Liberalism, republicanism, and the distinction between state and society

The rise of liberalism began in the sixteenth century as a counterreaction to absolutism. Its proponents considered the sovereign state as an independent legal entity separate from the law, if not above the law, which exercised its authority over citizens from a hierarchically superior position by means of public law. By contrast, relations among the citizens were presumed to be governed by the non-statal, non-positive corpus of private law.^{Footnote 23} The normative basis of private law was believed to reside in natural law, like in Locke’s account of a natural right to property;^{Footnote 24} or in practical reason, like in Kant’s post-natural law theory of right;^{Footnote 25} or in Savigny’s idea of a ‘Volksgeist’ (spirit of the people).^{Footnote 26} In all cases, liberalism emphasised the ability of society to regulate itself and assigns to the state a residual function like, for example, the enforcement of judgments.^{Footnote 27}

With some delay, liberalism has had an enormous impact on legal practice and scholarship in both civil law and common law jurisdictions (in Germany only since the eighteenth century due to the belated rise of a bourgeoisie).^{Footnote 28} In jurisdictions governed by absolute rulers, it provided the basis for an increasing division between private and public law.^{Footnote 29} On the continent, the idea of subjective, enforceable rights dominated private law,^{Footnote 30} whereas public law became the field of sometimes nearly unfettered governmental authority, immune to judicial review.^{Footnote 31} In common law jurisdictions, where the distinction between public and private law never emerged with the same clarity,^{Footnote 32} liberalist ideas led to legislative restraint in private law matters.^{Footnote 33}

This strand of thought has remained influential until today. It found a powerful reformulation in Hayek’s theory of spontaneous orders.^{Footnote 34} The trust in private self-regulation and mistrust towards traditional politics reached another level with systems theory and its claim that society is divided into functionally differentiated, self-reflexive systems like law and politics which cannot directly communicate with each other.^{Footnote 35} Even though such approaches are not based on methodological individualism and even though their goal is different from Hayek’s in that they seek a reconstruction of the modern welfare state instead of its destruction,^{Footnote 36} they tend to sing swansongs on democracy.^{Footnote 37} This brings liberalism on a collision course with the idea of democracy as society’s self-government through law.^{Footnote 38}

The republican (holistic) strand of thought does not equal the distinction between private and public law with the one between state and society. According to Rousseau, there is no law prior to the foundation of a society. All law derives from the social contract.^{Footnote 39} But Rousseau’s work does not yet distinguish between state and society. The decisive step was taken by Hegel, who coined the concept of civil society.^{Footnote 40} According to him, civil society is the place where citizens may exercise their freedom to pursue their private interests in order to meet their individual needs, while the state constitutes the realisation of the ethical idea, i.e. the collective entity which is more than just the sum of private interests.^{Footnote 41} Both private law, which lends protection to transactions in civil society, and public law are positive, statal law (‘Gesetz’).^{Footnote 42} In a republican perspective, the state enables freedom. Civil society is thus not opposed to the state, but essential to its operation.^{Footnote 43}

This strand of thought, a historic novelty at the time,^{Footnote 44} took account of the rising role of positive, statal law for the organisation of society. The Napoleonic codifications of private law epitomise the idea of sovereign control over private law.^{Footnote 45} But like liberalism, this strand of thought features certain flaws and risks. The trust in governmental capacity and mistrust in private self-regulation might not only bring about inefficient regulation. If taken to the extreme, on the basis of a pre-constitutional idea of the state and a rejection of methodological individualism, it might amount to an argument for the total state.^{Footnote 46} Similar, but different risks might arise from later theories belonging to this strand like those of Hans Kelsen. Although his neo-Kantian constitutionalism might make him an uneasy fit within the republican camp, his Rousseau-inspired approach to democracy justifies such a classification.^{Footnote 47} For Kelsen, the state is nothing but a legal entity. Private and public law do not reflect a difference between state and society, but merely a difference within the legal order.^{Footnote 48} As a consequence of his pure theory, law as such provides no justification for public authority.^{Footnote 49} Rather, such justification must derive from democratic government – a fact most often overlooked by adaptations of Kelsen in authoritarian regimes, which overemphasise and abuse the idea that the legal is legitimate.^{Footnote 50} While these risks are external to Kelsen’s theory and result from a misreading, he actually did not answer the question how social integration should be achieved, i.e. why the minority should obey the decision of the majority if compromise cannot be achieved given that the state is simply understood as a legal construct requiring no further attachment of its members.^{Footnote 51} In the absence of a discursive understanding of public law, strict forms of positivism lack explanatory capacity as to how society actually works.^{Footnote 52}

The preceding analysis reveals that, first, the distinction between the public and the private has a different function in liberal and republican theories of democracy; second, that the difference reflects diverging understandings of the relationship between state and society; and third, that both liberalism and republicanism might give rise to extreme versions which endanger individual rights or democracy. I believe that the second point is a crucial factor for the confusion surrounding the public/private law distinction, whether in a domestic or global context. We might fundamentally disagree on the relationship between state and society, so we are likely to disagree on that distinction, too. It is over such disagreement that the first point, the function of the public/private distinction for democracy, might get out of sight. It therefore seems advisable to base further reflection on the public/private distinction, especially in a global context, on a theory that attempts to combine liberalism and republicanism.

Discourse theory and the distinction between the public and the private

Discourse theory as developed by Jürgen Habermas combines liberalism and republicanism in a skilful way. It acknowledges that both the liberal and the republican understanding of the relationship between state and society, as well as the associated differences in the function of the public/private divide, reflect important aspects of modern societies.^{Footnote 53} It may therefore serve as a basis for a public/private distinction in global governance.

Instead of overemphasizing either individual or collective self-determination, Habermas parts from the idea that both fundamental rights and popular sovereignty are constitutive for modern societies and, in fact, precondition each other: in the (hypothetical) formative moment of a political community, where a people begins to understand itself as a people, the members of that community implicitly grant each other certain fundamental rights. They are the basis of the community’s life. They enable a free public discourse that feeds into law-making processes by representative institutions as well as law-application by the administration and courts. Fundamental rights are thus a condition for processes of public reasoning which ultimately lead to decisions that the members of the community have reason to accept.^{Footnote 54}

The mutual preconditioning of popular sovereignty and fundamental rights is the point where the public and the private spheres come together in discourse theory. Accordingly, the interaction of the public and the private spheres, of state and society, are essential for modern statehood. Without the former, there would be no democracy, while there would be no freedom without the latter. The political institutions of democratic states constitute society’s integrative centre. On the one hand, this is where ethical discourse in civil society about a good society is transformed into legally binding decisions reflecting society’s common interest.^{Footnote 55} This constitutes the republican element of the theory. While liberal theories claim that common interests are nothing but the aggregate of private interests, Habermas argues that strategic action in the pursuit of self-interest (‘bargaining’) is just one side of the medal in human communication. There is also communicative action whereby people seek a form of understanding that transcends the level of mutual self-interest (‘arguing’).^{Footnote 56} This enables ethical discourse that becomes constitutive for the convictions of a community. On the other hand, the political institutions may also absorb pragmatic discourse among self-interested individuals. They give rise to negotiations whereby self-interested individuals seek a compromise.^{Footnote 57} This is discourse theory’s liberal element. Thus, the democratic process is able to integrate different types of discourse: ethical self-reflection; pragmatic negotiations; as well as universal moral arguments.^{Footnote 58} Consequently, the freedom prevailing in the private sphere is not opposed to the democratic institutions at the apex of the public sphere.

Habermas deems the dividing line between the public and the private spheres, i.e. between the pursuit of collective self-determination and the realm of individual freedom, to be variable and subject to change. In principle, every issue might move into the public sphere if it is important enough to a large enough group of people.^{Footnote 59} Nevertheless, the distinction between the public and the private spheres as such remains a crucial precondition for the coexistence of democracy and individual freedom, including the freedom to pursue economic activities.^{Footnote 60} In sum, by combining elements of republicanism and liberalism and by holding the public/private distinction to be contingent and not fixed to any rigid understanding of the relationship between state and society, discourse theory lends itself as a basis for a reconceptualisation of the distinction on a global scale.

A further argument for relying on discourse theory is that it can inspire a relatively clear-cut distinction between public and private law, the source of much of our recent conceptual confusion. Accordingly, public law is tied to the political process. It has the twin functions of enabling the exercise of public authority as emanations of collective self-determination, and of limiting it out of respect for individual rights. In line with legal practice, public law makes public authority both legitimate and effective.^{Footnote 61} Like in any theory of democracy deploying the legitimizing potential of rational discourse, including those by John Rawls and Amartya Sen, this requires putting in place basic requirements of the rule of law and of democratic government. They include the separation of powers, the principle of legality, and legal remedies protecting the rights of those who disagree.^{Footnote 62} Authoritative acts respecting these public law requirements do not merely claim to be legitimate^{Footnote 63} – they may be rationally presumed to be legitimate as long as that claim is not rebutted. The conditions of legitimacy and their applications are not absolute givens but can be questioned in political discourse.^{Footnote 64}

Private law, by contrast, is composed of legal acts emanating from private individuals, not from the state. Areas traditionally termed as ‘private law’ like the law of contracts may deserve this designation because they regulate private activity, but not because they are private in an ontological sense.^{Footnote 65} Rather, as legislative acts, they constitute exercises of public authority.^{Footnote 66} Consequently, Habermas emphasises the political character of provisions determining the scope of legitimate private authority, including market activity, promoting public policy goals, or addressing market failures, such as structural inequalities among market participants.^{Footnote 67}

While these advantages make discourse theory an attractive option for tackling the problem of the private/public distinction in global governance, one should be aware of two particular aspects of discourse theory that might create problems in that respect. First, discourse theory claims that ‘hard’, enforceable law is the only form of legitimate public authority. The reason for this resides in the possibility of rational disagreement. Habermas quite realistically assumes that political processes will not always lead to a result with which everyone agrees. Indeed, decisions often only receive the support of the majority. According to what has been said above, those in the minority have good reasons to accept the decision nevertheless. But they will not always do so. It is for this reason that Habermas emphasises the dual function of the law originally devised by Kant: one might follow the law because it is legitimate, or one might do so because disobedience might entail negative consequences such as enforcement action.^{Footnote 68} ‘The point of law is to enable us to act in the face of disagreement.’^{Footnote 69} Second, discourse about non-universal, ethical questions might reveal areas of fundamental disagreement among the participants. According to Habermas, social integration through rational discourse will therefore only work in communities sharing a common ‘life-world’ characterised by a background consensus about the idea of a good, ethical life. Such a background consensus will reduce the risk of rupture as a consequence of disagreement over specific political questions.^{Footnote 70} The following section will show how, when turning to global governance, these two aspects require some recalibrations of the discourse theoretical framework devised by Habermas.

Public and private authority, not hard law

Habermas’ theory of democracy is based on the understanding that hard, binding law is a form of authority, and in fact, the only relevant one, because it can be enforced by the government. Other types of influence are not equally coercive and therefore do not necessarily require democratic legitimacy. This view, which I call the ‘constraint theory’ of law, has been very popular in legal scholarship and political thought since early modernity. It might originate in an anthropology of early modernity that resonated especially well with Protestantism.^{Footnote 71} In late antiquity, Augustine had built his theory of society on the idea that every person is an imago dei.^{Footnote 72} Medieval scholastics had presumed human nature to be amenable to both good and evil purposes and ends.^{Footnote 73} The Protestant Reformation, however, firmly established the idea of the natura corrupta, of the sinful character of human nature since the fall.^{Footnote 74} This is a core theme in Luther’s theory of justification. Without man’s sinful nature, there would have been no need for the passion of Christ.^{Footnote 75} Theories like that of Hobbes secularised this anthropology. He concluded that society requires a Leviathan to keep in check the rational egoists of which it is composed.^{Footnote 76} From there, the ‘constraint theory’ of law made its career through legal and political theory. It found a prominent representation in Kant’s philosophy.^{Footnote 77} Later, the idea of the homo oeconomicus reinforced a narrow concept of self-interest that could only be modified by rules that were enforceable. It maculated theories as diverse as the ones proposed by Smith,^{Footnote 78} Austin,^{Footnote 79} or even Marx and Engels.^{Footnote 80} Only Max Weber began unpacking the relationship between law and physical enforcement. His concept of command, albeit state-centric, includes orders corroborated by psychological pressure.^{Footnote 81}

The reality of contemporary global governance suggests that hard, enforceable law is no longer the only relevant means for the organisation of international society, maybe not even the predominant one. The international level has recently experienced a development which Michel Foucault observed to have taken place on the domestic level since the end of the eighteenth century, namely the diversification of the instruments and methods by which governments effectively influence society.^{Footnote 82} The concept of global governance epitomises the fact that states, international organisations, and non-state actors use many new types of instruments that do not have the status of a source of international law in the sense of Article 38 of the Statute of the International Court of Justice.^{Footnote 83} These instruments take the form of all kinds of soft law, but also of entirely non-legal instruments, such as information, including indicators and indexes.^{Footnote 84} Many of these instruments seem to have an impact on individual freedom or collective self-determination and might therefore require democratic legitimacy.

A discourse theoretical approach to global governance that wants to ensure its democratic legitimacy cannot gloss over these phenomena. It needs a wider concept of authority, one that encompasses not just hard law, but all kinds of instruments by which global governance affects people’s lives. This wider concept of authority should become the reference point for the public/private distinction. It will enable the public/private distinction to operate effectively and single out from a larger array of authoritative instruments those which require democratic legitimacy. Importantly, one cannot reach such a wider definition of authority by simply adding further sources to the concept of hard international law, or by extending the scope of existing ones.^{Footnote 85} For the concept of authority also needs to be able to map the influence of information. If the concept of law (hard or soft) is to make any sense, it should not comprise simple information that addresses our cognitive, not our normative expectations.^{Footnote 86} One should therefore decouple the concept of authority from the concept of law (hard or soft) and understand authority more broadly as the law-based capacity to legally or factually limit or otherwise affect other persons’ or entities’ use of their freedom.

The definition contains a few elements which require further clarification. First, what does it mean to be ‘affected’ by a governance instrument? This is a search for relevant forms of influence beyond governmental enforcement. One might argue that this requires a new anthropology. The most sublime forms of influence cannot count as authority, because then everything would count as authority and probably require democratic legitimacy. Rather, as the Kantian idea of the dual function of the law endorsed by Habermas shows, authority is characterised by the specific property that it does not rely only on voluntary compliance on the part of its addressees. They might always disagree. Therefore, authoritative acts are in a position to trigger extrinsic constraints which reduce the risk that the addressees might disobey in case of disagreement. Beyond governmental enforcement, which (potential) extrinsic constraints suffice for an act to be considered authoritative? The answer might benefit from the insights of motivational psychology. Accordingly, extrinsic motivation is not confined to physical force. Ryan and Deci distinguish four forms of extrinsic motivation. They enjoy some empirical plausibility for the explanation of human behaviour and might therefore serve as ideal types which guide the development of a wider concept of authority.^{Footnote 87} These four forms of extrinsic motivation range in between the poles of complete demotivation and pure intrinsic motivation. Each form assumes that a person’s behavior is influenced by external events. The forms vary with respect to the type of impact and the degree to which a person perceives such impact as imposed on her from the outside:^{Footnote 88}

1. – The first form of extrinsic motivation is called ‘external regulation’. It describes motivation induced by positive or negative sanctions which people perceive as external, such as physical force, economic incentives, or the potential of reputational consequences. External regulation is the usual way in which binding law triggers extrinsic motivation. But even soft law might motivate its addressees in this way to the extent that it influences their reputation.^{Footnote 89}

2. – The second form, called ‘introjection’,^{Footnote 90} presupposes a certain self-understanding, a predisposition on the part of the addressee, which allows certain external events to have the effect of a positive or negative sanction. The addressee feels motivated by external constraint, though only partially. Examples from global governance include the discursive constraints to which participants in international courts and other fora are exposed due to their wish to be and remain a member of a certain group. This resembles what Goodman and Jinks call ‘acculturation’.^{Footnote 91} In the context of legal argumentation, one could speak of ‘semantic authority’.^{Footnote 92}

3. – The third form is called ‘identification’ and captures situations where the addressee considers external influence as conducive to his or her own goals and partially internalizes it. Examples include governance by information, which triggers learning processes.^{Footnote 93}

4. – The fourth form is called ‘integration’ and describes situations where the addressee completely internalizes some external influence and is not aware of the fact that his or her reasons for acting have been induced externally. Examples include cognitive framing through terminology, categories, and ideas.^{Footnote 94}

The qualification of an instrument as an act of authority requires a comprehensive assessment of all forms of extrinsic motivation that it might trigger. However, since only the first and the second form of extrinsic motivation correspond to situations where the addressee feels compelled by external constraints, a discourse theoretical understanding of authority always presupposes the first or second form of extrinsic motivation, at least to some degree. The third and fourth forms of extrinsic motivation might corroborate or reinforce the authority of the instrument in question. They do not suffice alone for qualifying an instrument as an act of authority.

Another element of the definition of authority requiring clarification is the requirement that it be ‘law-based’. This has two implications. First, affectedness as the sole criterion would be incompatible with discourse theory: affectedness only emerges ex post. Discourse theory, like the pedigree thesis in legal positivism,^{Footnote 95} requires that the legal qualification of an act does not to hinge on its actual ex post effects, but that law makers determine their legal status ex ante. ^{Footnote 96} One might solve the supposed paradox by using standardised ex-ante descriptions of acts which typically entail effects that justify considering them as authoritative. Hence, it is necessary to develop ‘standard instruments’ of global governance which legally define categories of authoritative acts by a set of formal parameters.^{Footnote 97} Such a broadened concept of authority might level off much of the critique accusing public law approaches of being out of touch with empirical reality,^{Footnote 98} without, however, committing itself to a purely sociological perspective on the law.^{Footnote 99}

Second, authority in whatever form (hard or soft law, information) needs to be ‘law-based’, i.e. based on a legal mandate to exercise such authority. This requirement distinguishes authority from brute force on the one hand, and from mere morality on the other. Gunman-style brute force does not claim, nor receive, any legal justification. It posits itself outside the law. The authority of morality claims to be universal, not to depend on the law. Authority should be considered as law-based inasmuch as the author may reasonably claim the necessary legal mandate.^{Footnote 100} Law-based authority claims legitimacy, although the claim can be rebutted.^{Footnote 101}

The shift from hard law to a broader concept of authority that includes soft law, information and other, more subtle forms of influence thus provides a point of attachment for the public/private distinction. This implies that the concept of authority applies to both private and public authority. Private and public authority might cause similar effects. They only require different justifications. It will be the task of the public/private distinction to identify those authoritative instruments which require democratic legitimacy. But before proceeding to this distinction, another problem faced by discourse theory when transferred to a global context requires consideration.

The case for cosmopolitan pluralism

Shifting discourse theory to the global level raises another important question. The public in the original, nation-state-centred version of discourse theory is the state. Its citizens exercise collective self-determination. Beyond the state, there is no defined community which would constitute the subject of collective self-determination and legitimise international public authority.

Such legitimacy would be all the more necessary since the authority exercised by international institutions might have distributive effects which transcend the borders of their members. Especially acts of non-universal organisations like the OECD might affect third states or the competence of other institutions. Some institutions can only provide limited legitimacy for the potentially far-reaching effects of their decisions, such as private standard setters or dispute settlement tribunals. Legitimizing international public authority by state consent is miserable comfort from a discourse theoretical point of view. It does not even come close to the idea that each person should have one vote.^{Footnote 102} So what should be the relevant public, the community which might legitimately exercise public authority for the purpose of self-determination on the international level? And how should it be determined?

Radical pluralist theories suggest giving up the idea of universal principles guiding the exercise of public authority in a pluralist global environment.^{Footnote 103} But then again, it seems difficult to imagine how legitimate decisions should emanate from uncoordinated processes of mutual contestation, unpredictable not only in a substantive, but also a procedural sense; or how international courts should develop an ‘ordre public international’ without the possibility of determining the ‘public’.^{Footnote 104} Proposals for a responsive pluralism based on decentralised, loose mechanisms of coordination among the various communities within a pluralistic setting might not do the trick, either.^{Footnote 105} They merely avoid the question of the universal principles that need to guide such coordination should it be successful and legitimate.^{Footnote 106} On the other side of the spectrum, cosmopolitan theories suggest taking the individual as the be-all and end-all of collective self-determination.^{Footnote 107} However, neither the institutional, nor the social prerequisites for worldwide democracy exist. Cosmopolitan approaches might therefore end up concluding that there is no legitimate global public actor.^{Footnote 108} This appears unsatisfactory given that, first, many international institutions claim to promote public interests, and second, this corresponds to claims for effective international regulation voiced by public opinion at many occasions and on a worldwide scale.^{Footnote 109} It might therefore be worthwhile looking for alternatives between the extremes of a largely unstructured world society and a cosmopolitan world state.

Pluralistic approaches which recognise the political role of communities other than the state are anything but new.^{Footnote 110} In what might appear as a cosmopolitan reinterpretation of such approaches, and following the European Union’s model of dual legitimacy,^{Footnote 111} Habermas has suggested a dual strategy, which considers both individuals (world citizens) and states as the relevant subjects from which the legitimacy of international public authority needs to derive.^{Footnote 112} Besides individuals and states, one might also consider other entities as possible subjects of legitimacy, such as international and supranational organisations, or non-governmental organisations.^{Footnote 113} Accordingly, one and the same person might be a member of various, partly overlapping communities, commensurate to the multiplicity of identities of men and women in modern pluralistic societies.^{Footnote 114} This amounts to the idea of a pluralistic cosmopolitan (or constitutional) order which has found much support in the literature.^{Footnote 115} The cosmopolitan underpinning within an overall pluralist structure provides the ground for a thin layer of common principles, the ‘trinitarian commitment’ of democracy, human rights and the rule of law.^{Footnote 116}

But the question is whether groups beyond the level of the state are actually communities in a discourse theoretical understanding. What characterises a community as opposed to, for example, the parties to an exchange contract pursuing their mutual self-interest? Why should the parties to an exchange contract not be considered a community? This question is far from trivial, since liberal theories claim that common interests are nothing but the aggregate of private interests. At a key point in ‘Between Facts and Norms’, Habermas, referring to Kant, emphasises the difference between an exchange contract based on a convergence of the self-interests of rational egoists, and a social contract, which establishes a common interest, thereby becoming an end in itself.^{Footnote 117} Hence, for discourse theory, a community is characterised by its ability to express a common interest by way of communicative action (or ‘arguing’). By contrast, private interests reflect mainly the outcomes of strategic action.^{Footnote 118} There is thus a qualitative difference between common interests and aggregate private interests. This provides an epistemological basis for the identification of communities beyond the state. Indeed, the point has been made that states engage in communicative action when they meet in international organisations and similar fora.^{Footnote 119} Empirically, it is not easy to clearly distinguish communicative and strategic action in such settings.^{Footnote 120} But discourse theory recognises that practical discourse always consists in a mix of pragmatic (i.e. strategic action), moral (i.e. universal communicative action), and ethical (i.e. particular communicative action) discourse.^{Footnote 121} I consider communities beyond the state to be characterised by the prevalence of ethical discourse. The subsequent section will elaborate how their identification can be facilitated.

In a discourse theoretical perspective, the problem with such a cosmopolitan pluralist order is that the institutional and communicative structures of most of these communities might not meet the requirements for a traditional discourse theoretical justification of the exercise of public authority.^{Footnote 122} They often lack bodies that would represent the citizens concerned, or a corresponding public sphere for opinion-formation. From the viewpoint of discourse theory, global governance risks resulting in a technocratic form of authority, removed from the citizens affected, characterised by non-inclusive decision-making in fragmented institutions.^{Footnote 123} Habermas has therefore suggested that apart from the UN Security Council, whose decisions he considers as embedded in a thick layer of converging moral convictions, international institutions should refrain from decisions with distributive consequences.^{Footnote 124}

Yet, given contemporary global economic, ecological, security, and other challenges, the chances implied in such decisions might outweigh the risks. Inaction often does not seem to be an option which any responsible decision-maker would prefer.^{Footnote 125} This calls for the acceptance of less than perfect solutions. The lack of representative structures and public spheres might be mitigated on the international level by mechanisms of participation, transparency, and accountability.^{Footnote 126} And maybe there is more of a public sphere on the international level than common wisdom assumes. Empirical research has traced the emergence of transnational cleavages which amount to elements of world public opinion.^{Footnote 127} Even if one assumes that there is no global (or cosmopolitan) public like there is a domestic one,^{Footnote 128} one might tap the potential of domestic, regional and functional communities. They might contribute to debates about global issues to the extent that their respective public spheres are not only inward-oriented.^{Footnote 129} These proposals are far from perfect and probably not in line with an orthodox reading of discourse theory. But they might render international institutions as good as it gets and legitimate enough in order to satisfy the pressing need for cooperation in a world society characterised by enormous social, economic, environmental and other challenges of a global scale.

The concept of community as a question of identity

The concept of community elaborated in the preceding section comprises some degree of vagueness that might render its application difficult. But the legal nature of the actor may function as a proxy facilitating the identification of a community. In the following, I will first argue that communities beyond the state are characterised by a common identity rather than by a shared life-world. Second, I will show that the legal status of a certain community allows for a good guess about the ‘thickness’ of its common identity. It may therefore serve for the classification of communities according to their legal status.

Regarding the first point: discourse theory rejects any kind of foundational concept of community that would be held together by substantive bonds reaching back to times immemorial like a common history or ethnic or linguistic affiliation, as traditional accounts of nationhood have it.^{Footnote 130} The only requirement for a community is a communicative one: in a community, ethical discourse prevails and enables social integration. According to Habermas, one can only engage in meaningful ethical discourse if one shares a common life-world, i.e. if one holds shared basic convictions about an ethical life resulting from processes of opinion-formation in the public sphere of an open society.^{Footnote 131} Despite what has been said about world public opinion,^{Footnote 132} the members of the fragmented communities of a cosmopolitan pluralist order do not share a common life-world which is as thick and rich as that shared by the citizens of many states. They might only possess a rudimentary array of shared convictions. The latter is perhaps better called ‘identity’ instead of life-world, since it does not amount to the thick layer of common basic convictions usually associated with a life-world. Identity in this respect does not refer to sameness, but to shared elements in the self-understanding of the members of a community on the supranational level.^{Footnote 133} This concept of identity is not to be confused with the idea of a constitutional identity in the debate about the delimitation of domestic and supranational authority.^{Footnote 134} Rather, this concept of identity refers to a social phenomenon that is independent from the law and presupposed by it.^{Footnote 135} Identity designates the frames necessary for value judgments, which hinge on a person’s self-understanding. That in turn depends on how she sees herself in relation to others. Identity is therefore an intersubjective construction characterised by the dynamic interplay of the identities of the self and of our respective communities.^{Footnote 136} In the case of a community composed of abstract entities such as states, the relevant community identity will always include the individual members of these entities such as citizens. This follows from the idea advocated above that individuals are the ultimate subjects of legitimacy.^{Footnote 137}

As to the second point: by virtue of their structure, some legal entities might be more predisposed to bring about ethical discourse than others. They might therefore be presumed to constitute a community. This is based on the following consideration: the idea of the identity of a community has a factual and a normative dimension. On the one hand, the members of a community may as a matter of fact share a stronger or a weaker form of community identity; on the other hand, community membership might require its members to share a stronger or a weaker form of community identity. A strong community identity resembles a life-world.^{Footnote 138} The stronger the community identity is as a matter of fact, the better will ethical reasoning work. The more a community requires its members to share a common identity, the less inclusive it will be and the more externalities it will produce. As a result, the ‘ideal’ community may rely on a strong shared identity of its members (i.e. a common life-world) which facilitates communicative action, but it may not require its members to share much of that identity, thereby keeping it inclusive.

It is not difficult to derive from this that the modern liberal state assumes a predominant role for the exercise of public authority. It has few identity requirements and protects the freedom of its citizens to make up their minds about their idea of a good life.^{Footnote 139} At the same time, most states may rely on the broad, shared life-world of its members. The same applies for substate territorial entities, although the shared identity of their citizens may vary by degrees.

With respect to international organisations, one needs to distinguish the United Nations as a universal organisation with broad powers from other more specialised or regional organisations. According to the preamble of its charter, the United Nations requires even less commitment from its member states or their members than a state. However, the citizens of the world are far from sharing a common life-world. Only some rudimentary shared convictions might exist, which are at best coextensive with the principles of the preamble. Thus, while the United Nations is an extremely inclusive actor, communicative action is hard to organise. It may therefore very well exercise public authority, but such authority must not be very intrusive, unless it corresponds to the few shared convictions that might find worldwide acceptance.

Specialised international organisations like the WTO or regional organisations like the Council of Europe require a specific kind of common identity on the part of their members. In case of the WTO, they need to subscribe to the idea of free trade, and the Council of Europe requires commitment to rather specific human rights obligations. This should in principle facilitate the potential of these organisations to bring about communicative reasoning, although that might be higher in regional organisations like the OECD than in organisations with broad membership like the WTO. At the same time, they have a huge potential to create externalities both for non-member states and for policy areas outside their jurisdiction over which their shared identity does not extend.

Non-profit corporations established under domestic law usually require their members to share a strong community identity like, for example, the purposes of the World Wildlife Fund. This alleviates the need for deliberative structures in their internal organisation. Thus, under German law, the internal regulations of non-profit corporations only need to guarantee a minimum of procedural fairness to its members. However, the structure of organisations with an important economic or social position (i.e. a quasi-monopoly) need to respect democratic principles.^{Footnote 140} In line with this, the Model Nonprofit Corporation Act of the American Bar Association proposes a number of provisions on the internal organisation which ensure that governance of the organisation is member-driven. Corporations may not derogate from some of these provisions.^{Footnote 141} The flip side of the strong community identity is, again, the risk that they might create externalities for non-members. On the domestic level, the rules of private law keep these externalities in check. On the international level, there might be lacunae. In this respect, so-called hybrid corporations which count governmental institutions among their members might have an advantage since this membership structure broadens the basis of the relevant community, sometimes including the citizens of the respective territorial entities in a more direct fashion. This is one of the reasons why the concept of hybridity needs to be ‘unpacked’.^{Footnote 142}

A crucial question is whether groups composed of for-profit corporations are communities, such as professional associations or the invisible group of global players from which the lex mercatoria emerged. Habermas considers economic activity as the exclusive domain of strategic action of self-interested actors. It does contribute to social integration, but not through communicative action. Only their legal framework establishes a relationship between markets and the integrative forces of communicative action.^{Footnote 143} While it is true that economic activity and market forces are ultimately geared towards mutual self-interest, I find it difficult to imagine that market participants would not occasionally switch to communicative action.^{Footnote 144} Indeed, the fact that transnational market participants have created a lex mercatoria,^{Footnote 145} or that professional organisations routinely engage in norm-setting^{Footnote 146} provides strong hints that market participants might every now and then switch from bargaining to arguing about what is best for the community in order to maintain or improve the functioning of the market as a whole.^{Footnote 147} On the domestic level, the predominance and specificity of positive law relegates such communicative action among market participants to a back seat, perhaps with the exception of a few principles of honest trade. The situation is different for the international level. Here, the more substantial the shared identity of economic actors is, the more they might be expected to successfully engage in communicative action – and the more externalities they might create for non-participants. But it depends on the individual case whether such groups are to be considered as communities.

A different thing is the relationship between for-profit associations and their members. While partnerships might occasionally witness communicative action (although strategic action might prevail), communicative action is extremely unlikely to come about among the shareholders of a public company. There is no identity requirement except for the ownership of at least one share, and no shared community identity may be presumed. Therefore, this is the ideal-typical model of a legally constituted group which regulates its relationship by way of private authority, the antipode of the state.^{Footnote 148}

By way of an intermediate conclusion, it turned out to be possible to transfer the discourse theoretical understanding of legitimate authority to global governance, provided that one focuses on authority, not on hard law. The present cosmopolitan pluralist order has the capacity to legitimise such authority through its various communities. The communities which form part of this order can largely be recognised by their legal nature.

What the distinction is not

Under the Westphalian model, the distinction between public and private authority was relatively straightforward: in the end, all public authority derived from the state.^{Footnote 149} The idea of citizenship legitimised state authority to a sufficient degree, since the citizens of the state were presumed to be by and large identical with those affected by its exercise of authority. This approach still pervades contemporary conceptions of the public/private divide. Accordingly, in German administrative law, public authority is defined as authority exercised on a legal basis empowering an institution established for the exercise of state authority in that capacity. These institutions and their powers may ultimately be derived from the federal or state constitutions.^{Footnote 150} Certainly, this model always comprises an element of fiction.^{Footnote 151} Alien residents have never had a right to vote, while emigrants sometimes have. Transborder externalities of the exercise of state authority have often affected the citizens of other states. Nevertheless, the model worked as long as those empowering public authority were by and large identical with those affected by it.

Under the conditions of cosmopolitan pluralism, the Westphalian model no longer holds. Sticking to distinctions between private and public law, authority, or actors, leads to enormous conceptual complexities.^{Footnote 152} Many overlapping communities on different levels exercise authority whose effects regularly transcend the borders of the respective communities. What is ‘the public’ which may define the public interest? One might avoid the issue – but never resolve it – by referring to notions of ‘hybridity’. Instead, some try to operate with a material definition of publicness which defines certain issues as matters of public interest.^{Footnote 153} But a fixed distinction between public and private interests or subject matters contradicts discourse theory, which holds that the delimitation is politically contingent and subject to rational discourse.^{Footnote 154}

Alternatively, one might claim that private authority is consensual and public authority non-consensual. However, in most modern theories of democracy, public authority is ultimately based on the consent of those governed, whether that consent is expressed in some – fictitious – social contract or in continuous political participation. Consent in private-law relationships is often no less fictitious. One only needs to think of standard terms, or the decisions of corporate executive boards which generate effects for shareholders.^{Footnote 155}

Others advance ‘affectedness’ as the decisive criterion identifying public authority. In her impressive account of a pluralistic global stakeholder democracy, Macdonald states that ‘power should be designated as “public” and subject to democratic control whenever it impacts in some problematic way upon the capacity of a group of individuals to lead autonomous lives’.^{Footnote 156} However, this solution involves major problems, both practically and theoretically. First, which persons should count as ‘affected’ since one act may affect a myriad of people in at least an infinitesimal way?^{Footnote 157} Also, different acts will almost never affect exactly the same group of people, which would make it necessary to define a new group of stakeholders in each case.^{Footnote 158} Second, to rely on affectedness as a criterion for distinguishing private and public authority amounts to giving up that very distinction. If private authority is defined by its lack of effects, it seems that it would not qualify as authority at all under the terms of the definition of authority stated above.^{Footnote 159} However, this is entirely out of step with the role of private authority in a purely domestic setting where the distinction between the private and the public is relatively straightforward: domestic private authority may very well ‘affect’ other persons or groups, i.e. constitute an external constraint on their freedom. Contracts, their terms and enforcement might turn out to contravene the changing interests and preferences of parties and thereby compromise their freedom – even in the absence of unfair contractual clauses or power asymmetries among the contracting parties. Third, the group of ‘those affected’ is not a community intended and constituted for the exercise of collective self-determination on the basis of ethical discourse. It is not even a community, but only a more or less random group. Approaches based solely on the effects of authority would thus be tainted by a liberal bias which takes private autonomy and the exercise of private authority as the be-all and end-all of the organisation of society and does not recognise such a thing as the collective self-determination of communities. Therefore, the criterion of ‘affectedness’ does not allow for a meaningful distinction between private and public authority.

Public and private authority as different perspectives

My proposal is based on the concept of community set out earlier. The difference between private and public authority neither resides in the author, nor in the effects of such authority. Rather, considering discourse theory’s roots in a theory of action, I believe the difference to reside in the mode of action pursued by such authority. What distinguishes public from private authority is the presence of a community that pursues a common interest determined in an act of collective self-determination.^{Footnote 160} By contrast, private authority pursues the (mutual) self-interest of the actor(s) as part of their individual self-determination. When translating this distinction between common and private interests to a pluralistic, cosmopolitan environment characterised by multiple communities with overlapping membership, public authority becomes a matter of perspective: public authority is the authority exercised within one such community. On that basis, I propose the following definition: public authority is an act of authority whose actor reasonably claims to be mandated to act on behalf of a community of which the observer is a member, or a member of such member.

Some crucial explanations are in order. First, one needs to carefully distinguish the descriptive dimension of the definition and its normative consequences. The definition has a descriptive purpose, namely that of identifying acts which require a specific kind of legitimacy because the actor claims to promote the interest of a community on its behalf, i.e. to act in furtherance of the community’s collective self-determination. Whether an act thus identified actually has that kind of legitimacy is a different question. This is where the normative upshot of the above definition comes into play. Accordingly, public authority needs to be equipped with a legal framework that ensures it enjoys a presumption of legitimacy. In a discourse theoretical perspective, this requires structures of adequate representation, participation, transparency, accountability, as well as human rights protection that will bring about ethical and moral discourse based on communicative action, not only pragmatic compromise based on strategic action. The assumption of legitimacy is rebuttable.^{Footnote 161}

To clarify what I mean here, my approach can be contrasted with that of Kingsbury and Donaldson. While I agree with Kingsbury and Donaldson that public authority is the authority adopted in the name of a public (i.e. a community),^{Footnote 162} they establish the additional requirement that such exercises of authority need to respect certain defined substantive and procedural principles in order to be considered as public authority. By contrast, the present definition of public authority rests on the conviction that it is better not to fuse legality with legitimacy. Public authority should comprise acts which claim to be legitimate because they meet the applicable legal requirements – and nothing more. This move gives the concept of public authority its particular emancipatory thrust. Classifying as public authority all acts which claim a certain legitimacy (the descriptive dimension of the concept) immediately engages the responsibility of their authors to ensure that the act effectively respects individual and collective self-determination (the normative dimension of the concept). By contrast, the proposal of Kingsbury and Donaldson leaves open what should happen with acts that do not meet all public law principles. Should one classify them as ‘private authority’ (or private law, since they do not use the concept of authority)? Or as ‘non-authority’ (non-law)? If one does not classify such acts, one runs the risk of defining the problem away. By way of example: I do not see who would be served by categorizing the most egregious decisions, such as the UN Security Council’s listing of terror suspects, as private authority (or as non-authority) instead of public authority, albeit a potentially illegitimate version of it. The Kingsbury/Donaldson proposal features the reverse problem of Kelsen’s pure theory of law and democracy: while Kelsen did not want to tie his concept of law to any requirement of democracy, Kingsbury and Donaldson define as law (authority in the terminology used here) only that which is democratically sound. Kelsen’s theory ran the risk of beatifying authoritative acts with problematic impacts and doubtful legitimacy, Kingsbury/Donaldson run the risk of losing them out of sight. And while Hayek overestimated private self-coordination to the detriment of democratic forms of interaction, they might overestimate the latter to the detriment of the former. Possibly in recognition of these risks, Kingsbury and Donaldson recently clarified that they did not intend to draw a sharp distinction between private and public law (or authority).^{Footnote 163} In any case, the present definition takes a middle road and establishes a connection between authority and legitimacy that is stronger than in Kelsen’s theory because it claims that legality and legitimacy should ideally converge, and more flexible than Kingsbury’s and Donaldson’s original definition since it recognises that legality and legitimacy do not always converge as a matter of fact.

Second, public authority is always exercised within a community. This is why I refer to the observer: one and the same act may affect different people or groups, some of which may be members of the community on whose behalf the actor is mandated to act, others may not. For the non-members, the act might be considered as one of private authority. Accordingly, the rules of private law apply to it, be it domestic contract and tort law or their international law equivalents like the Vienna Convention on the Law of Treaties and the customary law of state responsibility as codified in the Articles on State Responsibility. They ensure that such authority is exercised on the basis of consent, the justificatory backbone of the private law paradigm. But the exercise of private authority does not require democratic legitimacy. To give an example: authoritative acts of non-universal international organisations such as the OECD are acts of public authority for their members. This means that they need to derive their democratic legitimacy from the member states and their citizens. However, they do not need to derive democratic legitimacy from non-member states and their citizens, even if they cause indirect effects for them, such as an act of the OECD causing policy changes in OECD member states that might affect their trade with non-member states. Only if the OECD could credibly claim to have a legal mandate to act also on behalf of non-members, the latter would see themselves confronted with an act of public authority.

The case is different if the impact of such private authority on the self-determination of those affected becomes too strong to be justified by consent, for example, because it causes enormous, non-remediable externalities or is exercised in the context of huge, inacceptable power disparities. This would be the case if an OECD policy establishes a global standard that has enormous consequences for non-member states, such as the OECD policy on tax havens.^{Footnote 164} In such a case, two options remain: either a competent public authority such as the state or an international institution needs to regulate the exercise of such private authority. Fundamental rights might afford additional legal protection and mitigate power disequilibria to the extent that they have horizontal effects.^{Footnote 165} Or, the exercise of such authority needs to be transformed into public authority to be exercised by an institution with a mandate allowing it to act on behalf of a more inclusive community, a community comprising those so severely affected.^{Footnote 166} Ideally, this would lead to the convergence of the group of those more than only marginally affected by an authoritative act, with the community in whose name such authority is being exercised. Such a shift from private to public authority might require the creation of new international institutions.^{Footnote 167} It would be illiberal to prescribe an existing group to expand its membership. It may only do so of its own. If one allows people to form communities for the purpose of self-determination, there will be exclusion. Public law can only impose restrictions on whether and how such communities may exercise authority.

Certainly, it is a political question, a question of legitimacy and not one of legality, to determine the point at which someone needs to intervene and regulate private authority, or to transform it into public authority by shifting it to a more inclusive institution. This issue is therefore up for contestation, and absent a world parliament, there is no forum that could ultimately decide this question in a legitimate manner. However, the United Nations General Assembly provides a relatively inclusive forum for the discussion of such matters, although it only comprises states.^{Footnote 168}

Third, it is not sufficient that the actor subjectively has the intention to pursue the community’s interest. The actor also needs to claim to be mandated to act on behalf of a community. It belongs to the community’s collective self-determination to select the persons or entities that may pursue its common interest, and how. Also, only those who claim to be empowered by the community face claims for democratic legitimacy, not those who voluntarily align their action with what they think is the common interest.

It is irrelevant whether the mandate has a legal basis in soft or hard law. As set out above, I consider both soft and hard law as potential acts of authority, hence as acts that might empower other exercises of authority. Likewise, the legal basis may be contained in private or in public law as traditionally understood. Associations governed by domestic private law such as standardisation organisations or professional associations might very well exercise public authority over their members (and even beyond, if acting upon a mandate by an international institution). This aspect of the definition might refute one of the strongest criticisms advanced by the school of transnational legal pluralism against the publicness approach.^{Footnote 169} Under which circumstance private law associations may exercise public authority hinges on the concept of community and requires some further consideration, which the next section will provide. Finally, public authority might be based on international or domestic law. In the former case, we refer to international public authority.

Fourth, the actor exercising public authority needs to make a reasonable claim that he or she is mandated to act on behalf of a certain community. Such a claim is reasonable if the actor may with some plausibility invoke a legal basis entitling him to act on behalf of the respective community.^{Footnote 170} This is an intersubjective criterion which is meant to exclude gross abuses of power. Striking ultra vires acts should not be beatified by according to them the status of public authority, since public authority implies a claim to at least potential legitimacy, not only in the understanding of discourse theory.^{Footnote 171} However, this has nothing to do with the question whether such egregious ultra vires acts may trigger the responsibility of the agent or institution. One should not mix up questions of iustitia distributiva (public authority and its legitimacy) with questions of iustitia commutativa (damages) or criminal law sanctions.

Fifth, the above definition not only applies if the affected entity is a member of the community in whose name the act has been adopted, but also if it is a member of such member. The latter case concerns in particular acts of international or supranational institutions which directly affect private individuals of their member states. Examples comprise arrest warrants of the International Criminal Court, judgments of international courts and tribunals where one of the parties is a private individual, or acts of information such as Interpol red notices. Each of these examples has direct legal or factual consequences for private individuals. States granted the respective international institution the necessary powers to exercise such authority directly over their citizens. This necessitates considering such authority as public, even though the individual affected is not directly a member of the acting institution.

Operationalizing the distinction

This section operationalizes the distinction set out above and points out some of its ramifications for international law in general. One of the most striking revelations of this definition is that certain parts of classical public international law should best be understood as exercises of private authority – at least from an international viewpoint. This includes all international legal acts which do not derive from an actor acting for a community, but by which rational actors pursue their egoistic interests. Declarations of war are an obvious case, and a large amount of bilateral agreements might also fall in this category – the gravel of the law of coordination.^{Footnote 172} Of course, international treaties are authoritative. They bind the parties legally and create obligations for them which constrain their freedom. States parties may not easily discard them, even if their preferences change, unless they are ready to bear the consequences, such as reputational losses, court procedures, or any kind of countermeasure. However, bilateral treaties are not necessarily public – at least from the perspective of international law. From the perspective of domestic law, such acts may very well be classified as exercises of public authority, since they oblige the entire community which constitutes the state.

Equally striking might be the insight that certain acts of private non-profit corporations are to be considered as exercises of public authority from the internal viewpoint of their members. However, this does not mean that the law governing non-profit corporations should be taught in classes on constitutional or administrative law. From the perspective of domestic law as applied by a domestic court, internal acts of non-profit corporations are acts of private authority because they are not adopted in the name of the state. Again, the classification hinges on the perspective. The same considerations might apply to some legal relationships in family law.

Some cases are quite hard to classify. The distinction between acta iure gestionis and acta iure imperii evokes the private/public law divide. Indeed, one might argue that in the case of the former, the state acts as a rational egoist, whereas in the latter, it acts on behalf of a community, even though the addressee of such acts does not have to be a member of that community, but I am not certain whether this leads to practical solutions.

Another hard case is immigration. Immigrants and refugees are often exposed to forms of authority which I hesitate to qualify as private. Certainly, from the viewpoint of domestic law, all government acts concerning immigration are clear cases of public authority since they are exercised in the name of the community. Naturalisation touches upon the self-definition of the community.^{Footnote 173} However, I doubt whether they should be qualified as public authority vis-à-vis the immigrant. That would have the paradoxical consequence that the decision about immigrant status would need to derive democratic legitimacy from a community which includes the immigrant, although the question at stake is exactly whether the immigrant should be included in that community. This is the result that follows descriptively from the application of the above definition. Whether this result is normatively satisfactory is an entirely different question.^{Footnote 174} Such decisions create huge externalities for the immigrant. In this respect, the Refugee Convention provides at least a few, perhaps insufficient, external constraints on this exercise of private authority (‘private’ only from the viewpoint of the immigrant). In any case, once a certain status has been accorded to immigrants, such as a residence permit, further elements of democratic inclusion might be required (such as a right to vote, which might be limited to certain specialised assemblies like councils of immigrants, or territorial assemblies like town councils). In that respect, one and the same government may act on behalf of different communities – the community of its citizens, the community of all residents, and even looser communities held together by only rudimentary traits of shared identity.