Who’s in charge? Corporations as institutions of global governance

In most accounts of global governance, where corporations are included, they are seen as either subject to various international organisations’ regulatory impact or are identified as having (benign or malign) influence over agenda setting around the scope and practices of global regulation. However, here I examine a third dimension that has hitherto been under recognised: this article starts to develop an analysis of the terrain that global corporations govern themselves, sometime singularly, sometimes collectively and sometimes collaborating with the more “normal” institutions of global governance. I seek to develop an account of how the corporation governs this terrain and the mechanisms that businesses have developed (or utilised) to maintain their authority. I suggest that it makes sense to understand global corporations as directly analogous to more “normal” institutions of global governance, and that discussion and analysis of global governance needs to integrate this third dimension if it is to examine the full spectrum of governance beyond the state. This article is published as part of a thematic collection dedicated to global governance.


Introduction
I n this article I am not going to discuss how various corporations have shaped the agenda of particular institutions of global governance, nor am I going to set out the impact that globally focussed regulatory institutions have on particular business sectors or corporations. These two common perspectives might be regarded as conceptualising the relationship between corporations and global governance as either input (helping shape the agendas of global governance institutions) or output (corporations impacted by institutional decisions) (Brühl and Hofferberth, 2013: 354). Rather, I propose a third perspective, one that is missing from the alternatives set out by Brühl and Hofferberth in their survey of approaches that see companies as social actors: that the global corporation itself has a significant governance role within its own international structures, and we can usefully consider this as a form of global governance in itself. 1 This complements the emerging account of corporate constitutionalism that extends the analysis of corporate governance (Bottomley, 2007), to suggest that global corporations' governance function reaches well beyond the internal relations between managers and shareholders. Indeed, while the United Nations Global Compact (UNGC), the Extractive Industries Transparency Initiative (EITI) and other organisations indicate there is a growing recognition in (global) policy circles that improving the governance function of corporations can have beneficial effects across their supply networks, there has been less discussion of how this governance function might be best understood as a political economic issue, rather than merely a technical question of effective economic control. Coe and Yeung's (2015) recent work on global production networks also maps out much of the space I am concerned with, and offers a dynamic account of supply/value chain development, but their account lacks any significant non-economic dimensions to how governance might be understood. Indeed, given that "governance" is often presented as the importation of business values and processes into politics (Eagleton-Pierce, 2014), it is perhaps surprising that corporations' (extra-economic) governance of their own networks has remained underremarked. Therefore, this article is concerned with the wider network of relations represented by the supply chain assembled and governed by contemporary globally active corporations.
Patterns of organisation vary widely between corporations (Dicken, 2003: 212-219), not least as the legal systems in which they are incorporated may be different (common or civil law systems) and have different national regulatory expectations. Furthermore, in some sectors such as resource extraction or infrastructure there are often a number of major corporations that may be owned partly or completely by the state. Nevertheless, here I will focus on the question of governance in general rather than look at actual corporate networks (as will be required to further develop this approach). What follows is different from much discussion of (global) supply chains that focusses on their management, 2 as it suggests there is a governance function being undertaken by corporations, which is characterised by explicit and implicit power relations, not only the technical management of efficiency in network interactions. These forms of authority do not necessarily reach beyond a corporation's own network(s), but because of the global reach of such networks their political economy parallels many of the characteristics more usually identified in the realm of global governance. Hence, rather than examine corporations as either influencing global governance institutions, or negotiating regulatory interactions with these institutions, both of which remain important issues, here I argue that our understanding both of global governance and the political economy of the global corporate sector can be enhanced by taking a third (complementary) view. This is to say, it will be enlightening to assess global corporations as institutions of global governance themselves.

A permissive (albeit brief) definition of global governance
The first step is to establish a definition of global governance that is both acceptable and permissive of inclusion of corporations. The first aspect-that of scale-is relatively easy to resolve: when we talk of global governance the two distinctions we are making are between domestic and non-domestic, and between the international and the global. The former is a question of state sovereignty, borders and national politics; global governance may have an impact on this realm of politics but is not limited to the territory delimited by any particular state's borders. The distinction between international and global draws attention to those aspects of governance that are not merely manifest between states on a bilateral or (explicitly) multilateral basis; "global" is intended to convey something about the (potentially) comprehensive character of this emerging complex of governance institutions (Weiss, 2013: 28). Although not all institutions of global governance have a comprehensive reach-many are limited by membership and almost all are limited by mission or sector-the institutions commonly included within the term do not suffer such limits to their collective globality. The term "global" conveys the potential rather than actuality of the scale of activities of any individual institution and is more accurate if all institutions of global governance are taken as a single collective class, developing a global complex of governing norms.
The second term, governance, was introduced into political analysis to identify governing as a process which was not limited to formalised state-constituted institutions. There is immediately some analytical difficulty around the interchangeable use of institution and organisation: the latter having a formal existence (headquarters, secretariat, rules, members) while the former has come to be seen as considerably more permissive, used to also include laws, customs and social practices that are unorganised in the sense of the latter term (Weiss, 2013: 29-31). This leads Weiss (2013) to offer the following definition: global governance is collective efforts to identify, understand or address worldwide problems that go beyond the capacities of individual states to solve … the capacity within the international system at any moment to provide governmentlike services and public goods in absence of a world government. Thus, it is the combination of informal and formal values, rules and norms, procedures, practices and policies, and organisations of various types that offer a surprising and desirable degree of global order, stability and predictability. (32) Weiss suggests that the provision of "order, stability and predictability" is delivered through global governance by a diverse set of (formal) organisations and (more informal) institutions. If in this conception the particular types or institutions or organisations is (relatively) permissive, based on effect rather than character, then the key issues must be scope/scale-globaland the provision of order, stability and predictability in this realm of interactivity. Therefore, while Weiss does not draw this explicit conclusion, these aspects of governance could be delivered or enacted by a corporation, rather than a regulatory organisation of some sort.
We also need to consider the forms of socio-political power that have informed the development of the institutions of global governance, an issue sometimes obscured by more organisational, functional or technical accounts of specific institution's history (Pierre, 2015: 190-202). The sources of power deployed through global governance are no more or less complex than social analyses of power would be in other political economic realms. In Michael Barnett and Raymond Duvall's (still) influential typology of forms of power in global governance we find various forms of power arrayed across four dimensions: compulsory (through the deployment of resources); institutional (through rules); structural (via the control of agendas of practice); and productive (or as it is often referred to constitutive-the ability to define the political realm and practices themselves) ( Barnett and Duvall, 2005; see also, Lukes, 2005). This typology of dimensions of power across global governance is reflected in Kennedy's (2009) observation that the globalisation of law, the legalisation of politics and economics, have brought with it a tremendous dispersion of law. All manner of rules, enforced and unenforced, may, as a matter of fact, affect any global transaction. And as a matter of law there are conflicting and multiplying jurisdictions, asserting the validity or persuasiveness of their rules, with no decider of last resort … We will need to assess the dynamic and distribution effects of one or another attitude toward the disorderliness of global governance. (55-56) The elements of this dispersion and disorderliness that interest me here are those that we can identify within the overall organisational structure of global corporate extended and networked supply chains. This is to say that global governance happens in a wider field of social relations than is often explicitly mapped.
Therefore in this article I suggest that if we examine global corporations' networks including their affiliated subsidiaries and their contractors we will be able to recognise a realm that looks remarkably similar to global governance as usually presented. Moreover, corporations' governance function underpins the competitive advantage gained by globally active corporations: corporations not only seek to manage the impact of a globally integrated political economy on their operations, but also the very opposite; they seek to shape, (re)form and govern the political economic terrain in which they operate to their advantage (Picciotto, 2011: 68). To establish the utility of an approach that sees corporations as institutions of global governance, I will now set out how the character of global corporate networks can be included within the dimensions of global governance explored briefly above. 3

Global corporations as institutions of global governance
To some extent the approach herein parallels Lynne Dallas' notion of the power model of corporate governance (counterposed to the efficiency model). Here, the corporation actively seeks to structure its environment to serve its needs of autonomy and discretion. While certain concrete realities exist, such as the scarcity of resources, how these resources are used and distributed is a matter of social organisation, dependent on cultural, historical, social and power factors. (Dallas, 1988: 114) The decisions made by the corporation are not necessarily directly concerned with efficiency or profit maximisation but rather with the perpetuation of control and the values stressed by the central management group(s). Moreover, these values are likely to vary from sector to sector: corporations in the pharmaceutical or high technology sectors will have considerable more interest in the manner in which their networks treat, use and protect intellectual property rights (IPRs) (May, 2010) than a corporation in the food sector might do. Likewise, corporations in capital intensive industries such as energy generation will confront different governance issues from clothing retailers with extensive and complex supply chains. For instance, a hightechnology corporation will be concerned not just with the manner in which partners use specific IPRs but also the national legal context, such as a state's government's willingness to enforce such rights; Chinese (lack of) willingness to enforce IPRs has in the past been identified as an issue for the development of some supply chain relations. However, for clothing and/or assembling networks, the impact on the brand of labour relations will be a larger concern with contractors seeking to either distance themselves from unwelcome developments (such as the Rana Plaza fatal building collapse) or by seeking more direct control of labour standards in the supply chain though forms of relational contracting.
As these examples illustrate it is likely that the corporation may be effectively managed by a coalition of stake-holding groups focussed on a myriad range of outcomes including immediate profitability but by no means limited to it, and therefore as corporations have become embedded in more fragmented supply chains so these coalitions have often needed to resolve internal conflicts over various competing outcomes (Dallas, 1988: 40-42). As this suggests we immediately need to differentiate between the control of the corporation and its beneficial ownership.
Certainly, much of the discussion on corporate governance in recent years has been concerned with aligning the interests of corporate managers with the (often presumed) interests of the corporation's shareholders (its beneficial owners). Ways are sought that ensure managers focus on and prioritise "shareholder value", a frequent proxy for rising share prices and maintaining dividend payments. That there has been considerable policy and legislative work on alignment suggests that hitherto owners and managers have not necessarily shared a view on how corporations should be managed. There is much to be said about this aspect of corporate governance (not least the effective extent of the division between ownership and control) but here I want to emphasise that even if we accept some form of alignment is necessary, it does indicate that the governance and control function of the corporation is not fully articulated to ownership (when firms are incorporated at least) and thus management groups within corporations are often governing (to some varying extent) autonomously. Taking an optimistic view of managerial interests, the chief outcome of the governance function is the direction and implementation of a chosen competitive strategy across its supply network. A more pessimistic view, which has driven much of the discussion of alignment and incentivisation, is that managers manage the corporation in their own interests and not the shareholders (presenting a classic principal-agent problem for resolution). This is further complicated by corporations themselves being owners of subsidiaries and thus across a range of different relationships are both principal and agent.
Partly due to the frequent requirement for local operations to be incorporated under local states' laws, contemporary global corporations are often actually groups of separate companies controlled through various means from a central point. These networks are frequently arrayed around a hub, even if the formal and legal arrangements are organised in such a way as to partly obscure this fact, and the states where the hub is located have little effective regulatory control over the entire (dispersed) network (Veldman, 2013: S25). However, the corporation needs to establish a working balance between centralisation (direct control) and the existence of a network of (formally) autonomous partners and contractors, which is where governance comes in.
Governance from this central point is seldom entirely hierarchical but is more often a process of negotiation and engagement. The central corporation may have distinct advantages but subsidiaries, contractors and others may also have forms of leverage that can be deployed within supply chain decisionmaking. 4 Thus, where the core corporation is contracting with partners to access particular technologies, or localised resources or knowledge(s), the contractor may have significant leverage opportunities within any contractual relation. The central corporation may also seek to benefit from "reverse diffusion" of practices and innovations where these have been developed in one part of the network; here governance is concerned with how to establish the new practice/approach around the network. In addition, the corporation at the centre of the network may have multiple interests and thus present competing positions to partners, shifting and changing the focus of potential leverage. For instance, particular sub-contracting partners in the network may get a very different steer from the Corporate Social Responsibility (CSR) team that visits their plant to ascertain their compliance with various standards and requirements, than they will from the buying team, whose emphasis is likely to be on price and quality. 5 Focusing on corporations as institutions of global governance suggests the reasons for differences between high-level commitments and what happens within the supplychain itself; like other governance institutions, global corporations have complex bureaucratic managerial arrangements which produce difficulties of coordination between policy and practice (a not uncommon observation within global governance analyses).
Following David Ciepley, we can posit that "within its jurisdiction, the business corporation exercises powers analogous to those of government, if more limited, including the right to command, regulate, adjudicate, set rules of cooperation, allocate collective resources, educate, discipline and punish" (Ciepley, 2013: 142). However, while Ciepley and others are right that when compared to the sovereign state there is considerable difference in the scope and range of such legitimate capabilities (Crane et al., 2008: 72-76), when we compare the global corporation to the typical institution of global governance these differences are considerably less significant. The corporation can actually be in a stronger position as regards its network than a globally-focussed institution or organisation is as related to its issue area (or constituent state members). Corporations may well be able to more effectively sanction and govern network members than some international organisations can govern state members.
Moreover, global corporations have developed processes that often mimic legal structures. This then negates many unwelcome consequences of globality; in the absence at the supranational laws of property or contract, the global corporation's network of subsidiaries and affiliates respond to the internalised law-like rules and regulations the core corporation puts in place (Gessner, 2012: 151), and less to differences in national legislation. Within its own network (or effective jurisdiction) the corporation is a form of governing body, adopting as Ceipley suggests many of the attributes of government; the authority it is able to mobilise, however, is not completely separate from government, but rather flows from the legal mechanisms (such as incorporation and property law) that facilitate corporations' operational modes in specific jurisdictions. The corporation's governance function is not against the state, but rather is partly facilitated by the state (Robé, 1997), and where differences in national legislation are to the corporation's advantage, the relationship between globalised internal rules and national laws is finessed to ensure both elements serve the corporations' needs.
Therefore, although global corporations certainly interact with the other institutions of global governance, they also govern significant realms of the global political economy, sometimes in conjunction with the regulations and guidelines of other institutions, sometimes as single governing authorities themselves. While using forms related to, and indeed constituted by states, their actual character and effect within the corporation's "internal" relations (that is within its networked supply chain not only within its incorporated core) are defined by the corporation not the (or a) state (Robé, 1997: 66-67). Where corporations are constituted differently from the dominant Western model, perhaps incorporated in states across the Global South, or are State Owned Enterprises, although the formal/legal issues may vary considerably, prompting in some cases considerably different governing practices, the underlying requirements for efficacy are likely to be similar (even if this would need to be confirmed by research into specific cases). This is to say: to be effective (even if state sanctioned) forms of supply chain, or network governance need to be regarded as legitimate by network partners, contractors and participants.

The legitimacy of corporations' governance function
The construction of the legitimacy of corporate network governance needs to take place across three initially negative aspects of power relations: differences in circumstances; disparities in interest; and requirements for compulsion (Beetham, 1991: 59-63). First, the difference between the core/organising corporation and its partners needs to be legitimated: this may be achieved through recognition of the ownership of technology or through appreciation of brand value(s), through access (albeit indirectly) to established sales networks and/or via other resources or assets that can only be utilised through the network. Second, the subordination of decision making (governance) which results from involvement in the network has to be demonstrably in the interests of participants; how involvement allows partners to effectively access various assets within the network is key. Third, to remain legitimate the relation must be (or at least plausibly be seen as) consensual; negotiations around "incomplete contracts" and conditions of supply will have significant impact on whether the governance of the network is regarded as legitimate or exploitative. 6 Differences between actors might also be justified through "cognitive" legitimacy (linked to the acceptance of market relations as themselves being generally legitimate), but where relations go beyond the one-off spot market contract, legitimacy makes supply-chain governance more effective. The core corporation needs to manage these networks more subtly than merely attempt to construct command and control arrangements that are more costly in time and other resources.
Legitimacy may be less of an issue where network partners are not actually independent of the core corporation. As corporations are now mostly legally constituted to allow them to also own companies themselves, in these cases governance is legally established (and not required to be consensual). Corporations can benefit from the limited liability afforded to investors/ shareholders, shielding the central (governing) corporation from some risks and liabilities in the network, and establishing a range of political economic firewalls to enhance the security of the central organisation. Indeed, part of the purposes of global corporations' governance function is the control of risk from outlying areas of the supply chain while harvesting the economic and other benefits of engagement with such partners (low wage costs; new technologies; new markets and so on). However, the core corporation then has its own principal/agency issue to resolve, in that it needs to ensure subsidiaries are working in the interests of their (corporate) owners.
These complex networks of both subsidiaries and independent companies are tied together through a web of contracts (Picciotto, 2011: 132-133). While the character of these supply ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057DOI: 10. /palcomms.2015 chain networks is perhaps specific to contemporary capitalism, management itself, as James Burnham pointed out in a remarkably prescient analysis over 70 years ago, is always about controlling (and benefitting from) the access to resources (Burnham, 1941(Burnham, [1972; see also Ciepley, 2004). Defining the managerial revolution as the shift in society from a privileging of ownership, to a society where increasingly power stemmed from control and/or management of institutions, Burnham suggests to us how and why global corporations' managers have become defacto agents of global governance. 7 Specifically, when we think of global corporations' governance function the key issue is not necessarily the ownership of assets and resources, but rather the ability and abiding interest in controlling access to those resources (Rajan and Zingales, 1998). While significant parts of the supply chain network are outside the ownership structure of any particular global corporation, the ability to control access to the central resources of the network and also (through noncompetition clauses in contracts, for instance) an ability to constrain access by competitors to key "independent" elements of the network is an important element of corporations' governance function. Although the ownership structures of global corporate networks are often complex, it is the management and control of these networks that produces identifiable governance effects. 8 This management interest is not limited to the issue of prices or market conditions but rather given their network character, it is also (crucially) an issue of managing time and space; a global corporation's managers seeks to control its own political geography to maximise its returns and remain a going concern (Dixon, 2014: 62-63). As Dallas contends the chief, and perhaps only really effective, lever that corporations have is their ability to mobilise and allocate resources across the space they seek to control. Here we should understand "resources" widely to include "social products" including influence, legitimacy and preferences that the central managerial coalition governing the supply-chain can deploy to shape more micro-level decision making (Dallas, 1988: 83-84). This can also be seen as the institutionalisation of particular networkspecific relations, habits and socialised knowledge about what would be regarded as normal activity. Furthermore, a key outcome of corporations' governance of their networks is the transfer of risk (another social resource) away from the central corporation and onto other stakeholders and/or partners in the supply chain. Seeing corporations as global governance institutions requires a focus on ongoing relations within these networks and not merely a view of a series of periodic intranetwork transactions.
For example, global corporations set standards and rules in many market sectors directly via the contracts that are used to pattern the supply chain, but also because many global sectors actually function as effective oligopolies (Harrod, 2006: 25), and as such are controlled by a small group of corporations whose managers even if not colluding with each other share a broad set of interests. 9 The acceptable and normal practices in often quite closely bounded markets segments are set by a small number of corporations. The corporation(s) at the centre of these supply chains seek to perpetuate their positon by re-creating this environment where and when possible, and as change/ innovation is required, shaping it to maximise their capture of any associated advantages or benefits (Dallas, 1988: 98), while shifting the risks associate with change into the supply chaincosts of transition are often borne by contracting partners not the core corporation itself.
Within these networks corporations construct regimes of private law to govern the relations between the various elements, while also seeking to influence public law institutions; private law can be manifest through standard setting as well as through contract law more generally. The use of private law (contract provisions and arbitration agreements) often utilises public international law as a background justification but equally is crafted to serve the needs of the particular corporate network in which it is deployed (Gessner, 2012: 158-159), rendering this part of that corporation's governance functionality. However, as Danielson (2005) has pointed out the decisions and actions of corporations have social consequences largely indistinguishable from those created by public regulators, but... corporate decision-making [i]s largely insulated from public participation, engagement or scrutiny... If corporations are significant institutions in the transnational governance regime, then policymakers and activists will need to find ways to affect the decision-making of these corporate institutions. (424) The democratic deficit often identified as compromising the legitimacy of global governance, is repeated and is likely more serious where corporations by their very actions are regulating economic interactions, even if these are also shaped by various other regulatory regimes and to a large extent internal to contracted networks across their global supply chains. The choice between competing standards, differential corporate governance regimes and the incorporation of national rules into standard corporate practices allow corporations to decide which regimes they might use, how they interpret them and if, as in the face of no acceptable standards or rules they can and need to set their own for their network's internal relations.
A key challenge of governance for many corporations is that theirs may only be one among a number of supply chains serviced by a particular supplier, and their ability to govern the actors involved is therefore incomplete; moreover there is much less clarity (than for states for instance) about the legitimate reach of corporations' jurisdiction as regards the other organisations in the supply chain, and hence their need to work to maintain legitimacy within the network. One might expect that the more any sub-contracting network partner was dependent on the core corporation the more compliant they would be, although there is also an issue of specificity of their own technology and/or other assets that might mitigate the asymmetry of power prompted by a narrow range of network engagements. This leads Macdonald (2014) to characterise the governance of supply chains as exhibiting "decentralised, non-hierarchical dynamics, without established deliberative processes or other norm-or proceduregoverned mechanisms to determine outcomes or resolve conflict" (179). For our purposes this then demonstrates the utility of developing a critique of (global) governance in these circumstances, identifying how parallel problems in other issues areas have played out. Thus for instance, where the governance of labour standards has become an issue for the core corporation, then there may be a political need for the governance function, and the regulations to which it is applied to become a more collective (which is to say "democratic") endeavour among a number (sometimes quite a large number) of corporations/ contractors and civil society actors. Here the internal governance function of the corporation and the realm of private authority, as exemplified by civil society organisations (and civil regulation) interact.

Global governance by global corporations
If the supply chains of global corporations are effectively a realm of global governance then how might we understand the practices that characterise this governance function? A range of supply chain management mechanisms that are relatively widely recognised includes: contractual contingency planning: recognising that contracts will be incomplete as they are unable to definitively predict all eventualities between contracting partners; target setting: may be part of the initial contract or a continuing set of negotiations between the contracting supplier and the core corporation (or a designated intermediate node in the supply chain); operational reviews: ranging from issues around quality of product through to internal (or external) reviews related to process and conduct, often linked to core corporations' CSR undertakings; information sharing: including data utilisation, revised and new technological specifications as well as advice on efficient processes/practices; joint problem solving: building on data/knowledge sharing, core corporations may seek to develop solutions with particular partners to enhance efficiency or compliance with adopted standards; supplier support: here corporations may aid partners develop their businesses through the provision of guidance and direct aid of various forms (utilising organisations like the Supplier Excellence Alliance for instance). 10 While these are all important to the profitable and effective management of the supply chain, the overarching governance function is intended to ensure that these practices remain uncontentious and acceptable to the contracting suppliers.
This network governance function is manifest in four key dimensions: firstly, the history of interactions between the corporation and its supply chain "partners" will have some influence over the manner in which the current relationship is governed-current arrangements will reflect prior engagements; secondly, a range of economic issues from tooling costs to investment in labour preparation (training, recruitment) and other (potentially) relationship specific costs shape relations; thirdly these are linked to technological issues around access to, and development of, new technologies; and finally a range of political issues can shape any relationship (from external state political concerns to internal micro-political considerations) (Cousins, 2013). Governance therefore will be concerned both with the generalised needs of the supply chain alongside the particular requirements of specific interactions within it. Indeed, popular demands that corporations act responsibly under the rubric of CSR already recognise that corporations are effectively governance institutions able to prompt shifts and changes in practices (albeit imperfectly) within the political economic spaces they govern (Anderson, 2006: 34;Macdonald, 2014: 169). It is relatively common then to recognise that corporations are political institutions, as well as economic ones, but this is seldom pushed to assess the more general governance function that such recognition must imply! Gary Gereffi and Joonkoo Lee have very helpfully developed a typology of general governance forms that pattern global supply chains; their analysis aims to understand the manner in which governance practices allow the capture of value across the supplier network, often referred to as a global value chain analysis. Their five types of supply chain governance range from hierarchical at one end (essentially fully integrated corporations with wholly owned subsidiaries) to market governance at the other, where all interactions in the chain are conducted via the market and governance is merely through the price mechanism (spot market). Between these two extremes a new set of "specific exoskeletal institutional arrangements" have been developed, 11 facilitating a range of balancing points between competition and cooperation, while also allowing for periodic re-organisation of production networks to deal with new challenges.
It is the three intervening types that reveal the sites of where the (global) governance function of corporations (and large firms) is evident. These are: Modular: Suppliers work to the standard specifications of purchasing agents in the chain; while these can be complex, communication between suppliers and purchasers allows these intermediate goods to be relatively easily specified. Relational: Relations are deeper, involving negotiations about burden sharing and knowledge transfer; these will be increasingly based on social ties between partners and burgeoning trust relations. Captive: Relations are (perhaps) pathological, with suppliers increasingly dependent for their ongoing profitability on a few (or even single) buyers/networks operating under conditions that are relatively rigid and/or intrusive and may be specific to a particular supply chain network (Gereffi and Lee, 2012: 25-26). 12 Crucially, in extended supply chains or production networks, different parts of the network may be characterised by different types of interaction. 13 In the past these general forms of supply chain were mobilised around core production-focussed corporations but increasingly global supply chains (especially in textiles and consumer technologies) have become organised by and around key buyers, controlling global brands. The particular relations within networks fall into one or more of these broad categories at least partly on the basis of the possibility (or lack of possibility) of easily codifying requirements for the supplier. The more complex the requirements (the more tacit knowledge and coordination required) the more likely relations will move from modular, to relational to captive at that point in the network (Gereffi et al., 2005). Thus, governance failings and the lack of ability to achieve required supply outcomes may shift relations into more explicitly integrated forms of governance and by doing so increase the direct exercise of power by the central coordinating corporation (but also its costs in time, organisation and resources).
When requirements are easily codified or even standardised (commodity-like) supply requirements, governance of relations may be mostly through a combination of contracts and/or tendering standards, that at the limit are completely undertaken as spot market interactions. However, as the intermediate goods required in the supply chain become more complex then increasingly technical standards, quality control mechanisms, service level agreements and other more developed contractual elements may be deployed to govern the relations. Seldom can a contract effectively deal with the intricacies of ongoing supply chain relationships on its own, and thus effective governance is crucial to the success of any network. This recognition that contracts are always incomplete (by virtue of the future being indeterminate) has often been treated as an argument for some (mission critical) functions to be brought within the formal legal borders of the corporation to allow for a flexible but hierarchical control. However, the need to control risk still may produce a preference for inter-firm contractual relations, leading to a need for a normative element to governance, to allow the corporation to achieve its particular strategic requirements from the supply chain. Moreover, while seeking to govern the internal network relations, corporations may also seek to obscure these relations from various external regulatory agents (Seabrooke and Wigan, 2014), leading to information asymmetries between them. Thus, the control of codification, alongside the definition and interpretation of acceptable practice all become important elements of the global corporate governance function.
Many corporations therefore have developed detailed codes of conduct for their network "partners". These codes reflect the direct needs and requirements of the core corporations around process, and often can reflect concerns about the perception of how a corporation enacts is purported social responsibility. CSR can be seen as a signalling function, to consumers (around brand differentiation), to workers (to attract and retain certain sorts of prospective employees) and to governments (to forestall more legalistic and formal regulation); corporations promoting CSR are trying to say something about themselves even if some critics regards such invocations as being articulated in bad faith, and a mask for the operation of a rapacious global capitalism. Indeed sometimes it seems that the assessment of risk has moved from a primary concern with impact of actions on others, to a concern for the risk that the publicity around adverse impacts may have economic costs to the corporation itself. Where this aspect of risk becomes central, critics understandably see CSR as merely marketing or a cynical interest in brand values. Thus, to add credibility to these codes (sometimes voluntarily, sometimes under political pressure) the auditing of compliance can be outsourced to other organisations in what Vogel (2010) has referred to as "civil regulation". Responding to perceived publicity effects and issues of brand integrity, some global corporations such as Levis or US Whole Foods have quite significantly shifted their practices (most often related to their supply chain) in response to codes of conduct emanating from civil regulatory organisations. The approach of marrying internal codes of conduct to external regulatory competence has perhaps been taken furthest in the (albeit controversial) UNGC, 14 and the development in the mid-2000s of the UN's "Norms on the Responsibilities of Transnational Corporations and Other Business Enterprises with Regard to Human Rights".
While there may be differences on the effectiveness of these norms, what is notable is that they imply a clear recognition of a direct governance function for the core-corporation over its complex networked global supply chain. The required administrative processes and evaluative mechanisms are at the very least analogous to, and in many ways directly parallel, the more often recognised institutions of global governance.

Global governance research and global corporate networks
Perhaps the most obvious link between an analysis that sees corporations as institutions of global governance and previous political economic analysis is the work on private authority (see for instance Hall and Biersteker, 2002;Büthe and Mattli, 2011). Here the focus is on the manner in which private actors are able to shape and inform the practices and structures of global governance. However, this approach seeks to understand global corporations as network actors deploying their authority (built up through their market activities) to shape existing institutions of global governance. Global corporations may be deploying their authority to influence existing institutions or through standard setting (and other normative mechanisms) seek to establish new institutions that reflect their collective interests, for instance by normalising market-based decision making. Global governance remains something corporations seek to inform, influence and shape beyond their own networks. Conversely, as I have developed above, global governance is something global corporations are doing within their own networked relations, suggesting that they have less need to rely on external private authority and are often directly governing their own networked supply chain. 15 The first element of this refocussing is to recognise that the relations of the global corporate supply chain are like other social networks; they are patterned by power relations. In the first instance, this is the core corporation's ability to deny access to the supply chain itself. In this sense, the networked supply chain resembles the pro-actively assembled group of partners that pattern almost all institutions of global governance: joining the network requires the explicit acceptance of institutional rules or conditions; prospective members must commit to the conditions of the network, but also to accept the governance of the lead corporation, as members of global governance institutions expressly accept the leadership of the institutional secretariat (however constituted). Once a firm has joined the network, apart from the threat of expulsion (for whatever reason), there is also the question of the allocation of resources, tasks and activities which are the subject of the governance function and are the result of power relations within the network (adapted from Baudry and Chassagnon, 2012). However, the final sanction of removal from the network is hardly the basis of a nuanced governance regime, and thus a more complex analysis is required.
Moving away from the threat of network-expulsion, and following the analytical approach adopted by Fuchs (2007), the power of corporations in global governance can be disaggregated into three elements: instrumental power-encompassing direct influence, based on the deployment of corporations' extensive and diverse resources; structural power-revolving around rulemaking and the establishment of agreed rules of practice; and discursive power-by which certain settlements are politically legitimated outside the formal processes of rule adoption, and corporations have been able to normalise various aspects of their practices and needs (56-58). The first dimension Fuchs identifies relates to the exit-threat issue, and to corporations' ability to control access to their producer/supplier networks. The second dimension plays out in the supply chain through the multifaceted negotiation and surveillance activities that I have discussed above. This element is clearly strengthened by a (perhaps variable) ability to utilise the exit-threat in specific instances (although when there have been significant sunk costs in an on-going network relation a break would not be costneutral for the core-corporation, and contractors are well aware of this). The discursive element in the governance of the supply chain is evident in the manner in which the corporations' codes of conduct seek to normalise specific behaviours and practices, but most importantly through supplier support and contractor engagement intending to socialise partners into the particular ways of working of the supply chain. 16 This can be then linked to Michael Barnett and Raymond Duvall's frequently cited typology of forms of power in global governance, which as noted above, encompasses four elements: compulsory, institutional, structural and productive power. While Barnett and Duvall allow that the resources that may be deployed as compulsory power may include symbolic and normative resources, for our purposes this maps relatively easily on to Fuch's instrumental power. The second element, institutional power seeks to capture the manner in which certain actors can further their interests and end via sets of institutionalised rules and practices (Barnett and Duvall, 2005: 16-17). If we regard the market as an institution, then market imperatives fit into this element, as do issues around legal regulation and (perhaps) codes of conduct, although that is perhaps more ambiguous. Barnett and Duvall's third element, which they call structural looks a little different from Fuch's conception, primarily due to their focus on how this element of power constitutes socio-economic capacities by allocating these capacities differently, and influences actors perception of their place and possibilities for action (Barnett and Duvall, 2005: 18). To some extent then this aspect of their model brings in the discursive elements that Fuchs is keen to emphasise, with the ability to normalise specific relations and network positions as a crucial manner in which corporations seek to govern their supply chains. However, this discursive constitution of relations then also works for Barnett and Duvall as their fourth element of power in global governance, the productive (where they explicitly include discursive elements) ( Barnett and Duvall, 2005: 21).
Bringing this together to think about the global corporation as an institution of global governance suggests that such a focus would not only require an analysis to take account of instrumental or compulsory power differentials in supply chain networks, but to focus more fully on the manner in which corporations seek to constitute and produce the socio-economic and spatial terrain that the supply chain encompasses. Specifying the required ends, however, requires more than an easy presumption of profit maximisation; it is seldom if ever the case that a corporation has a single directing mind or a closely knit management team with a single and defined set of governing principles and norms. While incorporation and contract relations might often give credence to a reified view of the identity/ personality of the corporation, the corporation like many institutions of global governance is actually subject to "bureaucratic politics" which is to say, internal bargaining, negotiation and (internal) political resources actually shape decision making.
Thus, and again following Dallas, it might be better to understand the governance function of the global corporation as being located within a "power coalition"; here power is exercised by, and on behalf of a coalition of actors and groups. Therefore, Dallas (1995) argues: while corporate goals derive from individual and group goals, no one individual or group is usually powerful enough to impose its goals on the corporation. Through corporate processes, corporate goals emerge that satisfy the parties or parties comprising the dominant coalition; these goals are distinct from the goals of any one individual or group. Furthermore, goals result from the communication among individuals and groups concerning opportunities and dangers represented by or to the organisational structure itself. (51)(52) Taking this approach the governance of global corporations' supply chains starts to resemble the forms of on-going (re)negotiation and alliance building that typify many of the (usually identified) institutions of global governance. This is not to say that no corporations are driven by a singular focus on profit maximisation when governing their supply chain, but rather that other goals of governance may frequently arise from different power coalitions within global corporate networks (Dallas, 1995: 58-59). At the very least there are likely to be considerable differences on how to maximise profit! The approach outlined above suggests a number of aspects of a critical perspective on the governance of global supply chains. If governance is to be seen as legitimate by those who are within the network, this approach implies that for governance practices to be effective they must be regarded as transparent and have modes of accountability to deliver the value that can be derived from these relations, over and above any basic market-based form of contractual relations. However, the closer that the relations between network "partners" approximate market relations (and are governed via market mechanisms), the less legitimacy will be required. The notion of the market will still need to command legitimacy but this is a wider issue and not immediately effected by one corporation's action within its supply chain. While it may seem a little odd to refer to a democratic deficit in supply chain governance (although in itself this may be the result of the discursive normalisation of market relations as being "outside" democracy), where governance is not limited to basic market mechanisms the reciprocity captured by this phrase is likely to be of some importance. In global governance the response to such deficits has often been to empower, or seek to empower civil society, which directly parallels the moves in CSR monitoring and auditing (and indeed is often used as an example in mainstream debates). Therefore, like other global governance institutions a fall-off of perceived legitimacy may prompt exit from the network-suppliers will seek other more amenable supply chains where plausible. Where lock-in issues are evident, then other resistance activities may be deployed, and civil society organisations may be drawn in to engage with the legitimacy of specific elements of the governance function.
However, in supply chain networks the actors that might benefit from empowerment are not only civil groups (labour, environmentalists, those seeking to promote human rights) but also the smaller contractors themselves, whose interests and requirements may often be quite different to the usual civil society actors identified in accounts of global governance. Thus, and interestingly, as Richard Locke points out in his discussion of the manner in which labour standards have been promoted in supply chain networks, whereas in early engagements auditing and inspecting was seen as a process for gauging the need (or otherwise) for sanctions for non-compliance, more recently auditors (internal and contracted-external) have regarded their work as developmental. It now often encompasses not merely judgement of the fulfilment of codes' requirements, but also, when failings are identified, prompts support to enhance existing and introduce new practices that are code-compliant (Locke, 2013: 181 and passim).
The negotiation and socialisation of network participants and at least a partial recognition of their interests may be the manner in which corporations most effectively govern these networks. This approach has been evident (and well publicised) in the garment supply chains of various large branded clothing corporations (of which Gap, Levis and Nike are perhaps the most obvious examples) leading to a more direct engagement with suppliers in the chain to develop and promote schemes intended to enhance the workers' experience and build less onerous (but more efficient) local working practices. However, as Locke discovered in a comparison of two of Nike's strategic partners in Mexico, individual plants/contracted suppliers can have quite different experiences within the same supply chain; in this case one supplier had a developmental relationship with Nike staff, gaining support and advice for up-skilling and better labour conditions; the other had a more formalised, distant and lesssupportive engagement and as such retained many of the shortcomings phased out from the first site. Interestingly, while Nike's intent seemed to be the same with both, the greater distance to travel to the second site, leading to a great reliance on phone/email communication and fewer visits resulted, Locke (2013) concludes, in very different responses to the "demands" of Nike's supply chain governance; one site moving towards preferred practices, the more distant adopting a minimalist (partial) compliance position (Chapter 5). Again such practical and spatial issues are well documented in mainstream global governance analysis, and suggest that as there, global supply chain governance will vary not merely by intent but through accessibility and communicative effects, alongside surveillance opportunities and "diplomatic" relations.

Conclusion
Global corporations are often engaged in the management of extremely complex networks of subsidiaries, affiliates and contractors, and thus the realm over which they have influence ARTICLE PALGRAVE COMMUNICATIONS | DOI: 10.1057DOI: 10. /palcomms.2015 (and power) is often global. As the UNGC, EITI and other moves to regulate labour standards (Locke, 2013) suggest, there is a clear recognition in policy circles that corporations are governing these networks not merely managing them. However, as production networks have fragmented, as markets have become globalised and technological developments have become less linear, hitherto hierarchical power relations between core/lead corporations and the companies in the (global) production networks have become less stable, less settled and perhaps less unbalanced (Herrigel and Zeitlin, 2010: 553-554). As such, the governance of production networks has become less like direct command and control, and more like models of governance that have developed in other areas of the global system.
Given that we already understand global governance as a diverse set of institutions across a fragmented and discontinuous political domain, it follows that we might also regard the management of a corporation's global complex as fulfilling a similar (partial, but global) governance function. Therefore, the purpose of the argument above has not been to suggest that the recognition of the global governance function fulfilled by corporations should replace the analysis of the impact corporations have on other institutions of global governance, nor the discussion of how corporate practices may be shaped by interactions with various international organisations. Rather, by adding a third dimension to the account of corporations in global governance the account of how the contemporary global political economy is ordered and governed is made more comprehensive. This account is meant to go beyond approaches that recognise corporations as "partners" in global governance (Lamrad, 2012) to suggest that much of the time, within their own networks they are the primary agent of governance and not necessarily dependent on other actors or social institutions for their legitimacy.
As Macdonald (2014) suggests a focus on the political economy of the supply chain allows analyses to recognise "the distinctive forms of social power and organisation that exist within these supply chains, and the contested distributional consequences of such power". Moreover, "simple assumptions about the effective subordination of economic power to state authority in global supply chains are becoming increasingly implausible" (11, emphasis in original). Understanding global corporations' control of these supply chains as global governance immediately opens up this realm of the global political economy to an interesting set of analytical tools already developed to examine more normally conceived forms of global governance, while also offering a complementary analysis that indicates how global governance may be experienced across the global system. By utilising the insights from the study of global governance, we can develop a critical approach to assessing the manner in which corporate rule over their supply networks is justified and the limits to legitimacy this governance function might confront.
Finally, however, the claim that this third dimension is a useful site of further research and analysis is not a claim that all corporations are equally able to govern their networks to the same degree or level of effectiveness. Indeed, as with the constellation of institutions and organisations that represent the usual scope of global governance studies we are likely to find some corporations achieve a level of control as good as, or even beyond that which the most effective global governance regimes exhibit; equally we will find other global corporations that are effectively unable or unwilling to establish strong governance functionality in the same way that we can identify weak or partly ineffective global governance institutions. Most importantly, we can usefully include global corporations within our analysis of global governance more generally to develop a better account of both global governance and the practices or global corporations themselves.
Notes 1 For other ways of seeing global corporations, see for instance: Forsgren (2013) or Collinson and Morgan (2009), both of which offer a range of other "lenses" an analyst could adopt.