The Local in the Global: Regions, Employment Systems and Multinationals

This article examines how the international competition to attract and retain foreign direct investment shapes the governance of business and employment systems. Through an examination of global production networks and the changing role of the state in economic governance, it highlights the sub&#8208;national regional space as an important level of institutional adaptation. Specifically, it explores how regions organise themselves to compete for inward investment, and the potential role of industrial relations actors within this. It argues that both research into multinational companies, and of the governance of employment systems more generally, need to incorporate an analysis of regional competition for productive investment more fully into their analysis.


INTRODUCTION
Work and employment relations across the world are conditioned by the increasing ability of firms to choose where to locate. Indeed, the long-standing 'crisis of industrial relations' in much of the Global North is intimately linked to this face of globalisation (Thelen and Kume, 1999;Baccaro and Howell, 2011). Some faces of the challenges posed by internationally mobile firms are well known to labour-oriented researchers. For example, the phrase 'social dumping' is an established part of European-level discourse on industrial relations (Cooke andNoble, 1998, Hyman, 1999), given the well-known predilection of 'regime shopping' firms for locations with low labour costs, lax regulation and weak trade union organisation (Streeck, 1991;Logan, 2002). The more general challenges that the mobility, and foreign ownership, of productive capital causes for national collective bargaining systems have also been extensively researched (Bélanger and Edwards, 2006;Marginson, 2016), within a substantial literature on the coordination of industrial relations by multinational companies (MNCs) (for a review see Collings, 2008).
This article attempts to complement these literatures by developing an analysis of the struggle between locations to attract and retain jobs within global networks of the production and exploitation of value, and how this shapes the governance of business and employment systems. These struggles involve state and other governance actors at a wide range of geographical levels. Here, we focus on sub-national, regional levels of governance, as these often constitute arenas within which social actors of various kinds-governmental actors, the managers of firms, trade unionists and a variety of civil society actors-try to 'tailor' (Phelps and Fuller, 2001;MacKinnon, 2012) labour markets and business systems to attract productive capital. Our aim is to identify forms of agency of social, political and economic actors at regional levels of governance, as they attempt to attract and retain investment from MNCs. In doing so, we hope to contribute to the comparative political economy of industrial and employment relations in general, and work on MNCs in particular, by deepening its analysis of sub-national geographical levels of governance. In order to do this, although the current article should be read as a conceptual rather than empirical contribution, we will make some reference to the empirical findings of our cross-national comparative project examining MNC-region state relations in Canada, Ireland, Spain and the UK, which forms the basis of the subsequent empirical articles in this issue.
The article proceeds as follows. The first main section briefly locates the international competition for the location of production within the broader regulation of contemporary capitalism. The second section introduces the ways in which the organisation of firms, and in particular MNCs, in global production networks (GPN) has altered their relations with host economies. The third section analyses the role of sub-national regional governance actors in attempting to shape their labour market and broader business systems in response to these pressures, before exploring international similarities and differences in how this space is constructed, with regard to the specific challenges of foreign direct investment (FDI). We argue that this space, in different places, can involve, to different extents, the deployment of market, state or associational governance, but that each of these forms of governance operates in dynamic tension with the others. The concluding discussion explains the particular impact of integrating regional governance within an industrial relations analysis of MNCs, before attempting to draw out wider implications of our work.

THE CONTEXT OF INTERNATIONAL COMPETITION FOR THE LOCATION OF PRODUCTION
MNCs have an increasingly wide choice of locations in which to situate their productive activities. The context of the competition for investment between locations that this causes can be analysed either from a variety of levels-the productive dynamics of the firm, the overall dynamics of contemporary capitalism and/or the role of the state and transnational institutions in constituting and reproducing patterns of competition between geographies.
At one level, specific factors causing increased international competition for production can readily be identified. Thus, researchers have highlighted: the fuller enrolment of China, India and other parts of the Global South, into global circuits of capital (Gereffi and Sturgeon, 2004); the establishment of continental markets governed by supranational legislation, particularly the EU and NAFTA, alongside a broader process of liberalisation of the regulation of global trade (Duina, 2006); the financialisation of the firm, and the related ideology of the lean enterprise, making decision-makers in firms increasingly unwilling or unable to tolerate redundancy of capacity, or to maintain an organisational presence in parts of its value network where returns on investment are deemed insufficient (Rutherford, 2004;Morgan, 2014); and the multiple ways in which technology, often allied to product market liberalisation, has enabled markets, both for physical products and services, to be serviced remotely, and enabled much closer and more accurate surveillance of the international operations of firms (Yamin and Sinkovics, 2007).
All these factors are related to a broad transition at state level from protecting national productive capitalisms to securing positions in international contests for mobile investment. Geographies-cities, sub-national regions, nations, and to an extent supranational trade blocs-engage in competition to attract private-sector investment. More broadly, the policy decisions of states, at all these geographical levels, are inflected by efforts to develop or maintain international or global 'competitiveness' (Pedersen, 2010). In other words, the contemporary international organisation of production and the nature of state regulation are intricately intertwined: the state, at regional, national and supranational levels, has been and remains a constitutive actor in enabling and reproducing patterns of international competition, through active processes of re-regulation, both of product markets and of factors of production, including labour. Theorists of the state's relationship to international regimes of capitalism have variously conceptualised this as 'the competition state' (Cerny, 1997), 'the Schumpeterian workfare state' or related visions of a post-Fordist state (Jessop, 2013), or analysed the material and ideological processes of the construction of neo-liberalism as an organising principle for globalising capitalism (Harvey, 2007).
The effects of these developments on the possibilities available for domestic business and employment system regulation and coordination have of course been widely debated. Within the field of industrial relations, the implicit debate on these possibilities has often gone hand-in-hand with a debate on international convergence/divergence of national systems (Hyman, 1999;Colvin and Darbishire, 2013). On the one hand, IR scholars tend, rightly, to reject exaggerated claims about the decline of the national state as a coordinating actor. On the other, and despite the continued popularity of the varieties of capitalism argument as a device for labelling national economies, Hall and Soskice's (2001) vision of globalisation leading to a relatively benign process of nations building upon their path-dependent institutionalised strengths is likely to seem optimistic to most industrial relations scholars, given the extent to which liberalising pressures have been felt even in the core 'coordinated market economies' such as Germany (Holst, 2014), and the broader difficulties faced in persuading firms to tolerate 'constraints', even if their effects may be beneficial at an aggregate, national level (Marsden, 2015).
A more mitigated argument, implied by a number of theorists of the political economy of the state (Cerny 1997;Jessop 2004), is that contemporary economic globalisation has gone hand-in-hand with changes in thinking on the functions and forms of states, in their role as economic coordinators. In particular, this refers to a move away from Fordist/Keynesian-style coordination and regulation of the political economy by nation states, towards one in which states effectively compete with each other to provide supports to innovation and competitiveness (Jessop, 2004). This transition has been marked by a decline in nation states' capacity, and in some cases interest, in controlling the lead firms within their territories: former national champions were privatised and/or 'deregulated', and often partly or wholly bought out by international capital; and state supports to firms became increasingly regulated at supra-national levels. Thus, since at least the 1980s, states have, in general, placed less emphasis on the defence of national productive capital-whether in the form of the state-owned monopolistic firms of Fordist-era Western Europe, or private-sector national champions. Instead, more emphasis has been placed on attempts to capture relatively mobile investment, and more broadly at integrating domestic firms within GPN (see below).
In summary, there has been a broad process of transition in highly developed economies from a 'Keynesian welfare state'-the sort of 'Fordist' state particularly likely to introduce social democratically inspired 'constraints' as part of national compromises between capital and labour, towards the institutions of political economy being reconceived as international competitive factors (Cerny;Jessop, 2004;2013). This has important implications for those who study industrial relations and labour market regulation, which have sometimes been masked by a tendency to see 'neo-liberal' reforms to labour markets as simply being about deregulation (for previous arguments in this vein, see MacKenzie and Martinez Lucio, 2005). Particularly, it is important that the field of industrial relations is able to ask questions about what sorts of competitiveness are favoured by state policy and by labour market governance actors, with what legitimacy, and with what role for associational governance and industrial relations actors.

THE 'MICRO' CONTEXT: FROM THE 'FORDIST' FIRM TO THE GLOBAL PRODUCTION NETWORK
The discussion above reflects the still-continuing fallout from the long decline of 'Fordist' or 'Keynesian' methods of stabilising capitalism at national levels. Equally, and relatedly, the context of the competition for productive investment has been shaped, at micro-level, by the progressive decline, over a period of several decades, of what is variously labelled the 'Chandlerian' or 'Fordist' firm. For our purposes here, these constituted the large firms, often benefiting from monopoly rents at the national level and sometimes with a history of state control and ownership, which coordinated integrated production systems with internalised supply chains (Herrigel and Zeitlin, 2010). These firms historically benefited from semi-closed national markets, which provided the protection under which they tolerated-and indeed benefited from-the social compromises of the post-war Fordist era (Boyer, 1986). Fordism, both as a productive system and as a nationally delineated system of the regulation of industrial capitalism, developed unevenly, and with different levels of overall coherence in terms of its relations with pre-existing national trajectories of social regulation, across different countries of the global North (Zysman, 1996). This allowed for the co-existence of a number of somewhat different national patterns of coordination of the political economy, as reflected in the comparative capitalisms literature (Whitley, 1992).
Chandlerian/Fordist firms, particularly those from dominant states, frequently sought to expand their productive activities beyond national borders. As critical labour-focussed scholarship at the time argued, such firms did attempt to exploit national differences in industrial relations and wage levels (Malles, 1971): where market access permitted, subsidiary units would be opened in lower labour-cost countries (cf. Fernandez de Sevilla, 2010). However, typically the division of labour between plants was relatively unsophisticated, often being primarily based on the need to avoid tariffs and other barriers to trade, and to reduce transportation costs. Very often, foreign subsidiary units essentially replicated production in the MNCs' country of origin (Doremus et al., 1998). This gave rise to a literature on the 'branch plant', the subsidiary unit with relative protection from wider global competition, but sometimes poorly integrated into local production systems in host economies, offering relatively few possibilities for economic development. Sonn and Lee (2012), in reviewing this literature, usefully detail and evaluate some of the perceived issues of such plants: limited employment effects in, or spillovers to, the wider economy, poor quality jobs, a lack of R&D activity and exploitation of government incentives.
While such branch plants continue to exist where product market access requires their existence, the current era has been marked by a pronounced fragmentation of the value chains of lead firms. This fragmentation is both organisational and geographical (Herrigel and Zeitlin, 2010). Organisational fragmentation initially mainly concerned the outsourcing of relatively peripheral parts of firms' value chains, but has advanced to the point where core parts of the production networks such as R&D may be outside the hierarchical scope of leading firms (Spithoven and Teirlinck, 2015). At the same time, firms are increasingly able to locate their production across a wider range of geographies. Herrigel and Zeitlin usefully refer to the 'fineslicing' of the activities of lead firms' value networks. In other words, firms are increasingly able to-and are under intense financial pressure to-make fine-grained decisions about the most appropriate geographical and organisational locations for different elements of their production processes.
These processes have given rise to extensive literatures on the coordination of global value chains (Gereffi et al., 2005), and more recently the somewhat more theoretically sophisticated concept of GPN, as an increasingly important structural feature of the organisation of production, as Murray et al. (this issue) reflect. These literatures are widely used in labour-focussed geographical research (Selwyn, 2012;Cumbers et al., 2008;Rutherford and Holmes, 2007). They have also had some reflections in the field of industrial relations (Lakhani et al., 2013;Helfen and Fichter, 2013), although more typically in the literature on the global South than in mainstream discussion of the coordination of firms in the global North.
The GPN literature gives analytical priority to how lead firms coordinate international production through relations with a wide range of actors. These include dependent firms, the state at various geographical levels, education, training and research infrastructures, and a wide range of associational actors, including employers' associations and trade unions. It therefore attempts to locate an analysis of the roles of such actors within 'fragmented', yet still actively coordinated, international networks of the production and exploitation of value. In particular, GPN analysis seeks to discuss the ways in which the sub-national region-its preferred geographical unit of analysis -can attempt to maintain and develop relations that will enable it to take and preserve privileged positions in production networks.
Importantly, for GPN scholars, the strategies regional actors must take to develop favourable positions in GPN go beyond the more established notions of 'comparative institutional advantage' often found in the comparative capitalisms literature. In other words, comparative institutionalists often posit relatively 'static' forms of comparative advantage, generally at the national level, such as the familiar argument that German business and employment system institutions created a comparative advantage in more sophisticated areas of the manufacturing sector through high-quality vocational training and participative patterns of work (Sorge, 1991). GPN, meanwhile, sees the relationship between regions and firms as a more dynamic relationship of 'coupling' (Coe et al., 2004). This core concept implies that there is, or needs to be, a continual process of adaptation of regions' human and other resources to the constantly evolving needs of firms within GPN. These processes of adaptation are not limited to occasional institutional reforms: rather, they refer to continual, mutually adaptive, relations between international firms and geographically embedded state and other governance actors.
In arguing for an explicit consideration of how regional actors activate networks in order to respond to the demands of international competition, the GPN literature builds upon, and extends, earlier arguments within regional studies and economic geography that the global fragmentation of production was likely to cause the local network embeddedness of the subsidiary operations of MNCs to increase (e.g. Morgan, 1997). Routine branch plants, serving national markets and operating to wellestablished corporate production methods, could often function with limited interactions with local business system actors. However, where lead firms develop the capacity to make fine-grained decisions about the geographical and organisational location of their activities, subsidiary units increasingly compete to exercise functions, or 'mandates' on an international basis (Birkinshaw, 1996), which may be global or at the level of the supranational 'region'. Units which are unsuccessful in these contests are increasingly marginalised, and threatened in cases where access to markets no longer requires the existence of local/national 'branches'.
In analysing what this means for 'host' institutional actors, on the one hand, as more conventional industrial relations analysis has long recognised, the power of MNCs to 'shop' between the regimes and resources of nations and regimes has greatly increased. The fact that much productive investment is mobile makes it much more difficult for regional governments, businesses, employees and trade unions to reach long-term relationships with inward investors. However, this marked shift in the balance of power does not mean that the role of host institutions has been reduced to a simple avoidance of 'constraints'. Rather, host institutions have been repurposed to at least some extent, as the attraction and retention of the investment and jobs associated with the high value-added components of GPN requires that regional and national actors seek, through the agency of their business system actors, to provide embedded resources.

REGIONAL BUSINESS SYSTEM COORDINATION
The GPN literature argues that locations have become, with greater or lesser degrees of willingness, subject to international competition to provide embedded resources while at the same time minimising costs. Sub-national, regional governance actors, as generally relatively low-power levels of the state, often bear the brunt of efforts to develop and retain links to GPN, as it is at local and regional levels that specific contests for investment have most social and economic impact.
At a minimum, regional competition is manifested through the existence of inward investment bodies and economic development agencies. These explicitly attempt to differentiate the 'offer' of their region to geographically mobile firms (Phelps and Fuller, 2001). Such efforts may be confined to place marketing and (direct or indirect) subsidies. However, as GPN analysis implies (Coe and Yeung, 2015), more transversal attempts at building competitiveness also occur. Relevant regional actors-including governmental actors, training institutions, research bodies, employers' associations, development agencies and trade unions-may seek to gain the capacity to engage in what is variously labelled 'institutional entrepreneurship' (Crouch, 2005) or 'institutional experimentation' (Kristensen and Morgan, 2012): building on, or in some cases departing from, elements of their national business systems in order to seek localised advantages. This means that for a particular subsidiary unit of an MNC, the 'host' business and employment system may offer resources in a different manner, or to different extents, than what an overall picture of the national organisation of capitalism might reveal. It includes the localised agency of conventional business system actors, and of those institutions which have the specific purpose of encouraging FDI. Indeed, assuming that the national system is capable of achieving basic requirements (such as a sufficient degree of macro-economic stability, access to national and international markets, and industrial peace), managers in subsidiary units of MNCs are often much more concerned with what labour market and skills institutions actually deliver on the ground within the labour market(s) in which they operate, that with what a portrayal of a national system might reveal.
In practice, the relevant geographical scope of subsidiary units' labour markets varies quite radically depending on the occupational profiles required by the firm or business unit; the labour market for post-doctoral research scientists in the field of life sciences may well be genuinely international, for example. However, for the majority of occupations, the relevant labour market is fairly local, meaning that MNC units that require non-generic skills depend on the outputs of the skills ecosystems present at local and regional levels. New inward investors have the luxury of choosing between the outputs of rival ecosystems, and can choose between regions as much as they can between nations. The competition for replacement investment in existing plants, meanwhile, is generally between specific existing subsidiary units, embedded in specific local and regional labour market ecosystems, rather than only between nations.
It is also important that the analysis of industrial and employment relations in MNCs takes account of sub-national levels of governance in order fully to understand the power that MNCs have on host business and employment systems. Both the effects of firms' location decisions, and the broader impact of MNCs on employment and organisational practices in host economies, are most keenly felt at local and regional levels. Threats of large plant closures, in particular, may be bad news for national governments, but clearly have much more acute impact at the level of local labour markets. Prior research in the field of economic geography has argued that this leads to the establishment of a region state-MNC 'nexus' (Phelps and Fuller, 2001), in which sub-national state actors, and often other local actors including trade unions, become incorporated into the needs of local subsidiary units (Christopherson and Clark, 2007). This leads to a risk of what might be termed 'company town effects', particularly where regional governance actors are faced with a small number of large MNCs, as tends to be the case in peripheral industrial regions (see Gonzalez et al., this issue).
Additionally, it has sometimes been argued that the shifts in the functions of economic governance actors associated with a focus on international competitiveness have been, or should be, accompanied by shifts in the levels of governance. The nation state, in other words, becomes increasingly 'leaky' as a 'container of governance' (Brenner et al., 2003). Its coordinating capacities are fragmented, with some capacity transferred upwards (e.g. supranational or transnational institutions providing competition rules), and some downwards (with sub-national regions or localities charged with developing flexible infrastructure, and tailoring to local economic and social needs). Influential ideas on industrial policy, not least of which is the current EU institutionalisation of 'smart specialisation' through Regional Innovation Strategies (RIS3, see European Commission, 2014) clearly follow this script: regional governments are charged with identifying their potential competitive strengths, and how these might be operationalised, in association with producer and user stakeholders (see also Gonzalez et al., this issue).
However, the balance between geographical levels of governance, and thus the resources available to state and other governance actors at the regional level, remains heavily contingent on national systems of political and economic governance. Unsurprisingly, the extent to which regional institutional actors have the capacity to develop autonomous means of coordinating business systems tends to be greater in federal political systems than in more unitary or centralised ones. In other words, in examining the similarities and differences in regional responses to the competition for investment, it is important to take account of the fact that there are clear national differences in the extent to which these is a governance space below the national level which has sufficient autonomy to shape a 'regional business system', or at least to significantly modulate particular spheres of the national business system. Thus, among the cases examined in the remainder of this issue, we would argue that in Spain, the regional autonomous communities have the capacity to put into place somewhat different modes of economic governance within the overall national business system. While the industrial relations system remains national, region states, in conjunction with employers' associations and trade unions, have non-trivial capacities to make different choices about the extent to which their governance of economic development follows a somewhat more 'coordinated' or more 'liberal' pattern, depending on electoral coalitions and the localised strength of organised capital and labour. Region states are responsible for economic development, and work in conjunction with EU-level governance actors on the development of plans such as the 'smart specialisation' strategies on which European regional funding is based (Camagni and Cappello, 2013). The analysis of Gonzalez et al. shows that the degree to which organised labour has a role in regional economic development differs considerably between more social democratically-oriented and more liberally oriented regions.
The federal state of Canada allows for even greater differences in sub-national economic coordination at the provincial level. These have been particularly enthusiastically seized in the case of Quebec (Murray et al., this issue). The development of a Francophone 'national' identity since the 1960s has been intrinsically linked to a considerably less 'liberal' approach to economic development than is to be found in the rest of Canada. This has, to various extents over time, incorporated union and business elites in a wide range of mechanisms for social and industrial concertation. In fact, while at least at the level of a labelling device, Canada can be characterised as a 'liberal market economy', it is at best dubious whether Quebec can (cf. Bélanger and Trudeau, 2009). This degree of pursuit of alternative approaches below the level of the sovereign nation poses significant questions for cross-national comparative research in general, which we return to in the concluding discussion.
Elsewhere, as is well known, US states can effectively choose between continuing to apply New Deal derived industrial relations regulation, and pursuing anti-union 'Right to Work' legislation. While the anti-labour animus of governance actors in the Southern United States is powerful enough to exist independently of the need to compete for FDI, it is very clear that such legislation is frequently used as part of a low-wage, high-subsidy FDI regime (Brady and Wallace, 2000;Young, 2015). Equally, a different form of sub-national variety, of particular relevance to a discussion of MNCs and GPN, arises in the form of 'Special Economic Zones' or Export Processing Zones in a large number of developing and transition economies (cf. Hammer, 2010).
Regions of England (Almond et al., this issue) clearly lack the coordinating capacities of Spanish autonomous communities or Canadian provinces. While there may be latent possibilities in the remainder of the UK, particularly in Scotland, English regions are, to borrow Brenner et al.'s (2003: 16) formulation, 'regions in themselves' but not 'regions for themselves'. While the current UK national government has been keen to re-cast regions as 'powerhouses' or 'engines' (HM Treasury, 2015), it is fairly clear that talk of devolution to city-regions in England is not likely to be accompanied by the development of autonomous regional level approaches to the governance of business systems. Ireland (Lavelle et al., this issue) is a somewhat different case; in a small and highly centralised state, regions lack significant coordinating capacities. However, from the perspective of FDI attractiveness, Ireland scarcely needs to construct regional state-MNC nexuses, given its small size, the national state's unwavering commitment to a development strategy based mainly around the attraction of American FDI, and the subordination of large elements of the coordination of the national business system to this.
The extent to which regional actors have the capacity to inflect national characteristics of the coordination of business systems is thus heavily related to path dependencies in national systems of government. It is also not necessarily stable: at moments of crisis, regional efforts at coordination may be disrupted by the reassertion of national or supranational interests of the state or of powerful fractions of capital, as, for example, occurred in Spain when the regime of collective redundancies was centralised (and liberalised) at the national level as a response to the 2008 crisis, removing the right of regional governments to impose somewhat different requirements on firms seeking to engage in collective dismissals. We are not, therefore, arguing that the changing nature of international economic competition following the demise of Fordism leads automatically to greater levels of autonomy of regional business systems. However, the geographical concentration of specific firms and specific types of economic activity in particular places does mean that, at business unit level, managers are likely to be primarily concerned with what resources the business and employment system creates within specific geographical labour markets. Equally, the specific power that mobile firms have over governance actors is to an extent geographically concentrated in the places they operate. For both these reasons, it is important that the industrial relations analysis of the construction of 'host' economy effects takes full account of institutional agency at regional levels, whether or not this is accompanied by strongly autonomous regional governance actors.

Forms of regional governance of FDI
The above arguments imply that sub-national levels of analysis are important in attempting to understand how the international competition to host mobile production shapes business and employment systems. Here, we attempt to build on this by reflecting on what regions may do in order to try to capture mobile investment, whether through their direct efforts to attract FDI (e.g. through inward investment agencies, etc.), or through attempts to shape their business and employment systems in ways that build competitiveness in GPN.
Comparative research in industrial relations rests quite heavily on typologies which seek to categorise national differences in political economy into a small number of ideal types. The most prominent schema is the varieties of capitalism framework (Hall and Soskice, 2001). In order to situate our conceptualisation of the governance of FDI, it is perhaps useful to briefly explain some of the problems with this (see also Almond and Gonzalez, 2014).
While it is clearly convenient to be able to present the broad characteristics of national political economies using the shorthand of liberal, coordinated or mixed market economies, it is a very blunt tool to use in discussions of governance. For example, as Piore (2016) points out, the postwar construction of the archetypal 'liberal market economy' of the USA was 'not…directed by the market at all, but by large corporate organisation, trade unions, and government agencies working directly with one another…it operated at the micro level in a way that was consistent with the Keynesian macroeconomy'.
In reality, the social construction of business and employment regimes involves some combination of market, associational and state-led governance. While the relative weight of these may differ, the construction of more or less coherent regimes requires elements of all three (see Crouch's (2005) arguments on how 'complementarities' require the supplementing the predominant mode of economic coordination with other forms). Equally, it is important to recognise that all attempts to 'fix' the regulation of capitalism are crisis-prone (Boyer, 1986;Jessop, 2013), and the result of conflict, whether implicit or explicit. For this reason, while imagining ideal types is a useful process in theory-building, one should be very cautious before ascribing actually-existing cases to these. In a literal sense, all national and regional variants of capitalism are in fact 'mixed market economies'. Thus, here we present categories of action (market, state or associational-centred), with regard to sub-national state actions taken to attract FDI, to retain and attempt to embed MNCs already present within regions, and more broadly actions taken specifically to develop international competitiveness. We are not, however, arguing that empirical cases will follow any one of these categories of action in an exclusive or pure manner.
Market-centred governance of FDI would, in principle, imply a rather passive dependence on the competitive advantages of a location. Here, the (region) state would effectively abstain from questions of actively coordinating industrial policy, in the belief that such coordination is unhelpful, unnecessary or beyond its capacity. The 'state-MNC nexus', in this case, would be limited to attempting to regulate in ways that permitted firms to choose their own routes to competition as efficiently as possible. Regional development institutions would be limited to some kind of inward investment marketing office, shorn of any links to wider economic development. This route would exclude a meso-level role in economic development for trade unions, and would also be exclusionary of organised capital and other civil society actors and institutions.
It is possible, in principle, to combine a passive, non-interventionist approach to inward investment attraction with an institutionally dense regional or national business system. However, because regions that favour market-centred governance of FDI tend to be governed by those that espouse a market-centred approach to political economy in general, this approach often consists of marketing the advantages associated with overtly neo-liberal approaches, such as low taxation, low labour costs, or employer-favourable labour regulation. It is also associated with the rhetoric of 'not picking winners' (see Almond et al., this issue).
Regions can and do compete on the basis of being the 'most free market region' within their national economy. This may consist of relatively poor regions competing in a race to the bottom on wages and labour regulation, where national legislation permits this, or of metropolitan regions trying to compete on the basis of reducing institutional complexity and challenging what they perceive as barriers to the operation of the free market. The Madrid region, for example, has attempted to compete on the latter basis within the context of a relatively densely institutionalised national economy (Gonzalez et al., this issue), while the Thatcher-era UK competed on this basis at a European level. However, there are at least three important qualifications to this. First, at a conceptual level it is important to recognise that anything approaching a 'free market' approach can only be created by active state regulation and interpretation of what a 'market' approach actually implies within a pre-existing social context. Second, and more concretely, few governments, regional or national, are genuinely market-centred enough to resist the temptation to provide direct or indirect subsidies to firms. 'Employer-friendly' or low tax approaches are generally accompanied with incentives to firms, which are in some cases very extensive (for the USA see Brady, 2015), and represent very clear 'distortions' from any conceptualisation of a free market. Finally, even in political contexts where the 'liberal' right is dominant, there are social limits to non-interventionist policies. In particular, the closure of local plants is clearly bad news with potentially dangerous political consequences. This tends to mean that national or regional states governments which espouse free-market approaches tend to come under pressure to engage, however reluctantly, in more active policies where there are genuine threats to local investment.
State-coordinated governance implies significant interventions into the regional business system, specifically targeted at increasing the international competitiveness of the private sector. This is accompanied by active attempts by the state (usually but not necessarily at regional level), to engage with existing foreign direct investors in order to build relationships aimed at developing some sort of coherence between the needs of specific MNCs and the development needs of the region. This is done both to attempt to secure further investment from these firms, and to attempt to encourage MNCs to develop linkages within the region-with local firms, with the research and education sectors, etc.-in order to 'couple' the regional economy with higher value-added portions of GPN (Coe and Yeung, 2015). Such attempts may involve direct intervention by agencies of the region state. It may also entail the political leadership of associational governance-that is, attempts to bring business system actors together in the interests of the regional economy's international competitiveness.
One possibility frequently used in attempts to 'upgrade' regional economies is for the state to put in place fairly encompassing regional development agencies. These effectively seek to ensure the supply of whatever economic coordination is perceived as necessary. This may consist of the direct provision of services to correct for 'market failures', particularly in more liberally-oriented economies (Huggins, 2001;Pike and Tomaney, 2009). The English Regional Development Agencies under New Labour were a clear example of this. Here, a centrist government in a liberal economy created state-led regional agencies in response to a perceived need for active coordination in order to secure international competitiveness at regional levels, closely following contemporary scripts on 'knowledge-driven' competitiveness within a globalising economy. The English RDAs were responsible for the coordination (and funding) of other institutional actors to build internationally competitive skills and innovation systems. For MNC unit managers, such agencies also played an important role in reducing institutional complexity, acting as the face of the state at regional levels.
In particular, through their direct interactions with international firms, Regional Development Agencies sought to perform an interest aggregation role, bringing the skills needs of diverse firms together in order that mechanisms could be sought to fulfil an aggregate demand. Such a role is much less necessary in regions and countries where industrial relations actors are more capable of articulating such needs, and where managers and collective labour actors at firm or unit level have a greater degree of social conditioning that points them towards collective action in these areas.
As the abandonment of RDAs in England shows, such agencies are crisis-prone, and face serious problems of legitimacy, particularly in more overtly neo-liberal polities where state institutions in the productive sphere come under question. Such problems may be somewhat mitigated if representative collective actors are involved in the governance of such institutions. Equally, where such institutions exist, it is important that representative industrial relations actors have the capacity to be seen as central actors with which the development agencies have close coordinating relations. For development agencies to contribute to regional business and employment system construction in a sustainable way, their role would need to be interpreted as attempting to build up the capacity of associational governance actors, rather than simply substituting for them.
Associational governance in the domain of FDI concerns the willingness and ability of local business system actors-including local firms, social partners, and actors in the education and research sectors, etc.-to engage with locally established MNCs in ways that leverage location-specific advantages. It might also include networking between managers of different MNCs within the region. It thus includes both formal and informal associationalism.
The formal side of this involves meso-corporatist arrangements of various kinds. Examples include sectoral workforce committees in Quebec, which, at least in industries where trade unions are relatively powerful, can be seen not only as attempts to create a better coordinated skills system, but also as an important institution in bringing together senior managers of firms and trade unionists outside the arena of collective bargaining. In the highly collectivist regional environment of Asturias in Spain, neo-corporatist arrangements are represented by a peak-level collective agreement. This goes well beyond 'industrial relations' questions such as skills and training, as it effectively underwrites the social and economic development policy of the regional government, including the operations of a regional development agency which itself has tripartite governance.
As with broader national level neo-corporatism, such institutions may respond to the needs of regional governments to seek legitimacy, as well as attempting to discipline industrial relations actors in the needs of regions subjected to international competition within GPN. Regions may also seek to build formal associationalist governance-beyond the 'average level' of their national states-as a form of expression of belief in social democracy, as a response to the concentration of productive investment in heavy industry, and/or as an expression of a territorial social identity.
Less formal types of associationalist governance include attempts by region-states or development agencies to encourage productive relations between firms in related sectors through cluster or 'smart specialisation' strategies, which aim to ensure that firms and other regional actors participate in collective thinking about the developmental needs involved in their areas of economic activity. Such initiatives may either be an attempt at building dialogue, in cases where this is lacking, or at channelling existing associations and local/regional networks towards efforts at building international competitiveness.
Finally, if the attraction and retention of FDI involves the provision of financial aid to firms, then it is possible for grants regimes to seek to bolster associational governance of one form or another. That is, aid can be accompanied by requirements to embed firms within the region. Typical examples include cooperation with local education and research institutions, and attempts to develop relations with the regional SME sector within production networks. The case of Quebec represents a particularly articulated version of a subsidy regime oriented towards embedding firms and developing regional economies.
In summary, the balance between market, state and associational governance differs between regions and nations. The specific governance of the FDI-state nexus may follow broader patterns of coordination of the wider national or regional political economy, or may attempt to complement them (Crouch, 2005), by compensating for their perceived weaknesses. The processes of exclusion of collective actors in market-centred governance, the shaping of labour markets involved in state-coordinated governance, the potential role of trade unions and other industrial relations actors in the associational governance of FDI, and the tensions between these forms of governance, all merit further attention within the field of industrial relations.

CONCLUDING DISCUSSION
There has been much discussion of how employment relations in MNCs is structured both by the selective transfer of practices drawing upon country-of-origin repertoires (Ferner, 1997), and by the resistance or hybridisation that takes place when HQ-derived policies are transferred to operations in foreign host economies (Collings, 2009). However, while the analysis of firm-level transfer, and its role in shaping social space within domestic employment systems, remains necessary, the arguments above imply that it is not sufficient to an understanding of the interplay between international firms and domestic employment systems. Work on industrial and employment relations within MNCs (and internationally exposed firms in general) needs to integrate a more adequate treatment of the context in which contests for investment take place, and of the role of actors both within and outside the firm in this.
Within the MNC, the processes of fineslicing indicated above mean that subsidiary managers, often alongside worker representatives and others with an interest in defending jobs within communities, must increasingly compete for supra-national mandates. Arriving at competitive positions in these 'global games' (Kristensen and Zeitlin, 2004) often requires various coalitions between the MNC subsidiary unit and locally embedded actors, which may involve collective labour, alongside other actors in vocational education and training, the regional innovation system, and, as we have indicated, the local and regional state (Kristensen and Morgan, 2012;Geppert and Dorrenbacher, 2014).
MNC subsidiary-unit actors differ in their capacity to do this. Different management structures, and different host business system configurations, may lead subsidiary managers either to tend to perceive their interests as being in simply following the orders of headquarters, or to perceive that they should attempt to use links with local actors outside the firm to boost their competitive position in a more entrepreneurial way. Morgan and Kristensen (2006) argue that the former, more reactive, outcome is the more likely possibility in centralised MNCs, and/or in cases where the local business system's resources are essentially limited to the absence of constraints-that is, where the basis of competitiveness is light regulation, low costs, flexibility, etc. The more proactive subsidiary unit management approach is more likely to emerge in more 'federal' firms, and where the local business ecosystem possesses unique or scarce resources which can be drawn upon by those with specific local expertise. Morgan and Kristensen use this as the basis for their distinction between 'boy scouts' and 'subversive strategists', with the strong implication that the latter are more likely to favour positive forms of regional development. Relatedly, the work of Bélanger et al. (2013), explores, using survey data, the relations between the embeddedness in parent MNC networks, on the one hand, and local economy networks, on the other, of foreign MNCs' subsidiaries in Canada, and finds that local network embeddedness is associated with greater degrees of subsidiary unit autonomy from central MNC control.
An analysis of how subsidiary unit managers, but also trade unions in their joint role as unit-level, sectoral and regional actors, interpret the specific nature of intracorporate competition within their subsidiary units, and attempt to activate the potential advantages of local and regional networks, is an important avenue for future research. More generally, understanding location and labour market structures on a case-by-case basis represents a significant research agenda which can better link studies of industrial relations and human resource management in multinationals with an analysis of the territorial social dynamics of globalising production. This involves going beyond a standard institutionalist analysis based on how industrial relations and human resource practices relate to the norms of host and host country systems, and integrating an analysis of the purposes of firms' investments in specific places, the scale of the labour markets it is drawing from, to what extent subsidiary unit competitiveness relies on interactions with local business and employment system actors, and what skills are necessary in order to access such resource networks.
The attempts of communities and societies to capture the high-value added parts of GPN should not be seen as necessitating 'deregulation'; rather, in order to cohere with such networks, the formal regulation of business and employment systems is increasingly supplemented or inflected by actions taken by the state and other governance actors, often at relatively local levels, to attract and embed internationally mobile capital, and more broadly to develop 'competitiveness'. As we have indicated above, and as the other articles in this issue explore in more detail, this may be done through a range of specific mechanisms, which may be more or less socially inclusive, and with more or less role for traditional industrial relations actors.
In some places, sub-national regions have significant political and social capacity to attempt to govern these relations, making the regional analysis of business and employment systems crucial to an understanding of broader institutional dynamics. This is clearly the case in federal states, and may increasingly be the case in other places that seek regional autonomy or indeed national independence. In some of these cases, it is worth asking the question as to whether 'sub-nations' merit more serious consideration as business systems. In other words, if comparative research consists, following the argument of one standard manual, of the 'attempt to study particular issues or phenomena in two or more countries with the express intention of comparing their manifestations in different socio-cultural settings, using the same research instruments' (Hantrais and Mangen 1996: 1), then sub-national territories with significantly different socio-political and/or national identities to those of their national states may acquire the status of having 'different socio-cultural settings' to at least some extent.
More generally, though, processes of articulation between territories and GPN take place, in one form or another, whether or not sub-national regions represent an important level of political governance. They involve a wide range of actors-institutional industrial relations actors, institutions governing skills development, all geographical levels of government, education and research institutions, sectoral associations and development agencies, to name but a few. These are organised differently in different places, with logics that may differ according both to the degree of institutional density of the political economy, and to the national pattern of relations between local, regional, national and supranational levels of political governance. Understanding these, in order that industrial relations research can inform actors' responses, requires in-depth analysis of how places' articulation with these networks is formulated.