INTRODUCTION

MNEs are nested in multiple domains (Heidrenreich, 2012; Kostova, Roth, & Dacin, 2008; Meyer, Mudambi, & Narula, 2011). Most obviously, they originate in a particular country-of-origin or national business system (NBS) from whence they developed their original structures and strategies and from where they are managed and controlled (Hall & Soskice, 2001; Lane, 2000). They are also nested in the institutional domains of their host countries in which they further grow and are shaped, but from which they may take flight if certain inimical circumstances come to prevail (Hayden & Edwards, 2001; Pauly & Reich, 1997; Tempel, Edwards, Ferner, Muller-Camen, & Wächter, 2006). Further they are nested in international or transnational domains wherein they encounter trade agreements, international regulations, competitive pressures and managerial “best practices” (Morgan, 2011; Tempel & Walgenbach, 2007). Finally, they are nested in a complex of industrial relations (IR) institutions and processes in which managerial workforce relations are played out at multiple levels and points of interaction: subsidiary – HQ, MNE – state (home and host) authorities, and MNE – global hegemonic influences (Ferner, Edwards, & Tempel, 2012; Ferner & Tempel, 2006; Ferner et al., 2011). Thus MNEs are embedded in a complex constitution of varying ecosystems and are inherently constituted as “transnational social spaces” (Morgan, Kelly, Sharpe, & Whitley, 2003).

The key question for international business (IB) and international human resource management scholars is how does such a complex of multiple – and perhaps non-complimentary – institutional contexts constrain and/or resource MNEs in their attempts to maintain internal organizational coherence and in the coordination and transfer of common management policies? We address this question by examining the intriguing case of a large MNE headquartered in a major emerging economy (Brazil) and its transfer of a US-sourced global HRM model to its subsidiaries located in more robust mature institutional spaces (Canada, the United Kingdom, Norway and Switzerland). There has been little work on the movement of firms from developing to developed economies. There has been considerably more written on MNEs from advanced countries moving into developing ones. Therefore the empirical material is novel. Our theoretical starting point is to inquire whether the subsidiaries in the developed countries were able to resist the transfer of foreign practices which risked undermining traditional practices and norms and which threatened institutional reconfiguration. We were mindful, however, of the risk of assuming a strong bifurcation between business systems from developed and developing countries. We thus segment advanced societies on the basis that they are not all equally integrated and robust, and that there is also wide variation in the levels of development and complexities of business systems within one nation. We thus explore the case of transfer to four variously configured host country systems, two of which might broadly be conceived as being institutionally “liberal” – Canada and the United Kingdom – and the other two – Switzerland and Norway – “robust” or “coordinated” (Hall & Soskice, 2001), to assess the variable “collision” of different institutional effects. The parent country’s NBS (Brazil) is a particular manifestation of a “hierarchical market economy” (Schneider, 2013). It is also characterized by weak formal institutions and by “negative complementarities”. This has prompted many of its MNEs to develop company-specific organizational practices (Khanna & Palepu, 2006) by drawing upon “foreign” best practices that are then transferred to their overseas subsidiaries (Fleury, Fleury, & Borini, 2013; Tanure, Evans, & Cançado, 2010).

The case industry (mining and metals refining) has been subject to major internalization pressures in recent years (mergers, acquisitions and consolidation), which has had significant implications for the industry’s structure, specifically in the rise of large global players. This in turn has inhibited the robustness of local institutional settlements, specifically the power of the state and organized labor. We find that these secular trends (internationalization, industrial restructuring, and the rise of “strong” dominant industry players and the attendant dependence of host states on foreign firms) were significant in facilitating the actions of the case MNE in pushing through its US-derived HRM practices in its overseas subsidiaries. This outcome was intriguing and would have been difficult to predict solely from a reading of the existing literature and from our a priori understanding of the interaction of management practices between national business systems of developing and developed economies. In seeking to address this puzzle we chose to adopt a contextualized analysis, which involved conceiving of the institutional context as a “contested terrain” (Edwards & Bélanger, 2009) in which different actors within and beyond the MNE seek to identify and exploit the “institutional space” of the host context, deploying a variety of resources as they do so.

As such, this article is positioned as a contribution to the growing stream of IB research on power and micro-politics in the MNE (Dörrenbächer & Geppert, 2011; Ferner et al., 2012; Geppert & Dörrenbächer, 2014). Hitherto, a great deal of the literature on the cross-national transfer of management practices has focused on the notion of “institutional duality”; that is, the extent to which MNEs’ practices are shaped both by the institutional terrain of the MNE itself and that of the host context in which it operates (Kostova, 1999; Kostova & Roth, 2002). This influential literature, often referred to as neo-institutionalism, argues that these various institutional contexts exert rival isomorphic pressures on the transfer of practices to a given subsidiary. The greater the “institutional distance” in the “institutional profile” of the home- and host-country institutions, the more challenging transfer is perceived to be. There are two interrelated weaknesses in this literature. They are first its inadequate treatment of politics and power relations; and second, its under-theorization of host country effects. In particular, there is little consideration of the manner and the conditions under which MNEs use their power resources to marshal host country institutions to influence the transfer process, and, in turn, how subsidiary management and employees may also seek to recruit local institutional resources to resist or modify transfer (Almond, 2011; Ferner et al., 2012). Another influential literature that of comparative institutionalism places particular store on the influence of country-of-origin effects (Almond & Ferner, 2006; Whitley, 2009), but it can be criticized for its static and over-socialized conceptualization of power and its failure to bring to life the dynamic political processes involving MNEs’ managers and other stakeholders (Geppert & Dörrenbächer, 2014).

In brief, then, this article is concerned to explore how a mix of influences derived from the institutions of an emerging NBS, the institutions of variously configured host countries, countries’ varying dependence on foreign investment, together with the size and place of a subsidiary in the organization’s value chain interact to shape the transfer of HRM practices within a MNE. This is an enormously complex matter to address; such effects cannot be read-off in any predetermined manner. Neither can transfer be seen to be derived from some “rational” or “free” decision-making process, whereby the transferred HRM practices are somehow deemed to be appropriate or the “best.” They may indeed be that, but we seek to impress, as have others in recent contributions informed by a similar perspective (e.g., Geppert & Dörrenbächer, 2014), that their shape ultimately arises from the relational interplay of various actors operating at different levels with diverse interests, ideologies and identities. Further, we endeavor to stress how the complexion of such effects change and evolve; they are not fixed or invariable. In turn, it matters a great deal that we understand how such circumstances of flux and fluidity are perceived by, and influence the power resources and postures of, the principal actors. The article makes an important contribution to the evolving literature in IB in its adoption of an actor-centered perspective together with a context-sensitive mode of explanation.

We focus on HRM practices – and in particular on pay and performance management – for a number of reasons. First, the transfer of common HRM practices can be viewed as a specific instance of HQ – subsidiary relations (Ferner et al., 2012). Many MNEs tend to regard pay and performance management as central to their employment systems and are likely to adopt a highly centralized approach (Almond et al., 2005). Thus its study is an important element in our analysis of how MNEs manage across national boundaries. Second, the shaping of HRM practices within MNEs is conducted at the intersection of home and host country environments, and as such provides a test-bed for examining the relative influence of parent-country and local isomorphic pressures. Third, and by corollary, they provide a means for understanding how MNEs act as agents of change by introducing novel practices into their subsidiaries and thence into the host business systems (Ferner & Quintanilla, 2002). Finally, given the traditional links between the regulation of pay and union representation, the transfer of common pay and performance management systems can be expected to demonstrate a variety of diverse and complex political processes.

LITERATURE REVIEW

A great deal of the existing institutional literature on the diffusion of practices within MNEs conceives of country-of-origin and country-of-operation effects as discrete influences – the former is available as a resource for the MNE to exploit, the latter a resource for the local subsidiary to seize; and that they can be predetermined a priori and that they are (largely) fixed and stable. We are critical of such assumptions as they provide little purchase in capturing the complex and multilayered nature of power dynamics in relations between an MNE’s HQ and its subsidiaries. Moreover, such previous contributions have tended to see power as a repository of institutional and organizational structures (Geppert & Dörrenbächer, 2014). Our preference is to lay stress on the action of power and not just its structuring. However, we do hold that to fully understand power relations we must understand the constant dialectic between these two elements (Clegg et al., 2006). Thus we argue for a relational, dynamic and multilevel explanation of institutional effects. Such an approach permits an analysis of the transfer of management practices as being fashioned through a relationship of power involving the interplay of interests, identities and ideologies of various actors within and beyond the MNE. Ultimately, we argue that the shape of MNCs’ HR practices must be understood in the context of the wider macro-political setting (size of MNE, nature of the industry, distribution of assets and employment, and the degree of the dependency of the host country) and the conduct of micro-politics within the MNE.

We begin with a consideration of the two dominant theoretical frameworks in the field, the neo-institutionalist and comparative institutionalist perspectives. Although the division between the two may be somewhat crude, it has been commonly used to differentiate the idea of duality and interaction highlighted respectively by American and European scholars (Ferner & Tempel, 2006; Kostova, 1999; Whitley, 1999).

Neo-Institutionalism

In broad terms, neo-institutionalists argue that subsidiaries encounter competing isomorphic pressures within the particular “organizational fields” in which they operate. Such pressures are seen to be “dual” in that they include both the institutions of the host-country and the institutional terrain of the MNE itself (Kostova & Roth, 2002). The extent to which subsidiaries might be pulled between these so-called institutional “pillars” (Scott, 1995) is perceived to be determined by the “institutional distance” between the “institutional profile” of the MNE’s parent-country and the institutional context of the host-country. Where this distance is small, the transfer of HRM practices from the MNE’s HQ is seen to be relatively unproblematic and any new practices can easily be “internalized” or assimilated by the subsidiaries’ staff.

Together with such “coercive” or “regulative” forms of institutional diffusion, institutionalists identify two other forms of isomorphism, “mimetic” and “normative” (DiMaggio & Powell, 1983). With the former organizations model themselves, particularly in contexts of uncertainty, on other firms they perceive to be more “legitimate” or successful. Models of HRM “best practice” are then diffused as companies imitate one another’s policies, perhaps by drawing on the expertise of consulting firms. Normative isomorphism stems primarily from the professional formation of management together with the growth of professional networks that span organizations and across which new models of HRM can diffuse rapidly. As such, neo-institutionalists believe MNEs imitate particular HRM practices less because they are nested in NBSs and more because they are promoted and diffused through dominant management models and professional networks. However, they do not propound necessarily that such “best practices” are more efficient, à la more rationalistic approaches (i.e., the eclectic OLI paradigm of Dunning, 2000; Dunning & Lundan, 2008), but rather that their adoption helps to ensure an organization’s legitimacy and thereby its future survival (Geppert & Dörrenbächer, 2014).

In a later paper, Kostova et al. (2008) query many of the assumptions of their earlier work. They suggest that MNEs are capable of escaping the clutches of their external institutional context and of developing an “intra-organizational institutional field.” Isomorphism is now seen to emerge from a managerial desire to disseminate and reinforce a shared global HRM model and not from any necessary quest for legitimacy. They argue such selected practices may be sourced from a MNE’s parent-country or from organizations that are identified as “global class.” Any observed similarity between firms’ HRM practices is seen then to arise less from compliance with external isomorphic pressures and more from the strategic choices pursued by MNEs.

Neo-institutionalist approaches have been criticized principally on two counts: first for their neglect of considerations of power, interests and competing value systems; essentially of what is “at stake” for actors when a MNE transfers its HRM practices to its subsidiaries (Ferner et al., 2012; Geppert & Dörrenbächer, 2014); and second, for their disregard of country-of-origin effects. We turn now to consider the influence of the latter by reviewing the contribution of comparative institutionalist theory before returning to address the issue of politics and power relations.

Comparative Institutionalism

Comparative institutionalism is derived largely from economic and political sociology (Jackson & Deeg, 2008; Morgan, Campbell, Crouch, Pedersen, & Whitley, 2010; Whitley, 1999, 2009). It seeks to understand MNEs’ behaviors in the context of the interaction of home- and host-country influences. Home-country effects refer to the manner in which institutional and ideological legacies generated by crucial episodes in a country’s historical development – including processes of industrialization combined with remnants of pre-modern forms of social organization – act to shape companies’ business strategies and HRM practices. Such effects are seen to be rooted in constellations of influences derived both from “hard” regulative structures and “softer” cognitive or normative understandings (Ferner, 1997; Hall & Soskice, 2001; Hayden & Edwards, 2001; Whitley, 1999). MNEs are said to act in a distinctively “national” way. For example, MNEs from the US are identified as sharing particular sets of characteristics that include strong centralized management structures (reflecting the primacy of shareholder interests), the relatively weak influence of lower levels of management and employees, and the transfer of standardized HRM practices to their subsidiaries (Almond & Ferner, 2006).

Echoing earlier IB research (Pauly & Reich, 1997; Ruigrok & van Tulder, 1996), recent evidence does indeed suggest that MNEs remain firmly rooted in their country-of-origin. The presence of strong NBS centralizing influences persists, for example, with corporate boards remaining largely the preserve of home-country nationals, and with key decisions in respect of capital allocation, control mechanisms and the remuneration and career advancement of subsidiary managerial personnel being overseen by corporate management (Dörrenbächer, 2000; Dörrenbächer & Geppert, 2009; Egelhoff, 2010; Ferner et al., 2012; Harzing & Sorge, 2003). However, in matters relating to IR and HRM, such as wage determination, performance management and employee participation, MNEs have been found to be more susceptible to the imprint of a host-country’s institutional framework. In the specific case of performance management MNEs are advised to be “culturally sensitive” (Schuler & Rogocsky, 1998); that is, to adapt their policies to suit local traditions and norms. While such cultural sensitivity may be less urgent in liberal market economies where individual performance pay and appraisal are often taken-for-granted, in other institutional contexts they have been challenged on grounds of social equity, fairness and solidarity (Ferner et al., 2012: 167). In Liberman and Torbiörn’s (2000) study, for example, while the transfer of a MNE’s standardized performance management system was accepted in the United Kingdom, Poland, Turkey and Russia, in Germany the works council rejected individualized employee assessment, and, in France, assessment was reduced to the use of impressionistic criteria and had no impact on compensation, with similar outcomes evident in Sweden.

Within a comparative institutionalism perspective such varied outcomes are linked to different actors’ interests and political contests. While this is important and represents an important advance on neo-institutionalism, there is a tendency to view such processes as being ingrained within the institutional features of the MNE. Power and political dynamics are thus conceived in a static and over-socialized manner, whereby they are seen to follow, as Geppert and Dörrenbächer (2011: 22) put it, the “logic” of relatively stable and coherent host (and home) country “effects.” They are much more than that. We turn now to consider the role of power and politics in MNEs.

Considerations of the Political Sphere and the Shape and Tenor of Power Relations within the MNE

We begin with a discussion of the interaction of home- and host-country effects as constituting a distinct (macro) political sphere. In circumstances where, as Streeck and Thelen (2005: 14–29) put it, MNEs have been constrained by the host context to act as “rule-takers,” they may still possess the means to “bend” host institutions to their purpose either by engaging in institutional “layering” (by-passing existing structures through the establishment of parallel institutions) or institutional “conversion” (redirecting institutions to serve new ends) such that their managerial practices are very different from the “ideal pattern of a rule” as originally enacted by policymakers. Accordingly, there is evidence, even in rigid labor markets, of MNEs being able to evade host-country constraints by exploiting “spaces of uncertainty” (Ferner, 2010), wherein a variety of HRM strategies prove feasible (Tempel et al., 2006). Thus adaptation to the host country context may evolve in nationally distinctive ways, where particular elements of MNEs’ home-country practices are preserved. Ferner, Quintanilla, and Varul (2001), for example, observe that while German MNEs’ attempts to develop cooperative employment relations led them to bypass unions in Britain, in Spain they were guided towards working with the comité de empresa. Two implications from this research warrant highlighting: one, there is the importance of management choice and orientation; and two, the room that is available to management to negotiate the manner in which regulatory constraints shape HRM policy-making in practice.

Streeck and Thelen (2005) also conceive of MNEs as “rule-makers,” where, by actively lobbying and influencing policymakers, they endeavor to recast the design of institutional settings to their liking. The evidence suggests that the opportunity for MNEs to exercise such rule-making influence hinges on their “locational flexibility” (Almond et al., 2005) and whether they are of a mind to engage in “regime arbitrage” or “regime shopping” (Streeck, 1997) – that is, whether they have the capacity to threaten to go elsewhere if some institutional accommodation is not made to their operations’ requirements. Consider, for example, the influence exercised by US-owned MNEs in Ireland over public policy formulation in respect of union recognition (Geary & Roche, 2001). Such dynamics should not be overstated, however: MNEs’ locational flexibility is likely to vary depending on their expansion strategy (market- or efficiency-seeking), on their subsidiaries’ roles and influence within the firm’s international division of labor and the associated skill levels and performance of local workforces (Edwards, Tregaskis, Collings, Jalette, & Susaeta, 2013; Meardi & Tóth, 2006), and the strategic capacity and intent of MNEs’ actions (Kristensen & Zeitlin, 2005). In such circumstances, then, where power resources are dispersed across the MNE, national subsidiaries and host institutions may be better positioned to shape a MNE’s employment and HRM practices (Morgan et al., 2003). But, as Bélanger and Edwards (2006) and Bouquet and Birkinshaw (2008b) observe (the latter in the case of “low-power actors”), it is one thing for local actors to possess critical power resources, but quite another for them to be aware that they possess them and have the capacity, including the legitimacy, to deploy them in a bargaining context. In summary, therefore, the interaction of home- and host-country effects constitute a distinct (macro) political sphere whereupon the capacity to transfer HRM practices must be seen to be highly contingent on the shape and tenor of power relations.

Interests, Power and Power Dynamics at the Micro-Level

We turn now to examine power relations at the “micro” level of the MNE. We draw on Lukes’ (2005) conceptualization of the different dimensions of power (see also Hardy, 1996 and Ferner et al., 2011; the italicized terms below are derived from the former). The first dimension refers to instances where power is used to attain one’s observable aims in opposition to those of another (power of resources). It refers to actual decisions made. The second dimension is the power of “non-decision-making.” Here powerful actors determine which issues are open to discussion and by what means and under what rules (power of processes). Conflict may not be overt in the sense of involving a direct clash such as a strike, but it is still observable in behavior, for example vocalized grievances (Edwards, 2006). The third dimension concerns the manner in which power is used to shape the preferences and interests of those with little power such that they may not express or even be conscious of their own interests (power of meaning). We begin by considering the power relations between corporate management and subsidiary management, before turning to relations between managements and subsidiary workforces. In both instances the orientations and loyalties of subsidiary management are critically important. These are likely to be shaped by their nationality (whether they are home-, host- or third-country nationals), their career ambitions, and their social embeddedness in local communities and networks.

The means by which subsidiary management influence corporate management may include a diverse range of power of processes, including “issue-selling,” “attention attraction,” “lobbying” and the “exercise of voice” (Bouquet & Birkinshaw, 2008a; Cantwell & Mudambi, 2005; Dutton, Ashford, O’Neill, & Lawrence, 2001; Gammelgaard, 2009; Ling, Baldridge, & Floyd, 2005). The merits of their case is also likely to depend on their ability to exploit their subsidiary’s “material” resources such as its place in the corporate value chain, its ability to deliver profits, meet targets, exercise control over critical resources, or perhaps simply from their possessing the wherewithal to navigate through a complex host environment (Bouquet & Birkinshaw, 2008b; Dörrenbächer & Gammelgaard, 2011; Dörrenbächer & Geppert, 2011). It is also likely to hinge on how corporate management interpret local management’s behaviors; whether they are seen as being cooperative or hostile, and on whether they adopt retaliatory postures to weaken the subsidiary management’s negotiating hand. We are thus drawn to consider corporate management’s power of resources, which include, for example, their authority to recast a subsidiary’s mandate, to reshape value chain configurations, to reduce their dependence on a subsidiary, or to dismiss local management.

The second form of micro-politics is that which Burawoy’s (1985) refers to as the “politics of production”; that is, the relations of power and the dynamics of control and consent in interactions between managements and subsidiary employees (Bélanger & Edwards, 2006). The starting point is the transfer of HRM practices may differentially impact on workers and managers’ interests. Resistance may emerge, as Ferner et al. (2012) put it, under conditions of “criticality,” where the transfer of a practice disrupts existing accommodations between management and employees and/or affects workers’ economic security. Consider, for example, the tensions associated with the transfer of HRM practices to the British subsidiaries of Japanese MNEs studied by Elger and Smith (2006), which, in large part, were generated by employees’ perception of the differential distribution of benefits. Resistance has also been found to be greater where subsidiaries are acquired or merge and where imported HRM practices are seen to jar with pre-existing norms; where there is distrust and suspicion of a new management and where relations and identification with the MNE’s HQ are weakly embedded (Geppert & Dörrenbächer, 2014); and where local management act as “subversive strategists” (Kristensen & Zeitlin, 2005).

Finally, while it is necessary to highlight that managements and employees’ interests may vary and come into conflict across the MNE, it is also important to emphasize that their interests overlap and are shared. This is the value of Edwards and Bélanger’s (2009) concept of the “contested terrain”: although management and employees may have distinct sets of interests they can be expected to relate in complex and contradictory ways. Similarly, that which is deemed by one party to be worth defending or pursuing may change over time, as does actors’ ability to marshal and use power resources both within and beyond the firm to good effect. Therefore terrains of contest can be expected to be multiple and shifting. As Morgan and Kristensen appreciate, the MNE is a “highly complex configuration of ongoing micro-political power conflicts at different levels in which strategizing social actors/groups inside and outside the firm interact with each other and create temporary balances of power that shape how formal organizational relationships and processes actually work in practice” (Morgan & Kristensen 2006: 1473; our emphasis).

From this review of the literature the following conclusions are derived. First, the process of the transfer of HRM practices within a MNE cannot be read-off in any predetermined manner from the institutional features of its home- or host-country. Neither can it be seen as the outcome of some “rational” or “free” decision-making process where given practices are deemed to be appropriate or the “best”; it is much more than a matter of mere organizational design. It is above all else, a relationship of power, a political “game” (Geppert & Dörrenbächer, 2014) involving the interplay of interests and the possession and deployment of power resources by a variety of actors within and without the MNE (Ferner & Tempel, 2006: 31). Second, these political processes are played out at multiple levels such that there is scope for kaleidoscopic variation in the definition, interplay and pursuit of organizational interests (Ferner, Quintanilla, & Sánchez-Runde, 2006). MNEs are thus best conceived of as social constructs enacted by powerful actors, which bring diverse and sometimes contradictory “contextual rationalities” into play when contesting and negotiating the methods and degree of local adaptation (Geppert & Dörrenbächer, 2014: 235). By corollary, political processes – be they macro- or micro-political processes – are not separate from structural forces, but represent the working out of responses to them (Edwards, Ferner, & Sisson, 1993). And, to reiterate, their complexity is accentuated by their being played out across multiple levels of the organization and across different institutional regimes. Any consequent political accommodations are therefore likely to cut across the boundaries of the firm. Political tensions are pervasive to the MNE (Morgan et al., 2003; Kristensen & Zeitlin, 2005; Bélanger & Edwards, 2006). Third, the relationship between MNEs and the institutional contexts which they inhabit is dynamic. They co-evolve in consort (Cantwell, Dunning, & Lundan, 2010); MNEs are at once “rule-takers” and “rule-makers” (Streeck & Thelen, 2005). Finally, the implications of adopting such a view for the study of the dynamics of HQ – subsidiary relations and the transfer of HRM practices are manifest – it requires a “bottom up” and actor-centered perspective, which obliges us to inquire “how different identities and interests of key actors stabilize and destabilize established institutional, cultural and organizational structures” (Geppert & Dörrenbächer, 2014: 236).

RESEARCH APPROACH

Our research approach is informed by a critical realist epistemology. Within IB studies critical realism (CR) remains a relatively novel approach where otherwise the positivist paradigm predominates (Piekkari, Welch, & Paavilainen, 2009; Saka-Helmhout, 2014). CR draws primarily from the works of Bhaskar (1978). Critical realists criticize positivists for producing “thin” accounts of research phenomena and for focusing on empirical regularities, which fail ultimately to account for why things occur as they do (Edwards, 2005). A key insight is that reality is more complex and multiply sequenced than may be apparent in “raw” observations and that actual causes of events may neither be manifest nor any longer observable. Ultimately reality is seen to be “multiply determined” and that multiple causes must be teased out from detailed explorations of the setting (Bhaskar, 1978). Reality is also identified as being stratified and emergent; hence CR’s call for a depth or layered ontology (O’Mahoney & Vincent, 2014) whereby organizations are viewed as existing within layered and open systems, and must be studied in their context (Bhaskar, 2014). Complexes of influences can be expected to change both temporarily and geographically, and in unexpected ways. In being consistent with this epistemological positioning of our study, we are interested in identifying causal processes that affect the transfer of HRM policies from the MNE’s HQ to its subsidiaries. Such effects, CR stresses, are taken to be contingent; that is, causal mechanisms may or may not be activated and their activation will depend on the context in which they operate.

Thus our explanatory approach is to view the case MNE and its subsidiaries in their varying contexts and as interpretable wholes (Ragin, 2000). We attempt to produce holistic and particularized causal explanations (Piekkari et al, 2009). Our mode of theorization is that of “contextualized explanation” (Welch, Piekkari, Plakoyiannaki, & Paavilainen-Mäntymäki, 2011; Locke & Thelen, 1995), which we believe is ideally suited to explain phenomena in diverse national and institutional contexts. It is, as Harzing and Pudelko (2016) put it, by comparing contextual particularities across different locations that we are enabled to identify common phenomena.

Our theorizing process is best described as abduction whereby our understanding of events is derived from a continuous dialog – to-and-fro – between theoretical frameworks, data sources and analysis. From our review of the literature, we identified a theoretical prior understanding of the phenomena we were interested in, which was then confronted with our empirical observations. In some instances, the theory and data fitted together well, while in others they did not.Footnote 1 When the latter occurred, we were pressed to engage in a process of retroduction which entailed “reasoning backwards” (Reed, 2005: 1631) behind the observed outcomes to identify what produced them. As Reed puts it: such a logic of explanation has “to identify and account for the highly complex ‘compound effects’ of the interactions between, often contradictory and conflicting, powers and influences operating at different levels of analysis and drawn from a wide range of structures and mechanisms operating in dynamic situations and contexts.”

In our choice of case sites we conceded to a decidedly “ambiguous” or “unsettled” position. Although we strove to select matched-pair case sites in host countries that would most likely represent particular theoretical types – permissive host institutional regimes in the case of the United Kingdom and Canada, and constraining in the cases of Norway and Switzerland – and by so doing, set out to examine whether particular institutional settings might be expected to give rise to certain effects – we did not a have a fixed or definitive conception a priori as to what our cases were “cases of.” Rather, each was “permitted” to evolve and was not fully “cased” until the research was completed (Piekkari et al., 2009; Ragin, 1992). For example, at the outset of our research we focused in particular on identifying and examining the influence of institutional effects and, only after some time in the field, did we come to fully appreciate the import of what we term the “political relations of dependence.” In adopting this flexible approach we sought to avoid, in Burawoy’s (2013) terms, locking our case MNE and its subsidiaries into their national containers and to assume that influences pertaining to international competitive pressures, national business systems, host labor market institutions and the state were unchanging. And although our approach is less purely ethnographic in its orientation than that advocated by Burawoy (1998), we do adopt his extended case study method. As such, we attempt to locate the complexities of social relations within the case MNE’s subsidiaries in their political, economic and geographical context. By so doing, we seek to demonstrate how multiple, layered and interacting contextual influences mediate and ultimately refract “common” challenges into divergent (domestic) struggles. In short, as Locke and Thelen (1995) emphasize, seemingly similar challenges confront (subsidiaries in) different countries with varying degrees of intensity and/or at different historical moments. Thus seemingly parallel external forces cannot be expected to be of similar import in different national contexts. This alerts us to consider countries’ subsidiaries different “starting points,” as well as the significance of any change for the identities and power resources of the various actors involved. Thus in contextualizing a MNE’s transfer of common HRM practices, our attention is drawn to the possibility that such a process will set in motion fundamentally different struggles in different contexts.

Our attempt to adopt a matched-pair research design was also confounded by our subsidiaries’ varying roles and sizes. Holding contingencies constant, as Edwards, Almond, and Colling (2011) note, is not as straightforward an exercise as it might sometimes appear, and is often made more difficult in studies of single MNEs whose subsidiaries often assume specific and dedicated roles. Thus we were compelled to “compare” proverbial “apples with oranges”; that is, to assess outcomes in subsidiaries of different sizes performing different roles. In acknowledging such difficulties, Edwards, Almond, and Colling (2011) argue that researching MNEs compels one to grapple with the challenges of “comparing the incomparable.” In confronting such real-life complexities we had to be very cautious in ascribing effects to host-country regimes that might otherwise reside in other structural contingencies or influences.

In summary, we emphasize the following aspects of our research approach. First, while we sought to adopt a matched sample of subsidiaries in countries with contrasting institutional arrangements in anticipation that such arrangements might plausibly be expected to moderate the effect of corporate management’s intentions, the full context of our research was difficult to anticipate a priori. Second, our explanatory approach draws heavily on the interaction and interplay of effects both within and without the case. Finally, the tracing of causal processes in a context of complex and contingent relations led us to pursue a retroductive critical realist approach which involved some serendipity and required some intellectual agility, and, in relaying the findings of our research, some honesty in explaining its research design.

RESEARCH METHODS

MiningCo and its SubsidiariesFootnote 2

The case company, MiningCo, was chosen for its size, geographical reach and rootedness in the Brazilian NBS. It is among the largest and most international of Brazilian MNEs with operations in almost 40 countries. It is a former state-owned company. It was insulated from foreign competition for many decades, which afforded it the space to develop a large domestic presence before expanding overseas. Since its privatization in the late 1990s it has grown significantly, principally on foot of a series of foreign acquisitions to become one of the largest mining and metals conglomerates in the world. It has a diversified portfolio, operates across five continents and employs over 150,000 workers. Its headquarters are located in Brazil as is over 80% of its workforce and 60% of its assets. Its HQ management and board of directors are composed mainly of Brazilian nationals. Although publicly traded, the Brazilian State continues to have a stake in the company through investments made on its behalf by BNDES (the state-owned development bank), along with investments by a state-owned pension fund. Together they account for a 20% holding.

MiningCo’s state enterprise legacy provides a clear conceptual benefit. It serves to amplify two institutional effects – Brazil’s hierarchical-centralist business culture and that of a large sectorally-dominant “national champion.” In these respects, MiningCo can be fairly claimed to be “a case of” a Brazilian MNE as well as a “strong firm” in the manner as conceptualized by Royle (2006). Hence the case affords an ideal proving ground for an assessment of the nature and consequences of Brazilian influences.

Following its privatization, MiningCo adopted a centralized corporate model and introduced new mechanisms of financial and budgetary control. A Department of Budget was established in 2004 together with a global report initiative involving 84 indicators. A global HRM department was also established. With the help of US-based consultancy firms it reviewed all its HRM processes and strategies in respect of its international subsidiaries. A common global HRM model was prepared and transferred to its subsidiaries. It identified four key policy areas as constituting the “company’s DNA.” These included leadership training, career development, pay-for-performance, and performance evaluation. Leadership training encompassed the MNE’s seven management levels. For the first and second levels, which are respectively analyst and supervisor positions, staff undertook a week-long course in “what it means to be a MiningCo leader.” Management in L3 positions upwards undertook a 1-week course at MIT, which was designed to teach participants the complexities involved in managing people in different countries. There was a further week-long leadership course at IMD business school in Switzerland. The global career and development policy used a common performance evaluation mechanism to evaluate over 10,000 managers (from L1 to L5) annually across the entire MNE’s operations. This was designed to facilitate staff recruitment, development and promotion. It was also used to assess individual employee performance and was linked to the pay-for-performance scheme. The company perceived itself as operating a “meritocratic system.” The policy stipulated that 70% of non-managerial staff annual earnings be fixed with the remaining 30% variable. (It was a 50:50 split for management). The variable component was determined on the basis of three indicators: individual performance (50%), departmental performance (25%) and the company’s overall performance (25%).

MiningCo acquired a presence in Canada and the United Kingdom through its acquisition of a Canadian-owned MNE (NickelCo) involved in the extraction and refining of nickel in 2006. Its Canadian operations included a number of sites including a regional office. We focus on its production operations, the largest of which employed 3200 workers. Its other three sites employed 1870 employees. Prior to its acquisition, Canada accounted for 47% of NickelCo’s revenues; post-acquisition, it accounted for a mere 4% of MiningCo’s revenues. Thus NickelCo went from being, as one interviewee observed, “a large fish in a small pond to being a small fish in a very big pond.” Union density was high at 70%. Although management – union relations were adversarial under NickelCo’s ownership, the company was paternalistic in its orientation to employees. The British facility was a refining plant and employed 260 workers. It was unionized, but the union had exercised little, if any, influence over management decision-making.

The Norwegian facility was located in the north of the country. It was a former state-owned company and had been closed for a number of years before the facility was bought and reopened by MiningCo. Some former employees were re-recruited. It employed 78 workers. Union density was 90%. It produced manganese ferroalloys mainly for the British and German steel industries. Overall, it accounted for a mere 1% of MiningCo’s revenues and employed just 0.13% of its total workforce. It had recorded a loss of €200 million in 2009 mainly because of the economic crisis in Europe. It was of modest strategic significance to the MNE. MiningCo’s Swiss subsidiary acted as a trading and regional business services center. It was established in 2006 and employed nearly 80 people, most of whom were Brazilian expatriates.

Of the four case countries, the weight of local reliance on MiningCo was considerably greater in Canada than in the other countries. One of the Canadian subsidiaries employed almost 65% of the active labor force in its locality. The figures for the two other production subsidiaries were lower at 26 and 15% respectively. This dependency was further accentuated by high levels of local unemployment which varied between 7 and 11%. The United Kingdom subsidiary, too, was located in an area of high unemployment (8%), and mining and associated industries had become a small and increasingly diminished presence. While unemployment in Norway and Switzerland was around 4%, the Norwegian subsidiary was located in a remote region with moderately high unemployment (7%).

Finally, there is the timing of our research. As the international economic crisis deepened in 2008/2009 (immediately prior to our research) mineral prices plunged. Producers, including MiningCo, were forced to cut production, accept significant price reductions and cuts in financial returns. In an effort to cope with the vagaries of recession, MiningCo reduced its capital expenditure, temporarily closed some mining operations, reduced job numbers and demanded concessions from its workers. Other companies in the industry followed similar strategies. This had the desired effect of reducing supply. Late 2009 and early 2010 witnessed a recovery in commodity prices as Western economies slowly stabilized and as demand in China began to recover. MiningCo’s share prices improved and by 2009 its market capitalization had returned virtually to pre-recession values.

Data Sources

The data derive from in-depth semi-structured interviews and documentary sources. The interview guide is provided below in an Appendix. Preliminary interviews were conducted in July 2009 with two HRM global managers in the company’s HQ. Between July 2009 and February 2011 a further 49 interviews were undertaken. Of these, 16 were conducted in Brazil, 8 in Switzerland, 12 in Canada, 8 in Norway and 7 in the United Kingdom. These were supplemented by email communications as well as two additional telephone interviews with HRM managers in the MNE’s HQ in January 2013. Excluding these latter interviews and four other interviews (one each in Switzerland, Canada, the United Kingdom and Brazil) all were tape-recorded, giving a total of over 36 hours recording. These were transcribed by one of the article’s authors who is a native Portuguese speaker. The latter had a key methodological benefit in that it ensured that interviews with Brazilian nationals were conducted in the language with which they were most at ease. In the Swiss subsidiary, too, most employees spoke either fluent English or Portuguese. Our linguistic (in)competence, however, did not permit us to interview the Norwegian employees in Norwegian, but English proved a second-best shared language.

The majority of informants held managerial positions and, in the case of the Swiss and Canadian subsidiaries, expatriate managers were also interviewed.Footnote 3 At the headquarters, the global HRM Director, three global HRM managers, and various senior managers in international training, workforce planning, rewards and industrial relations were interviewed. The former global HRM Director was also interviewed in an effort to attest for the veracity of the testimonies given by current executives. In each of the subsidiaries the site manager, the HRM manager, finance manager and operations manager were interviewed. We also sought employees’ views to comprehend their influence in shaping local HRM practices and to better understand the intersections between host and home-country influences (Edwards et al., 2011). This was achieved by interviewing trade union representatives in Brazil (CUT, Força Sindical, and UGT), Norway (LO), Canada (USW) and the United Kingdom (Unite). There was no employee representative present in the Swiss subsidiary. From conducting interviews with respondents across managerial specialisms and hierarchies together with interviews of employee representatives, we were able to triangulate actors’ interpretations and observations. By this we do not assume that reality can be reconstructed by simply employing a variety of methods, but rather that it has multiple aspects, that it is interpreted through uncertain processes and that different actors understand that reality in different ways (Edwards, Vincent, & O’Mahoney, 2014).

We also examined various documentary sources including company annual reports from 1942 to 2010, the company’s website, 16 papers on MiningCo presented at the Encontro Nacional de Programas do Pós-Graduação em Administração from 1995 to 2009, seven books recording the company’s history, and a documentary film. The latter provided important information on the management of the company under state ownership as well as under its subsequent privatization. Print media sources were also analyzed, including articles in Veja (a weekly Brazilian magazine) and in Folha de São Paulo (probably Brazil’s most highly-regarded daily newspaper) from 1968 to 2010 and 1994 to 2010 respectively. In Canada, articles in the Toronto Star from 1985 to 2010 were reviewed together with materials in the USW website as well as various news reports and videos of the strike action (of which more below). In Switzerland, various government reports and articles in Le Temps and Tribune de Genève from 1998 to 2012 were examined. In Norway, we reviewed various LO, NHO (the employer’s confederation), and development agencies’ documents. Finally, in the United Kingdom we reviewed company collective agreements and articles in the Western Mail and South Wales Evening Post from 1980 to 2011. The information derived from these sources was used at the outset to identify key issues to explore with interviewees and subsequently to cross-check interviewees’ recollection of critical events.

The study’s data were transposed into a computer file for the conduct of a deductive content analysis. The data analysis was conducted in two steps. First, the material was reviewed in its original language. At this point we desisted from translating the materials into one language given the problems associated with such “translations” and the fallacy of achieving some form of exact equivalence (Chidlow, Plakoyiannaki, & Welch, 2014; Xian, 2008). No computer-aided software was used; instead a qualitative form of content analysis was conducted (Krippendorff, 2004; Welch et al., 2011), in which the basic unit of analysis was the sentence related by key words grouped as categories to produce a “cluster.” Four aggregate dimensions were identified: style of management; subsidiaries’ purpose and structure; source and transfer of reward policies; and structures of control and local constraints. In order to confirm the internal consistency of the data analysis and the reliability of the identified categories, the files were coded twice by the same researcher, a process which Krippendorff (2004) has termed “stability reliability.” No significant differences were found with each categorization. Once this stage was completed, the key materials were translated by the Portuguese-speaking member of the research team into English and proofread by an English native speaker. Then, they were given to another Brazilian national for a back translation. When the original and the back translation diverged, a discussion was conducted with the English and Portuguese native speakers with the researchers. Popular sayings were the most challenging to reach a consensual translation on. Notwithstanding the risks of reducing translation to a mere technical procedure, we thought this process of back translation important as full-language fluency in two or more languages is a rare gift. However, we did not pursue a “mechanistic application of a lexical rulebook” (Chidlow et al., 2014: 576), but rather strove to be sensitive, to the extent that we could be, to the variable cultural contexts and meanings of the “foreign” language.

The study’s preliminary findings were returned to the company for “member-checking” in two phases. First, they were reviewed with the global HRM director in a face-to-face interview and over the phone with the Swiss and Canadian HRM managers. Then, a draft report was sent to the site manager of each of the four subsidiaries for factual verification.

Finally, we make the case for the study’s quality by reiterating that we made every effort to ensure that the right data was collected by framing a research question which was derived from the existing literature. We sought to enter the field with a deep and broad theoretical appreciation so that we might develop our theoretical repertoires throughout the research process (Timmermans & Tavory, 2012). Thus in line with abductive logic, our early findings were considered as precisely that, and were challenged against rival explanations in an effort to establish a clear chain of evidence. Thus data collection was identified as a discovery process, which led to the research problem being reconceived in a new light as the research unfolded. Multiple sources of evidence and types of data were used, interviews were taped, two researchers interpreted the data and key informants were asked to review our first findings.

In an attempt to adequately conceptualize the nature of our research puzzle and as a prelude to our data analysis we turn next to examine particular aspects of the institutional contexts of the case MNE’s parent country, Brazil, together with those of the four host countries, Canada, the United Kingdom, Norway and Switzerland. We also consider the development of Brazilian MNEs and their adoption and diffusion of HR practices to their subsidiaries.

INSTITUTIONAL PARTICULARITIES

The Home Country Institutional Context

Schneider’s (2013) identification of the distinctive attributes of business systems in Latin America, which he labels hierarchical market economies (HMEs), neatly captures many of the core attributes of the Brazilian NBS. HMEs are characterized by the dominance of diversified business groups, the significant presence of foreign MNEs, the lack of investment in education and training, atomized labor relations, weak (workplace) trade unions, large informal labor markets, and highly centralized and hierarchical styles of management. The last of these is seen to be rooted first in a long tradition of state intervention in labor markets that served to discourage employers from engaging in bilateral negotiations with unions and instead inclined them to pursue their interests directly with the state; and second, in companies’ ownership (many family ownership) and size (a small number of very large companies, mainly MNEs, that account for a large proportion of economic activity and dominate their given sectors). Together these influences gave rise to unequal relations – “imbued with a hint of coercive hierarchy” – with competitors, clients and suppliers (Schneider, 2009: 8). Schneider further argues that the pervasive presence and reach of hierarchies has locked Latin American countries, including Brazil, into a particular pattern of institutional development in which hierarchy has come to constitute the “default preference” for the state and business elites, and from which it is difficult for the parties to extricate themselves.

While Brazil fits well into this stylized account, other features of its NBS warrant highlighting. First, the state has played a key role in sponsoring the development of industrial companies in key economic sectors, including natural resources (Diniz, 1997; Hopewell, 2014). Second, Brazil has two parallel labor markets: one formal and highly regulated, the other poorly regulated wherein approximately 40% of workers exist outside the protections normally afforded by the legal code (Barros, Cruz, Foguel, & Mendonça, 1997). Labor turnover is high and is attributed to the lack of institutional constraints on firing workers and the low costs of dismissal (Carvalho Neto, 2003). Third, unions are weak and fragmented, and struggle to gain recognition from employers (Cardoso & Gindin, 2009). The vast bulk of unions are perceived to be largely ineffective in representing workers’ interests. Collective agreements where they exist generally cover working hours and wages, but rarely extend to other matters (Carvalho Neto, 2003).

In tracing the origins of Brazil’s authoritarian management style, scholars give emphasis to the country’s historical legacy of slavery, colonialism, the early formation of an elite class, the persistence of a highly class structured society, and the rise of “bureaucratic authoritarianism” during periods of military rule (Carvalho Neto, 2003; Ribeiro, 1995). However, as pertinent as these influences might indeed be, it would be unnecessarily crude to portray Brazilian employers as being “cultural dopes.” Local scholars now acknowledge that traditional styles of management have been tempered in recent years, particularly within large well-resourced Brazilian companies, by management’s training in leading international business schools or by companies availing of the expertise of US-based management consultancy firms (Chu & Wood, 2008; Nicholls-Nixon, Davila Castilla, Sanchez Garcia, & Rivera Pesquera, 2011).

Brazilian MNEs

Brazilian companies internationalized late. This was due to state policies of economic substitution; underdeveloped capital markets, high capital costs and high interest rates; poor investment in research and development; and the dominance of family-owned enterprises that traditionally were risk-averse and favored development in domestic markets over foreign expansion (Casanova & Kassum, 2013; Rocha, Carneiro, & Silva, 2007; Tavares & Ferraz, 2007). Many large Brazilian companies operated in the natural resources sectors, whose products, at least up until the mid-1990s, attracted low prices in international markets. There was thus little incentive to move abroad, and certainly not to developed economies where production costs were considerably higher (Tavares & Ferraz, 2007). With the return to liberal democracy (military dictatorship came to an end in 1985) and with the introduction of economic reforms (Plano Real), fresh stimulus was given to creating the conditions for economic growth and company expansion. Impetus came, too, from the support and investment provided by BNDES, whose role was central in the government’s strategy of financing industrial champions. Further, the State’s abandonment of its former protectionist policies (with pressure from the IMF and World Bank) together with the privatization of state-owned enterprises forced many of the better performing companies to restructure at home and to seek efficiencies and foreign capital by expanding overseas (Casanova & Kassum, 2013). In addition, the creation of the commercial zone, Mercado Comum do Sul, comprising of Brazil, Argentina, Paraguay and Uruguay in 1991, encouraged Brazilian companies to expand into neighboring countries. Their motivations, in part at least, were defensive; that is, to secure and increase their regional market-share in the face of uncertainty and economic volatility at home (Coutinho, Hiratuka, & Sabatini, 2008; Fleury & Fleury, 2008). Finally, soaring commodity prices and high growth rates in foreign markets helped Brazilian MNEs to expand overseas, mainly through aggressive company acquisition.

While the bulk of Brazilian FDI remains within Latin and North America and in the Caribbean, in recent years expansion into Europe and Africa has increased (Caseiro & Masiero, 2014; Tanure, Cyrino, & Penido, 2007). Brazilian MNCs’ route to multinationality is distinctive. Many of their foreign acquisitions, particularly in North America, operated in sunset industries, and were either in financial difficulty or were part of a company’s operations that were deemed to be of little worth as its parent sought to shift production into higher value-added products. The motivation in acquiring such facilities resided less in the incorporation of physical (plant facilities) or intangible assets (technology or brand name) and more in the savings that were to be gained in time and resources in acquiring a presence in local markets. As such, a market-seeking rationale trumped an asset-seeking one (Fleury & Fleury, 2014). Importantly, too, there would appear to be little or no evidence of Brazilian MNEs moving overseas in an attempt to engage in a form of “institutional arbitrage”; that is, moving particular activities overseas in order to secure the advantages a host’s country’s institutional framework.

The foreign expansion of large Brazilian MNEs is reflected in their listing in the Global Fortune 500 ranking of large companies, which increased from 3 in 2005 to 8 in 2012. Of these, only two are in private ownership, the remainder are either in state ownership or are a mix of private/family-owned companies. Some are now major international players, such as Petrobras which is among the largest companies internationally by capital capitalization, and JBS which is a world leader in the international beef industry (Casanova & Kassum, 2013). Many other Brazilian MNEs are small to average-sized companies. However, the degree of internationalization, as measured by the number of foreign subsidiaries, tends to be inversely related to company size: many of Brazil’s largest MNEs are not necessarily the most internationalized.

The research to date on the management of HQ – subsidiary relations within Brazilian MNEs suggests that subsidiaries tend to be highly integrated with their headquarters and possess limited local autonomy and particularly in respect of HRM policy-making (Borini, Fleury, Fleury, & Oliveira, 2009; Muritiba, Muritiba, Albuquerque, Fleury, & French, 2012). The transfer of personnel from HQ to subsidiaries during the early stages of firms’ expansion is a common practice. While the evidence suggests that Brazilian MNEs rate HRM as an important competence in pursuing international expansion (Fleury & Fleury, 2011), few seem to dedicate sufficient resources to developing a global HRM competency (Muritiba et al., 2012), or in providing sufficient training for expatriate management to cope with, or to be sensitive to, cross-cultural management (Tanure, Barcellos, & Fleury, 2009), or that they necessarily appreciate subsidiaries’ management competence in HRM (Fleury & Fleury, 2011; Muritiba et al., 2012). In a recent review of the available evidence, Thite (2015: 105) further suggests that HRM would appear to be “the weakest link” in Brazilian MNEs, which he attributes to their lack of experience as global players. This is widely seen to have led Brazilian companies to import US-styled HRM policies (Chu & Wood, 2008: 986), especially in respect of career development and performance management systems (Tanure et al., 2010). Such “US practices” are diffused through the foundation of Brazilian business schools based on American models of management education (Fischer, 1984), the resourcing of the expertise of US-based consultancy firms and the influence exerted by the presence of American subsidiaries in Brazil (Chu & Wood, 2008; Tanure et al., 2010).

The notion of there being a “US model of HRM” deserves some reflection. In the empirical literature, US MNEs have been found to rely to a considerably greater extent than MNEs of other national origin on formal, codified, and standardized global policies and monitoring systems that afford little policy discretion to local management (Almond & Ferner, 2006; Ferner, Bélanger, Tregaskis, Morley, & Quintanilla, 2013). Their HRM policies are often associated with the use of sophisticated non-union or union substitution strategies wherein emphasis is given to the development of individualist and direct relationships with employees through techniques of employee involvement and a range of contingent reward practices, particularly performance-related-pay (PRP) (Edwards & Ferner, 2002). In low cost sectors, US firms have been found to adopt cruder practices informed by a strong anti-union animus but where a focus on individualist approaches prevails involving, for example, the inculcation of market-driven relationships (Royle, 2000), or in strongly unionized sectors appending individualist policies to collective industrial relations practices.

In summary, the evidence points towards Brazilian MNEs’ adopting an approach to the management of HRM, which bears a close resemblance to US models and that has much in common with Perlmutter’s (1969) ethnocentric style in which subsidiaries are managed as a cultural extension of the parent by expatriate managers who act as “enforcers” of HQ policy and home country values.

The Host Countries’ Institutional Contexts and Sectoral Influences

In classifying our host case countries we use Hall and Soskice’s (2001) “varieties of capitalism” as a means of distinguishing between broad categories of IR regimes (rather than complete systems of the reproduction of capitalism). Canada and the United Kingdom are typically portrayed as being liberal and flexible, while Norway and Switzerland are seen to be strategically coordinated and more constraining. On Hall and Gingerich’s (2009) index of coordination, Canada and the United Kingdom record low values, 0.13 and 0.07 respectively on a scale of 0 to 1 indicating that market modes of coordination are more important than strategic forms.Footnote 4 The values for Norway and Switzerland (0.76 and 0.51 respectively) suggest that there are relatively highly coordinated. In recent decades successive administrations in Canada and the United Kingdom have deregulated their labor markets in an effort to augment managerial authority and autonomy, as well as to enhance their countries’ appeal as a location for foreign direct investment (FDI) (Godard, 2009; Howell, 2007). The OECD’s Strictness of Employment Protection Index (2013) scores Canada 0.92 and the United Kingdom 1.10. By contrast, Norway’s score is 2.33 and Switzerland’s 1.6 (the OECD average is 2.04). Both the United Kingdom and Canada are also regarded as being open to the importation of novel HRM practices, which is attributed to their long history of hosting MNEs, particularly from the United States (Ferner et al., 2013).

Switzerland and Norway’s IR regimes are distinguishable. While it is perhaps more accurate to use the term “national(ly) coordinated” when we speak of Norway – as opposed to a sector-coordinated regime which characteristically applies to Germany – neither could be said to apply to Switzerland (Thelen, 2012). Switzerland is more properly conceived as a “soft” or “liberal” variant of a coordinated economy (Börsch, 2008). Trade unions are relatively weak, principally due to the country’s decentralized economic structure, linguistic and religious divides and the presence of a large foreign workforce. The state’s regulatory capacity is considerably more constrained than in other European countries arising from the country’s canton structure. Labor law is restrained in its reach and most aspects of employment regulation are governed by collective agreements. At the workplace, works councils are a minor influence. The Norwegian IR system has tended to be highly coordinated, highly regulated and characterized by considerable state intervention (Dølvik, 2007). While union density (52%) might be relatively low in comparison to other Nordic countries, wage bargaining coverage extends to two-thirds of the workforce. The dominant form of collective bargaining is a two-tier system of negotiations where national agreements are supplemented by negotiations at enterprise level.

We turn now to consider the industrial and employment dynamics of the mining sector and associated industries globally and then in the individual countries. It is an industry, which is characterized by the consolidation of ownership through company mergers and acquisitions; substantial organizational restructuring and changes in the contours of labor relations; an increase in outsourcing and in the use of temporary and contract labor; the continued dominance of male employment; the extension in working hours with compressed and extended shifts; union membership decline; the decentralization of collective bargaining and the increasing number of individualized workplace agreements, which have weakened the bonds of collectivism in traditional mining communities; and the increased focus on individual market-driven HRM practices (Ellem, 2006; McAllister, 2006; McDonald, Mayes, & Pini, 2012).

Canada depends heavily on the production and export of raw materials. Mining employs 320,000 workers and accounts for around 3% of the country’s GDP. Most companies are located in peripheral areas and are responsible for employing significant proportions of the economically active local population (Mining Association of Canada, 2012). In recent decades, large sections of the industry have merged with or been acquired by foreign-owned companies. Workers in the industry are organized by the largest private sector union in North America, the US-based United Steelworkers (USW). Wage rates are high and considerably more generous than that paid to workers of comparable skill levels in other industries. Relations between employers and unions were traditionally adversarial and acrimonious. Strikes were a well-established feature of the industry. (One of the case company’s subsidiaries witnessed seven strikes between the mid-1960s and mid-2000s.) While unions often exert considerable influence over the wage-effort bargain, management rights in respect of the implementation of organizational changes are generally well-established and earnestly defended (Godard, 2009). Unionization rates declined significantly over the past 25 years, from 48 to 24%. Collective bargaining is generally conducted at company or subsidiary level and is supported by a complex infrastructure of conciliation and mediation. In an attempt to respond to the growing global influence of large MNEs, the USW has worked with other unions internationally to develop global alliances, negotiate international framework agreements and engage in transnational bargaining. Such campaigns, however, are in their relative infancy when compared with that of other unions, notably that of the dock workers’ unions.

Traditionally, unions in Britain’s mining industry were well organized, had high membership and were entwined in antagonistic relations with employers. Their decline was precipitous following the defeat of the miners’ strike in the mid-1980s together with the subsequent closure of large numbers of mines and associated industries in the north and south west (where the case company subsidiary is located). After the miners’ strike, workers’ terms and conditions (including new payment schemes linked to output) were often unilaterally recast by employers (Golden, 1996).

In Norway’s metal industry, there is a strong tradition of company-level bargaining. Company-level negotiations cannot be negotiated with recourse to strike action, although other forms of industrial action, including work-to-rule, are permitted. The labor market is characterized by strict employment protection. At workplace level, union representatives play a key role in representing employees and enjoy consultation rights rarely bestowed on their counterparts in liberal regimes (Dølvik, 2007). The case of the Swiss subsidiary is distinct from the other subsidiaries. It operates as a trading and regional business services center.

RESEARCH FINDINGS

This article examines whether and how a mix of influences derived from a diverse range of institutional influences (the institutions of an emerging NBS, the institutions of variously configured host countries), interacted with the various features of the subsidiaries (their size, mode of establishment, history, pattern of industrial relations, location in the firm’s value chain, as well as the host economies’ dependence on foreign investment) to shape the transfer of HRM practices within the case MNE. We frame the analysis of our findings around three questions. First, to what extent was MiningCo’s style of management Brazilian in its hierarchical centralism and its forceful assertion of management’s prerogative to manage? Second, from where did it source its performance management policies? Third, how might we account for employees’ response to the transfer of the new performance management practices?

MiningCo’s managerial hierarchy was composed of seven levels. The most senior posts – L7 (CEO) and L6 (vice-presidents) – were occupied by Brazilians, as were the majority of L5 (global directors) and L4 (global managers) positions. Levels of reporting up the hierarchical chain were dense and frequent. While Brazilian management saw this as being “normal,” subsidiary management found them excessively demanding and unnecessary. There was also a clear expectation that management would assert their managerial prerogative in the conduct of their functions:

Strict control is part of our culture. (Corporate Brazilian HR manager, based at HQ)

Foreigners are surprised with this Brazilian culture; it’s extremely command oriented and it’s authoritarian. The key point was to identify the lowest common denominator of a policy and not to give in under any circumstance. (Former Corporate HR Director)

Senior management in Brazil accredited this style of management to two influences. First, they pointed to the development of a political and social system in which hierarchy and centralized and bureaucratic forms of decision-making predominated; and second, to MiningCo’s status as a large international company, which faced little if any constraints on its activities at home and which was now prepared to be equally assertive in its overseas operations.

At subsidiary level, the ostensible influence of hierarchy was perhaps most evident in the Swiss facility where most of the L5 and L4 positions were occupied by Brazilian expatriates and where Portuguese was the lingua franca. Management in all four subsidiaries observed they had little, if any, influence over the implementation of MiningCo’s HRM policies. They further believed that any questioning of Brazilian management was ill-advised and risked being interpreted as a direct challenge to their authority:

Never question a Brazilian face-to-face; never say why, never ever do that; it’s rule number one. (Senior British manager)

The first question they (Swiss managers) typically ask is why and while it’s not a misplaced why, for us it sounds aggressive. They just want to understand, but we do not have the same approach in Brazil. (Corporate Brazilian HR manager, based at HQ)

Brazil said, ‘you have to have them’ [the HRM policies]. We can’t challenge them and we’ve got our hands slapped when we tried to. (Middle-ranking Canadian manager)

The loss of local managerial autonomy – both generally and in respect of HRM policies – was acutely felt in Canada and the United Kingdom. Following NickelCo’s acquisition, MiningCo moved quickly to impose a direct and centralized structure of management control. This was exemplified by the recruitment of 25 Brazilians (almost a third of the total workforce) to the regional Canadian HQ. Strictures in terms of reporting were tightened and any decision in respect of the introduction of new HRM practices had to be approved in advance. While there was no expatriate management present in the United Kingdom subsidiary, the existing managers were retained and all communications continued to be conducted through the Canadian office, including those coming from Brazil. This might suggest that local managers enjoyed considerable autonomy and freedom of maneuver. This was not the case, however, as evidenced by these views.

They want to control everything. In NickelCo it never used to be like that. That is a Brazilian influence for sure, that is control and command. (Senior British manager)

NickelCo always encouraged people to do their own thing and not to be afraid of failure. I think MiningCo is the opposite; we don’t do it unless they know (in advance) that we are doing it. There is a feeling of being less empowered. (Middle-ranking British HR manager)

We find it quite frustrating if we are told “I don’t want to talk about it, just do it” and we don’t have much opportunity to input into decisions. (Middle-ranking British manager; emphasis expressed by interviewee)

The Norwegian managers, too, complained of MiningCo’s “excessive” reporting demands, which were seen to distract them from their main duties and were interpreted as a lack of trust.

So many hours are used in reporting; we do not see the reason for it. (Middle Norwegian manager)

Corporate management confirmed that they had indeed imposed new HRM policies in a top-down fashion and that their approach at subsidiary level had not been well received.

We took NickelCo’s policies and practices and we changed them. Today they are 100% ours. (Middle-ranking Brazilian HR manager, based at HQ)

Our (HR) policies are not changing for any country. (Corporate Brazilian HR manager, based at HQ)

We have an aggressive attitude regarding pay-for-performance. It is more aggressive than that which is usually found in the marketplace. Sometimes it is not well accepted in other cultures, but we still implement it. (Middle-ranking Brazilian HR manager, based at HQ)

While local management expressed considerable unease and objection to the manner in which MiningCo introduced its policies across the subsidiaries, and particularly in the Canadian subsidiaries, they eventually resigned themselves to accept them. This had wider ramifications: it created little, if any, scope for management – worker alliances to develop in the face of corporate management’s introduction of a new performance management system.

Our findings in respect of where MiningCo’s sourced its policies can be dealt with expeditiously. Management in the four subsidiaries confirmed that a standardized global HRM model was introduced and that it bore a distinct American character particularly with its emphasis on individual PRP. MiningCo’s corporate management confirmed that its performance management model bore little resemblance to anything that might reasonably be considered “Brazilian,” but rather was developed and implemented with the assistance of a number of US-based consultancy firms, including Booz Allen, Accenture and McKinsey.

Employees’ Response

We turn now to account for employees’ response to the imported performance management policy. The response of employees in the Swiss subsidiary varied. Foreign staff expressed few, if any, reservations. However, locally recruited staff did not welcome it. They complained that the base (fixed) component lagged market rates and the variable component was too big and was much larger than that of other local companies. However, Swiss management was unable to prevail upon their superiors in Brazil to reduce the variable proportion to single figures as they would have preferred. Only a modest 3% reduction was achieved.

It (PRP) has been strongly defended by the Brazilians. It is not the one we would have adopted here… In this market, it is too high; there’s too much uncertainty… The Swiss, in particular, prefer to be sure what they are getting. (Middle-ranking Swiss manager)

Two factors account for the Swiss subsidiary’s employees’ acquiescence. First, the majority of staff was either of Brazilian or of Latin origin. They were recruited on the basis that would likely conform to the values of MiningCo’s corporate culture. Together with the use of Portuguese in daily work interactions, the Swiss subsidiary was cast as a microcosm of a Brazilian subsidiary. The staff felt compelled to accept the new practices and gave little thought to raising any concerns they might have had. The Swiss employees, of whom there were only 16, were fearful of mounting any resistance in the absence of an employee representative forum. We suggest this is an illustration of Lukes’ second dimension of power wherein power is exercised in a manner in which potential conflicts of interest are handled such that they do not become matters of open decision. This is the power of “non-decision-making,” or in Hardy’s terms power of processes. Conflict is present, but it is not directly observable in workers’ behaviors as it was in the other subsidiaries.

The Norwegian workforce and unions enjoyed some considerable success in altering aspects of MiningCo’s pay policy. The variable component for production workers was reduced from 30 to 8 and 20% for shop-floor workers and engineers respectively and from 50 to 30% for managers. The union was also successful in prevailing upon management to re-institute an incremental salary scale that had been removed. These represented significant gains for the local workforce and were rare examples of concessions made by Brazilian management. However, it must be emphasized that the Norwegian workforce and union had originally set themselves resolutely against the introduction of PRP. In this respect they did not succeed; ultimately they were compelled to accept the principle of a PRP scheme, albeit at a much reduced ratio to their fixed base earnings. In their negotiations, corporate management let it be known that the staff would not receive any salary increase if they continued to oppose PRP’s introduction. In conceding to the introduction of the PRP scheme, the workforce was mindful of the cost of their labor (five times that of Brazil), the high levels of local unemployment and the poor financial performance of the subsidiary.

We are very concerned about how this company is doing. We see that the company has to survive, has to make money and we don’t get our payment if the company doesn’t earn money. It is the way it is, you can’t just demand and demand, so the company goes bankrupt. (Norwegian union representative)

In Canada after acquiring NickelCo, MiningCo moved quickly to introduce its performance management system and recast a nickel bonus scheme. The latter, which had been paid to employees in both Canada and the United Kingdom, was linked to the market price of nickel; as it rose, the bonus payment rose. Initially the bonus payments were modest and cost NickelCo little. However, shortly after MiningCo acquired NickelCo the price of nickel rose significantly to the extent that the miners’ salary almost doubled, in some cases in excess of US$60,000 per year. MiningCo sought to introduce an annual bonus cap of $15,000. In the face of workforce and USW objections, the company moved to act unilaterally.

There is no negotiating, there is no negotiating whatsoever. They told us that, since we weren’t ready to concede with the nickel bonus, there were to be no further discussions. (Senior USW representative).

The workers, supported by the USW, engaged in what was to become the largest private sector strike in Canada in over 30 years lasting for 12 months in two subsidiaries and 18 months in the other.Footnote 5 In the face of the strike, the company hired a leading Canadian law firm and public relations consultancy firm to conduct its dealings with the union and to frame its legal and communications strategy. With its help the company won court orders to limit the number of striking workers on picket lines; it took civil contempt hearings against union leaders; it contested Labor Relations Board decisions; and it levied multi-million law suits against the USW for infringements on picket line protocol, for information posted on the union’s website, and for a blockade staged by community and individual union members. By such means the MNE acquired the capacity to mobilize local resources and expertise on the fine workings of the local IR system and, in so doing, identified the “institutional spaces” to exploit within the Canadian institutional fabric. In an effort to maintain production capacity the company redeployed supervisors, clerical workers, engineers, geologists and maintenance workers. Seven months into the strike, MiningCo hired replacement workers. (In Canada, once a strike persists for more than 6 months an employer is legally permitted to recruit replacement employees.) This had never occurred under NickelCo’s ownership. This departure in management’s strategy was perceived by the striking workers as a significant escalation of the dispute and placed them in a vulnerable position.

It is very, very disturbing. (In) all the history of NickelCo, the 100 years of NickelCo they had never started an operation with scabs or with replacement workers. (Senior USW official)

Usually when you are in the union you feel protected, but this is the first time that they are probably feeling vulnerable. (Senior Canadian manager)

We are not giving in, we are not. I do not blame NickelCo for having giving in (in the past), because they had 35% of their business in Canada and this placed the company in a weak situation. We are not in a weak situation. (Senior Brazilian corporate manager, based at HQ)

MiningCo was now legally entitled to offer a permanent job to any replacement worker, and there was no legal obligation on the company to reinstate any striking worker, even if a new collective agreement was eventually arrived at. The union had acquired – as it transpired – a false sense of security from the forbearing disposition assumed by NickelCo in previous conflicts. In the face of the current dispute, however, MiningCo stuck to the regulative letter of the labor code rather than its normative spirit (as NickelCo had done), and in doing so used provisions of the institutional context to its advantage. Thus MiningCo demonstrated its unwillingness to play by the old implicit rules of the game.

The workers’ case was disarmed by two developments, one structural, and the other political. Prior to the strike, the demand for, and the price of, nickel fell appreciably. This created the “market space” for the company to confront a strike. The political response to the strike varied in its form and direction depending on the level of political representation. Political representatives at provincial and national levels sided with the company. Support for the striking workers resided only among local political representatives. While the latter condemned the company’s use of replacement workers, the former exerted pressure on the workers to cease their protest arguing that, as in the case of a senior federal cabinet minister, MiningCo’s takeover of NickelCo had prevented the local region from becoming a “valley of death.” Senior political representatives otherwise stayed silent on the tactics deployed by the company’s management and, in doing so, created the “political space” within which the company could seek to compel the workers to accept its terms. When the strike action ceased, MiningCo proceeded to introduce the new pay policy, albeit with a slight adjustment by fixing the nickel price bonus component at 8% and by reducing the PRP component from 30 to 22%.

In the United Kingdom, the IR context is important in explaining outcomes. In the years prior to NickelCo’s takeover, the workers’ union, Unite, exercised little influence over management decision-making. The last collective agreement was signed in 1993. It established a mechanism by which staff’s wages were automatically linked to the median point in wage settlements across the chemical sector. Subsequent salary negotiations became ritualistic and had little, if any, effect on workers’ terms and conditions of employment. Following the facility’s acquisition, staff management relations remained good, with most employees considering MiningCo to be a fair company (as observed by the trade union representative and findings from the company’s own employee climate survey). Collective negotiations were seen by both parties to have been “painless,” and conditions of employment, including salary levels, generous.

There was no resistance to the changes introduced to the nickel price bonus or to the PRP scheme. Managers and employees preferred the new bonus scheme as its payment would now be determined exclusively by their subsidiary’s performance and not, as previously had been the case, by the financial performance of the Canadian mines.

This way it’s better because we have goals. If you achieve these goals we get the bonus and if there is a profit made as well, the bonus increases. (Unite union representative)

In addition, the new bonus scheme enhanced employees’ pay relative to other workplaces in the locality.

There aren’t a lot of other heavy industries in this area. So you’re comparing MiningCo with work in the local shops and the public sector. Do you think people working in the public sector get 22% bonuses? (Senior British Manager A)

In terms of the local environment we are quite well paid, in terms of the national environment we’re still competitive, and conditions of work here are very good. (Senior British manager B)

Thus the UK subsidiary’s staff response to the imported policies was framed around their perceived desirability relative to those which existed under NickelCo’s ownership and those prevailing elsewhere within their own employment locality.

The British workers’ support for the revised pay scheme had the effect of undermining any prospect of cross-national resistance to the new policies. (There was little, if any, appreciable sense of the Canadian strike in the Swiss and Norwegian subsidiaries.) Indeed, the British subsidiary increased its production to make up for the shortfall in output in Canada. The company also sourced nickel from other company mines in Asia. While the British workers were prevented by law from engaging in secondary industrial action, when we enquired why they had not supported their Canadian colleagues in some, if only symbolic, manner, they replied that, if it had been the other way around, the Canadians would not have supported them and, further, they feared any such action would have jeopardized their job security.

The boys here are not going to lose any sleep over what happened in Canada, because Canada would not have lost any sleep over what happens to us. There’s no question about it; if we had gone out in sympathy, you and I wouldn’t be having this conversation now. (Unite union representative)

Thus the British workers stood to gain from their acceptance of the new pay scheme (greater predictability and control over bonus payments), as well as the enhanced reputation gained from corporate management in their efforts to meet the production shortfall caused by the Canadian strike. As a consequence, they say themselves as being better positioned to secure their long-term interests.

DISCUSSION

The article’s findings point to the case company exercising considerable power in transferring its HRM practices to its subsidiaries. That it was able to do so is explained by a number of influences. We highlight three. First, there is the company’s style of management. We make the case that this was linked to its nationality and its roots in a particular national business system. MiningCo’ style of management bore distinct Brazilian hallmarks. It was highly centralized and coercive and forceful in a manner consistent with Schneider’s (2009) conceptualization of firms originating in HMEs, albeit the specific content of its HRM policies was informed by US “best practice.” Further, as a former state-owned company, MiningCo was given the space to attend to and develop a large domestic market which, for decades, was closed to foreign competition. It was thereafter well placed to assume a sectorally dominant position internationally. MiningCo thus emerged as a “strong firm” which was predisposed towards pursuing its interests in a dogged and determined fashion.

Second, the political context mattered. We argue that, while MiningCo’s coercive hierarchical style of management is deeply rooted in Brazilian culture, its expression was facilitated by a particular pattern of economic and political relations in the host countries. The case of the United Kingdom is perhaps the most straightforward. The deregulation of its labor market, together with the degradation of its mining and associated industries and the undermining of union influence, disarmed the UK’s subsidiary’s workforce of any significant institutional resources to resist or reshape MiningCo’s policies. In Canada, the complex interaction of market and political influences trumped workers’ efforts to exploit their institutional resources. While management faced a confident and well-organized workforce, a particular market conjuncture (the drop in the demand for and price of nickel) afforded management the opportunity to confront the workers’ strike and wait-out a market recovery. Further, the Canadian political establishment’s championing of FDI created a context of economic dependence by which the State was pressed to support the interests of the employer and to act in a manner which rewrote the-rules-of-the-game in its favor. In Norway, despite the presence of a strong trade union and legislation which gave it significant powers, the workforce complied with management’s request to recast their remuneration scheme, although they were able to achieve some modifications. Here again the particular product and labor market circumstances of the subsidiary were important: the local workforce was dependent on MiningCo’s investment and their high labor costs placed them in a precarious position within the MNE’s global supply chain. Matters were somewhat different in Switzerland. The Swiss subsidiary was effectively colonized as a “Brazilian island” and was largely insulated from any significant pressures from the local institutional and labor market context. The transfer of corporate HRM practices encountered little, or no, resistance. Thus we argue the market and political context of the MNE’s subsidiaries is vital to explaining the outcomes witnessed in the case subsidiaries. As a “strong firm” MiningCo came to emasculate elements of the host countries’ IR regimes and was capable of establishing and operating – in large part – within its own intra-organizational field.

Third, there is the strategic capacity of the MNE to use its power resources to good effect. The evidence shows how MiningCo was able to manoeuver its way successfully through the constraints of the host institutional contexts by the adroit deployment of its resources. In the Canadian case this is exemplified by the company’s harnessing of the support of the political class, its use of replacement workers and the recruitment of local legal expertise in its campaign against its striking workers. The latter equipped corporate management with the hard-headed local institutional know-how by which it was able to “game” the very institutional context that might otherwise have been marshaled by others to constrain it. This contrasts with the inability of the subsidiaries’ workforces to operate as successfully, partly due to objective structural factors, such as the low market price of nickel, as in the Canadian subsidiary; the constitution of the Swiss subsidiary as a Brazilian island and the espoused commitment of the political establishment, particularly in Canada, but also in the United Kingdom, to FDI; and partly due to failings of strategy, such as the lack of skill or capacity on workers’ part to take advantage of their potential power resources – both as derived from the institutional context and their collective power. In the Canadian case, while it might be harsh to adjudge that the workers and the union were foolhardy in opposing MiningCo, it is probably fair to claim that they misjudged the consequences of the state’s increasing dependence on foreign investment and the consequent delimiting of its capacity to intervene politically on the side of labor. As such, the union failed to appreciate the wider significance of the political context and relations of dependence which existed between state authorities and MiningCo, and the manner in which they had been recalibrated following MiningCo’s acquisition of NickelCo.Thus we stress that, as the MNE might be properly conceived as a “contested terrain” (Edwards & Bélanger, 2009), so too might the host institutional context, as different actors within and beyond the MNE seek to identify and exploit any available “institutional space” for their advantage, deploying a variety of resources as they do so.

That MNEs are indeed capable of evading institutional constraints has been observed in the IB literature, but rarely, has research been so detailed in accounting for the dynamics and processes involved and, in particular, in showing how the influence of institutional factors must be understood in the context not only of the micro-politics of the firm but also the wider macro-political terrain. From this latter observation, it is important to emphasize the complexities of host institutional configurations and their varied and mixed effects. In Canada, we witnessed how political support for the workers’ case altered across different political levels with the key group of federal ministers supporting the company. Thus access to institutional resources was in practice highly differentiated depending on which local actor is being considered and what is perceived to be at stake – in this case workers’ efforts to preserve particular terms and conditions of employment as against the State’s inclination to side with a major incoming investor that brings employment and investment. This political relationship of dependence created a context within which the State’s support for the case as advanced by the workers and their union remained at best muted and worst absent. It was by such inert actions – in a manner that resembles Lukes’ (2005) power of “non-decision-making” – that the State came to stand alongside the MNE and to be leagued against the workforce. While such dynamics were less obvious in the other case subsidiaries, they were arguably no less present. MiningCo in Switzerland had its way by being able to exploit other institutional resources (the facility of recruiting a Brazilian/Latin workforce) and in Norway and the United Kingdom on foot of the local communities’ dependence on it for employment.

From these explanations, we make the following observations in respect of the existing literature. First, we emphasize that, while the literature to date has rightly laid stress on the manner in which MNEs are capable of exploiting resources of their parent NBS, this case study explains how a MNE from an emerging economy devoid of a rich repertoire of home-country resources was adept both in absorbing US hegemonic management practices and in manipulating host countries’ institutional resources to its own advantage. Second, a great deal of the institutional literature has tended to perceive country-of-origin and host-country effects as discrete influences – the former is available as a resource for the MNE to exploit and the latter a resource for the local subsidiary to seize. Such a neat binary division does little to capture the complex and multilayered dynamics of MiningCo’s relationship with its subsidiaries. Ultimately the extent to which the institutional resources of the host country came to be captured by the MNE was dependent on its prior power resources which rested on its size, dominant market position and, in turn, on the host states’ dependence on it for employment and community prosperity. Finally, much of the literature to date has tended to concentrate on actors and relationships internal to the MNE. We make the case that to develop an adequate understanding of MNEs one has to better understand their “territorial embeddedness” (Almond, 2011) or “the political economy of employment relationships” (Watson, 2004); that is, MNEs’ relations with actors beyond the firm, and in particular, governance actors. Such an admission inevitably behooves a class analysis of MNEs that has been absent from much of the IB literature, and particularly the institutional perspective. By this we mean the manner in which the material interests of different groups (in this case capital and labor) become intertwined with political forces to produce distinct political power alignments. Consider again the Canadian case: as the state became ever more dependent on FDI so the political space to intervene on the side of the local workforce was commensurately reduced. And while it might be objected that the state would likely have acted in a similar fashion regardless of the firm’s nationality, such consideration points even more forcefully to the class basis of the terms of the conflict. In these terms, then, an overreliance on a “local”−“foreign” institutional analysis misses the overarching class basis of the terms of the “contested terrain.”Footnote 6

CONCLUSION

This article is motivated by a desire to untangle the complex causal factors in the observed patterns of HRM in the subsidiaries of a MNE from one of the emergent economic powers, Brazil. The case is of a Brazilian mining MNE that expands into Canada, the United Kingdom, Switzerland and Norway and imposes a new pay and performance management system, contrasting with existing norms. It uses this to examine interrelated questions about the influence of an emerging-economy parent-business-system and how this interacts with the well-developed – albeit variable – institutional regulation of the host economies in a context of complex relations of dominance and economic dependence. Hence we are forced into the interesting realm of multilevel analysis about MNEs, power relations and institutional change. It considers a variety of factors and the position of actors within and beyond the MNE and their relational interplay. These include the macro-political terrain of the MNE – the home and host institutional influences; the context of the subsidiaries (their size, mode of establishment, history, pattern of industrial relations and location in the firm’s value chain); and the host economies’ dependence on foreign investment – on which the interests and identities of actors are formed and represented, and on which the ensuing micro-political relations are played out. The article is positioned at the cross roads of two developing and important debates in IB: the internationalization strategies of emerging market MNEs and the responses of host country actors to MNE entrants in general. The article integrates two theoretical approaches – institutionalism and a power relations perspective.

The article is situated in the critical realist epistemology. It adopts Burawoy’s (1998) extended case study method together with a context-sensitive and an actor-centered approach. Its major contribution is its untangling of a complex of influences and processes that interact at multiple levels to explain the transfer of HRM practices within the MNE. To date IB scholars have viewed institutions largely as constraints on the activities of MNEs. Such a tendency has been particularly pronounced where analyses are conducted through the lens of the neo-institutional perspective and where there is a focus on institutional distance. We have adopted a different theoretical lens and have found it more useful to examine not only how institutions might constrain MNEs’ strategic choices, but also to examine how institutions might come to shape and support the capabilities and preferences of MNEs. We go yet further in seeing such processes as involving not only a “willing facilitation” where institutions are re-modelled (by state agencies) to enable MNEs’ capabilities, but that institutions might also be captured and colonized by the preferences of MNEs. In fact, the key theoretical point is that institutions exist simultaneously as constraint and opportunity and their coexistence is not fixed or immutable, but is a matter of context and circumstance and opportunity (see also Dörrenbächer & Geppert, 2011). The processes associated therewith are inevitably enormously complex and involve a variety of actors (MNE’s local and corporate management, workforces, unions and political representatives) who possess shared and conflicting interests and who interact at multiple levels. We therefore argue that, while institutional theory remains important to our understanding of the behaviors of MNEs, it needs to be more explicitly combined with a power relations perspective and a consideration of how actors act to avoid, capture or refashion institutions to their ends.

We find that the case MNE was able to use a variety of institutional resources to achieve its objectives. In the case of Canada this involved a blunt and pragmatic exercise of power in a manner consistent with Lukes’ (2005) first dimension of power. Institutional resources were captured to serve its ends against the institutional constraints marshaled by the workers’ union (the USW strike). Further, the MNE’s use of these Canadian resources was more varied than anticipated in theory. However, there wasn’t a great deal of creativity or perseverance on the part of the MNE to remold the institutional setting to its liking, in a manner which is consistent with Streeck and Thelen’s (2005) concept of MNEs being “rule-makers.” Neither can it be said to have been a “rule-taker.” Rather it rewrote the practice of rules to its liking in a most pragmatic and instrumental of fashion, and moved on. The MNE had its way in the other subsidiaries too although the exercise of power was less evident, in manner that resembles Lukes’ (2005) second face of power (the power of “non-decision-making” in Norway) and third face of power (the power of meaning in the Swiss subsidiary). In all cases norms surrounding pay and reward were re-written. We thus interpret our data to argue that national business systems are not fixed or immutable; rather, they are fluid, flexible and adaptable and allow much scope to actors to exploit their potential in ways that are not fixed a priori.

In the specific context of our case MNE, then, NBSs (be they of a liberal or coordinated market variant) are not inevitably or invariably perceived as constraints on MNEs’ activities. We stress, however, that this foregoing argument is not made in a general sense. Rather in making the case that seeing institutions as hindrances that can be circumvented by the use of other more favorable institutional resources or institutional capture, our findings point towards a need to understand the institutional specificity of firms better, and that there is a requirement to determine what firms are trying to do, and an assessment of the conditions under which actors might, for instance, be able to “capture” key institutions.

In this study the case MNE was able to implant a distinctive parent-country (Brazilian) management style and transfer its (US-sourced) HRM policies by a conjuncture of power relations whose elements included the MNE’s leading global sectoral role; its extensive resources; its dominant position as a large foreign investor in FDI-dependent host economies of modest size and as a large investor in regional/peripheral economies; and the contingency of host institutional resources/constraints, where their potential influence hinged on the skills and capacities of the parties concerned to mobilize or exploit such resources. In the case of Canada, and to a degree also of the United Kingdom and Norway, the deal “negotiated” was that the company be afforded the space to conduct its internal affairs as it thought best. Arising from such political processes MiningCo was enabled to adopt its own company-specific practices in a manner consistent with Kostova et al.’s (2008) notion of the development of “intra-organizational fields.” Thus consideration of an evident Brazilian influence must be placed in the wider context of economic and political relations, which themselves were shifting and being recast. The manner in which such a complex of influences was played out in variable institutional contexts represents a significant contribution of this case study to the literature.

From this we make an important observation in terms of method and mode of generalizing. Context-sensitive explanatory approaches have not been widely used in the IB literature. As this case study demonstrates they provide a powerful means for understanding the manner in which phenomena – as causal factors and mechanisms – interact, and how their operation may vary across different contexts. Thus the “same” phenomenon (the transfer of a similar performance management system) had different meanings and was responded to differently depending on context. We thus make the case, as have others (see especially Welch et al. (2011) and Edwards (2005)), for a context-sensitive mode of explanation to be combined with a more actor-centered approach (Blazejewski & Becker-Ritterspach, 2001; Geppert & Dörrenbächer, 2014).

Finally, we argue for the retention of an institutional perspective but of a form that desists from engaging in a static reading-off of institutional influences. We further argue for the adoption of a dynamic multilevel institutionalist framework wherein an analysis of the micro-politics of the MNE is located within a wider macro-political terrain (cf. Almond et al., 2005; Morgan, 2011). The challenges are considerable but not insurmountable.