Today it is widely accepted that political legitimacy is a desirable quality of international political entities, such as international organizations, and that we have good reasons to want them to have more of it. The goal of strengthening the legitimacy in the exercise of political power in the global domain has become a major concern among national governments, international institutions, civil society organizations, and concerned citizens alike. However, despite this broad consensus on the value of political legitimacy in the global domain there is still little agreement on what the specific regulative content of the principles of legitimacy ought to be – or even what are the basic normative concepts that best capture the category of principles that we ought to employ in the regulation of global public affairs. To explore this further is important not only for political theorists but also for empirical researchers. Empirical work on global politics is shaped by normative commitments to specific ideas of what are the appropriate solutions to problems of political conflict, arbitrary rule, and deep disagreement, which remain not only unarticulated but also undertheorized. Before we can pursue successful strategies for strengthening the political legitimacy in global politics and for framing empirical enquiry about its political and social preconditions, the development of clearer conceptual and normative frameworks is called for to better understand what political legitimacy in the global domain requires. It is as a response to this task that the present paper is meant as a contribution.
Research on global (or international) political legitimacy is still in its infancy, but two main paths have thus far been taken in the normative-theoretical literature to respond to the legitimacy deficit in the global domain: one via the ideal of justice, the other via the ideal of democracy. However, both seem to have limitations when theorizing political legitimacy at the global level. Although I am far from the first to point out that global political legitimacy (GPL) should not be conflated with either justice or democracy (Buchanan and Keohane Reference Buchanan2006; Caney Reference Christiano2009; Besson Reference Besson2009a), it is often still theorized in terms of a less demanding notion of justice and/or democracy in the literature, and above all, it is still far from clear what (if anything) makes GPL conceptually and normatively distinct (for exceptions, see Buchanan Reference Buchanan and Keohane2011, Reference Caney2013; Macdonald Reference Millermanuscript). A major limitation of the ‘justice path’ is that it tends to neglect the procedural aspects of legitimacy, which cannot be fully responded to via substantive demands such as through the redistribution of basic human rights. Legitimacy is something that must be achieved rather than received, through justifiable procedures. A major limitation of the ‘democratic path’ is that the ideal of democracy, the rule by the people, on a modern understanding concerns law- and policy-making. But numerous global governance arrangements have other functions in global politics, such as international organizations concerned with policy enforcement, or those concerned with interpreting and applying law, such as international courts (ICs) (Buchanan Reference Caney2013). While it seems unacceptable to demand of such organizations to become democratic, it seems equally unacceptable to allow such political entities to exercise political power without any requirements of legitimacy. The overall purpose of this paper is to examine these two paths in the endeavour to explore the possibilities of a third path, which investigates GPL as a special kind of normative good for political entities (powers, agents, arrangements, decision bodies, and institutions) in the global public domain – a value that is at a basic level distinct from the values of justice and democracy in that it is not reducible to either of them.Footnote 1
More specifically, the paper aims to fulfil two tasks: one conceptual and one normative. The conceptual task consists in identifying some characteristics of the concept of GPL that makes it distinct from political legitimacy generally, as well as arguing for its usefulness for normative theorizing. The normative task is twofold: first, to demonstrate that the value of GPL is reducible neither to justice nor to democracy; and second, to develop the contours of a dual account of GPL, in which both justice and democracy play essential roles.Footnote 2
Admittedly, justice and democracy have not been the only concerns when theorizing legitimacy in international politics and there are many alternative ways to approach GPL. In the International Relations (IR) literature, for example, the justification of international organizations was mainly measured in terms of effectiveness and efficiency until the beginning of the 1990s. The motivation for focussing on justice and democracy in theorizing GPL is to do with the nature of the argument that will be pursued, which may be described as ideal-theoretical (for an overview, see Valentini Reference Valentini2012b). It is not ideal-theoretical in the sense that GPL is theorized under an assumption of ‘full compliance’ or in the form of an ‘end-state theory’ (Rawls Reference Reidy1999). Neither is the project pursued here ideal-theoretical in the ‘utopian’ sense working without feasibility constraints (Cohen Reference Dahl2008). Rather, it is best described as ideal-theoretical in that it adopts permissive feasibility constraints on principles of GPL. According to these constraints, such principles should be compatible with the basic features of human nature as we know it as well as be possible to achieve from the status quo (Buchanan Reference Buchanan2004; Gilabert and Lawford-Smith Reference Goodhart2012).Footnote 3 Apart from these feasibility constraints, the defended account adopts the accessibility constraint that GPL is desirable and worth pursuing if it is not morally unapproachable in the sense of involving extreme moral costs (Buchanan Reference Buchanan2004).
To take a comprehensive and ideal-theoretical approach to such a complex matter as GPL is not without risks, since such a philosophical exercise is abstract and therefore will bracket numerous concerns, not least pertaining to questions of realizability, institutional design, and non-ideal theorizing of current social practices.Footnote 4 But my hope is that a broad outlook that systematically theorizes GPL could be beneficial for the debate as a whole, since GPL is rarely addressed as a distinct concept or as a distinct value, but is more often discussed indirectly through other concerns of international politics or through domestic (nation-state) lenses.
The structure of the paper is straightforward. The first section is devoted to the conceptual task, outlining the general understanding of GPL as it is applied in the paper and demonstrating its usefulness for normative theorizing. The subsequent two sections are devoted to the normative task. Here I begin by taking a closer look at attempts to theorize GPL from the ideal of justice and of democracy, respectively, and discuss their limitations. Against the backdrop of these theoretical considerations, the third section develops the contours of a dual account of GPL – which consists of two principles: ‘the equal say principle’ and ‘the principle of public legitimacy’ – as well as makes some remarks on how this account may be applied in practice. The final section discusses the strengths and weaknesses of the account in relation to competing accounts.
GPL and conditions of applicability
In the theoretical literature, political legitimacy is generally described as a virtue of political arrangements and the rules (laws and policies) made within them. It refers to the justification of coercive power or the justification and sanctioning of political power or authority, usually signifying the right to rule and on most accounts entailing political obligations (Christiano Reference Christiano1996; Wellman Reference Wellman1996; Buchanan Reference Buchanan2002). It is further commonly presumed that principles of political legitimacy are intended to regulate the relationship between what may be broadly categorized as ‘rule-makers’ (political entities) and ‘rule-takers’ (political subjects), that is, between political entities that make, apply, and enforce rules and the subjects to whom these rules apply (Buchanan Reference Buchanan2010; Valentini Reference Valentini2012a). Hence, a principle of political legitimacy is a special kind of normative principle distinguished by its function to regulate rule-makers by specifying standards for ruling, that is, the conditions under which these political entities are entitled to exercise political power, and are worthy of compliance and/or support (Rawls Reference Reidy1999; Buchanan Reference Buchanan2002; Buchanan and Keohane Reference Buchanan2006; Hurrell and Reference MillerMacdonaldReference Kingsbury, Krisch and Stewart2012; Valentini Reference Valentini2012a). While there are disagreements between individual theorists concerning how to best characterize this in detail, there is wide agreement on this general understanding of the term, which suffices for the present purposes.
The context in which political legitimacy has been theorized is primarily that of the domestic domain. However, in response to the extent to which severe problems increasingly transgress nation-state borders – such as world poverty, migration, inequalities, and trade – ever more attention has been directed to what political legitimacy may mean in the global domain. It is far from obvious that the principles providing political legitimacy within states are necessarily the most appropriate ones for providing it within global arrangements. Several tendencies in the last decades have reinforced the need to look at the specific circumstances of global politics, such as the emergence of new forms of arbitrary rule, relations of domination, conflict and deep disagreements, and the exercise of unchecked powers, as well as new kinds of powerful actors (Bohman Reference Brighouse and Fleurbaey2004; Reference MillerMacdonaldReference Millerand Ronzoni Reference Macdonald2012).
Despite the many differences between the domestic and global domains, however, a general characterization of the concept of GPL should, similar to the concept of political legitimacy, be permissive enough to allow a rich variety of specifications. This way we are able to compare and critically assess the strengths and weaknesses of different conceptions of GPL.
Like political legitimacy, GPL is supposed to tell us when a political entity is entitled to exercise political power. But is there anything that would make GPL a distinct concept of legitimacy? An important step towards identifying some distinguishing properties is taken by Buchanan. In his view, the predominant philosophical view of political legitimacy, which employs a very strong understanding of ‘the right to rule’ suitable for a nation-state context, is ill-fitting for global politics. While such a statist view emphasizes the exclusive right to use coercion to secure compliance, there is no reason to be so restrictive about which institutions could count as legitimate or not when concerned with the global domain – in particular since none of the existing international institutions rule or claim to rule coercively in this robust way (Buchanan Reference Buchanan2010, 82). Therefore, a weaker coercive element seems sufficient for GPL. Following Buchanan, I think we have strong reasons to interpret the element of ‘being morally justified in exercising political power’ in terms of issuing rules and seeking compliance or non-interference when concerned with legitimacy in the global domain, where benefits and costs are attached to, for example, compliance and non-compliance. Such a modification would still hold on to the relational component of the statist view of the right to rule in the form of substantial, content-independent moral reasons for complying with the rules or for not interfering with the efforts to govern (Buchanan Reference Buchanan2010, 83–84).
Indeed, it is important to recall that concepts in contrast to, say, normative principles, do not have propositional content and cannot therefore be true or false (correct or incorrect). Concepts classify objects and can be more or less successful in doing so (see List and Valentini Reference Lopez-Guerra2016). Consequently, it is not a fruitful aim of philosophers to try to determine which conceptual account of GPL is the right one, since this would presumably depend on the purpose for which it is to be applied, such as what kind of building block it is supposed to constitute in a theory or a principle. Buchanan’s modification in the form of a weaker coercive element is one important way of making the concept of GPL useful for theorizing normative standards for many kinds of international institutions.
Yet, this modification seems insufficient since it still applies only to rule-making, where legitimacy is seen as the right to rule. On the commonsensical understanding of political legitimacy in the philosophical literature, both the idea of ‘the right to rule’ and the idea of ‘rightful authority’ equate rule and rule-making with law and law-making (Buchanan and Keohane Reference Buchanan2006). Moreover, legitimacy is almost exclusively associated with legitimate authority and is discussed in legal terms, such as in terms of the capacity to impose (legal) duties, as justified (legal) coercion, or as the right to rule (make laws) (Reference CohenChristianoReference Christiano2013). However, if political power that does not manifest itself as authority in this legal sense is not a proper object of the concept of GPL, it would be problematic, since most global governance arrangements and international institutions are not law-makers. Even when this condition is relaxed to include governance activities generating ‘law-like’ rules and policy-making (Buchanan Reference Buchanan2010, 91), it still seems too restrictive for a global context. If the task of principles of GPL, similar to principles of political legitimacy in a domestic context, is to regulate the relationship between political entities and those over whom they exercise or project power – setting out standards for when this exercise is justified – focussing on law- and policy-making would allow political entities to exercise power through other ways of conduct, for example, in the form of domination and general norms, while escaping the demands of legitimacy. This seems undesirable.
Therefore, it would be beneficial to divorce the concept of GPL from the concept of authority, even if a normative account of GPL in the end will place greater demands of legitimacy on authorities with legal capacities. What distinguishes the domestic from the global context is that while different kinds of political entities have different functions domestically (courts, administrative bodies, military power, etc.), they are characterized by a certain kind of unity, where the legislature (in democratic states) has a kind of supremacy (though limited by the constitution established by the constituent power) in that it lends legitimacy to the other entities to carry out democratic functions. In the global domain, however, these different kinds of entities are not united but to a large extent dispersed. For this reason, we need principles of GPL to be able to regulate a wide range of ‘freestanding’ entities – with, for example, administrative, judicial, and executive functions – independent of whether they are established and mandated by democratic decision-making or not. This diversity is rarely explored and theorized in the literature on GPL, where legitimacy is, if not fully reduced to legitimate (legal) authority, at least intimately related to it (von Bogdandy et al. Reference Bohman2010; Zürn et al. Reference Zürn and Joerges2012).
Insofar as we do not want political entities to be able to steer clear of any requirements of legitimacy by simply exercising power through non-legal means, we have reason to broaden the domain of application of the concept of GPL to include not only objects concerned with law- and policy-making, but also those that exercise political power in other ways. Buchanan (Reference Buchanan and Keohane2011, Reference Caney2013) has taken this path in recent writings, where he extends the applicability of legitimacy in a global context beyond rule-making. In a similar manner, I suggest that the set of objects of which it is meaningful to ask whether they are legitimate or not are not only rule-makers (law- and policy-makers), but all those entities that exercise power over subjects in the global public domain.
This needs to be specified further, so let us move from the concept’s domain of application to its extension, that is, from exploring of which objects it is meaningful to ask whether they are legitimate or not, to specifying which kinds of political entities fall under the concept and thus are candidates for being subjects of principles of GPL. Here too it is useful to adopt a permissive framework in light of the aim of elaborating the contours of an ideal-theoretical account of GPL, not least because narrowing the focus to one specific kind of political arrangement would leave political powers from which we ought to demand legitimacy outside the scope of political control. Rawls is right to note that we need to know about the nature of the object that is supposed to be regulated by specific normative principles in order to properly theorize those principles (Rawls Reference Rawls1971, 29). But what we need is not a full-fledged conceptual account but a characterization of the object’s core properties. Insofar as normative principles are supposed to regulate the conduct and structure of a practice, this practice puts limits on what the regulative principle may be, such that any candidate principle must satisfy a condition of applicability (Beitz Reference Beitz2014, 227). With regard to GPL, it seems reasonable to presume that not any object that exercises power over subjects in the global public domain would fall under the concept and thus be proper subjects of principles of GPL. Even on a broad understanding, two defining features stand out as essential.
First, it seems sensible to require that power in the public domain (political power) is exercised in a purposeful way, that is, it is not the result of unintended patterns of behaviour. Thus, objects of the concept of GPL must be agent centric (see Hurrell and Reference MillerMacdonaldReference Kingsbury, Krisch and Stewart2012). To the extent that GPL concerns the relationship between political entities and those over whom they exercise power, these entities must be agents in the sense of acting in intentional ways in order for the subjects to be able to hold them to account. If accountability is not possible, legitimacy is not applicable. Second, their exercise of power has to be systematic. For while the legitimate regulation of political power calls for common legal and political institutional structures, we cannot set up an institution as soon as, say, a group of people subjects a person to arbitrary rule. What is constitutive of legal and political institutions is their sturdiness, that is, their relatively stable and sluggish structure (Erman Reference Erman and Möller2014b).
Hence, on this broad outlook, political entities that exercise purposeful and systematic power over subjects in the global public domain are proper subjects of principles of GPL.Footnote 5 In addition to the exercise of purposeful and systematic power, I see no reason why political powers must also have the capacity to effectively protect core political values to qualify as a proper subject of GPL, as suggested by some theorists (Hurrell and Reference MillerMacdonaldReference Kingsbury, Krisch and Stewart2012, 559–60). The principle of ‘ought implies can’ is indeed complex and contestable, but to stretch it to ‘ought implies effectively capable of’ not only seems unnecessarily strict, it also seems indefensible in view of the present purposes. On such a reading, if an entity intentionally and systematically exercises political power over others, but does not have the capacity to protect key political values, for example, by advancing ‘goods that are grounded in consensus and common interests’ (Hurrell and Reference MillerMacdonaldReference Kingsbury, Krisch and Stewart2012, 560), it may well continue doing so without any demands of legitimacy until it happens to acquire this capability. The capability requirement thus seems arbitrary from a normative point of view.
Apart from specifying conceptual conditions for the ‘exercise’ side of GPL, something should be said about the ‘recipient’ side, that is, about those over whom power is exercised – which concerns what we may call the ‘uptake conditions’ of affected parties. On the traditional understanding of political legitimacy as the right to rule, those over whom the rules apply are often claimed to have a duty to comply (what we may call a strong uptake condition), or at least not interfere with the authority’s ruling (a weak uptake condition) (Buchanan Reference Caney2013; Reference CohenChristianoReference Christiano2013). However, apart from these strong and weak uptake conditions, it may also be the case that no particular proattitude is required towards the political entity on the part of the subjects as long as certain normative criteria are fulfilled in the very exercise of power. In such cases, the legitimacy of a political entity is merely a ‘justification’ right (see Reference CohenChristianoReference Christiano2013).
As will become clear in the third section where my account of GPL is developed, I think it would be a mistake to restrict principles of GPL to one of these uptake conditions (strong, weak, or no uptake condition) if we wish those principles to be applied to a wide range of purposeful and systematic power projection in the global domain. For while a strong condition of compliance (in the modified sense described above) seems fitting for democratic rule-making, neither a strong nor a weak uptake condition may turn out to be suitable for non-institutionalized practices.
Justice, democracy, and GPL
Since concepts are used as building blocks for normative principles and theories, conceptual and normative aspects of an evaluative concept such as GPL are deeply intertwined. On the one hand, we seem to need at least a shared idea of the general meaning of GPL – locating some basic characteristics, as it were – in order to make comparisons between different normative accounts of GPL and be able to evaluate their normative criteria. On the other, the motivation behind the above attempt to broaden our understanding of GPL is normative: to be equipped to evaluate situations as legitimate/illegitimate that do not fall under the traditional concept of political legitimacy. A good test of whether the suggested conceptual framework is plausible is thus whether it is proven useful for theorizing a normative account of GPL, which is what I will try to show in the rest of the paper. In this section, the first part of the twofold normative task is undertaken, notably, to demonstrate that the value of GPL is reducible neither to justice nor to democracy.
The ideal of justice and GPL
On the standard response to the question of what justice is, principles of justice are said to determine ‘who owes what to whom’, where entitlements are commonly expressed in terms of a set of rights. Furthermore, in contrast to moral theory in general, justice on most accounts primarily concerns the moral quality of basic institutions rather than individual actions. Even if there is disagreement on how to specify this in detail, there is broad agreement on the general characterization that principles of justice establish when institutions give their subjects what they are entitled to, that is, when they respect their rights (Valentini Reference Valentini2012a, 595). On contemporary liberal accounts of justice, which are committed to the principle of equal respect for persons, institutions are just in as much as they secure basic rights and realize a fair distribution of burdens and benefits among citizens (Rawls Reference Rawls1971; Dworkin Reference Erman2000; Tomasi Reference Valentini2001).
According to justice-based approaches to political legitimacy, the moral purpose of political power is foremost to achieve justice. Given the coercive character of political power, the argument goes, nothing short of justice could justify it (Buchanan Reference Buchanan2002, 709). On such a standard, a ‘wielder of political power’ – in the ‘making, application, and enforcement of laws’ – is legitimate if and only if it protects ‘at least the most basic human rights of those over whom it wields power’, and does so through actions and processes that themselves respect these basic rights (Wellman Reference Wellman1996; Buchanan Reference Buchanan2002, 703). Hence, on this and similar justice-based approaches, political entities are legitimate when they sufficiently meet the demands of justice by protecting certain fundamental rights (Buchanan Reference Buchanan2002; Reidy 2007). Focus is thus directed at minimal standards rather than full justice (Buchanan Reference Buchanan2010, 81).
In other words, a justice-based approach to political legitimacy is typically theorized in terms of specific substantive demands, expressed through basic rights.Footnote 6 However, independent of whether or not political legitimacy requires that such substantive conditions be met, I argue that procedural conditions are a necessary requirement of political legitimacy.Footnote 7 To this end, even if justice and legitimacy on a liberal view may both express the (same) value of equal respect for persons, as some would claim, they are still distinct values (cf. Valentini Reference Valentini2012a).Footnote 8 From the general understanding of GPL outlined so far, the primary relationship is not between individuals (or citizens) simpliciter, accounting for who owes what to whom, but between political entities and those over whom they exercise power. Therefore, the task of political legitimacy is not to tell us when a political power gives its subjects what they are owed, but when its exercise of power meets a certain normative standard such that it is worthy of, for example, compliance or non-interference (Valentini Reference Valentini2012a, 595).
In contrast to justice then, legitimacy consists of a two-way relation rather than a one-way relation: it is not foremost about the distribution of entitlements and rights and about subjects being the recipients of goods, but about the justifiability of political relations. Since legitimacy is concerned with regulating the relationship between entities exercising power and those over whom it is exercised, it is something that must be achieved – rather than received – through justifiable procedures (Erman Reference Erman2011, Reference Erman2013). Admittedly, the distinction between procedural and substantive conditions is controversial and sometimes not even logically sound, since we may re-describe a procedural condition in substantive terms (as it sets up boundary conditions for substantive outcomes). But here I apply the distinction in a straightforward way: to capture the difference between ‘how’ decisions are made (procedural), and ‘what’ is the substantive outcome/decision (non-procedural).
Even if it seems reasonable to favour a minimalist justice approach to GPL over a full justice approach, it is unsatisfactory to reduce GPL to justice. Since the criterion of minimal moral acceptability is a substantive demand, expressed through basic human rights, it does not accommodate a procedural criterion for regulating the relationship between political entities and those over whom they exercise power. For even though the minimalist view would require that the protection of basic human rights be achieved through processes that are themselves minimally just, that is, do not violate these rights (Buchanan Reference Buchanan2002, 708, 719), rather than being independently valuable for satisfying procedural demands, the role of such processual guarantees is to satisfy substantive demands.Footnote 9 Yet, it is the fulfilment of a procedural criterion that gives political legitimacy its content-independent character, i.e., gives those to whom the decisions apply content-independent reasons to, for example, comply, support, accept, or not interfere with the functioning of the political entity making these decisions, regardless of their positive or negative view of the contents (see, e.g. Buchanan and Keohane Reference Buchanan2006).
In contrast to liberals like Wellman, Buchanan does suggest a criterion that must be fulfilled independently of the protection of basic human rights, namely, the condition of non-usurpation. This condition is meant to secure that a political entity ‘does not come to wield power by unjustly displacing an entity that is politically legitimate’ (Buchanan Reference Buchanan2002, 708). This is indeed a process demand rather than a substantive demand, and it seems crucial from the standpoint of justice and political legitimacy alike. However, it cannot replace a procedural criterion for regulating the relationship between entities exercising political power and those over whom it is exercised, for the reason that non-usurpation refers to the origins of power, that is, the coming to power by a wrongful or illegitimate process, whereas principles of political legitimacy are supposed to offer standards for the rightful exercise of power by an already existing political entity.
The conclusion drawn in this subsection is simple: substantive demands fulfilled for example through basic human rights are not sufficient for GPL. As will become evident in the subsequent sections, there have been exceptions to the general neglect of theorizing procedural conditions in the more recent philosophical literature (Buchanan and Keohane Reference Buchanan2006; Buchanan Reference Buchanan2010, Reference Buchanan and Keohane2011, Reference Caney2013). Before developing my account in relation to these proposals, however, let us take a look at the relationship between the ideal of democracy and GPL.
The ideal of democracy and GPL
Apart from theorizing political legitimacy through the ideal of justice, the most common way to do so has been via the ideal of democracy. There is growing literature on democracy in a transnational and global context. Democracy, ‘the rule by the people’, is generally understood as a normative ideal of a form of collective self-determination or self-rule through equal decision power.Footnote 10 This is not the place to go into detail about different proposals for global democracy in the theoretical literature. Instead, the aim here is to pinpoint some problems involved in theorizing GPL from the standpoint of democracy. In order to do so in a structured manner, two interrelated questions are addressed. The first concerns conditions of democratic legitimacy: what are the minimal conditions of democratic legitimacy? The second concerns conditions of applicability: under what conditions is democracy an applicable ideal?
Indeed, similar to most (if not all) normative concepts, democracy is highly contested. Still, it is commonly agreed that democracy is a political organization or decision-making body that is considered legitimate if the rules that govern it are taken by those to whom the rules apply. The rule by the people alludes to ‘equal decision power’ or ‘equal political power’. If we unpack the idea of equal decision power, one condition comes to the fore as fundamentally important and seems hard to dismiss for any modern account of democratic legitimacy: political equality. What distinguishes democracy from other forms of political organizations, such as dictatorship, monarchy, or aristocracy, is that anyone to whom rules of a decision-making body applies should have an equal opportunity, secured through an equal right, to participate (directly or indirectly) in the decision-making about them (Reference CohenChristianoReference Christiano1996). Now, of course, satisfying the condition of political equality would in turn require that several additional conditions be met, such as accountability, transparency, and impartiality in the application of rules.Footnote 11 But these are inseparable from political equality in the sense that they cannot be assessed independently from the standpoint of democratic legitimacy.
In view of the condition of political equality, it does seem as if the ideal of democracy fares better than the ideal of justice for theorizing GPL, since it lays stress on procedural aspects (above all, procedures in which participants have an equal say over the decision-making), which is what the regulation of the relationship between wielders of power and those over whom they exercise power requires. So why do I want to argue that GPL is not reducible to democracy?
This brings us to the second question about the applicability of the ideal of democracy. As noted earlier, democracy on the established understanding alludes to a political entity in which the rules that govern it are taken by those to whom the rules apply. On this view, principles of democratic legitimacy are supposed to regulate the relationship between rule-makers and rule-takers and set up standards for ruling, that is, for legitimate law- and policy-making. Again, and not surprisingly, this has often taken place within a nation-state framework even if we today witness democratically structured entities beyond the domestic context, such as in the European Union.
But the exercise of purposeful and systematic power in the public domain comes in many forms apart from rule-making, for example, through general norms or forms of conduct such as domination and power abuse (see also Buchanan Reference Buchanan and Keohane2011). And it seems indefensible to claim that political entities would not have to live up to any standards of political legitimacy in their exercise of power just because they are not law- and policy-makers. Provided that we agree on this, two conclusions follow. First, following conceptually from the established understanding of the term, according to which principles of political legitimacy are supposed to regulate the relationship between rule-makers and rule-takers, democracy is not an applicable ideal in all political contexts, but only in what we may call ‘contexts of rule-making’, that is, contexts in which law- and policy-making takes place. In other political contexts, where power is exercised in the public domain through other forms of conduct, other normative criteria of legitimacy are called for. For this reason, second, an account of GPL should respond to two different contexts of political legitimation: ‘contexts of rule-making’ and what we may call ‘contexts of power projection’.
While contexts of power projection are relatively straightforward – since they are defined in the negative as the systematic and purposeful exercise of power in the public domain that is not rule-making – contexts of rule-making raise concerns about ‘statehood’ and ‘stateness’. Traditionally, core properties of statehood have been a monopoly of the legitimate use of force and bureaucratic control over a territorial jurisdiction (Koenig-Archibugi Reference Landwehr2011). However, the relevant question for an account of GPL is not statehood but stateness, that is, what is minimally required to constitute a context of rule-making. For example, empirical research on the EU systematically demonstrates that a central monopoly of force is not necessary to ensure high levels of compliance with the law (Zürn and Joerges Reference Zürn, Binder and Ecker-Ehrhardt2005). What seems needed, though, is some degree of political centralization (Koenig-Archibugi Reference Landwehr2011) or what we may call ‘sufficient stateness’. In this paper, it is assumed that contexts of rule-making possess sufficient stateness, which includes the following institutional features: (a) a legislative apparatus; (b) a sufficiently effective coercive apparatus or mechanisms that indirectly secure equivalent functions; and (c) an executive apparatus or mechanisms that indirectly secure equivalent functions.Footnote 12 An example of ways of indirectly securing equivalent functions is for an international treaty body to ‘outsource’ enforcement to states and to ‘softer’ mechanisms such as NGOs for promoting compliance. This functional integration with domestic institutions and entities, utilizing their powers, is common among certain kinds of international institutions (see Buchanan Reference Caney2013, 209–10).
To sum up, in view of the general description of GPL outlined in the first section, it is concluded that democracy – similar to justice – is not a sufficient condition for GPL, since procedural criteria need not necessarily take a democratic form. Before moving to the dual account developed in the next section, let me bring up three potential objections to the idea of law- and policy-making as a condition of applicability for the ideal of democracy. The first worry is conceptual, having to do with the fact that we have witnessed democracies in history, such as Ancient Greece, in which democratic decision-making was not a legislative practice and democracy had little or no connection to law-making. Would that not be problematic for the view of GPL defended so far? I think not.Footnote 13 The view of conceptual change adopted here is the uncontroversial Wittgensteinian idea that a principle or rule does not wear its application ‘on its sleeve’, but only has a determinate meaning given a background practice or custom; a way of applying it that is implicit in practice rather than explicit in the form of further background rules (Wittgenstein Reference Wittgenstein1953, §241). In line with this pragmatist view, the general understanding of democratic rule-making as law- and policy-making, upon which my account of GPL relies, is a distinctly modern notion of democracy (Landwehr Reference Letsas2010). Thus, the account makes no claims to be applicable to historical contexts.Footnote 14
The second worry is normative and would go something like this: claiming that the ideal of democracy is only applicable to contexts of rule-making (i.e. law- and policy-making) is self-defeating considering the many forms of political power that have historically been subject to demands for democratization whilst not (yet) constituting contexts of rule-making. The claim is thus problematic since it prevents us from making democratic demands on an absolute tyrant as long as he or she acted above and outside the law. This worry, however, neglects an important distinction between applicability and desirability. The argument so far has tried to define the ‘nature of the thing’ – to speak to Rawls – to which principles of GPL are intended to apply, such as principles of democratic legitimacy in contexts of rule-making (similar to the basic structure of society for Rawls’ principles of justice). These core properties constitute normatively relevant facts, that is, non-normative features that are relevant for the normative status of the political entity as a proper subject of principles of democratic legitimacy. But the proposed condition of applicability for democracy is fully compatible with democracy being a desirable ideal in all kinds of contexts. Let us assume that Anna desires to follow the principles guiding car drivers in traffic. As it happens, these principles do not apply to her since she does not have a driving license and is thus forbidden to drive. However, this does not prevent her from finding them desirable and therefore striving to become a proper subject of them by taking driving lessons to get the license. Similarly, there is nothing in the condition of applicability defended here that prevents us from demanding democratization in failed states or other chaotic political contexts, for example, by demanding basic rights and liberties, rule of law, accountability, and equality. Neither does the condition of applicability prevent us from trying to democratize political power by transforming it through legal means.
The third worry is explanatory and resembles the second. The argument is that it would be challenging for my view if it turned out, as a matter of fact, that rule-making was a consequence of democracy rather than the other way around. Let me respond to this in two steps. To begin with, explanation and justification are generally speaking distinct notions: it does not follow that if P explains Q, then it also justifies Q. A causal explanation of, for example, an action typically does not by itself constitute a justification for the action.Footnote 15 Hence, this empirical fact would not be problematic for my normative account. That said, if it turned out that rule-making could only come about through democracy, then my account – at least the principle applicable to contexts of rule-making – would indeed be rather toothless (even if not false). Fortunately, however, we have sufficient evidence from the political world that this is not the case.
Towards a dual account of GPL
As noted in the introduction, the aim of this paper is twofold. So far focus has been directed at the conceptual task of identifying some characteristics of the concept of GPL that makes it distinct from political legitimacy generally, as well as the first part of the normative task of demonstrating that the value of GPL is reducible neither to justice nor to democracy. Within this theoretical framework, the second part of the normative task will now be undertaken: to develop a normative account of GPL in which both justice and democracy play essential roles. Insofar as we accept the general descriptive characterization that GPL is concerned with the exercise of political power (purposeful and systematic power in the public domain) and that the task of principles of GPL is to regulate the relationship between entities exercising this power and those over whom it is exercised through justified procedures, as well as find plausible that political equality is a minimal condition of democratic legitimacy from the standpoint of the ideal of democracy, what normative account of GPL should we opt for?
In what follows, I will defend a dual account of GPL, consisting of two principles of legitimacy that specify the standards for when political entities are legitimate: the so-called ‘equal say principle’ as a criterion of democratic legitimacy in contexts of rule-making, and the so-called ‘principle of public legitimacy’ as a criterion of political legitimacy in contexts of power projection. As will be argued below, each principle has its distinct justification and scope. In short, the proper subjects with regard to entities exercising purposeful and systematic power will depend on whether they do so through rule-making or not. The proper subjects with regard to those over whom this power is exercised differ such that the equal say principle is applicable to agents subjected to laws and policies, whereas the principle of public legitimacy is applicable to individuals significantly affected by the exercise of political power. The principles also accommodate different uptake conditions and procedural demands.
The equal say principle in contexts of rule-making
Irrespective of how democracy is ultimately grounded, a question of concern for any account of democracy is referred to as the ‘boundary question’ (or ‘boundary problem’) in democratic theory, which deals with who is to be considered rightfully included in the group of people (‘demos’) that are supposed to rule over themselves democratically (‘kratos’) (Whelan Reference Whelan1983; Dahl Reference Dryzek1989; Habermas Reference Held1996; Arrhenius Reference Arrhenius2005; Goodin Reference Gould2007, 43). It seems as if a coherent principle of democratic legitimacy must complete two tasks: it must not only accommodate the minimal condition(s) of democratic legitimacy, that is political equality on the view proposed here, but also a criterion of rightful inclusion that is compatible with this condition. For if it is not compatible, the criterion is not a solution to the specific boundary problem in question, notably, the boundaries of the people in ‘the rule by the people’, but to some other boundary problem (and there are indeed numerous boundary problems in normative space, concerned with who are the rightful subjects of normative ideals and principles).
While the predominant answer to the boundary question in the philosophical and empirical literatures alike is the so-called ‘all-affected interests principle’, the so-called ‘all subjected principle’ is another candidate in the debate. In what follows, I will try to make the case that the all subjected principle in its general form can be made compatible with political equality in a way that the all-affected interests principle cannot. But let me first specify the equal say principle, and thereafter motivate why it takes the form of an all subjected principle.
The equal say principle states that ‘all agents that are subjected to a political entity’s rules (laws and policies) should have an equal say in the decision-making’ (Erman Reference Erman and Möller2014b). Note here that agents may not only be individuals but also, for example, states. Let me make some further specifications. First, ‘equal say’ on this principle at the minimum requires robust participation in both formal decision procedures (e.g. through electoral vote) in which a major and not fixed part of the members takes part, and informal processes (e.g. in civil society and the public sphere).Footnote 16 However, it is important that the principle stays neutral with regard to different kinds of formal decision rules, for while we are used to individual-majoritarian rule in domestic contexts, a variety of voting rules are applied in global governance arrangements (majoritarian, weighted, unanimity, and so on). For example, while the WTO exhibits majoritarian ‘one-state, one-vote’ decision rules, the World Bank applies a weighted procedure. And in international law institutions the consensus method dominate.Footnote 17
Second, ‘subjected to’ has a twofold structure on this principle. It has a ‘legal interpretation’ in the sense that those subjected to the laws have legal obligations to comply with them (Beckman Reference Beckman2009; Owen Reference Parfit2012). But it has also a ‘coercive interpretation’ in the sense that those subjected to the decisions and laws are those coerced by them, physically – through force or threats of disciplinary actions (Abizadeh Reference Abizadeh2008) – as well as socio-psychologically – through symbolic processes of socialization (Smith Reference Song2008; Abizadeh Reference Abizadeh2012). While neither of the two conditions is necessary, each of them is sufficient. This means that I disagree with Arash Abizadeh that the legal interpretation of the all subjected principle is ‘perverse’ and should be rejected just because it may allow political authorities to exercise coercive power ‘without imposing legal obligations’ (Abizadeh Reference Abizadeh2012, 878, n. 25). The fact that it is not necessary does not give us reason to reject it. On the contrary, the legal interpretation is needed to elucidate the normative significance of the two-way structure of democratic legitimacy. As illustrated by David Owen, even in cases where a law does not have any coercive power over a person, we would still want to claim that she is bound by it because she has authorized it (Owen Reference Parfit2012, 145).
So why is the equal say principle taking the form of an all subjected principle and not an all-affected interests principle? The all-affected interests principle roughly states that ‘all whose interests are significantly affected by a decision should have a say in the decision-making’ (Whelan Reference Whelan1983; Held Reference Hurrell and Macdonald1995; Shapiro Reference Smith1999; Benhabib Reference Benhabib2004; Arrhenius Reference Arrhenius2005; Goodin Reference Gould2007).Footnote 18 There are of course as many versions of this principle as there are contestations about how to best interpret ‘significantly affected’. There is also disagreement as to whether the principle should refer to those ‘possibly affected’ or those ‘actually affected’ (Goodin Reference Gould2007; Owen Reference Parfit2012). But these internal disputes are bracketed here since our concern is how the general structure of the principle feeds into the ideal of democracy. Despite the normative attractiveness of the all-affected interests principle, it is argued that it is ill-fitting for offering a criterion of rightful inclusion as part of a principle or an account of democratic legitimacy, since it is not compatible with the condition of political equality, which is an essential property of the ideal of democracy (Erman Reference Erman and Möller2014b).Footnote 19
It is commonly claimed that the all-affected interests principle is fundamentally egalitarian precisely since it counts all interests equally, and equal political power is the cornerstone of democracy (Goodin Reference Gould2007; see also Beitz Reference Beitz1989). However, it is not clear how it is able to take us from a conception of moral equality in terms of counting all interests equally, to political equality in terms of the equal decision power. For political equality is not only premised on the idea that members of a constituency are morally equal and as such have an opportunity (secured by a right) to participate in the decision-making to the extent they are significantly affected, but more specifically, that they have an equal opportunity to do so (Erman Reference Erman and Möller2014b). The difference is crucial: the all-affected interests principle allows for a proportional view of affectedness, according to which those whose interests are more affected by a decision should have more say than those less affected (Gould Reference Griffin2004; Reference MillerMacdonaldReference Macdonald and Ronzoni2008). In fact, supporting proportional influence seems sensible since it is affectedness that motivates a right to participate in the decision-making in the first place. However, this would undermine majority voting as a defensible procedure from the standpoint of democracy, since voting on an issue would generate clear winners and losers in light of the fact that it will never be the case that people are equally affected.
On the established understanding, the ideal of democracy alludes to a political system in which members rule over themselves and shape their common institutions through law- and policy-making on a wide range of political, social, legal, and economic issues. They do not decide upon a separate ready-made issue but define ‘packages of aims’ by identifying what should count as a political issue to be decided upon in the first place (Reference CohenChristianoReference Christiano2012, 33). Insofar as this system involves decision-making in terms of a proportional say due to proportional affectedness, it is at set levels (e.g. local municipality) within a legal-institutional framework that secures equal decision-making power on each level.Footnote 20
Compare the all-affected interests principle with the all subjected principle, which in its general form states that ‘all those who are subjected to the rules (laws and policies), i.e. those whose actions are governed by them, should have a say in their making’ (Dahl Reference Dryzek1989; Habermas Reference Held1996; Lopez-Guerra Reference Macdonald2005; Abizadeh Reference Abizadeh2008; Erman Reference Erman2014a, Reference Erman and Möller2014b). While people might be (and presumably are) differently affected by a society’s laws and regulations, they are still equally subjected to them. Thus, on the standard version of the all subjected principle, the criterion of inclusion is not gradually but binary coded such that one is either a rightful subject or not. It is this binary structure that makes it fitting for theorizing democratic legitimacy (Erman Reference Erman2014a). First, the binary structure opens the possibility for securing the condition of political equality through legal equality and thus the equal right (opportunity) to participate in the political decision-making. Second, democratic legitimacy is about the regulation of political power through authorization, which foremost requires subjects as agents rather than subjects as bearers of interests, even if most agents are bearer of interests too (Saunders Reference Scholte2012, 280–81). Without subjects’ exercise of agency, by taking part in the decision-making on equal footing, there would be no authorization and hence no achieving of democratic legitimacy. It is for these reasons that the equal say principle on the dual account of GPL developed here takes the form of an all subjected principle rather than an all-affected interests principle.
One objection to this line of reasoning would be to claim that since not all laws in a political community apply to all members, the all subjected principle is on closer inspection not binary coded at all, but as gradually coded as the all-affected interests principle, and thus equally unfit to be compatible with the condition of political equality. For example, some laws that apply to policemen do not apply to citizens in general, and some laws that apply to people living in the south of France do not apply to those living in Paris. However, the proportionality exemplified here is not of the right kind to support this claim. The binary structure of the all subjected principle originates from the fact that the ultimate legal competence and legal force springs from one and the same place, on the same level of governance. In the case of nation-states, it springs from the parliament on the state level: the parliament may delegate decision power to municipalities for pragmatic reasons and for reasons of subsidiarity, but it may in principle take back this power at any point.
Moving from the ‘exercise’ side of political legitimacy to the ‘recipient’ side and the uptake conditions of affected parties discussed in the conceptual section, both the strong uptake condition in the form of a duty to comply with the rules (in Buchanan’s (Reference Buchanan2010, 82–93) modified sense) and the weak uptake condition in the form of a duty of not interfering with the entity’s ruling are demanded by the equal say principle.Footnote 21
The principle of public legitimacy in contexts of power projection
In contrast to those who claim that the ideal of democracy is applicable to all contexts in which political power is exercised or in which people’s interests are significantly affected (Held Reference Hurrell and Macdonald1995; Shapiro Reference Smith1999; Benhabib Reference Benhabib2004; Gould Reference Griffin2004; Arrhenius Reference Arrhenius2005; Reference MillerMacdonaldReference Macdonald and Ronzoni2008; Forst Reference Gilabert and Lawford-Smith2014),Footnote 22 it has been argued so far that democracy is applicable only in contexts of rule-making, where GPL requires the fulfilment of the equal say principle. However, since the task of principles of GPL is to target the exercise of power in the public domain and the different ways it is projected on subjects, contexts of rule-making are, as we have seen, only one context of relevance. So what does GPL require in contexts of power projection, that is, in contexts where the exercise of power involves political activities and forms of conduct apart from law- and policy-making? It will be argued below that GPL then requires the fulfilment of the principle of public legitimacy.
In contexts of power projection, the regulation of political power is not about achieving self-determination or self-rule but about regulating forms of conduct in ways that eliminate power abuse and domination, which prevent people from exercising their autonomous agency, that is, the distinctive human capacity to form and lead an autonomous life (see Griffin Reference Habermas2008; Forst Reference Forst2011).Footnote 23 The principle of public legitimacy defended here states that ‘all persons who are significantly affected by the exercise of political power must not have their basic human rights violated and may demand accountability through impartial procedures’. It accommodates two requirements: the substantive requirement that significantly affected persons be projected to power in ways that do not violate their basic human rights; and the procedural requirement that significantly affected persons have an opportunity to hold the wielder of power to account for alleged violations of human rights through impartial procedures, which secure that equal cases are treated equally.
Note that what is suggested here is not an all-affected interests principle, discussed above, since ‘affectedness’ is targeted at agency rather than interest. The motivation for this shift stems from the conceptual framework laid out earlier, according to which principles of political legitimacy are concerned not with protecting and redistributing primary goods such as rights (answering to ‘who owes what to whom’) but with procedurally regulating the relationship between two kinds of agents, notably, those exercising political power and those over whom it is exercised. The fact that this regulation may partly be made through the protection of fundamental interests is of second-order concern (for a theory of GPL).Footnote 24
There are several reasons for why a principle based on ‘affectedness’ is preferable to a principle based on ‘subjectedness’ for contexts of power projection. To begin with, there are no basic conditions of democracy, such as political equality, which set ‘limits’ to the criterion of rightful inclusion. Some conception of moral equality – such as the equal respect for autonomous agency, the equal concern and respect for persons, or some similar liberal egalitarian principle – seems to be all that we need to formulate a criterion when it comes to properly identifying those subject to the exercise of power in the form of domination or power abuse.
While it follows conceptually from the formulations offered here that a context of power projection cannot at the same time be a context of rule-making, contexts of rule-making are likely to be contexts of power projection. This means that insofar as general norms and forms of conduct in a democratic polity significantly affect people’s agency – in the sense of preventing them from leading an autonomous life – both principles of legitimacy apply. In fact, in such cases the principle of public legitimacy is more fundamental than the equal say principle in three important respects. First, having one’s agency protected through basic human rights is an empirical condition for exercising democratic rights via practical engagement in processes of law- and policy-making. Second, being dominated or being the victim of the abuse of political power is more harmful for leading an autonomous life than not having an equal say in the law- and policy-making, and thus a greater wrong. Third, the principle of public legitimacy sets up normative boundary conditions for the exercise of power within a democratic polity through the protection of basic human rights and through impartial procedures securing accountability.
With regard to uptake conditions, the proposed account distinguishes between two contexts of power projection. In the first context, purposeful and systematic power is exercised by institutions whose main aim is to perform a beneficial function of coordination. This idea draws on Buchanan’s recently developed Metacoordination View of political legitimacy, which takes as a starting point the distinctive practical role of legitimacy judgements serving to solve metacoordination problems (Buchanan Reference Caney2013, 178). It seems reasonable to require a duty of non-interference in these kinds of institutional contexts. A duty to comply, however, appears indefensible since affected persons have no chance of influencing the substance of the general norms or forms of conduct. To see this, compare a legislative assembly with a court. While citizens should comply with the assembly’s laws, they are not obligated to respect the court’s judgements about the law (see Reference CohenChristianoReference Christiano2013).
In the second context, power is either exercised by institutions that do not aim to perform a beneficial function of coordination or is exercised through other social practices in the public domain. For sure, there is no clear-cut boundary between institutions and social practices. But on a rather conventional view, social practices are not institutions but are – together with conventions, rules, norms, and rituals and so on – among the constitutive elements of institutions (Miller 2011, 2).Footnote 25 South Africans were presumably involved in purposeful and systematic practices of discrimination before Apartheid became institutionalized, let alone before this institutional system developed into performing a beneficial function of coordination. On the view defended here, we could have rightfully demanded legitimacy also before this system was fully developed. Or consider large transnational companies whose exercise of power significantly affects people’s lives. In these cases, however, neither the strong nor the weak uptake condition seems appropriate. In fact, no particular proattitude towards an entity projecting power in this way looks defensible. Instead, the wielder of power would merely have a ‘justification’ right insofar as no basic human rights are violated and impartial procedures are established for affected persons to hold the wielder of power to account for alleged violations of these rights.
From theory to practice: some reflections on application
In view of the broad conceptual and normative approach pursued in this paper, this is not the place to go into detail about how this dual account may be applied in practice. However, a few initial remarks may be illuminating to demonstrate how it can be used when assessing the legitimacy of international organizations. To begin with, the suggested dual account requires that we distinguish analytically between international organizations that interpret and apply law, such as courts – for example International Court of Justice dealing with general disputes, and the International Criminal Court dealing with criminal prosecutions – from international organizations that are law-makers, such as the EU. In ideal-typical terms, the former entities are (only) required to fulfil the principle of public legitimacy in their interpretation and enforcement of international law, whereas the latter must also fulfil the equal say principle for the law- and policy-making to be legitimate. Of course, insofar as political entities of the former kind are established, formed and mandated by entities of the latter kind, they are democratically legitimate in an indirect sense, similar to courts, administrative institutions, and other mandated decision bodies within democratic states. However, regardless of whether they are mandated or not, they still have to fulfil the principle of public legitimacy on the defended account.
Now, every researcher doing empirical work on international organizations knows how hard it is to uphold this distinction in practice. But the complexity of the world should not make us abandon the distinction, but instead apply it also to different functions within international organizations. It is important not to let the messy world make us lose sight of the distinction between a sociological notion of legitimacy and a normative one. In order to make (normative) assessments about legitimacy deficits, I have argued that the suggested framework offers justifications for certain ways of assessing law-making functions vis-à-vis other legal functions.
Consider international organizations such as WTO. One main function of WTO is dispute settlement (the dispute settlement system includes bodies such as the DSB panels, the Appellate Body, and the WTO Secretariat) for handling the violation of trade rules and agreements signed by representatives of its members. Another main function is to oversee and manage the implementation and administration of the covered agreements. But the WTO is also in important respects a policy-maker, since it sets the rules of trade policy by establishing a policy framework for all members (Anderson and Hoekman Reference Anderson and Hoekman2006). Thus, apart from being a wielder of power by significantly affecting people’s agency through the enforcement of trade rules on national economies, it also constitutes a context of rule-making. Since the primary agents subjected to these rules (by being coerced by them and/or legally bound by them) are states, the equal say principle would require that states, rather than individuals, are included as primary rule-takers in the policy- and law-making.
This conclusion goes against the common claim among academics and practitioners that NGOs and civil society organizations ought to be included in WTO decision-making to represent those people around the world whose fundamental interests are significantly affected by the outcomes (Scholte Reference Shapiro2005; Dryzek Reference Dworkin2006; Reference MillerMacdonaldReference Macdonald and Ronzoni2008; Stevenson and Dryzek Reference Southwood2012). On my view, however, they may still rightfully participate via the impartial procedures established to satisfy the principle of public legitimacy, through which they can hold the WTO to account for basic human rights violations and demand that equal cases be treated equally. If this picture is correct, the primary task for a non-ideal theory of GPL in these cases would be to focus on how to democratize non-democratic states.
Let me end this section by considering two examples that could be used to argue that the distinction between rule-making and power projection is too blurred in today’s globalized world. Consider ICs and global administrative law (GAL) in relation to law-making functions vs. other legal functions. There is a growing literature on the legitimacy problems following from law-making by ICs (von Bogdandy et al. Reference Bohman2010). It is not the case that ICs ‘happen’ to act beyond their authorization and competences; rather, they are open ‘by design’ to leave room for ‘independent’ law-making.Footnote 26 Similarly, there is an emerging field of GAL, constituting a regulatory ‘space’ that transcends interstate relations which calls for new legitimacy mechanisms (Kingsbury et al. Reference Koenig-Archibugi2005).
It is true that both kinds of entities are engaged in law-making in some sense. However, this is not law-making in the robust sense referred to in relation to democratic legitimacy. It would be more appropriate to call it ‘mandated law-making’ and ‘rule alteration’. ICs and global administrative bodies make rules within the framework of, for example, a treaty that establishes the institution and implement the general treaty law though mandated law-making and rule alteration. Although they have some kind of independence, they are constrained so that they effectively and genuinely pursue the aims and realize the principles established by the principal parties (ChriReference Cohenstiano forthcoming).Footnote 27
Strengths and weaknesses in relation to competing accounts
This final section contrasts the proposed account with what I see as the most promising account of GPL developed so far in the philosophical literature, namely, Buchanan’s account as it has developed in recent writings. Admittedly, Buchanan’s approach is much more comprehensive, offering not only a sophisticated argument for how political legitimacy relates to political authority, obligation, and authoritativeness, but also a grounding of political legitimacy in a robust natural duty of justice implied by the principle of equal concern and respect for persons (Buchanan Reference Buchanan1999, Reference Buchanan2002, Reference Buchanan2004). My ambitions here have been much more modest: I have not been concerned with the question of ultimate grounding, but only with developing the contours of a normative account of GPL (compatible, I take it, with all or most liberal groundings in an egalitarian principle of some sort). That said, both of us develop accounts of GPL that are intended to be applicable to a wide range of different institutions and wielders of power in the global domain. In what follows, I will discuss three main differences and the possible strengths and weaknesses of my account vis-à-vis Buchanan’s.
A first difference has to do with the scope of the two accounts. Both broaden the applicability of principles of GPL beyond entities that rule in the traditional sense. But whereas Buchanan’s account concerns institutional legitimacy for institutions aiming to perform a beneficial function of coordination, my account has a broader scope, including all forms of purposeful and systematic power projection in the global public domain. The distinctive function of legitimacy assessments, according to Buchanan, is ‘to determine when institutions are worthy of having the special standing that is usually required for their performing the tasks for which they are designed and doing so without unacceptable costs’ (Buchanan Reference Caney2013, 180). An advantage of Buchanan’s more limited scope is that he is able to offer not only more fine-grained analyses of existing international institutions but also presumably more action-guiding recommendations in concrete cases. However, it comes at a normative cost, as wielders of power may then eschew the demands of GPL by simply exercising power through non-institutionalized channels and by non-formalized means (see also Forst Reference Gilabert and Lawford-Smith2014).
Hence, here I disagree not only with Buchanan but also with so-called ‘practice-dependent’ theorists, who claim that principles of legitimacy apply only to concrete and actually existing institutions (Sangiovanni Reference Saunders2008). For it is one thing to claim that principles of GPL require certain institutionalized structures – which they do according to the defended view – quite another to claim that principles of GPL are only applicable to institutionalized structures.
Importantly, though, the wider scope of the defended account does not imply a weakening of the normative demands, since the principle of public legitimacy requires different uptake conditions in the two contexts of power projection: it allows us to keep the bar high for institutions whose main aim is to perform a beneficial function of coordination (demanding a duty of non-interference, similar to Buchanan), while renouncing any particular proattitude towards entities exercising purposeful and systematic power in other ways, since such a demand would be unreasonable. We should not expect that, for example, strong economic agents, which significantly affect peoples’ lives, establish impartial procedures and accountability mechanisms for basic human rights. Some form of ‘business ethics’ is not a property of political legitimacy. Instead, similar to compliance with and enforcement of human rights generally, this must be achieved by public institutions – human rights instruments on the national, regional, and international levels, including the quasi-legal enforcement mechanisms of the human rights bodies of the United Nations.
A second difference pertains to the procedural conditions of political legitimacy. In contrast to many liberal philosophers, Buchanan agrees that substantive criteria of minimal justice are not sufficient for principles of GPL, even in the form of substantial side-constraints on procedures, but that independent procedural criteria are also required. Apart from regarding democratic rule-making as an important procedural criterion, similar to my account of GPL, Buchanan focusses mainly on two process demands: non-usurpation (discussed before) and institutional integrity. Let us compare these demands with the procedural criterion of impartiality accommodated by the principle of public legitimacy. As we saw earlier, even if non-usurpation seems to be of great importance – since we would not want to allow entities to be able to exercise power legitimately if they came to power by a wrongful process or came about through seriously unjust actions (Buchanan and Keohane Reference Buchanan2006; Buchanan Reference Caney2013, 199) – it cannot replace a procedural criterion. For while the former is about ‘institutional origins’, the latter is supposed to offer standards for the rightful exercise of power by an already existing political entity. On the defended account, such a standard is what the criterion of impartial procedure – together with respect for basic human rights – is meant to supply.
The criterion of institutional integrity fares better in this regard. In Buchanan’s view, institutional integrity demands a ‘reasonable match between the institution’s most important justifying goals and principles and its actual performance’ (Buchanan Reference Caney2013, 190). Of course, the criterion of institutional integrity and the criterion of impartial procedure are in no sense mutually exclusive. However, I argue that impartiality captures something essential about political legitimacy that institutional integrity does not. Institutional integrity is an ‘internal’ standard in the sense that it is about the gap between the actual performance of an institution and its most justifying goals and principles. Whilst this seems to be important, it does not accommodate the crucial ‘external’ and relational core of principles of political legitimacy, which is supplied by the criterion of impartiality, demanding that impartial procedures are established so that those significantly affected by the wielder of power can hold it to account for alleged violations of human rights. In a nutshell, while institutional integrity is a process demand that is fulfilled through a series of actions, impartiality is a procedural demand that is fulfilled through a set of rules.
A third difference pertains to the nature of the arguments pursued. As I made clear in the introduction, the defended dual account of GPL is ideal-theoretical in the sense that it is premised on permissive feasibility constraints such that principles of GPL must be compatible with the basic features of human nature as we know it as well as be possible to achieve from the status quo. Furthermore, it was argued that this ideal is worth pursuing as long as it does not involve extreme moral costs. In light of these constraints, most cases of non-democratic rule-making in current global affairs may reasonably be demanded to strive towards fulfilling the equal say principle.
The nature of Buchanan’s account of GPL is less clear. While his (natural duty) account of justice seems to be theorized under similar ideal-theoretical feasibility constraints, the content, role, and status of the feasibility constraints with regard to political legitimacy are ambiguous. For example, democracy is only required for political legitimacy when institutional resources for democracy are already in place (Buchanan Reference Buchanan2002, 717). With regard to international institutions, Buchanan argues, ‘not being democratic need not be an obstacle to legitimacy, if democracy is not feasible in their case, so long as they supply morally important benefits’ (Buchanan Reference Caney2013, 189). Just because democracy is not feasible in the case of international institutions generally at this point in time, ‘it would be unreasonable to deem them illegitimate simply because they are not democratic’ (Buchanan Reference Caney2013, 222).
Whether or not such a demand of democracy is reasonable depends on what kind of argument is made. There are two ways to understand Buchanan’s view: if it is an ideal-theoretical argument that is made, one would like to know which feasibility constraints are placed on the content and justification of principles of political legitimacy, as well as why it would be unreasonable to demand the fulfilment of the equal say principle in rule-making international institutions even when proper ‘democratic’ institutional resources are not yet in place.Footnote 28 However, if it is intended as a non-ideal proposal – which seems more likely considering the demanding feasibility constraints on the desirability of democracy – it looks as if Buchanan’s account is fully compatible with the view proposed here. On such a compatibility view, Buchanan’s conditions of legitimacy could be seen as capturing some core values of the equal say principle in theorizing what steps to take from where we are now, under current conditions, such as accountability for acting in the interests of all subjected parties (Buchanan and Keohane Reference Buchanan2006; Buchanan Reference Buchanan and Keohane2011, Reference Caney2013). In more general terms, we may perfectly well use one set of feasibility constraints for justifying principles of legitimacy, responding to the question of what GPL requires, and another set of feasibility constraints for theorizing applied principles for here and now, responding to the question of what we should do in the current situation if we wish to draw nearer this ideal.
To conclude, against the proposed dual account of GPL it might be objected that it is too thin from the standpoint of justice. First, the basic rights needed to prevent domination and power abuse in the exercise of intentional and systematic power in the public domain would primarily be civil and political rights rather than socio-economic rights. Second, insofar as socio-economic rights were to be included in the package of basic rights, they would satisfy only a minimal threshold of socio-economic goods without being equipped to tackle severe socio-economic inequalities.
Both observations are correct as far as they go. But, again, political legitimacy is not about the redistribution and reception of primary goods but about the regulation of political power. The task of principles of GPL is to regulate the relationship between political entities and those over whom they exercise power and it is something that must be achieved rather than received. That said, precisely because principles of GPL have this limited political function, we might regard some moral matters as more important and thus have reasons under certain circumstances to prioritize them over political legitimacy. Indeed, just because the ideal of justice is not considered sufficient to theorize GPL, the argument here does not entail that commitments to principles of global justice may not trump considerations about GPL. Recall the accessibility constraint on my project, according to which the defended account of GPL is only considered desirable and worth pursuing if the moral costs are not too high. There may well be situations where the cost of striving towards fulfilling the principles of GPL is simply too high in terms of injustice. This could be the case, for example, in contexts of law- and policy-making where the stakes between the members are too disproportionately distributed, such that some systematically have high stakes in the decisions whereas others have very little; or in cases where non-institutionalized systematic and purposeful power projection is too discriminatory even when public institutions are set up to secure impartial treatment and accountability in relation to basic human rights.
A similar accusation of thinness may be raised from the standpoint of democracy. Securing that the exercise of purposeful and systematic power in the global public domain does not violate basic human rights and having the opportunity to require accountability for these rights through impartial procedures is a far cry from having an equal say in the law- and policy-making about their contents. Again, true as far as it goes. However, from what has been said so far, either it would have to be argued that we should demand political equality in the regulation of all forms of power projection in the public domain, which seems indefensible, or we would have to defend democracy as an ideal in which political equality had no role to play, which seems equally repugnant.
Acknowledgements
The author owes special thanks to Thomas Christiano, Bob Goodin, Rainer Forst, Gustaf Arrhenius, Leif Wenar, Laura Valentini, Thomas Pogge, Mathias Koenig-Archibugi, Darrel Moellendorf, and Luke Ulas for comments on earlier drafts of this paper as well as to the participants of conferences in Tromsö (2013), Bordeaux (2013), Gothenburg (2014), and Bad Homburg (2015) for valuable suggestions. The comments from the editors and anonymous referees of International Theory have also been tremendously helpful. Finally, the author wishes to thank Marianne and Marcus Wallenberg Foundation and the Swedish Research Council for the generous funding of the research.
