitive pressures, and diverse practices in the east and central European apparel industry. Environment and Planning A , 38: 2305– 2324. Google Scholar CrossRef Search ADS Pitelis C. N., Teece D. J. ( 2010) Cross-border market co-creation, dynamic capabilities and the entrepreneurial theory of the multinational enterprise. Industrial and Corporate Change , 19: 1247– 1270. Google Scholar CrossRef Search ADS Plank L., Staritz C. ( 2015) Global competition, institutional context, and regional production networks: up- and downgrading experiences in Romania’s apparel industry. Cambridge Journal of Regions, Economy and Society , 8: 421– 438. Google Scholar CrossRef Search ADS Sirmon D. G., Hitt M. A., Ireland R. D., Gilbert B. A. ( 2011) Resource orchestration to create competitive advantage: breadth, depth, and life cycle effects. Journal of Management , 37: 1390– 1412. Google Scholar CrossRef Search ADS Smith A., Pickles J., Bucek M., Pastor R., Beggy B. ( 2014) The political economy of Global Production Networks: regional industrial change and differential upgrading in the east European clothing industry. Journal of Economic Geography , 14: 1023– 1051. Google Scholar CrossRef Search ADS Somaya D., Teece D. J., Wakeman S. ( 2011) Innovating in multi-invention contexts: mapping solutions to technological and intellectual property complexity. California Management Review , 53: 47– 79. Google Scholar CrossRef Search ADS Stam E. ( 2010) Entrepreneurship, evolution and geography. In Boschma R., Martin R. (eds) The Handbook of Evolutionary Economic Geography . Cheltenham: Edward Elgar. Stones R. ( 2005) Structuration Theory . Palgrave: London. Google Scholar CrossRef Search ADS Storper M. ( 2010). Why does a city grow? specialization, human capital, or institutions? Urban Studies , 47: 2027– 2050. Google Scholar CrossRef Search ADS Sturgeon T. J., Kawakami M. ( 2010) Global value chains in the electronics industry: was the crisis a window of opportunity for developing countries? In Olivier Cattaneo O., Gereffi G., Staritz C. (eds) Global Value Chains in a Postcrisis World . Washington: World Bank. Teece D. ( 2007) Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance. Strategic Management Journal , 28: 1319– 1350. Google Scholar CrossRef Search ADS Walker R. ( 1989) ‘ A requiem for corporate geography: new directions in industrial organization, the production of place and uneven development’ Geografiska Annaler , 71: 43– 68. Google Scholar CrossRef Search ADS Werner M. ( 2016) Global Production Networks and uneven development: exploring geographies of devaluation, disinvestment, and exclusion. Geography Compass , 10: 457– 469. Google Scholar CrossRef Search ADS Whitley R. ( 2009). The multinational company as a distinct organizational form. In Collinson S., Morgan G. (eds) Images of the Multinational Firm . Oxford: Blackwell. Whittington R. ( 2001) What is Strategy - and Does it Matter ? London: Thomson Learning. Williams C., Lee S. H. ( 2011) Political heterarchy and dispersed entrepreneurship in the MNC. Journal of Management Studies , 48: 1243– 1268. Google Scholar CrossRef Search ADS Winter S. ( 2003) Understanding dynamic capabilities. Strategic Management Journal , 24: 991– 995. Google Scholar CrossRef Search ADS Yamin M., Sinkovics R. R. ( 2009). Infrastructure or foreign direct investment? An examination of the implications of MNE strategy for economic development. Journal of World Business , 44: 144– 157. Yang C. ( 2008) Strategic coupling of regional development in Global Production Networks: redistribution of taiwanese personal computer investment from the Pearl River Delta to the Yangtze River Delta, China. Regional Studies , 43: 385– 407. Google Scholar CrossRef Search ADS Yang Y., Hsu J., Ching C. ( 2009) Revisiting the Silicon Island? The geographically varied ‘strategic coupling’ in the development of high-technology parks in Taiwan. Regional Studies , 43: 369– 384. Google Scholar CrossRef Search ADS Yeung H. ( 2009a) Transnational corporations, Global Production Networks, and urban and regional development: a geographer’s perspective on multinational enterprises and the global economy. Growth and Change , 40: 197– 226. Google Scholar CrossRef Search ADS Yeung H. ( 2009b): Regional development and the competitive dynamics of Global Production Networks: an East Asian perspective. Regional Studies , 43: 325– 351. Google Scholar CrossRef Search ADS Yeung H. ( 2016) Strategic Coupling: East Asian Industrial Transformation in the New Global Economy . Ithaca, NY: Cornell University Press. Yeung H., Coe N. ( 2015) Toward a dynamic theory of Global Production Networks. Economic Geography , 91: 29– 58. Google Scholar CrossRef Search ADS Zollo M., Winter S. ( 2002) Deliberate learning and the evolution of dynamic capabilities. Organization Science , 13: 339– 351. Google Scholar CrossRef Search ADS © The Author (2017). Published by Oxford University Press. All rights reserved. For Permissions, please email: firstname.lastname@example.org
Journal of Economic Geography – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera