Business authority in global governance: Companies beyond public and private roles

International studies investigate the governance authority of state versus non-state actors in terms of their public or private authority. However, the public–private distinction does not sufficiently capture the variety of governance actors, or the forms of their authority, beyond that distinction. Focussing on businesses, this paper argues that certain governance actors assume public and private roles, as well as a third category of roles it calls ‘societal’ that transcend notions of public and private. To understand these roles and how they affect governance authority, this paper treats the public–private relationship as mediated and extends it with the ‘societal’ category, then translates it into the concept of business authority, which constitutes a particular form of governance authority alongside public and private authority. It does so by operationalising governance authority as a triadic concept composed of power, legitimacy and a connection to public interests. In all three components, business authority escapes the binary distinction between public and private without simply merging the two.


Introduction
A high variety of actors partake in and shape global governance. While this has motivated international studies to 'push the study of global governance beyond the notion of "add actors and processes into the international organization mix and stir"' (Weiss and Wilkinson, 2014: 213), analysing and distinguishing the forms of their authority remains a challenge. Authority in global governance refers to the ability to govern and regulate public matters, and to take part in developing, interpreting and implementing norms and laws. Governance studies investigate private and transnational actors as regulators, thereby extending the classic approaches of international law and international relations that treat states as the only governing actors.
Business companies represent one of the most relevant actors in global governance besides states (Hofferberth, 2019;May, 2015;Mikler, 2013). They are perceived not simply as economically powerful actors, but as actors endowed with the legitimacy to partake in global governance, and thus enjoy a form of political power. This is often referred to as private authority (Cutler et al., 1999), distinct from the public authority of states. However, as Fuchs (2013) summarises: 'While thus provided with a form of political authority and legitimacy as a political actor in the wider sense, [. . .] companies have not come to be seen as political actors in the narrow sense ' (p. 87, also cf. Cutler, 2018: 62;Flohr et al., 2010;Wettstein, 2009). This shows how the public-private distinction is insufficient to fully capture the complex forms of companies' governance authority, without simply equating them with states' political authority.
This paper argues that a major reason for this gap is that businesses assume roles that escape the binary distinction between public and private actors and the associated difference between a political public sphere and an apparently apolitical private sphere. By contrast, and resonating with widespread critiques of such a binary distinction, this paper understands the public-private relationship as mediated (i.e. mutually constitutive) and extends it with a third category that it calls 'societal'. To avoid simply conflating all types of non-state governors, the paper focusses on businesses and their business-societal roles. This lays the foundation for introducing the concept of business authority in global governance, capturing governance roles that are neither distinctively public nor distinctively private. Far from negating public and private roles -which remain pivotal to understanding governance authority -societal roles in business authority capture activities and their effects that do not fall into either category.
Although the notion of global governance has inspired considerable rethinking of the dichotomy between public and private, a third category is needed because the distinction between state actors (as public) and non-state actors including businesses (as private) remains pervasive. It continues to shape global governance in many ways, including international law and the human rights regime. Two vignettes illustrate this point.
First, Facebook was widely criticised in 2018 for not (immediately and comprehensively) removing hate speech and pro-genocide pages directed against the Rohingya minority in Myanmar. Remarkably, these pages had been launched by members of the state's military -that is, public actors (The New York Times, 2018). Second, Facebook caused controversy in the United States the same year when it deleted accounts and an associated event page that called for a march against a white supremacist rally in the aftermath of a violent demonstration in Charlottesville, Virginia the previous year. The company explained that it took this decision because the accounts were linked to disinformation campaigns connected to Russia. However, local civil society actors who also supported the event page accused Facebook of censorship and impeding their right to free speech (The Washington Post, 2018).
In both cases, a private company got involved in public matters. It reacted in exactly the two ways that are traditionally prescribed in international human rights law and its distinction between public and private actors. In the first case, Facebook (initially) fulfilled a private role by not regulating human rights violations committed by other actors (particularly state actors). In the second case, it acted in a public manner by regulating a (public) sphere of deliberation, getting involved with (and restricting) human rights. However, neither the public nor private reaction was sufficient from a normative democratic or human rights-based point of view.
Additionally, in a growing number of cases, businesses do not simply act in a classic private or public way. Rather, they become involved in regulating other actors or developing and interpreting human rights norms -that is, with global governance -while also pursuing private interests. Therefore, when assessing companies' governance authority, a third option beyond the public-private distinction is needed conceptually, empirically and normatively. In reaction to that need, a growing number of studies perceive businesses as hybrid, rather than purely private or public actors. These studies differ, however, in whether they consider the public-private distinction as blurred, see the two sides as a spectrum, add a new component or regard the two as mediated.
Discussing the strengths and the shortcomings of these analytical models, this paper proceeds from the model of mediation to argue that businesses as governance actors assume public and private roles, as well as a third category -societal roles. Defining governance authority as the power to participate in governance (i.e. to regulate matters that affect public interests) that strives or appears to be legitimate due to a connection to public interests, the paper distinguishes three interrelated components of authority: power, legitimacy and a connection to public interests. The paper detects businesses' public, private and societal roles in all three components, which builds the basis for introducing the concept of business authority alongside public and private authority.
Ultimately, this paper conceptually captures a phenomenon that fuels international studies as well as human rights discussions empirically and normatively: the governance authority of non-state actors. Introducing a third category to the classic public-private relationship, and applying it to the particular governance authority of businesses, avoids the binary distinctions between public and private, and between political and (apparently) apolitical realms, without simply merging them. Rather, the public, private and societal roles of global governors constitute the particular form of their governance authority. This approach helps lay the foundation for normative discussions and legal solutions regarding their responsibilities. Their strong effects on public interests, human rights and democracy require responsibility mechanisms beyond purely private regulation, on the one hand, and state-like duties, on the other.
The paper proceeds as follows. Section 1 reviews the literature on public and private authority and examines how multiple models (blurring, spectrum, novelty and mediation) portray businesses as hybrid, rather than purely private or public actors. Mediation serves as the basis for investigating the public-private relationship in Section 2, where I propose a third category -'societal' -to capture governance roles beyond public and private. Section 3 provides a definition of governance authority and operationalises it with the three components of power, legitimacy and connection to public interests. On this basis, Section 4 elaborates the concept of business authority by identifying businesses' societal roles with regards to their power, legitimacy and connection to public interests in global governance. Section 5 summarises the argument and revisits the two Facebook vignettes to suggest normative conclusions and areas for future research.

Public authority, private authority and beyond
International studies have established an elaborate link between public and private spheres in efforts to address the ways in which private, that is, non-state, actors become involved in matters of public interest. These studies have complemented the classic notion of public authority (wielded by states) with the concept of private authority (wielded by non-state actors). In addition, studies of companies as governance actors highlight their public, private or even hybrid roles. This section provides an overview of these approaches and discusses their strengths as well as some of their shortcomings to set the stage for proposing a mediated approach between public and private.
According to the classic notion of public or political authority (Pierson, 2015), the state is endowed with the power and legitimacy to safeguard public interests, which are (ideally) determined via democratic mechanisms (Best and Gheciu, 2014). This authority is designated as public, whereby the '''publicness' of authority is constituted by a distinction between private and public activities' (Cutler, 1999: 63). This division is based on the assumption that there is an apolitical private sphere (in which economic interactions, market, family, contracts and other supposedly private activities are conducted), juxtaposed with a public sphere of politics (critically Cutler, 1999: 64).
Transgressing the dichotomy of public versus private spheres, the concept of private authority has proven innovative by capturing situations in which private actors regulate public matters. The concept thus combines two ostensibly contradictory dimensions: interconnecting public matters (presumably confined to public authority) with private actors reveals a genuine association between the two spheres of public and private.
Similarly, research on corporate social responsibility and the human rights responsibilities of businesses discusses the public roles of private businesses (Deva and Bilchitz, 2017;Fuchs, 2013;Scherer et al., 2016). Other approaches stick to the concept of 'public authority, independent of the question whether it is carried out by state or non-state actors. It involves an element of "publicness"' (Zürn, 2018: 4). Hence, even approaches that hesitate to use the notion of private authority acknowledge that global governance connects private actors to the public sphere. Studies on the close interaction between the two spheres (Cashore et al., 2021) consolidate this point. They all transcend the assumption that there is a rigid dichotomy between public and private.
These contributions to breaking down the dichotomy between public and private cannot be overestimated, and serve as a pivotal basis for my argument. However, they preserve the public-private distinction in their two-sided division between public (state-related) authority, on the one hand, and private authority performed by non-state actors, on the other. In this context, the term 'private actors' continues to serve as a catchall category for governance actors 'with no formal ties to state authorities' (Cashore et al., 2021: 4). Going one step further, studies on the governance roles of businesses take one of three forms. They differ in whether they perceive businesses as (1) solely private, (2) rather public or (3) somehow hybrid actors.
The first approach emphasises the private character of businesses: they are private legal entities, legitimately acting in self-interest. As Friedman (2007) famously summarises: the only 'social responsibility of business is to increase its profits'. While this approach takes the private roles of businesses seriously, it cannot capture all the ways in which businesses transcend the private realm.
The second approach conceives of companies as (almost) equal to governments (Wettstein, 2009). It underlines their public character, derived from their public practices: The fact that what appear to be private actors [. . .] are increasingly engaged in public practices [. . .] is too easily viewed as simply another example of the rise of private authority -missing the crucial ways in which these practices are redefining those actors as public because of what they do, not where they are situated (Best and Gheciu, 2014: 17).
While this approach takes the public roles of businesses seriously, it risks equating businesses with states. This might neglect their pivotal differences, for example in relation to their democratic legitimacy or legal roles.
The third approach therefore entails a number of diverse solutions that aim to capture companies' concurrent public and private roles. 1 These hybrid perspectives treat corporations as 'partially private, partially public' (Bilchitz, 2016: 164, similarly Ciepley, 2013. Many (at least implicitly) resort to one of four analytic models of the public-private relationship: blurring, spectrum, novelty and mediation.
One very common analytic model assumes a blurring. In order to transcend the dichotomy between public and private, blurring indicates that the two sides dissolve into each other, making the distinction between the two permeable. A similar notion refers to a common sphere in which public and private elements intermingle. However, if taken seriously, this model implies a mere blending of private and public elements, which does not allow us to identify these elements individually and how they differ from each other.
The notion of a spectrum perceives public and private as opposite endpoints, with degrees of more or less public (and accordingly less or more private) actors or activities in between. A spectrum thus distinguishes between different forms according to their degree of publicness or privateness. However, it falls short in two respects. First, since it describes public and private as two opposite (and remote) points, it cannot account for how the two interact or even mutually constitute each other (i.e. their mediation, which I introduce below). Second, a spectrum as such can only capture degrees between public and private, but nothing that exists beyond the spectrum, such as new or different qualities.
By contrast, the novelty model seeks to capture the new roles of companies beyond a spectrum or a mere blurring. According to Ruggie (2004), 'in many instances of "private governance" there has been no actual shift away from public to private sectors. Instead, firms have created a new transnational world of transaction flows' (p. 503, emphasis added). Similarly, Abbott and Snidal (2009) call regulatory standard setting a 'new form of transnational "regulation"' (p. 45, emphasis added). However, business roles have shaped global governance long enough that they cannot truly be considered 'new' anymore. The novelty model thus indicates a persistent struggle to capture these roles within the frame of the public-private relationship. Against this background, this paper suggests a way to characterise companies' 'new' governance roles while taking their concurrent public and private roles seriously. Put differently, this paper advocates an approach that considers all the different business roles without merging or neglecting any.
To this end, this paper uses the mediation model, which emphasises the mediated relationship rather than a binary distinction between public and private. This model also appears in feminist approaches, which have been criticising the dichotomy between public and private spheres in liberalism and international law for decades (Chinkin, 1999;Pateman, 1983). Such studies point out the political nature of the private sphere, in that it is coined by public matters (e.g. power, gender inequality and public regulation), and underline how (traditionally women's) activities within the private sphere (household, care work, parenting) are not simply dichotomously detached from the public realm of politics, but constitute and enable the latter in the first place. The mediation model hence captures how public and private mutually constitute each other, which builds the foundation for my argument.

The public-private relationship and its extension with the 'societal'
This section elaborates the public-private relationship as mediated and proposes a third distinct, yet interrelated category, 'societal', which does not dissolve the public and private sides, but interacts with them. The notions of public, private and societal refer to actors as well as their roles or realms. Thus, a single actor may perform several concomitant roles -and may still be adequately captured as a public or private actor (depending on the context). For instance, a state that performs private roles most likely remains a public actor yielding public authority. By contrast, companies perform public and private as well as societal roles to such an extent that they cannot be sufficiently described as private actors exerting private authority. They instead assume business authority, as I argue below.
The model of mediation used here resonates with the aforementioned feminist approaches, but draws on a more nuanced assessment of internal and external mediation (based on Hegel and Adorno, as discussed in Müller, 2020;Ritsert, 2017). The external mediation between public and private means that each sphere's inclusion is the other sphere's exclusion, and vice versa. Based on this view, governance actors are either public or private. This echoes the common understanding in global governance that public and private are 'two faces of the same coin' (Graz and Nölke, 2008: 11), or that one sphere represents the residual of the other (Hall and Biersteker, 2002). Yet the mediation model between public and private goes further. The spheres are also internally mediated, which means that each (re)produces and (in)forms the inner scope, logics and content of the other. Each sphere has functions of (and effects on) the other, through regulation or even non-intervention. Their mutual constitution is so pervasive that each serves as a constitutive part of the other; neither would exist without the other. However, they remain distinct: they do not dissolve into each other, but retain their own logics. This is what distinguishes the model of mediation from those of blurring or binaries. Note that what is considered public or private in a given context may vary considerably, but their mutually constitutive relationship remains (Mende, 2020).
Nevertheless, even the mediated public-private relationship does not adequately apprehend the 'new' governance roles mentioned above. Therefore, this paper suggests broadening the concept to a three-sided relationship that includes the proposed societal category. I do not use 'society' here to indicate an overarching, all-embracing category. Rather, I borrow from sociology the habit of identifying sub-units within society (writ large) as societies as well. 'Societal' thus refers to practices and roles that are neither tied to states or state-like functions (public), nor confined to a realm of self-interest and individual pursuance without a connection to public interests (private). Borrowing again from sociology, the term can be hyphenated to refer to multiple societal roles, depending on the actor in question. Civil-societal roles, for example, capture how civil society actors transcend the public-private distinction. Since this paper investigates business actors, it focusses on business-societal roles.
I apply the mediation model to the three-sided relationship accordingly: public, private and societal are distinctive sides that simultaneously constitute and contain each other. For example, the family is at the heart of presumably all otherwise contingent definitions of the private realm. At the same time, gender relations and power inequalities within the family are strongly shaped by both public (e.g. state regulation of marriage, divorce and child care) and societal sides (e.g. businesses' wage politics, recruitment and promotion strategies). Thus each side affects (and is affected by) the other two. Yet the three sides do not simply dissolve or blur into each other. They remain distinct, with discrete features, roles and power relations.
This model of the public-private-societal relationship provides a basis for discussing the concomitance of different roles that governance actors can assume: public roles, private roles and roles that go beyond these categories without simply negating them. The shift from actors to roles makes that concurrency and variety of public, private and societal roles analytically accessible. The remainder of the paper applies this notion to business authority.

The triadic concept of governance authority
In order to apply the three categories of public, private and societal roles to the study of businesses as global governors, the paper operationalises governance authority with three related components: power, legitimacy and connection to public interests. This triadic concept draws on a definition of governance authority as the power to participate in governance (i.e. to regulate matters that affect public interests) that strives or appears to be legitimate by a connection to public interests. These three interconnected components constitute different forms of authority in global governance, as each component may have different weight and forms. 2 The triadic concept thus pertains to soft and hard, formal and informal, liquid (Krisch, 2017) and reflexive (Zürn, 2018), state and nonstate (Ruggie, 2004: 519) forms of governance authority, varying in the degree of institutionalisation and the scope of competences. The triadic concept thus facilitates an understanding of different forms of authority. It provides a common framework of analysis, reconciling competing approaches that pay different degrees of attention to each component. This section briefly outlines the three components and their intersection in the triadic concept of governance authority.

Power
Power is central to all definitions of authority, even those that disagree over whether power equals authority. According to the triadic concept of authority, power does not equal authority, but serves as one of its components. Power comes in one of three forms, material, agenda setting and ideational, as conceptualised by Lukes (2005: 20) and further developed in international studies. The distinction between the three forms is analytical; empirically, they closely interact.
Material power mostly appears in realist approaches. This form of power allows actors to advocate their interests in decision-making processes (Lukes, 2005: 16, based on Dahl, 1957). It is primarily concerned with influencing the output side of politics, based on material resources as well as enforcement instruments (Barnett and Finnemore, 2005: 176;Fuchs, 2004: 136).
Agenda setting power denotes the capacity to set agendas, frame knowledge and define problems and their solutions. It enables actors to shape the input side of decisionmaking and prevent other actors from formulating their own interests and agendas (Lukes, 2005: 20 based on Bachrach and Baratz, 1970, also cf. Barnett and Finnemore, 2005: 177;Fuchs, 2004: 137).
Ideational power lies at the heart of constructivist approaches in international studies (Carstensen and Schmidt, 2016;Holzscheiter, 2005). This notion captures the ability to form and shape (the perception of) ideas and interests even before they become part of a political agenda. It works through socialisation, internalisation and incorporation via norms, rules, standards, identities, discursive practices and institutions (Lukes, 2005: 139ff.). In a nutshell: it 'does not simply pursue interests but creates them' (Fuchs, 2004: 138).
Some scholars add a fourth form of productive power, based on Foucault's understanding of power as productive and producing (Digeser, 1992). Lukes rejects such an extension as it conflates power as domination with any form of socialisation (Lukes, 2005: 88ff.). Current international studies tend to blur the line between the third and fourth forms, as both operate through similar mechanisms such as socialisation and internalisation. For both reasons, I am referring to the three forms as outlined above when speaking about power as (overt or subtle) domination or framing.

Legitimacy
The second component of governance authority is legitimacy. Legitimacy indicates an actor's or institution's 'minimal moral acceptability' (Buchanan and Keohane, 2006: 219) and justifiability (Beetham, 2013: 11ff.). It thus empirically denotes 'the normative belief by an actor that a rule or institution ought to be obeyed. It is a subjective quality, relational between actor and institution, and defined by the actor's perception of the institution' (Hurd, 1999: 381, emphasis in original).
Using this common approach, authority concepts can be categorised into two strands based on the weight they give to legitimacy. The first suggests that authority equals legitimate power. According to this view, 'the phrase legitimate authority is, strictly speaking, redundant' (Hurd, 1999: 400, emphasis deleted, also cf. Lake, 2010, drawing on Weber, 2002. The second strand emphasises the difference between authority and legitimacy: 'authority is to be distinguished from legitimacy: authority implies a rebuttable claim to legitimacy' (Bogdandy et al., 2017: 140, referring to Raz, 1988. According to this strand, the authority of governance actors may lack legitimacy. While these studies maintain that authority is based on recognition (Zürn et al., 2012: 83) or legal sources (Bogdandy et al., 2017), this may or may not lead to legitimacy. Nevertheless, they do ascribe some importance to legitimacy, as governance authority cannot survive without it in the long run (Zürn, 2018: 9 and62ff., similarly Buchanan andKeohane, 2006: 407;Tallberg et al., 2018: 3).
Both strands thus envision a strong connection between power and legitimacy. The triadic concept of authority reconciles both strands by capturing legitimacy as one of three components with varying forms, weight and scope. Authority in global governance denotes power that is, strives, or appears to be legitimate. The legitimacy component can be contested or diminished, but not eliminated. Authority may rise or fall as legitimacy increases or decreases, but the strength of authority also depends on other factors, such as the substance of the shared norms and values that constitute legitimacy -particularly the connection to public interests.

Connection to public interests
The third component of governance authority is how governance power relates to public interests: 'The core demand of authority is to make the institutionalization of power in the best interests of the governed population' (Koppell, 2007: 194). The connection to public interests is a necessary component of governance authority, not least because authority may restrict its subjects: 'authority is the adoption of an act that affects the freedom of others in pursuance of a common interest' (Bogdandy et al., 2017: 117). Conversely, authority can be contested and limited with regard to its contribution to public interests (Avant and Haufler, 2014: 48). Thus, the component of public interests builds on normative terms, which complements the empirical category of legitimacy. 3 Accordingly, I use public interests in the plural, as their content may vary widely, being a matter of deliberation and contestation.
Global governance builds upon the assumption that regulation can no longer be undertaken only at the domestic level; cooperation and regulation beyond the state are necessary to solve common problems and contribute to public interests (Ruggie, 2004: 500;Scharpf, 2000;Zürn, 2018: 4). Thus, the claim to contribute to public interests is the normative core of global governance, either implicitly with the general intent to solve problems, or explicitly. While explicit orientations towards 'global public goods' (Zürn, 2018: 249) are also a matter of contestation, it is possible to identify shared global norms (Wiener, 2007). These include international peace and security as the founding principles of global cooperation in the United Nations, as well as the protection of human rights.
Even apparently neutral implicit references, such as efficiency or problem solving, connect to public interests because something can only be considered efficient with respect to a normative yardstick or aim, and problems can only be solved by defining and framing them based on normative assumptions of who, how or what they are affecting or regulating.
The triadic concept of authority captures the connection to public interests in two ways. First, it entails how an actor claims to contribute to public interests, and thereby seeks legitimacy (Reus-Smit, 2007: 159;Zürn, 2018: 4). Second, it captures an actor's effects on public interests, including detrimental ones. The component does not assert the fulfilment of public interests, or stakeholders' perception that they are being fulfilled; rather, the latter is representative of legitimacy.
The important point for my argument is that a connection to public interests is what takes governance authority out of a purely private sphere. A parent regulating their children's playing behaviour, for instance, does not constitute an example of governance authority. The demand to regulate all children's playing behaviour in the neighbourhood to safeguard public tranquillity, however, might do so if it represents an aspiration to assume power over others that strives for self-legitimation by claiming to contribute to public interests. In the context of this paper, a private company regulating other companies via codes of conduct designed to respect human rights, for instance, is not a purely private matter either, as it involves human rights and hence a connection to public interests.

Business authority in global governance
This section elaborates the concept of business authority by examining the multiple roles that businesses perform in global governance. The concept of business authority draws on how business power, legitimacy and connection to public interests include roles and functions that cannot be described as purely public or private; rather, they form a third category of business-societal roles, as introduced above. These roles do not simply substitute private and public roles; they supplement them. They are entangled (mediated) with both types, but escape the divide in important ways. Business authority therefore clearly differs from states' public authority. Advancing and adding to the notion of private authority, the concept of business authority offers an extended frame to deal with the intersections of public and private and the appearance of new roles in global governance.
Three additional conceptual clarifications help elaborate the concept of business authority. First, while business power, legitimacy and connection to public interests constitute business authority in global governance, each component may also exist outside the realm of governance authority. For instance, companies may wield power in informal markets or over enslaved people. This power is not necessarily related to or endowed by global governance.
The second clarification is that while the three components may strengthen or weaken each other respectively, other mechanisms outside the triad can also do so. These include competition and economic pressure, political and legal regulation, protectionism, the extent and complexity of global markets and value chains, as well as targeted marketing strategies. This explains why the three components do not deterministically influence each other. For example, more power does not necessarily equal greater legitimacy and hence stronger authority. Social businesses illustrate this point, since their orientation towards public interests does not endow them with more power than purely profit-oriented multinational companies. Furthermore, some companies can wield power without legitimacy -especially those that are less visible to the public, such as those that operate in the folds and gaps of global governance, in informal markets and the lower tiers of supply chains. By contrast, the more a company acts as an authority in global governance, the clearer is its connection to public interests, and the greater its need for legitimacy.
Third, the concept of business authority leaves room for considerable variation in how businesses perform various roles, which is demonstrated by the different weight that can be given to the three components of authority. This allows me to use the broad category of businesses as governance actors -an umbrella term that includes different companies and their functions. Further sub-differentiation (which is beyond the scope of this paper) could be based on differences in their size and global reach (most notably transnational companies vs small and medium-sized companies), sector, the form and institutionalisation of their participation in governance, their corporate cultures and forms of corporate governance. Businesses also vary with regard to differences between formal and informal sectors, publicly owned or privately held companies, the tiers of supply and value chains, sector-specific differences and the political and judicial system of a company's home or host state. The scope and extent of the individual components of business authority -as well as the concomitant public, private or business-societal roles -vary according to these differences. Hence, context and individual cases matter (Hofferberth, 2019: 3) in further sub-differentiating business authority, for which this paper provides a basis.
The remainder of this section makes the case for the concept of business authority by exploring the ways in which business roles resist being characterised as either public or private. It addresses this question by analysing companies' power, legitimacy and connection to public interests in global governance, and the relationships among these. It shows that in all three components of business authority, companies assume both private roles (their self-interest, their assignment to a state's jurisdiction and to civil law, material power) and public roles (their regulation of others, the provision of public interests). They also assume a third category of roles that cannot be described as either public or private: business-societal roles (the extensive effects of their actions on others, the double-edged access to and demand for legitimacy, the concurrency of private and public interests, agenda setting and ideational power).

Business power in global governance
More traditional accounts conceive of companies' power as economic power that is supposed to belong to the private realm and is distinct from political power in the public realm. This can be captured in terms of a company's material power, which is closely tied to its size, resources, networks and wealth. However, this type of power also manifests in the ability to directly influence others' behaviour and the output side of politics, traditionally via lobbying (Fuchs, 2013: 82ff.). Thus businesses also exercise political power using their economic power. In addition, companies' agenda setting and ideational powers not only represent an application of their economic power; they exceed it.
Companies' agenda setting power to influence policy input is particularly connected to global governance institutions when companies participate in international organisations (e.g. companies taking seats in the Universal Postal Union, a specialised UN agency), public-private partnerships, multi-stakeholder initiatives and roundtables. It also manifests in the roles that companies play in framing the norms, standards, certificates and rules they use to regulate their own and other companies' (but also state and other actors') behaviour (Flohr et al., 2010;May, 2015;Peters et al., 2009).
Companies wield ideational power when they shape ideas, norms and identities via their products, advertisements, and the ways they engage in society, such as when they participate in 'processes of sense-making as well as interpretive frames' regarding, for example, a certain human rights violation (Hofferberth, 2017: 138). Even before the rise of global governance, Bowen (2013Bowen ( [1953) described the extent to which business activities shape lives and society (p. 8ff.). He argued that business behaviour influences individuals' standard of living, economic stability and economic progress, but also societal order, national security, justice and freedom. Accordingly, he described business conduct as 'not only a means to human life and human ends but a large part of human life, and an end in itself' (Bowen, 2013(Bowen, [1953: 11). State laws and business codes of conduct targetting, for example, sexist stereotypes in advertisements, demonstrate a high awareness of business ideational power. Hence, while companies may strive to exercise ideational power on top of their usual activities (e.g. by advertising certain positions in public debates), it is already integral to their core business activities.
Ideational and agenda setting powers exceed economic power and reveal business roles in (and for) society. They depend on non-material factors and a company's reputation, such as its moral standing or (perceptions of) the quality of its expertise. These powers are thus most closely connected to a company's legitimacy. At the same time, they are also related to its material power, as in many cases 'a firm's ability to become politically active [. . .] also increases with its size' (Bernhagen andMitchell, 2010: 1177).
Business power evades the public-private relationship in two ways. First, the sheer scope and effects of material, agenda setting and ideational power transcend the business realm to such an extent that they outgrow solely private roles. They shape and (co-)constitute the private sphere, such as with their effect on gender relations within the family via hiring and wage politics (Prügl and True, 2014). Companies also affect other actors beyond the private realm, including states. For example, their codes of conduct may expand to a whole sector, and they may dictate (or forestall) state laws (Porter, 2005). Their products may widely influence the quality of living, for example, by affecting the nutrition and health of large groups of people. Yet although these roles extend beyond the private realm, companies are not simply public actors; they also retain private roles, such as pursuing their private interests.
Second, the ways in which business power is (or is not) restrained also transcend the public-private distinction. Although companies (or their legal representatives) are de jure subject to a state's jurisdiction and to civil law (marking them as private actors), companies (particularly transnational companies) de facto can wield their powers beyond domestic legislation on a global level. Significant regulatory gaps in international law and global governance allow companies to evade the power restraints that usually apply to private actors -without equating them with public actors. Recent developments in international human rights and investment law further contribute to business roles beyond public and private: while international law does not treat companies as subjects comparable to states, it acknowledges them as actors in international law (Mende, 2021;Wouters and Chané, 2015). Multilateral investment agreements and arbitration mechanisms treat companies as global subjects, which augments their power (Choudhury, 2009) and takes them out of the private sphere -but again does not equate them with public actors. The power of companies is not only private, but not simply public either. Rather, their power signifies a third category: business-societal.

Business legitimacy in global governance
According to Bowen's (2013Bowen's ( [1953) paraphrase of Abraham Lincoln's bon mot regarding democratic legitimacy, 'business, like government, is basically "of the people, by the people, and for the people" Such power [. . .] is given because the 'people" believe this decentralization to be desirable' (p. 5f.). This (very far-reaching) legitimacy is mirrored in similar (though less far-reaching) global governance mechanisms that base business legitimacy on the belief in a company's expertise (Cutler et al., 1999), the quality and effectiveness of its products or services (Brühl and Rittberger, 2001), or its social behaviour manifesting in codes of conduct, social engagement and dialogue with social and political partners (Voss, 2013: 30). As studies on private authority (Cutler et al., 1999) have pointed out, democratic mechanisms of legitimation that (ideally) apply to public actors' legitimacy are replaced by other sources of legitimacy for private actors. Apathy (Hall and Biersteker, 2002: 5), silent approval (Wettstein, 2009: 210) and the absence of contestation may also bolster companies' de facto legitimacy. Notably, business legitimacy does not necessarily involve a direct relationship with the affected. It may also rest upon the recognition of companies by other global governance actors, such as in governance fora or public-private partnerships.
Business legitimacy and power in global governance are closely related, in both an enabling and mutually restrictive way. If a company manages to strengthen its legitimacy, for instance by providing expertise or plausibly referring to human rights, it consolidates its power in global governance, and thus its authority. At the same time, increased civil society attention to business conduct with regard to human rights compels companies to legitimise their activities. Brand-sensitive or otherwise publicly visible companies can no longer choose to ignore critiques or remain silent (Brühl and Hofferberth, 2013). They must take action to avoid damaging their reputation and legitimacy, and thus their authority. This gives other actors an instrument with which to contest or even restrict business power.
Legitimacy in global governance is therefore a double-edged sword for companies: they have access to public legitimacy and they are pressured to publicly legitimise their behaviour. Accordingly, in the two introductory vignettes, Facebook was pressured to publicly react to the criticism. Purely private actors do not have to legitimise their private activities to the public; companies (at least publicly visible ones) do. At the same time, the demand for legitimacy gives companies access to public legitimacy, which drags them further out of the private realm. Referring to human rights (or other public interests) helps companies legitimise their activities and provides them with a 'social license to operate' (Ruggie and Sherman, 2017: 294) far beyond their private roles. Yet, neither empirical developments nor normative demands equate companies' legitimacy to that of states. The double-edged legitimacy (access and demand) instead takes on a particular business-social form.

Business connection to public interests in global governance
Business activities are connected to public interests in manifold ways that have been extensively covered in international studies. These connections are what gave rise to concepts of private authority and the hybrid roles of companies in the first place. Companies are involved in providing public goods, such as funding research at public universities (Best and Gheciu, 2014: 18f;Moon and Knudsen, 2018). They are even involved in matters of conflict, peace and security through the privatisation of security and armies (Deitelhoff and Wolf, 2010), which the Westphalian world order framed as the prerogative of states. The human rights regime has begun to take business roles and responsibilities for human rights into account (Deva and Bilchitz, 2017;Mende, 2021), and companies actively intervene in societal discussions to (re)frame certain public interests (Hofferberth, 2017).
How is the business connection to public interests different from public and private roles? For states and international organisations, this connection (even if it is contested or underdeveloped) serves as the foundation of their public authority. States are endowed with judicative, legislative and executive powers in order to safeguard public interests (Abrahamsen and Williams, 2014;Best and Gheciu, 2014: 32). For this reason, states are (or should be) democratically legitimised. Even though not all states embody that ideal, it serves as a normative yardstick that contrasts with dictatorship and totalitarianism.
Companies' activities clearly differ from these public roles. Instead, they are commonly perceived as legitimately following their private interests -just as any other private actor. 4 They do not have the same duties (or associated rights) to uphold public interests as states do. However, companies do have a large effect on public interests, and are 'seen as the primary expert, able to deliver in "the public interest"' (Fuchs, 2013: 87). Even if they do not claim or want to contribute to public interests, business decisions can have tremendous effects on these interests. The two Facebook vignettes illustrate this point: due to its functions, roles and the extent of its power, Facebook cannot help but affect public interests, regardless of whether it is active (as in deleting an event against a racist rally) or remains inactive (as in not deleting pro-genocide pages). Companies are so pervasively connected to public interests that their roles in global governance cannot be perceived as purely private. Their connection to public interests is not simply public either; it can be better captured with a third category of the business-societal.

Conclusion
This paper discusses the multiple forms of authority that non-state actors can assume in global governance by going beyond the classic public-private divide. It introduces a third category, 'societal', to capture new and hybrid governance roles. In order to avoid simply neglecting the continuing relevance of public and private roles, the paper uses the analytical model of mediation that captures how the public, private and business-societal sides mutually constitute each other. To investigate these roles in terms of governance authority, the paper operationalises authority as consisting of three components: power, legitimacy and connection to public interests. This framework is applied to companies as crucial governance actors. The paper introduces business authority as a concept that supplements the concepts of private and public authority: business authority clearly differs from public authority, which captures the public power, public legitimacy and public interest orientation of states. It also provides a more nuanced notion than that of private authority to capture how businesses exert power, enjoy legitimacy and affect public interests in concomitantly pursuing public, private and business-societal roles.
The triadic concept of governance authority therefore allows us to distinguish the manifold forms of authority that non-state actors in global governance can assume. However, this paper does not seek to replace one passe-partout category with another. On the contrary, it aims to create a basis for further differentiation in two respects. First, it allows us to differently weight the individual components of authority: power, legitimacy and connection to public interests. This also allows the necessary differentiation between companies of different size, type, agency, sectors and leverage. Second, the analysis of public, private and societal roles in each of the authority components can be extended to other global governors as well. The next most prominent type of actors to investigate is probably civil society actors. Their civil-societal roles differ from businesssocietal roles in decisive ways and accentuate a different form of governance authority than examined here, which may also function as a counterpart to business authority.
In sum, this paper illustrates a broader phenomenon that characterises global governance -governance roles that escape the public-private distinction. Labelling these as societal roles and underlining their mediation with public and private roles helps avoid the trap of equating civil society or business actors as global governors with the rights and duties of states, or with each other.
It is important to point out that the concept of business authority captures companies' existing governance roles. Hence, this paper does not normatively demand business authority; rather, it investigates its empirical presence. It thereby provides a basis for discussing the normative and regulatory questions that arise from such authority.
The two Facebook vignettes from the introduction demonstrate the potential for follow-on normative discussions. Facebook empirically assumes roles (and power) that make its intervention in human rights issues unavoidable. Yet the traditional public-private divide in human rights law offers only two solutions, neither of which helps protect human rights in these cases: refraining from action or performing far-reaching regulative functions in lieu of a state, but without a state's democratic legitimation and accountability mechanisms.
This impasse is not confined to the Facebook vignettes, but marks long-standing debates on the social, environmental and human rights-related responsibilities of businesses. As long as we maintain the two-sided dichotomy between public and private actors, we can only resort to two regulative solutions. But neither treating companies as purely private actors with no (or very limited) further responsibilities for public matters, nor associating businesses with the duties, powers and democratic legitimacy of states, suffices to protect public matters such as democracy and human rights. For this reason, normative and regulatory debates need to catch up with the empirical development of companies assuming roles beyond the private realm (without simply equating them to public actors). A clearer identification of these roles as societal provides a basis for developing new or hybrid regulatory mechanisms and forms of responsibility that evade the public-private divide as well.
Such developments are beginning to coin current legal and normative debates. With regard to Facebook's power, yet lack of democratic legitimacy, to regulate human rights issues (and more generally the digital sphere as neither a purely public nor private realm), they involve, for instance, democratic participation via joint governance platforms including independent experts and civil society actors (Fertmann and Kettemann, 2021), or new legal doctrines and norms related to the freedom of expression (Arun, 2021). Since Facebook's power makes its activities so relevant to public interests, that power might as well be used as a yardstick to require Facebook to contribute to public interests (Wood, 2012). 5 And vice versa, regulating a company's governance power, for example by granting a company access to governance fora, could be tied much more closely to its actual contributions to public interests, such as its human rights performance. Overall, these ongoing discussions illustrate how addressing challenges in global governance requires going beyond the public-private divide (however without dissolving it, since the realms of private law or public state authority, for instance, continue to assume important functions in global governance).
This paper responds to this need by highlighting a way of perceiving global governors that is not restricted by the dichotomous choice of perceiving companies (or certain other actors) as either public or private. Rather, it shows how companies can assume private and public as well as societal governance roles. This presents a way out of the impasse of whether to endow companies with only private responsibilities or to burden them with state duties. Ultimately, taking companies' empirical authority into account conceptually provides a basis for normatively addressing and regulating their roles, rights and duties in global governance.